AuroraMiddleware/skia2
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
4824cf4c1e
skia2/tests/PathOpsSimplifyTest.cpp

6020 lines
204 KiB
C++
Raw Normal View History

Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
/*
* Copyright 2012 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#include "PathOpsExtendedTest.h"
#define TEST(name) { name, #name }
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(2,0);
path.lineTo(1,1);
path.lineTo(0,0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine1x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(2,0);
path.lineTo(1,1);
path.lineTo(0,0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
static void addInnerCWTriangle(SkPath& path) {
path.moveTo(3,0);
path.lineTo(4,1);
path.lineTo(2,1);
path.close();
}
static void addInnerCCWTriangle(SkPath& path) {
path.moveTo(3,0);
path.lineTo(2,1);
path.lineTo(4,1);
path.close();
}
static void addOuterCWTriangle(SkPath& path) {
path.moveTo(3,0);
path.lineTo(6,2);
path.lineTo(0,2);
path.close();
}
static void addOuterCCWTriangle(SkPath& path) {
path.moveTo(3,0);
path.lineTo(0,2);
path.lineTo(6,2);
path.close();
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addInnerCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine2x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addInnerCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addInnerCCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addInnerCCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addInnerCWTriangle(path);
addOuterCCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addInnerCWTriangle(path);
addOuterCCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3b(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addInnerCCWTriangle(path);
addOuterCCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3bx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addInnerCCWTriangle(path);
addOuterCCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine4(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addOuterCCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine4x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addOuterCCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine5(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addOuterCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine5x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addOuterCWTriangle(path);
addOuterCWTriangle(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine6(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
path.moveTo(2,0);
path.lineTo(6,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine6x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
path.moveTo(2,0);
path.lineTo(6,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
path.moveTo(6,0);
path.lineTo(2,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
path.moveTo(6,0);
path.lineTo(2,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,0);
path.lineTo(4,0);
path.lineTo(2,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7b(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,0);
path.lineTo(4,0);
path.close();
path.moveTo(6,0);
path.lineTo(2,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine7bx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,0);
path.lineTo(4,0);
path.close();
path.moveTo(6,0);
path.lineTo(2,0);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine8(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(2,4);
path.lineTo(6,4);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine8x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(2,4);
path.lineTo(6,4);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine9(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(6,4);
path.lineTo(2,4);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine9x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(6,4);
path.lineTo(2,4);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine10(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(2,1);
path.lineTo(3,4);
path.lineTo(6,1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine10x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,4);
path.lineTo(4,4);
path.lineTo(2,2);
path.close();
path.moveTo(2,1);
path.lineTo(3,4);
path.lineTo(6,1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine10a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,4);
path.lineTo(8,4);
path.lineTo(4,0);
path.close();
path.moveTo(2,2);
path.lineTo(3,3);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine10ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0,4);
path.lineTo(8,4);
path.lineTo(4,0);
path.close();
path.moveTo(2,2);
path.lineTo(3,3);
path.lineTo(4,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
static void addCWContainer(SkPath& path) {
path.moveTo(6,4);
path.lineTo(0,4);
path.lineTo(3,1);
path.close();
}
static void addCCWContainer(SkPath& path) {
path.moveTo(0,4);
path.lineTo(6,4);
path.lineTo(3,1);
path.close();
}
static void addCWContents(SkPath& path) {
path.moveTo(2,3);
path.lineTo(3,2);
path.lineTo(4,3);
path.close();
}
static void addCCWContents(SkPath& path) {
path.moveTo(3,2);
path.lineTo(2,3);
path.lineTo(4,3);
path.close();
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine11(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addCWContainer(path);
addCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine11x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addCWContainer(path);
addCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine12(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addCCWContainer(path);
addCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine12x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addCCWContainer(path);
addCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine13(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addCWContainer(path);
addCCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine13x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addCWContainer(path);
addCCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine14(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
addCCWContainer(path);
addCCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine14x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
addCCWContainer(path);
addCCWContents(path);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine15(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine15x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine16(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 4, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine16x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 4, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine17(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine17x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine18(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 4, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine18x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 4, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine19(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 16, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine19x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 16, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine20(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine20x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine21(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 16, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine21x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 16, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine22(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine22x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine23(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine23x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine24a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(2,0);
path.lineTo(4,4);
path.lineTo(0,4);
path.close();
path.moveTo(2,0);
path.lineTo(1,2);
path.lineTo(2,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine24ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(2,0);
path.lineTo(4,4);
path.lineTo(0,4);
path.close();
path.moveTo(2,0);
path.lineTo(1,2);
path.lineTo(2,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine24(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine24x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine25(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine25x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine26(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine26x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine27(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 8, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine27x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 8, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine28(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine28x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine29(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 12, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine29x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 18, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 12, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine30(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine30x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine31(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 4, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine31x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 4, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine32(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine32x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine33(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine33x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine34(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine34x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine35(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 0, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine35x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 0, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine36(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 10, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine36x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 10, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine37(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 24, 30, 30, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine37x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 24, 30, 30, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine38(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCW_Direction);
path.addRect(12, 12, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine38x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCW_Direction);
path.addRect(12, 12, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine40(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 18, 24, 24, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine40x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 18, 24, 24, SkPath::kCW_Direction);
path.addRect(4, 16, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine41(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 24, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine41x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 24, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine42(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(8, 16, 17, 17, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine42x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(8, 16, 17, 17, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine43(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 24, 18, 18, SkPath::kCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine43x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 24, 18, 18, SkPath::kCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine44(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 32, 27, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine44x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 32, 27, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine45(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine45x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine46(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 0, 36, 36, SkPath::kCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine46x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 0, 36, 36, SkPath::kCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine47(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine47x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine48(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine48x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine49(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine49x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine50(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine50x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine51(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine51x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine52(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 30, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 20, 18, 30, SkPath::kCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine52x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 30, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 20, 18, 30, SkPath::kCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine53(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 20, 24, 30, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine53x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 20, 24, 30, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine54(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 0, 18, 18, SkPath::kCW_Direction);
path.addRect(8, 4, 17, 17, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine54x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 0, 18, 18, SkPath::kCW_Direction);
path.addRect(8, 4, 17, 17, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine55(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 6, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine55x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 6, 18, 18, SkPath::kCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine56(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 20, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine56x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(18, 20, 30, 30, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine57(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(20, 0, 40, 40, SkPath::kCW_Direction);
path.addRect(20, 0, 30, 40, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine57x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(20, 0, 40, 40, SkPath::kCW_Direction);
path.addRect(20, 0, 30, 40, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine58(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine58x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 0, 12, 12, SkPath::kCCW_Direction);
path.addRect(0, 12, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine59(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 6, 18, 18, SkPath::kCCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine59x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 6, 18, 18, SkPath::kCCW_Direction);
path.addRect(4, 4, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine60(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine60x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(6, 12, 18, 18, SkPath::kCCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine61(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 0, 24, 24, SkPath::kCCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine61x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 0, 24, 24, SkPath::kCCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine62(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine62x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine63(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 10, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine63x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 10, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 6, 12, 12, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine64(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 6, 30, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine64x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 6, 30, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine65(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 0, 36, 36, SkPath::kCW_Direction);
path.addRect(32, 6, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine65x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 0, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 0, 36, 36, SkPath::kCW_Direction);
path.addRect(32, 6, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine66(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 30, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 20, 24, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine66x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 30, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 20, 24, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine67(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine67x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68b(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68bx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68c(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68cx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68d(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68dx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 4, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68e(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68ex(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68f(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68fx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68g(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68gx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68h(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine68hx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 8, 8, SkPath::kCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(2, 2, 6, 6, SkPath::kCCW_Direction);
path.addRect(1, 2, 2, 2, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine69(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine69x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine70(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 24, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine70x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 24, 12, 12, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine71(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 0, 24, 24, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine71x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 20, 20, SkPath::kCW_Direction);
path.addRect(12, 0, 24, 24, SkPath::kCW_Direction);
path.addRect(12, 32, 21, 36, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine72(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(6, 20, 18, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine72x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 40, 30, 30, SkPath::kCW_Direction);
path.addRect(6, 20, 18, 30, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine73(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 40, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine73x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(0, 40, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCW_Direction);
path.addRect(0, 0, 9, 9, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine74(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(20, 30, 40, 40, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(32, 24, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine74x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(20, 30, 40, 40, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(32, 24, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine75(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 0, 30, 30, SkPath::kCCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine75x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCW_Direction);
path.addRect(10, 0, 30, 30, SkPath::kCCW_Direction);
path.addRect(18, 0, 30, 30, SkPath::kCCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine76(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(36, 0, 66, 60, SkPath::kCW_Direction);
path.addRect(10, 20, 40, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(32, 6, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine76x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(36, 0, 66, 60, SkPath::kCW_Direction);
path.addRect(10, 20, 40, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(32, 6, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine77(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(20, 0, 40, 40, SkPath::kCW_Direction);
path.addRect(24, 6, 36, 36, SkPath::kCCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine77x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(20, 0, 40, 40, SkPath::kCW_Direction);
path.addRect(24, 6, 36, 36, SkPath::kCCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine78(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 30, 60, SkPath::kCW_Direction);
path.addRect(10, 20, 30, 30, SkPath::kCCW_Direction);
path.addRect(18, 20, 30, 30, SkPath::kCCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine78x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 30, 60, SkPath::kCW_Direction);
path.addRect(10, 20, 30, 30, SkPath::kCCW_Direction);
path.addRect(18, 20, 30, 30, SkPath::kCCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine79(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 36, 60, 30, SkPath::kCW_Direction);
path.addRect(10, 30, 40, 30, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine79x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 36, 60, 30, SkPath::kCW_Direction);
path.addRect(10, 30, 40, 30, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine81(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(-1, -1, 3, 3, SkPath::kCW_Direction);
path.addRect(0, 0, 1, 1, SkPath::kCW_Direction);
path.addRect(0, 0, 1, 1, SkPath::kCW_Direction);
path.addRect(0, 0, 1, 1, SkPath::kCW_Direction);
path.addRect(1, 1, 2, 2, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(2, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate1x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(2, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate2x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate3(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate3x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(2, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate4(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testDegenerate4x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 1);
path.lineTo(2, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate1x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 1);
path.lineTo(2, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(0, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 2);
path.lineTo(0, 3);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate2x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.lineTo(0, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 2);
path.lineTo(0, 3);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate3(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 1);
path.lineTo(1, 1);
path.lineTo(0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate3x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 1);
path.lineTo(1, 1);
path.lineTo(0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate4(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(0, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 2);
path.lineTo(0, 3);
path.lineTo(1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testNondegenerate4x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.lineTo(0, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 2);
path.lineTo(0, 3);
path.lineTo(1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral5(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(3, 2);
path.lineTo(3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral5x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(3, 2);
path.lineTo(3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral6(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral6x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.lineTo(1 + 1.0f/3, 2.0f/3);
path.close();
path.moveTo(1 + 1.0f/3, 2.0f/3);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.lineTo(1 + 1.0f/3, 2.0f/3);
path.close();
path.moveTo(1 + 1.0f/3, 2.0f/3);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6b(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(6, 6);
path.lineTo(0, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6bx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(6, 6);
path.lineTo(0, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6c(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 3);
path.lineTo(3, 0);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6cx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 3);
path.lineTo(3, 0);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6d(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 3);
path.lineTo(3, 0);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(6, 6);
path.lineTo(0, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testFauxQuadralateral6dx(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 3);
path.lineTo(3, 0);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(4, 2);
path.close();
path.moveTo(4, 2);
path.lineTo(6, 6);
path.lineTo(0, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral6a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral6ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(6, 0);
path.lineTo(0, 6);
path.lineTo(6, 6);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral7(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.lineTo(2, 2);
path.lineTo(1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral7x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.lineTo(2, 2);
path.lineTo(1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral8(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(3, 1);
path.lineTo(1, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 1);
path.lineTo(0, 2);
path.lineTo(3, 2);
path.lineTo(2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral8x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(3, 1);
path.lineTo(1, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 1);
path.lineTo(0, 2);
path.lineTo(3, 2);
path.lineTo(2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral9(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 1);
path.lineTo(2, 1);
path.lineTo(1, 3);
path.lineTo(2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral9x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(1, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 1);
path.lineTo(2, 1);
path.lineTo(1, 3);
path.lineTo(2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine1a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 0, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine1ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 0, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine2ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 20, 20, 20, SkPath::kCW_Direction);
path.addRect(0, 20, 12, 30, SkPath::kCW_Direction);
path.addRect(12, 0, 21, 21, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine3aax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(18, 20, 30, 30, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine4ax(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(10, 30, 30, 30, SkPath::kCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(0, 32, 9, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic1x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(3, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic2x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(3, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic3(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic3x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic4(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic4x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic5(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic6(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic7(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(3, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 0, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic8(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic9(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(3, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(1, 2, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic14(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(3, 2, 3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic15(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(0, 1);
path.quadTo(1, 1, 0, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic17x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 3, 1);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(3, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic18(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic19(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic20(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic21(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic22(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 2, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic23(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 2, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic24(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(2, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic25(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic26(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic27(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic28(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 2);
path.quadTo(1, 2, 0, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic29(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 2, 1);
path.lineTo(0, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic30(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 2);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic31(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 2);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic32(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic33(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(2, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic34(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 0, 0, 1);
path.lineTo(0, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.quadTo(2, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic35(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 1, 1, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(3, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic36(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 1, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(3, 1);
path.lineTo(1, 2);
path.quadTo(3, 2, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic37(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 2, 1, 2);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(3, 1);
path.quadTo(0, 2, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic38(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.quadTo(0, 1, 1, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 2);
path.quadTo(2, 2, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic51(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(369.863983f, 145.645813f);
path.quadTo(382.380371f, 121.254936f, 406.236359f, 121.254936f);
path.lineTo(369.863983f, 145.645813f);
path.close();
path.moveTo(369.970581f, 137.94342f);
path.quadTo(383.98465f, 121.254936f, 406.235992f, 121.254936f);
path.lineTo(369.970581f, 137.94342f);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic53(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(303.12088f, 141.299606f);
path.lineTo(330.463562f, 217.659027f);
path.lineTo(303.12088f, 141.299606f);
path.close();
path.moveTo(371.919067f, 205.854996f);
path.lineTo(326.236786f, 205.854996f);
path.quadTo(329.104431f, 231.663818f, 351.512085f, 231.663818f);
path.lineTo(371.919067f, 205.854996f);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic55(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(303.12088f, 141.299606f);
path.lineTo(330.463562f, 217.659027f);
path.lineTo(358.606506f, 141.299606f);
path.lineTo(303.12088f, 141.299606f);
path.close();
path.moveTo(326.236786f, 205.854996f);
path.quadTo(329.104431f, 231.663818f, 351.512085f, 231.663818f);
path.lineTo(326.236786f, 205.854996f);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic56(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(366.608826f, 151.196014f);
path.quadTo(378.803101f, 136.674606f, 398.164948f, 136.674606f);
path.lineTo(354.009216f, 208.816208f);
path.lineTo(393.291473f, 102.232819f);
path.lineTo(359.978058f, 136.581512f);
path.quadTo(378.315979f, 136.581512f, 388.322723f, 149.613556f);
path.lineTo(364.390686f, 157.898193f);
path.quadTo(375.281769f, 136.674606f, 396.039917f, 136.674606f);
path.lineTo(350, 120);
path.lineTo(366.608826f, 151.196014f);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine80(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(4, 0);
path.lineTo(3, 7);
path.lineTo(7, 5);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 6);
path.lineTo(6, 12);
path.lineTo(8, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic58(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path ops work in progress BUG= Review URL: https://codereview.chromium.org/18058007 git-svn-id: http://skia.googlecode.com/svn/trunk@9908 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-07-08 17:17:02 +00:00
path.moveTo(283.714233f, 240);
path.lineTo(283.714233f, 141.299606f);
path.lineTo(303.12088f, 141.299606f);
path.lineTo(330.463562f, 217.659027f);
path.lineTo(358.606506f, 141.299606f);
path.lineTo(362.874634f, 159.705902f);
path.lineTo(335.665344f, 233.397751f);
path.lineTo(322.12738f, 233.397751f);
path.lineTo(295.718353f, 159.505829f);
path.lineTo(295.718353f, 240);
path.lineTo(283.714233f, 240);
path.close();
path.moveTo(322.935669f, 231.030273f);
path.quadTo(312.832214f, 220.393295f, 312.832214f, 203.454178f);
path.quadTo(312.832214f, 186.981888f, 321.73526f, 176.444946f);
path.quadTo(330.638306f, 165.90802f, 344.509705f, 165.90802f);
path.quadTo(357.647522f, 165.90802f, 364.81665f, 175.244537f);
path.lineTo(371.919067f, 205.854996f);
path.lineTo(326.236786f, 205.854996f);
path.quadTo(329.104431f, 231.663818f, 351.512085f, 231.663818f);
path.lineTo(322.935669f, 231.030273f);
path.close();
path.moveTo(326.837006f, 195.984955f);
path.lineTo(358.78125f, 195.984955f);
path.quadTo(358.78125f, 175.778046f, 343.709442f, 175.778046f);
path.quadTo(328.570923f, 175.778046f, 326.837006f, 195.984955f);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic59x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(3, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic59(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(3, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic63(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(3, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 1);
path.quadTo(2, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic64(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(2, 3);
path.close();
path.moveTo(1, 2);
path.lineTo(2, 2);
path.quadTo(0, 3, 3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic65(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(3, 2);
path.close();
path.moveTo(2, 1);
path.lineTo(2, 2);
path.quadTo(0, 3, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic67x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 1);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(1, 1, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic68(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 2, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic69(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(2, 0);
path.lineTo(1, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic70x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 2, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic71(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 1, 3, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic72(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 2);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic73(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 0, 3);
path.lineTo(0, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic74(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 3);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic75(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic76(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(2, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 2);
path.quadTo(1, 2, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic77(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 1);
path.lineTo(3, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic78(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 2);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic79(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 1, 2);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(0, 2);
path.lineTo(2, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 0);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight3(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 2);
path.lineTo(2, 0);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight4(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(2, 0);
path.lineTo(0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight5(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(1, 2);
path.lineTo(0, 2);
path.lineTo(2, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight6(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(2, 0);
path.lineTo(0, 2);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight7(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 1);
path.lineTo(2, 1);
path.lineTo(2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight8(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(2, 1);
path.lineTo(0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight9(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(1, 2);
path.lineTo(2, 1);
path.lineTo(0, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testEight10(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(1, 0);
path.lineTo(0, 1);
path.lineTo(2, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic80(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(1, 0, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(3, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic81(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic82(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(0, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic83(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 0);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 1);
path.lineTo(0, 2);
path.quadTo(2, 2, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic84(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(2, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.quadTo(0, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic85(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(3, 0, 1, 1);
path.lineTo(1, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(3, 0);
path.quadTo(0, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic86(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 1, 1, 1);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 1, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic87(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(0, 2, 3, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic88(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(2, 1, 0, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.quadTo(0, 2, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic89x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(3, 1, 2, 2);
path.lineTo(0, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 1);
path.quadTo(3, 1, 3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic90x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(3, 0, 2, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic91(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(3, 2, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(2, 1, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic92x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.quadTo(3, 0, 2, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(20, 0, 40, 40, SkPath::kCCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(24, 32, 33, 36, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82a(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82b(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82c(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82d(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82e(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82f(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82g(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine82h(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 0, 6, 10, SkPath::kCCW_Direction);
path.addRect(2, 2, 4, 4, SkPath::kCCW_Direction);
path.addRect(2, 6, 4, 8, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine83(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(10, 30, 30, 40, SkPath::kCCW_Direction);
path.addRect(0, 12, 12, 18, SkPath::kCCW_Direction);
path.addRect(4, 13, 13, 16, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine84(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(0, 12, 60, 30, SkPath::kCCW_Direction);
path.addRect(10, 20, 40, 30, SkPath::kCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine84x(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 12, 60, 30, SkPath::kCCW_Direction);
path.addRect(10, 20, 40, 30, SkPath::kCCW_Direction);
path.addRect(0, 12, 12, 12, SkPath::kCCW_Direction);
path.addRect(4, 12, 13, 13, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testLine85(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.addRect(36, 0, 66, 60, SkPath::kCCW_Direction);
path.addRect(20, 0, 40, 40, SkPath::kCCW_Direction);
path.addRect(12, 0, 24, 24, SkPath::kCCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 1);
path.lineTo(2, 2);
path.lineTo(2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testCubic1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0, 0);
path.cubicTo(0, 1, 1, 1, 1, 0);
path.close();
path.moveTo(1, 0);
path.cubicTo(0, 0, 0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic93(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(3, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(0, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.quadTo(1, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testCubic2(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,2);
path.cubicTo(0,3, 2,1, 4,0);
path.close();
path.moveTo(1,2);
path.cubicTo(0,4, 2,0, 3,0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad1(skiatest::Reporter* reporter, const char* filename) {
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
SkPath path;
path.moveTo(0,0);
path.quadTo(0,0, 0,1);
path.lineTo(1,1);
path.close();
path.moveTo(0,0);
path.quadTo(1,1, 0,2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral2(skiatest::Reporter* reporter, const char* filename) {
fix a few pathops test bugs exposed on Mac and Linux git-svn-id: http://skia.googlecode.com/svn/trunk@8601 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-10 18:53:01 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(2, 2);
path.lineTo(0, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(3, 0);
path.lineTo(0, 1);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
fix a few pathops test bugs exposed on Mac and Linux git-svn-id: http://skia.googlecode.com/svn/trunk@8601 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-10 18:53:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic94(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(8, 8);
path.quadTo(8, 4, 4, 4);
path.quadTo(4, 0, 0, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic95(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
SkPath path;
path.moveTo(8, 8);
path.lineTo(0, 0);
path.quadTo(4, 0, 4, 4);
path.quadTo(8, 4, 8, 8);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic96(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
SkPath path;
path.moveTo(8, 0);
path.lineTo(0, 8);
path.quadTo(0, 4, 4, 4);
path.quadTo(4, 0, 8, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadratic97(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
SkPath path;
path.moveTo(0, 8);
path.lineTo(8, 0);
path.quadTo(4, 0, 4, 4);
path.quadTo(0, 4, 0, 8);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testTriangles1(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress fix bugs in tests on 32 bit release Most changes revolve around pinning computed t values very close to zero and one. git-svn-id: http://skia.googlecode.com/svn/trunk@8745 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-18 15:58:21 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 2);
path.lineTo(1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress fix bugs in tests on 32 bit release Most changes revolve around pinning computed t values very close to zero and one. git-svn-id: http://skia.googlecode.com/svn/trunk@8745 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-18 15:58:21 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testTriangles2(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress fix bugs in tests on 32 bit release Most changes revolve around pinning computed t values very close to zero and one. git-svn-id: http://skia.googlecode.com/svn/trunk@8745 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-18 15:58:21 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(3, 3);
path.close();
path.moveTo(1, 1);
path.lineTo(2, 3);
path.lineTo(1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress fix bugs in tests on 32 bit release Most changes revolve around pinning computed t values very close to zero and one. git-svn-id: http://skia.googlecode.com/svn/trunk@8745 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-18 15:58:21 +00:00
}
cumulative pathops patch Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer. Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering. Unify t and point values so that small intervals can be eliminated in one pass. Break cubics up front to eliminate loops and cusps. Make the Simplify and Op code more regular and eliminate arbitrary differences. Add a builder that takes an array of paths and operators. Delete unused code. BUG=skia:3588 R=reed@google.com Review URL: https://codereview.chromium.org/1037573004
2015-03-26 14:52:43 +00:00
// A test for this case:
path ops -- handle non-finite numbers Op() and Simplify() do nothing if the input is non-finite. Add code and tests. Review URL: https://codereview.chromium.org/14407006 git-svn-id: http://skia.googlecode.com/svn/trunk@8882 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-26 19:51:16 +00:00
// contourA has two segments that are coincident
// contourB has two segments that are coincident in the same place
// each ends up with +2/0 pairs for winding count
// since logic in OpSegment::addTCoincident doesn't transfer count (only increments/decrements)
// can this be resolved to +4/0 ?
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testAddTCoincident1(skiatest::Reporter* reporter, const char* filename) {
path ops -- handle non-finite numbers Op() and Simplify() do nothing if the input is non-finite. Add code and tests. Review URL: https://codereview.chromium.org/14407006 git-svn-id: http://skia.googlecode.com/svn/trunk@8882 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-26 19:51:16 +00:00
SkPath path;
path.moveTo(2, 0);
path.lineTo(2, 2);
path.lineTo(1, 1);
path.lineTo(2, 0);
path.lineTo(2, 2);
path.lineTo(1, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(2, 2);
path.lineTo(3, 1);
path.lineTo(2, 0);
path.lineTo(2, 2);
path.lineTo(3, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- handle non-finite numbers Op() and Simplify() do nothing if the input is non-finite. Add code and tests. Review URL: https://codereview.chromium.org/14407006 git-svn-id: http://skia.googlecode.com/svn/trunk@8882 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-26 19:51:16 +00:00
}
// test with implicit close
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testAddTCoincident2(skiatest::Reporter* reporter, const char* filename) {
path ops -- handle non-finite numbers Op() and Simplify() do nothing if the input is non-finite. Add code and tests. Review URL: https://codereview.chromium.org/14407006 git-svn-id: http://skia.googlecode.com/svn/trunk@8882 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-26 19:51:16 +00:00
SkPath path;
path.moveTo(2, 0);
path.lineTo(2, 2);
path.lineTo(1, 1);
path.lineTo(2, 0);
path.lineTo(2, 2);
path.lineTo(1, 1);
path.moveTo(2, 0);
path.lineTo(2, 2);
path.lineTo(3, 1);
path.lineTo(2, 0);
path.lineTo(2, 2);
path.lineTo(3, 1);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- handle non-finite numbers Op() and Simplify() do nothing if the input is non-finite. Add code and tests. Review URL: https://codereview.chromium.org/14407006 git-svn-id: http://skia.googlecode.com/svn/trunk@8882 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-26 19:51:16 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad2(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 1);
path.close();
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad3(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad4(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(2, 0);
path.quadTo(0, 1, 1, 1);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(0, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad5(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(2, 0);
path.quadTo(0, 1, 2, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad6(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(2, 0);
path.quadTo(0, 1, 2, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.quadTo(0, 1, 1, 1);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad7(skiatest::Reporter* reporter, const char* filename) {
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
SkPath path;
path.moveTo(3, 0);
path.quadTo(0, 1, 1, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(3, 0);
path.quadTo(0, 1, 1, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- fix skp bugs This fixes a series of bugs discovered by running the small set of Skia skp files through pathops to flatten the clips. Review URL: https://codereview.chromium.org/14798004 git-svn-id: http://skia.googlecode.com/svn/trunk@9042 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-05-07 18:51:31 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadLineIntersect1(skiatest::Reporter* reporter, const char* filename) {
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(3, 1, 0, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 1);
path.quadTo(3, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadLineIntersect2(skiatest::Reporter* reporter, const char* filename) {
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(3, 1, 0, 3);
path.lineTo(0, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 1);
path.quadTo(3, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadLineIntersect3(skiatest::Reporter* reporter, const char* filename) {
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
SkPath path;
path.moveTo(0, 0);
path.quadTo(3, 1, 0, 3);
path.lineTo(1, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 1);
path.quadTo(3, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops -- rewrite angle sort This is a major change resulting from a minor tweak. In the old code, the intersection point of two curves was shared between them, but the intersection points and end points of sorted edges was computed directly from the intersection T value. In this CL, both intersection points and sorted points are the same, and intermediate control points are computed to preserve their slope. The sort itself has been completely rewritten to be more robust and remove 'magic' checks, conditions that empirically worked but couldn't be rationalized. This CL was triggered by errors generated computing the clips of SKP files. At this point, all 73M standard tests work and at least the first troublesome SKPs work. Review URL: https://codereview.chromium.org/15338003 git-svn-id: http://skia.googlecode.com/svn/trunk@9432 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-06-04 17:59:42 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void skphealth_com76(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress BUG= Review URL: https://codereview.chromium.org/18058007 git-svn-id: http://skia.googlecode.com/svn/trunk@9908 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-07-08 17:17:02 +00:00
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.moveTo(708.099182f, 7.09919119f);
path.lineTo(708.099182f, 7.09920025f);
path.quadTo(704.000000f, 11.2010098f, 704.000000f, 17.0000000f);
path.lineTo(704.000000f, 33.0000000f);
path.lineTo(705.000000f, 33.0000000f);
path.lineTo(705.000000f, 17.0000000f);
path.cubicTo(705.000000f, 13.4101496f, 706.455078f, 10.1601505f, 708.807617f, 7.80761385f);
path.lineTo(708.099182f, 7.09919119f);
path.close();
path.moveTo(704.000000f, 3.00000000f);
path.lineTo(704.000000f, 33.0000000f);
path.lineTo(705.000000f, 33.0000000f);
path.lineTo(719.500000f, 3.00000000f);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress BUG= Review URL: https://codereview.chromium.org/18058007 git-svn-id: http://skia.googlecode.com/svn/trunk@9908 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-07-08 17:17:02 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void tooCloseTest(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 1);
path.lineTo(1,-1);
path.close();
path.moveTo(0, 0);
path.lineTo(1,-2);
path.lineTo(1, 2);
path.lineTo(2, 0);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testRect1(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
SkPath path;
path.addRect(0, 0, 60, 60, SkPath::kCCW_Direction);
path.addRect(30, 20, 50, 50, SkPath::kCCW_Direction);
path.addRect(24, 20, 36, 30, SkPath::kCCW_Direction);
path.addRect(32, 24, 36, 41, SkPath::kCCW_Direction);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testRect2(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.moveTo(0, 0);
path.lineTo(60, 0);
path.lineTo(60, 60);
path.lineTo(0, 60);
path.close();
path.moveTo(30, 20);
path.lineTo(30, 50);
path.lineTo(50, 50);
path.lineTo(50, 20);
path.close();
path.moveTo(24, 20);
path.lineTo(24, 30);
path.lineTo(36, 30);
path.lineTo(36, 20);
path.close();
path.moveTo(32, 24);
path.lineTo(32, 41);
path.lineTo(36, 41);
path.lineTo(36, 24);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testTriangles3x(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(2, 1, 0, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad8(skiatest::Reporter* reporter, const char* filename) {
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
SkPath path;
path.moveTo(3, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(0, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(3, 0);
path.quadTo(1, 1, 2, 2);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress path ops work in progress BUG= Review URL: https://codereview.chromium.org/21359002 git-svn-id: http://skia.googlecode.com/svn/trunk@11291 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-09-16 15:55:01 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testTriangles4x(skiatest::Reporter* reporter, const char* filename) {
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@11585 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-10-03 07:01:37 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(2, 0, 0, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 1);
path.quadTo(3, 2, 2, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
path ops work in progress make more skps work remove edit files BUG= Review URL: https://codereview.chromium.org/23542056 git-svn-id: http://skia.googlecode.com/svn/trunk@11570 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-10-02 14:49:34 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad9(skiatest::Reporter* reporter, const char* filename) {
optimize pathops coverage Remove unused code from SkOpSegment.cpp and friends. Add new tests exposed by coverage. Fix a bug exposed by coverage -- removing the need to detect points that are nearby when intersecting. Add gyp rule for building coverage flavor on Mac. R=mtklein@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/75453003 git-svn-id: http://skia.googlecode.com/svn/trunk@12344 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-21 17:04:29 +00:00
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.quadTo(2, 1, 1, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
optimize pathops coverage Remove unused code from SkOpSegment.cpp and friends. Add new tests exposed by coverage. Fix a bug exposed by coverage -- removing the need to detect points that are nearby when intersecting. Add gyp rule for building coverage flavor on Mac. R=mtklein@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/75453003 git-svn-id: http://skia.googlecode.com/svn/trunk@12344 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-21 17:04:29 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad10(skiatest::Reporter* reporter, const char* filename) {
optimize pathops coverage Remove unused code from SkOpSegment.cpp and friends. Add new tests exposed by coverage. Fix a bug exposed by coverage -- removing the need to detect points that are nearby when intersecting. Add gyp rule for building coverage flavor on Mac. R=mtklein@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/75453003 git-svn-id: http://skia.googlecode.com/svn/trunk@12344 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-21 17:04:29 +00:00
SkPath path;
path.moveTo(1, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(3, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(2, 0);
path.quadTo(2, 3, 3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
optimize pathops coverage Remove unused code from SkOpSegment.cpp and friends. Add new tests exposed by coverage. Fix a bug exposed by coverage -- removing the need to detect points that are nearby when intersecting. Add gyp rule for building coverage flavor on Mac. R=mtklein@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/75453003 git-svn-id: http://skia.googlecode.com/svn/trunk@12344 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-21 17:04:29 +00:00
}
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad11(skiatest::Reporter* reporter, const char* filename) {
fix pathops quad line intersection git-svn-id: http://skia.googlecode.com/svn/trunk@12374 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-25 13:39:12 +00:00
SkPath path;
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@12389 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-26 07:02:21 +00:00
path.moveTo(2, 0);
path.quadTo(0, 1, 1, 2);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(1, 3, 3, 3);
path.close();
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
testSimplify(reporter, path, filename);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@12389 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-26 07:02:21 +00:00
}
fix pathops quad line intersection git-svn-id: http://skia.googlecode.com/svn/trunk@12374 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-11-25 13:39:12 +00:00
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuad12(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 0);
path.lineTo(0, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 0, 1);
path.close();
testSimplify(reporter, path, filename);
}
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void testQuadralateral3(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.lineTo(0, 1);
path.close();
testSimplify(reporter, path, filename);
}
static void testDegenerate5(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(1, 0);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(0, 1);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuadralateral4(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.lineTo(3, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 1);
path.lineTo(3, 1);
path.close();
testSimplify(reporter, path, filename);
}
static void testDegenerates1(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 2, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuad13(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 2, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuad14(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 1, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuad15(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 0, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads16(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads17(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 2, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads18(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 2, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads19(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 1, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads20(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(2, 1, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads21(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 0, 2, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads22(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 3, 2);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads23(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 3, 2);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads24(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(0, 1);
path.lineTo(0, 1);
path.quadTo(0, 2, 3, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads25(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 1);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 0, 2, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads26(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 3, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 3, 2);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads27(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(2, 0);
path.quadTo(3, 0, 1, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads28(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 0, 1);
path.lineTo(2, 2);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 0, 2, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads29(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 3, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(3, 0);
path.lineTo(3, 0);
path.quadTo(3, 1, 0, 2);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads30(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 0);
path.lineTo(2, 2);
path.close();
path.moveTo(2, 0);
path.lineTo(2, 0);
path.quadTo(3, 2, 1, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads31(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(0, 1);
path.quadTo(2, 1, 1, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads32(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(1, 2);
path.close();
path.moveTo(1, 1);
path.lineTo(1, 1);
path.quadTo(3, 1, 0, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads33(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 1);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(3, 0, 2, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads34(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(2, 0, 3, 3);
Sanitizing source files in Housekeeper-Nightly git-svn-id: http://skia.googlecode.com/svn/trunk@14196 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-15 03:04:18 +00:00
path.close();
testSimplify(reporter, path, filename);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
}
static void testQuads35(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(1, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(3, 1, 0, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads36(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(2, 0, 1, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(3, 0, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads37(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(1, 0);
path.quadTo(2, 0, 1, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(3, 0, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads38(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(3, 0, 0, 2);
path.lineTo(3, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(2, 1, 3, 1);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads39(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(3, 0, 0, 3);
path.lineTo(0, 3);
path.close();
path.moveTo(1, 1);
path.lineTo(0, 2);
path.quadTo(1, 2, 0, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads40(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(3, 0, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 1);
path.lineTo(2, 2);
path.quadTo(3, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads41(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 0, 1, 0);
path.lineTo(2, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 1, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads54(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(3, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(1, 1);
path.quadTo(1, 1, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads53(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(3, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(1, 1);
path.quadTo(2, 3, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads52(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(2, 0, 1, 1);
path.lineTo(3, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(1, 1);
path.quadTo(2, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads51(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 0, 2, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(3, 1);
path.quadTo(3, 1, 1, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads50(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 0, 2, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(3, 1);
path.quadTo(1, 2, 1, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads49(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 0, 2, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(2, 2);
path.quadTo(2, 2, 0, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads48(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 0, 2, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(2, 2);
path.quadTo(3, 2, 0, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads47(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 0, 2, 1);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(2, 2);
path.quadTo(0, 3, 0, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads46x(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(2, 0);
path.quadTo(0, 1, 3, 2);
path.lineTo(1, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(3, 2, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads45(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 2, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 2);
path.quadTo(3, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads44(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 2, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(0, 2);
path.quadTo(3, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads43(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(2, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 2);
path.lineTo(0, 2);
path.quadTo(2, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads42(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 2, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 2);
path.quadTo(3, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads56(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 1, 0, 2);
path.lineTo(3, 2);
path.close();
path.moveTo(3, 0);
path.lineTo(2, 1);
path.quadTo(2, 1, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads57(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(3, 0, 3, 1);
path.lineTo(2, 2);
path.close();
path.moveTo(2, 0);
path.lineTo(3, 1);
path.quadTo(2, 2, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads58(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(3, 0, 3, 1);
path.lineTo(1, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(3, 1);
path.quadTo(2, 2, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads59(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(3, 0);
path.quadTo(3, 1, 3, 1);
path.lineTo(2, 2);
path.close();
path.moveTo(2, 0);
path.lineTo(3, 1);
path.quadTo(2, 2, 3, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads60(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(2, 1);
path.quadTo(0, 2, 3, 2);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(1, 1, 2, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads61(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.moveTo(0, 0);
path.quadTo(0, 0, 2, 0);
path.lineTo(1, 1);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(1, 0, 2, 2);
path.close();
testSimplify(reporter, path, filename);
}
Enabling the canvas bit to turn the clip stack into a flat replace exposed around 100 failures when testing the 800K skp set generated from the top 1M web sites. This fixes all but one of those failures. Major changes include: - Replace angle indices with angle pointers. This was motivated by the need to add angles later but not renumber existing angles. - Aggressive segment chase. When the winding is known on a segment, more aggressively passing that winding to adjacent segments allows fragmented data sets to succeed. - Line segments with ends nearly the same are treated as coincident first. - Transfer partial coincidence by observing that if segment A is partially coincident to B and C then B and C may be partially coincident. TBR=reed Author: caryclark@google.com Review URL: https://codereview.chromium.org/272153002
2014-06-17 12:15:38 +00:00
static void testQuadralateral10(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kWinding_FillType);
path.moveTo(0, 0);
path.lineTo(0, 0);
path.lineTo(0, 0);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.lineTo(2, 2);
path.lineTo(1, 3);
path.close();
testSimplify(reporter, path, filename);
}
update pathops core and tests split out skpclip (the test of 1M pictures) into its own project TBR=reed Author: caryclark@google.com Review URL: https://codereview.chromium.org/400033002
2014-07-18 12:08:14 +00:00
static void testRect3(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 60, 60, SkPath::kCCW_Direction);
path.addRect(10, 30, 40, 30, SkPath::kCCW_Direction);
path.addRect(24, 6, 36, 36, SkPath::kCCW_Direction);
path.addRect(32, 6, 36, 41, SkPath::kCCW_Direction);
testSimplify(reporter, path, filename);
}
Now, path ops natively intersect conics, quads, and cubics in any combination. There are still a class of cubic tests that fail and a handful of undiagnosed failures from skps and fuzz tests, but things are much better overall. Extended tests (150M+) run to completion in release in about 6 minutes; the standard test suite exceeds 100K and finishes in a few seconds on desktops. TBR=reed BUG=skia:3588 Review URL: https://codereview.chromium.org/1037953004
2015-04-20 15:31:59 +00:00
static void testRect4(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType(SkPath::kEvenOdd_FillType);
path.addRect(0, 0, 30, 60, SkPath::kCCW_Direction);
path.addRect(10, 0, 40, 30, SkPath::kCCW_Direction);
path.addRect(20, 0, 30, 40, SkPath::kCCW_Direction);
path.addRect(32, 0, 36, 41, SkPath::kCCW_Direction);
testSimplify(reporter, path, filename);
}
static void testQuads62(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(3, 2);
path.quadTo(1, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(2, 0);
path.quadTo(1, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads63(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(3, 0);
path.quadTo(0, 1, 1, 2);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 1);
path.quadTo(0, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads64(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(3, 0);
path.quadTo(0, 1, 1, 2);
path.lineTo(2, 2);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 1);
path.quadTo(0, 2, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
Path ops formerly found the topmost unprocessed edge and determined its angle sort order to initialize the winding. This never worked correctly with cubics and was flaky with paths consisting mostly of vertical edges. This replacement shoots axis-aligned rays through all intersecting edges to find the outermost one either horizontally or vertically. The resulting code is smaller and twice as fast. To support this, most of the horizontal / vertical intersection code was rewritten and standardized, and old code supporting the top-directed winding was deleted. Contours were pointed to by an SkTDArray. Instead, put them in a linked list, and designate the list head with its own class to ensure that methods that take lists of contours start at the top. This change removed a large percentage of memory allocations used by path ops. TBR=reed@google.com BUG=skia:3588 Review URL: https://codereview.chromium.org/1111333002
2015-05-11 14:21:27 +00:00
static void testTriangle1(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(2, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 2);
path.lineTo(1, 0);
path.close();
testSimplify(reporter, path, filename);
}
static void testTriangle2(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.lineTo(1, 0);
path.lineTo(0, 1);
path.close();
path.moveTo(2, 0);
path.lineTo(0, 2);
path.lineTo(2, 2);
path.close();
testSimplify(reporter, path, filename);
}
deal more consistently with unsortable edges Improve line/curve coincident detection and resolution. This fixed the remaining simple failures. When an edge is unsortable, use the ray intersection to determine the angles' winding. Deal with degenerate segments. TBR=reed@google.com BUG=skia:3588,skia:3762 Review URL: https://codereview.chromium.org/1140813002
2015-05-13 15:23:48 +00:00
static void testArc(skiatest::Reporter* reporter,const char* filename) {
SkRect r = SkRect::MakeWH(150, 100);
SkPath path;
path.arcTo(r, 0, 0.0025f, false);
testSimplify(reporter, path, filename);
}
fix path op builder The rewrite of path ops caused the inner contour direction to be reversed. This exposed an existing bug in path ops builder, namely that the implicit winding of the internal sum path could hide inner contours if they ended up in the wrong direction. Setting the sum path's fill type to even-odd ensures that the inner contours aren't discarded. R=fmalita@chromium.org BUG=skia:3838 Review URL: https://codereview.chromium.org/1126193004
2015-05-14 21:18:13 +00:00
static void testIssue3838(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(220, 170);
path.lineTo(200, 170);
path.lineTo(200, 190);
path.lineTo(180, 190);
path.lineTo(180, 210);
path.lineTo(200, 210);
path.lineTo(200, 250);
path.lineTo(260, 250);
path.lineTo(260, 190);
path.lineTo(220, 190);
path.lineTo(220, 170);
path.close();
path.moveTo(220, 210);
path.lineTo(220, 230);
path.lineTo(240, 230);
path.lineTo(240, 210);
path.lineTo(220, 210);
path.close();
testSimplify(reporter, path, filename);
}
fix builder winding again Record the nesting level when finding the edge winding contribution so that inner edges can be reversed as needed. R=fmalita@chromium.org BUG=skia:3838 Review URL: https://codereview.chromium.org/1140383002
2015-05-18 19:56:57 +00:00
static void testIssue3838_3(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(40, 10);
path.lineTo(60, 10);
path.lineTo(60, 30);
path.lineTo(40, 30);
path.lineTo(40, 10);
path.moveTo(41, 11);
path.lineTo(41, 29);
path.lineTo(59, 29);
path.lineTo(59, 11);
path.lineTo(41, 11);
testSimplify(reporter, path, filename);
}
When three or more edges are coincident, the logic needs to compute the overlapping ranges and combine the winding into a single destination. This computes coincidence more rigorously, fixing the edge cases exposed by this bug. Also, add the ability to debug and dump pathop structures from the coincident context. TBR=reed@google.com BUG=skia:3651 Review URL: https://codereview.chromium.org/1182493015
2015-07-06 18:38:33 +00:00
static void testQuads65(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(1, 2);
path.quadTo(3, 2, 0, 3);
path.lineTo(1, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 2);
path.quadTo(3, 2, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
fix path ops fuzz buster Mark collapsed segments as done and remove collapsed segment references from the coincidence array. Also add test names to global debugging. R=fmalita@chromium.org BUG=512592 Review URL: https://codereview.chromium.org/1250293002
2015-07-23 19:40:22 +00:00
static void fuzz864a(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(10, 90);
path.lineTo(10, 90);
path.lineTo(10, 30);
path.lineTo(10, 30);
path.lineTo(10, 90);
path.close();
path.moveTo(10, 90);
path.lineTo(10, 90);
path.lineTo(10, 30);
path.lineTo(10, 30);
path.lineTo(10, 90);
path.close();
path.moveTo(10, 90);
path.lineTo(110, 90);
path.lineTo(110, 30);
path.lineTo(10, 30);
path.lineTo(10, 90);
path.close();
path.moveTo(10, 30);
path.lineTo(32678, 30);
path.lineTo(32678, 30);
path.lineTo(10, 30);
path.close();
path.moveTo(10, 3.35545e+07f);
path.lineTo(110, 3.35545e+07f);
path.lineTo(110, 30);
path.lineTo(10, 30);
path.lineTo(10, 3.35545e+07f);
path.close();
path.moveTo(10, 315);
path.lineTo(110, 315);
path.lineTo(110, 255);
path.lineTo(10, 255);
path.lineTo(10, 315);
path.close();
path.moveTo(0, 60);
path.lineTo(100, 60);
path.lineTo(100, 0);
path.lineTo(0, 0);
path.lineTo(0, 60);
path.close();
path.moveTo(10, 90);
path.lineTo(110, 90);
path.lineTo(110, 30);
path.lineTo(10, 30);
path.lineTo(10, 90);
path.close();
path.moveTo(10, 3.35545e+07f);
path.lineTo(110, 3.35545e+07f);
path.lineTo(110, 30);
path.lineTo(10, 30);
path.lineTo(10, 3.35545e+07f);
path.close();
path.moveTo(10, 90);
path.lineTo(110, 90);
path.lineTo(110, 30);
path.lineTo(10, 30);
path.lineTo(10, 90);
path.close();
testSimplify(reporter, path, filename);
}
compute split conic endpoints exactly The divide by w can generate slightly erroneous results even for t == 0 or t == 1. The error in turn defeats detecting a point in common for a pair of curves that travel in opposite directions. Instead, special case endpoints when the t is 0 or 1. TBR=reed@google.com BUG=514118 Review URL: https://codereview.chromium.org/1259513004
2015-07-28 12:12:19 +00:00
static void cr514118(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(SkBits2Float(0x42c80000), SkBits2Float(0x42480000)); // 100, 50
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 100, 0, 50, 0, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 0, 0, 0, 50, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 0, 100, 50, 100, 0.707107f
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 100, 100, 100, 50, 0.707107f
path.close();
path.moveTo(SkBits2Float(0x42c80133), SkBits2Float(0x42480000)); // 100.002f, 50
path.conicTo(SkBits2Float(0x42c80133), SkBits2Float(0x00000000), SkBits2Float(0x42480267), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 100.002f, 0, 50.0023f, 0, 0.707107f
path.conicTo(SkBits2Float(0x3b19b530), SkBits2Float(0x00000000), SkBits2Float(0x3b19b530), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 0.00234539f, 0, 0.00234539f, 50, 0.707107f
path.conicTo(SkBits2Float(0x3b19b530), SkBits2Float(0x42c80000), SkBits2Float(0x42480267), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 0.00234539f, 100, 50.0023f, 100, 0.707107f
path.conicTo(SkBits2Float(0x42c80133), SkBits2Float(0x42c80000), SkBits2Float(0x42c80133), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 100.002f, 100, 100.002f, 50, 0.707107f
path.close();
testSimplify(reporter, path, filename);
}
looking for fuzz failure These tests were isolated from their respective minimized test cases. The tests work fine and pass path ops internal validation; hopefully some more intensive x-san or valgrind test will help isolate the bug. Sheriff, please revert if it fails and I don't get to it first. TBR=reed@google.com,halcanary@google.com BUG=535127,535151 Review URL: https://codereview.chromium.org/1359263003
2015-09-24 01:07:51 +00:00
static void fuzz994s_11(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType((SkPath::FillType) 0);
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x42b40000)); // 110, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x41f00000)); // 110, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x46ff4c00), SkBits2Float(0x41f00000)); // 32678, 30
path.lineTo(SkBits2Float(0x46ff4c00), SkBits2Float(0x41f00000)); // 32678, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x4c000006)); // 10, 3.35545e+07f
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x4c000006)); // 110, 3.35545e+07f
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x41f00000)); // 110, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x4c000006)); // 10, 3.35545e+07f
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x439d8000)); // 10, 315
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x439d8000)); // 110, 315
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x437f0000)); // 110, 255
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x437f0000)); // 10, 255
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x439d8000)); // 10, 315
path.close();
path.moveTo(SkBits2Float(0x00000000), SkBits2Float(0x42700000)); // 0, 60
path.lineTo(SkBits2Float(0x42c80000), SkBits2Float(0x42700000)); // 100, 60
path.lineTo(SkBits2Float(0x42c80000), SkBits2Float(0x00000000)); // 100, 0
path.lineTo(SkBits2Float(0x00000000), SkBits2Float(0x00000000)); // 0, 0
path.lineTo(SkBits2Float(0x00000000), SkBits2Float(0x42700000)); // 0, 60
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x42b40000)); // 110, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x41f00000)); // 110, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x4c000006)); // 10, 3.35545e+07f
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x4c000006)); // 110, 3.35545e+07f
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x41f00000)); // 110, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x4c000006)); // 10, 3.35545e+07f
path.close();
path.moveTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x42b40000)); // 110, 90
path.lineTo(SkBits2Float(0x42dc0000), SkBits2Float(0x41f00000)); // 110, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x41f00000)); // 10, 30
path.lineTo(SkBits2Float(0x41200000), SkBits2Float(0x42b40000)); // 10, 90
path.close();
testSimplify(reporter, path, filename);
}
static void fuzz994s_3414(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType((SkPath::FillType) 0);
path.moveTo(SkBits2Float(0x42c80000), SkBits2Float(0x42480000)); // 100, 50
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 100, 0, 50, 0, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 0, 0, 0, 50, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 0, 100, 50, 100, 0.707107f
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 100, 100, 100, 50, 0.707107f
path.close();
path.moveTo(SkBits2Float(0x42c84964), SkBits2Float(0x42480000)); // 100.143f, 50
path.conicTo(SkBits2Float(0x42c84964), SkBits2Float(0x00000000), SkBits2Float(0x424892c8), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 100.143f, 0, 50.1433f, 0, 0.707107f
path.conicTo(SkBits2Float(0x3e12c788), SkBits2Float(0x00000000), SkBits2Float(0x3e12c788), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 0.143339f, 0, 0.143339f, 50, 0.707107f
path.conicTo(SkBits2Float(0x3e12c788), SkBits2Float(0x42c80000), SkBits2Float(0x424892c8), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 0.143339f, 100, 50.1433f, 100, 0.707107f
path.conicTo(SkBits2Float(0x42c84964), SkBits2Float(0x42c80000), SkBits2Float(0x42c84964), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 100.143f, 100, 100.143f, 50, 0.707107f
path.close();
path.moveTo(SkBits2Float(0x42c80000), SkBits2Float(0x42480000)); // 100, 50
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 100, 0, 50, 0, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x00000000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 0, 0, 0, 50, 0.707107f
path.conicTo(SkBits2Float(0x00000000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 0, 100, 50, 100, 0.707107f
path.conicTo(SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42c80000), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 100, 100, 100, 50, 0.707107f
path.close();
path.moveTo(SkBits2Float(0x4c00006b), SkBits2Float(0x424c0000)); // 3.35549e+07f, 51
path.conicTo(SkBits2Float(0x4c00006b), SkBits2Float(0xcbffffe5), SkBits2Float(0x43d6e720), SkBits2Float(0xcbffffe5), SkBits2Float(0x3f3504f3)); // 3.35549e+07f, -3.35544e+07f, 429.806f, -3.35544e+07f, 0.707107f
path.conicTo(SkBits2Float(0xcbffff28), SkBits2Float(0xcbffffe5), SkBits2Float(0xcbffff28), SkBits2Float(0x424c0000), SkBits2Float(0x3f3504f3)); // -3.3554e+07f, -3.35544e+07f, -3.3554e+07f, 51, 0.707107f
path.conicTo(SkBits2Float(0xcbffff28), SkBits2Float(0x4c00000c), SkBits2Float(0x43d6e720), SkBits2Float(0x4c00000c), SkBits2Float(0x3f3504f3)); // -3.3554e+07f, 3.35545e+07f, 429.806f, 3.35545e+07f, 0.707107f
path.conicTo(SkBits2Float(0x4c00006b), SkBits2Float(0x4c00000c), SkBits2Float(0x4c00006b), SkBits2Float(0x424c0000), SkBits2Float(0x3f3504f3)); // 3.35549e+07f, 3.35545e+07f, 3.35549e+07f, 51, 0.707107f
path.close();
path.moveTo(SkBits2Float(0x43ef6720), SkBits2Float(0x42480000)); // 478.806f, 50
path.conicTo(SkBits2Float(0x43ef6720), SkBits2Float(0x00000000), SkBits2Float(0x43d66720), SkBits2Float(0x00000000), SkBits2Float(0x3f3504f3)); // 478.806f, 0, 428.806f, 0, 0.707107f
path.conicTo(SkBits2Float(0x43bd6720), SkBits2Float(0x00000000), SkBits2Float(0x43bd6720), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 378.806f, 0, 378.806f, 50, 0.707107f
path.conicTo(SkBits2Float(0x43bd6720), SkBits2Float(0x42c80000), SkBits2Float(0x43d66720), SkBits2Float(0x42c80000), SkBits2Float(0x3f3504f3)); // 378.806f, 100, 428.806f, 100, 0.707107f
path.conicTo(SkBits2Float(0x43ef6720), SkBits2Float(0x42c80000), SkBits2Float(0x43ef6720), SkBits2Float(0x42480000), SkBits2Float(0x3f3504f3)); // 478.806f, 100, 478.806f, 50, 0.707107f
path.close();
testSimplify(reporter, path, filename);
}
Enabling clip stack flattening exercises path ops. Iterating through the 903K skps that represent the imagable 1M top web pages triggers a number of bugs, some of which are addressed here. Some web pages trigger intersecting cubic representations of arc with their conic counterparts. This exposed a flaw in coincident detection that caused an infinite loop. The loop alternatively extended the coincident section and, determining the that the bounds of the curve pairs did not overlap, deleted the extension. Track the number of times the coincident detection is called, and if it exceeds an empirically found limit, assume that the curves are coincident and force it to be so. The loop count limit can be determined by enabling DEBUG_T_SECT_LOOP_COUNT and running all tests. The largest count is reported on completion. Another class of bugs was caused by concident detection duplicating nearly identical points that had been merged earlier. To track these bugs, the 'handle coincidence' code was duplicated as a const debug variety that reported if one of a dozen or so irregularities are present; then it is easier to see when a block of code that fixes one irregularity regresses another. Creating the debug const code version exposed some non-debug code that could be const, and some that was experimental and could be removed. Set DEBUG_COINCIDENCE to track coincidence health and handling. For running on Chrome, DEBUG_VERIFY checks the result of pathops against the same operation using SkRegion to verify that the results are nearly the same. When visualizing the pathops work using tools/pathops_visualizer.htm, set DEBUG_DUMP_ALIGNMENT to see the curves after they've been aligned for coincidence. Other bugs fixed include detecting when a section of a pair of curves have devolved into lines and are coincident. TBR=reed@google.com Review URL: https://codereview.chromium.org/1394503003
2015-10-16 16:03:38 +00:00
static void fuzz_twister(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType((SkPath::FillType) 0);
path.moveTo(0, 600);
path.lineTo(3.35544e+07f, 600);
path.lineTo(3.35544e+07f, 0);
path.lineTo(0, 0);
path.lineTo(0, 600);
path.close();
path.moveTo(63, 600);
path.lineTo(3.35545e+07f, 600);
path.lineTo(3.35545e+07f, 0);
path.lineTo(63, 0);
path.lineTo(63, 600);
path.close();
path.moveTo(93, 600);
path.lineTo(3.35545e+07f, 600);
path.lineTo(3.35545e+07f, 0);
path.lineTo(93, 0);
path.lineTo(93, 600);
path.close();
path.moveTo(123, 600);
path.lineTo(3.35546e+07f, 600);
path.lineTo(3.35546e+07f, 0);
path.lineTo(123, 0);
path.lineTo(123, 600);
path.close();
testSimplify(reporter, path, filename);
}
fix pathops coincidence fuzz bug Simplifying a series of rects with very large bounds triggers a coincidence bug where, after one of the intersection points that marks a coincident range has been deleted, it is referenced. Both the deletion and reference is (probably) happening in the SkOpCoincidence::AddExpanded() phase of HandleCoincidence(), and may signify a bug that could happen with usable input data, but I haven't been able to determine that. For now, abort the Simplify() when the erroneous condition is detected. TBR=reed@google.com BUG=558281 Review URL: https://codereview.chromium.org/1463923002
2015-11-20 22:06:28 +00:00
static void fuzz_twister2(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(SkBits2Float(0x00000000), SkBits2Float(0x44160000)); // 0, 600
path.lineTo(SkBits2Float(0x4bfffffe), SkBits2Float(0x44160000)); // 3.35544e+07f, 600
path.lineTo(SkBits2Float(0x4bfffffe), SkBits2Float(0x00000000)); // 3.35544e+07f, 0
path.lineTo(SkBits2Float(0x00000000), SkBits2Float(0x00000000)); // 0, 0
path.lineTo(SkBits2Float(0x00000000), SkBits2Float(0x44160000)); // 0, 600
path.close();
path.moveTo(SkBits2Float(0x427c0000), SkBits2Float(0x00000000)); // 63, 0
path.lineTo(SkBits2Float(0x4c00000f), SkBits2Float(0x00000000)); // 3.35545e+07f, 0
path.lineTo(SkBits2Float(0x4c00000f), SkBits2Float(0x00000000)); // 3.35545e+07f, 0
path.lineTo(SkBits2Float(0x427c0000), SkBits2Float(0x00000000)); // 63, 0
path.close();
path.moveTo(SkBits2Float(0x42ba0000), SkBits2Float(0x00000000)); // 93, 0
path.lineTo(SkBits2Float(0x4c000016), SkBits2Float(0x00000000)); // 3.35545e+07f, 0
path.lineTo(SkBits2Float(0x4c000016), SkBits2Float(0x00000000)); // 3.35545e+07f, 0
path.lineTo(SkBits2Float(0x42ba0000), SkBits2Float(0x00000000)); // 93, 0
path.close();
path.moveTo(SkBits2Float(0x42f60000), SkBits2Float(0x00000000)); // 123, 0
path.lineTo(SkBits2Float(0x4c00001e), SkBits2Float(0x00000000)); // 3.35546e+07f, 0
path.lineTo(SkBits2Float(0x4c00001e), SkBits2Float(0x00000000)); // 3.35546e+07f, 0
path.lineTo(SkBits2Float(0x42f60000), SkBits2Float(0x00000000)); // 123, 0
path.close();
pathops coincidence and security rewrite Most changes stem from working on an examples bracketed by #if DEBUG_UNDER_DEVELOPMENT // tiger These exposed many problems with coincident curves, as well as errors throughout the code. Fixing these errors also fixed a number of fuzzer-inspired bug reports. * Line/Curve Intersections Check to see if the end of the line nearly intersects the curve. This was a FIXME in the old code. * Performance Use a central chunk allocator. Plumb the allocator into the global variable state so that it can be shared. (Note that 'SkGlobalState' is allocated on the stack and is visible to children functions but not other threads.) * Refactor Let SkOpAngle grow up from a structure to a class. Let SkCoincidentSpans grow up from a structure to a class. Rename enum Alias to AliasMatch. * Coincidence Rewrite Add more debugging to coincidence detection. Parallel debugging routines have read-only logic to report the current coincidence state so that steps through the logic can expose whether things got better or worse. More functions can error-out and cause the pathops engine to non-destructively exit. * Accuracy Remove code that adjusted point locations. Instead, offset the curve part so that sorted curves all use the same origin. Reduce the size (and influence) of magic numbers. * Testing The debug suite with verify and the full release suite ./out/Debug/pathops_unittest -v -V ./out/Release/pathops_unittest -v -V -x expose one error. That error is captured as cubics_d3. This error exists in the checked in code as well. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 Review-Url: https://codereview.chromium.org/2128633003
2016-07-18 17:01:36 +00:00
testSimplify(reporter, path, filename);
fix pathops coincidence fuzz bug Simplifying a series of rects with very large bounds triggers a coincidence bug where, after one of the intersection points that marks a coincident range has been deleted, it is referenced. Both the deletion and reference is (probably) happening in the SkOpCoincidence::AddExpanded() phase of HandleCoincidence(), and may signify a bug that could happen with usable input data, but I haven't been able to determine that. For now, abort the Simplify() when the erroneous condition is detected. TBR=reed@google.com BUG=558281 Review URL: https://codereview.chromium.org/1463923002
2015-11-20 22:06:28 +00:00
}
fix security bug This fix is a tradeoff. It changes intersection to treat a case where one coincident run is intersected at one point and the other edge is not as continuing to be a span. The old code tried to treat this as a single point. The old code is probably right, but this change alone made the data structures inconsistent. Later, extending the coincident runs would fail by incorrectly discarding the single point intersection. As a result, this fixes the security test and one other, but makes a different test fail. Isolating the failure uncovered a reduced case that fails with and without the change, so there are more serious problems here. Those problems are addressed in a separate CL. Many of the test edits below remove ill-thought out debugging messaging that fire off global state, which isn't usable in a multi-threaded test environment. In the end, with this fix, all existing tests (modulo one new failure and one new non-failure) pass in debug and in the extended release test suites. TBR=reed@google.com BUG=614248 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2018513003 Review-Url: https://codereview.chromium.org/2018513003
2016-05-26 16:01:47 +00:00
static void fuzz763_4713_b(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.setFillType((SkPath::FillType) 0);
path.moveTo(SkBits2Float(0x42240000), SkBits2Float(0x42040000));
path.quadTo(SkBits2Float(0x42240000), SkBits2Float(0x4211413d), SkBits2Float(0x421aa09e), SkBits2Float(0x421aa09e));
path.quadTo(SkBits2Float(0x4211413d), SkBits2Float(0x42240000), SkBits2Float(0x42040000), SkBits2Float(0x42240000));
path.quadTo(SkBits2Float(0x41ed7d86), SkBits2Float(0x42240000), SkBits2Float(0x41dabec3), SkBits2Float(0x421aa09e));
path.quadTo(SkBits2Float(0x41c80000), SkBits2Float(0x4211413d), SkBits2Float(0x41c80000), SkBits2Float(0x42040000));
path.quadTo(SkBits2Float(0x41c80000), SkBits2Float(0x41ed7d86), SkBits2Float(0x41dabec3), SkBits2Float(0x41dabec3));
path.quadTo(SkBits2Float(0x41ed7d86), SkBits2Float(0x41c80000), SkBits2Float(0x42040000), SkBits2Float(0x41c80000));
path.quadTo(SkBits2Float(0x4211413d), SkBits2Float(0x41c80000), SkBits2Float(0x421aa09e), SkBits2Float(0x41dabec3));
path.quadTo(SkBits2Float(0x42240000), SkBits2Float(0x41ed7d86), SkBits2Float(0x42240000), SkBits2Float(0x42040000));
path.close();
path.moveTo(SkBits2Float(0x4204f72e), SkBits2Float(0x41c56cd2));
path.quadTo(SkBits2Float(0x42123842), SkBits2Float(0x41c52adf), SkBits2Float(0x421baed7), SkBits2Float(0x41d7bac6));
path.quadTo(SkBits2Float(0x4225256d), SkBits2Float(0x41ea4aad), SkBits2Float(0x42254667), SkBits2Float(0x4202666b));
path.quadTo(SkBits2Float(0x42256760), SkBits2Float(0x420fa77f), SkBits2Float(0x421c1f6c), SkBits2Float(0x42191e14));
path.quadTo(SkBits2Float(0x421bff97), SkBits2Float(0x42193e89), SkBits2Float(0x421bdf6b), SkBits2Float(0x42195eb8));
path.quadTo(SkBits2Float(0x421bbff6), SkBits2Float(0x42197f32), SkBits2Float(0x421ba03b), SkBits2Float(0x42199f57));
path.quadTo(SkBits2Float(0x421b605e), SkBits2Float(0x4219e00a), SkBits2Float(0x421b1fa8), SkBits2Float(0x421a1f22));
path.quadTo(SkBits2Float(0x421ae0f1), SkBits2Float(0x421a604b), SkBits2Float(0x421aa09e), SkBits2Float(0x421aa09e));
path.quadTo(SkBits2Float(0x4211413d), SkBits2Float(0x42240000), SkBits2Float(0x42040000), SkBits2Float(0x42240000));
path.quadTo(SkBits2Float(0x41ed7d86), SkBits2Float(0x42240000), SkBits2Float(0x41dabec3), SkBits2Float(0x421aa09e));
path.quadTo(SkBits2Float(0x41c80000), SkBits2Float(0x4211413d), SkBits2Float(0x41c80000), SkBits2Float(0x42040000));
path.quadTo(SkBits2Float(0x41c80000), SkBits2Float(0x41ed7d86), SkBits2Float(0x41dabec3), SkBits2Float(0x41dabec3));
path.quadTo(SkBits2Float(0x41db19b1), SkBits2Float(0x41da63d5), SkBits2Float(0x41db755b), SkBits2Float(0x41da0a9b));
path.quadTo(SkBits2Float(0x41dbce01), SkBits2Float(0x41d9ae59), SkBits2Float(0x41dc285e), SkBits2Float(0x41d952ce));
path.quadTo(SkBits2Float(0x41dc55b6), SkBits2Float(0x41d924df), SkBits2Float(0x41dc82cd), SkBits2Float(0x41d8f7cd));
path.quadTo(SkBits2Float(0x41dcaf1e), SkBits2Float(0x41d8ca01), SkBits2Float(0x41dcdc4c), SkBits2Float(0x41d89bf0));
path.quadTo(SkBits2Float(0x41ef6c33), SkBits2Float(0x41c5aec5), SkBits2Float(0x4204f72e), SkBits2Float(0x41c56cd2));
path.close();
quad and conic do not intersect If a quad a conic intersect only where the end of one is contained by the convex hull of the other, and the curve contained by the hull is nearly a straight line, treating it as a line may move the end point to the other side of the curve. Detect this by checking to see if the end point is in the hull, and if so, continue to subdivide the curve rather than treating it as a line. This fixes several existing tests that were disabled earlier this year. A typo in SkDCurve::nearPoint() prevented detecting when the end of a line was nearly touching a curve. Also fixed concidence a bit to get the second half of tiger further along. All existing tests, including extended testing in Release and the first half of tiger, work. TBR=reed@google.com BUG=skia:5131 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2338323002 Review-Url: https://codereview.chromium.org/2338323002
2016-09-15 14:48:18 +00:00
testSimplify(reporter, path, filename);
pathops coincidence and security rewrite Most changes stem from working on an examples bracketed by #if DEBUG_UNDER_DEVELOPMENT // tiger These exposed many problems with coincident curves, as well as errors throughout the code. Fixing these errors also fixed a number of fuzzer-inspired bug reports. * Line/Curve Intersections Check to see if the end of the line nearly intersects the curve. This was a FIXME in the old code. * Performance Use a central chunk allocator. Plumb the allocator into the global variable state so that it can be shared. (Note that 'SkGlobalState' is allocated on the stack and is visible to children functions but not other threads.) * Refactor Let SkOpAngle grow up from a structure to a class. Let SkCoincidentSpans grow up from a structure to a class. Rename enum Alias to AliasMatch. * Coincidence Rewrite Add more debugging to coincidence detection. Parallel debugging routines have read-only logic to report the current coincidence state so that steps through the logic can expose whether things got better or worse. More functions can error-out and cause the pathops engine to non-destructively exit. * Accuracy Remove code that adjusted point locations. Instead, offset the curve part so that sorted curves all use the same origin. Reduce the size (and influence) of magic numbers. * Testing The debug suite with verify and the full release suite ./out/Debug/pathops_unittest -v -V ./out/Release/pathops_unittest -v -V -x expose one error. That error is captured as cubics_d3. This error exists in the checked in code as well. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 Review-Url: https://codereview.chromium.org/2128633003
2016-07-18 17:01:36 +00:00
}
static void dean4(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
// start region
// start loop, contour: 1
// Segment 1145.3381097316742 2017.6783947944641 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2017.0033947825432
path.moveTo(1145.3381347656250, 2017.6783447265625);
path.lineTo(1145.3381347656250, 2017.0034179687500);
// Segment 1145.3381097316742 2017.0033947825432 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6927231521568 2017.0033947825432
path.lineTo(1143.6927490234375, 2017.0034179687500);
// Segment 1143.6927231521568 2017.0033947825432 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1144.8640675112890 2018.1589246992417
path.lineTo(1144.8640136718750, 2018.1589355468750);
// Segment 1144.8640675112890 2018.1589246992417 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2017.6783947944641
path.lineTo(1145.3381347656250, 2017.6783447265625);
path.close();
// start loop, contour: 2
// Segment 1145.3381097316742 2016.3216052055359 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1144.8640675258462 2015.8410752863977
path.moveTo(1145.3381347656250, 2016.3216552734375);
path.lineTo(1144.8640136718750, 2015.8410644531250);
// Segment 1144.8640675258462 2015.8410752863977 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6927230811802 2016.9966052174568
path.lineTo(1143.6927490234375, 2016.9965820312500);
// Segment 1143.6927230811802 2016.9966052174568 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2016.9966052174568
path.lineTo(1145.3381347656250, 2016.9965820312500);
// Segment 1145.3381097316742 2016.9966052174568 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2016.3216052055359
path.lineTo(1145.3381347656250, 2016.3216552734375);
path.close();
// start loop, contour: 3
// Segment 1147.3323798179626 2014.3542600870132 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220239557 2014.8347900059885
path.moveTo(1147.3323974609375, 2014.3542480468750);
path.lineTo(1147.8063964843750, 2014.8348388671875);
// Segment 1147.8064220239557 2014.8347900059885 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220516883 2014.8347899786306
path.lineTo(1147.8063964843750, 2014.8348388671875);
// Segment 1147.8064220516883 2014.8347899786306 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.3323798179626 2014.3542600870132
path.lineTo(1147.3323974609375, 2014.3542480468750);
path.close();
// start loop, contour: 4
// Segment 1146.3696286678314 2013.4045072346926 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436708778083 2013.8850371497379
path.moveTo(1146.3696289062500, 2013.4045410156250);
path.lineTo(1146.8436279296875, 2013.8850097656250);
// Segment 1146.8436708778083 2013.8850371497379 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436709015571 2013.8850371263100
path.lineTo(1146.8436279296875, 2013.8850097656250);
// Segment 1146.8436709015571 2013.8850371263100 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.3696286678314 2013.4045072346926
path.lineTo(1146.3696289062500, 2013.4045410156250);
path.close();
// start loop, contour: 5
// Segment 1143.2063037902117 2016.5251235961914 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615802348 2016.0445936811461
path.moveTo(1143.2062988281250, 2016.5251464843750);
path.lineTo(1142.7322998046875, 2016.0445556640625);
// Segment 1142.7322615802348 2016.0445936811461 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615564860 2016.0445937045740
path.lineTo(1142.7322998046875, 2016.0445556640625);
// Segment 1142.7322615564860 2016.0445937045740 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.2063037902117 2016.5251235961914
path.lineTo(1143.2062988281250, 2016.5251464843750);
path.close();
// start loop, contour: 6
// Segment 1143.0687679275870 2016.7286419868469 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.5428101613127 2017.2091718784643
path.moveTo(1143.0687255859375, 2016.7286376953125);
path.lineTo(1143.5428466796875, 2017.2092285156250);
// Segment 1143.5428101613127 2017.2091718784643 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.7437679395080 2017.0109272411960
path.lineTo(1143.7437744140625, 2017.0109863281250);
// Segment 1143.7437679395080 2017.0109272411960 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.7437679395080 2016.7286419868469
path.lineTo(1143.7437744140625, 2016.7286376953125);
// Segment 1143.7437679395080 2016.7286419868469 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.0687679275870 2016.7286419868469
path.lineTo(1143.0687255859375, 2016.7286376953125);
path.close();
// start loop, contour: 7
// Segment 1143.2063037902117 2017.4748764038086 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615603032 2017.9554062991915
path.moveTo(1143.2062988281250, 2017.4748535156250);
path.lineTo(1142.7322998046875, 2017.9554443359375);
// Segment 1142.7322615603032 2017.9554062991915 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615746241 2017.9554063133189
path.lineTo(1142.7322998046875, 2017.9554443359375);
// Segment 1142.7322615746241 2017.9554063133189 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.2063037902117 2017.4748764038086
path.lineTo(1143.2062988281250, 2017.4748535156250);
path.close();
// start loop, contour: 8
// Segment 1146.3696286678314 2020.5954928398132 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436708977399 2020.1149629444303
path.moveTo(1146.3696289062500, 2020.5954589843750);
path.lineTo(1146.8436279296875, 2020.1149902343750);
// Segment 1146.8436708977399 2020.1149629444303 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436708834190 2020.1149629303029
path.lineTo(1146.8436279296875, 2020.1149902343750);
// Segment 1146.8436708834190 2020.1149629303029 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.3696286678314 2020.5954928398132
path.lineTo(1146.3696289062500, 2020.5954589843750);
path.close();
// start loop, contour: 9
// Segment 1147.3323798179626 2019.6457400321960 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220484741 2019.1652101374082
path.moveTo(1147.3323974609375, 2019.6457519531250);
path.lineTo(1147.8063964843750, 2019.1651611328125);
// Segment 1147.8064220484741 2019.1652101374082 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220383478 2019.1652101274185
path.lineTo(1147.8063964843750, 2019.1651611328125);
// Segment 1147.8064220383478 2019.1652101274185 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.3323798179626 2019.6457400321960
path.lineTo(1147.3323974609375, 2019.6457519531250);
path.close();
// start loop, contour: 10
// Segment 1145.3381097316742 2018.3533948063850 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2018.3533948063850
path.moveTo(1145.3381347656250, 2018.3533935546875);
path.lineTo(1156.6848144531250, 2018.3533935546875);
// Segment 1156.6848182678223 2018.3533948063850 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2017.0033947825432
path.lineTo(1156.6848144531250, 2017.0034179687500);
// Segment 1156.6848182678223 2017.0033947825432 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2017.0033947825432
path.lineTo(1145.3381347656250, 2017.0034179687500);
// Segment 1145.3381097316742 2017.0033947825432 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2018.3533948063850
path.lineTo(1145.3381347656250, 2018.3533935546875);
path.close();
// start loop, contour: 11
// Segment 1156.6848182678223 2018.3533948063850 0.3569631313191 0.0000000000000 -0.2645167304388 0.2609454237780 1157.6574279406423 2017.9723661860094
path.moveTo(1156.6848144531250, 2018.3533935546875);
path.cubicTo(1157.0417480468750, 2018.3533935546875, 1157.3929443359375, 2018.2332763671875, 1157.6574707031250, 2017.9724121093750);
// Segment 1157.6574279406423 2017.9723661860094 0.2653344079822 -0.2617520616521 0.0000000000000 0.3596905289350 1158.0474975705147 2017.0000000000000
path.cubicTo(1157.9227294921875, 2017.7105712890625, 1158.0474853515625, 2017.3597412109375, 1158.0474853515625, 2017.0000000000000);
// Segment 1158.0474975705147 2017.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6974975466728 2017.0000000000000
path.lineTo(1156.6975097656250, 2017.0000000000000);
// Segment 1156.6974975466728 2017.0000000000000 0.0028009248351 0.0403311981485 0.0118595244351 -0.0220843520393 1156.6941780622435 2017.0325257649940
path.cubicTo(1156.7003173828125, 2017.0402832031250, 1156.7060546875000, 2017.0104980468750, 1156.6942138671875, 2017.0324707031250);
// Segment 1156.6941780622435 2017.0325257649940 -0.0032637855860 0.0184860248562 0.0120617528380 -0.0065934603083 1156.7093435710913 2017.0113063061967
path.cubicTo(1156.6909179687500, 2017.0510253906250, 1156.7214355468750, 2017.0047607421875, 1156.7093505859375, 2017.0113525390625);
// split at 0.4496445953846
// path.cubicTo(1156.6927490234375, 2017.0407714843750, 1156.6981201171875, 2017.0360107421875, 1156.7033691406250, 2017.0289306640625);
// path.cubicTo(1156.7097167968750, 2017.0201416015625, 1156.7159423828125, 2017.0076904296875, 1156.7093505859375, 2017.0113525390625);
// Segment 1156.7093435710913 2017.0113063061967 -0.0070717276929 0.0122220954353 0.0203483811973 -0.0039136894418 1156.7268834554304 2016.9985353221975
path.cubicTo(1156.7022705078125, 2017.0235595703125, 1156.7471923828125, 2016.9946289062500, 1156.7269287109375, 2016.9985351562500);
// Segment 1156.7268834554304 2016.9985353221975 -0.0244396787691 0.0123649140586 0.0433322464027 0.0026558844666 1156.6848182678223 2017.0033947825432
path.cubicTo(1156.7023925781250, 2017.0108642578125, 1156.7281494140625, 2017.0061035156250, 1156.6848144531250, 2017.0034179687500);
// split at 0.4418420493603
// path.cubicTo(1156.7160644531250, 2017.0040283203125, 1156.7150878906250, 2017.0061035156250, 1156.7136230468750, 2017.0065917968750);
// path.cubicTo(1156.7116699218750, 2017.0070800781250, 1156.7089843750000, 2017.0048828125000, 1156.6848144531250, 2017.0034179687500);
// Segment 1156.6848182678223 2017.0033947825432 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2018.3533948063850
path.lineTo(1156.6848144531250, 2018.3533935546875);
path.close();
// start loop, contour: 12
// Segment 1158.0474975705147 2017.0000000000000 0.0000000000000 -0.3596905289350 0.2653344079822 0.2617520616521 1157.6574279406423 2016.0276338139906
path.moveTo(1158.0474853515625, 2017.0000000000000);
path.cubicTo(1158.0474853515625, 2016.6402587890625, 1157.9227294921875, 2016.2894287109375, 1157.6574707031250, 2016.0275878906250);
// Segment 1157.6574279406423 2016.0276338139906 -0.2645167304388 -0.2609454237780 0.3569631313191 0.0000000000000 1156.6848182678223 2015.6466051936150
path.cubicTo(1157.3929443359375, 2015.7667236328125, 1157.0417480468750, 2015.6466064453125, 1156.6848144531250, 2015.6466064453125);
// split at 0.5481675863266
// path.cubicTo(1157.5124511718750, 2015.8846435546875, 1157.3414306640625, 2015.7839355468750, 1157.1577148437500, 2015.7220458984375);
// path.cubicTo(1157.0062255859375, 2015.6711425781250, 1156.8460693359375, 2015.6466064453125, 1156.6848144531250, 2015.6466064453125);
// Segment 1156.6848182678223 2015.6466051936150 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2016.9966052174568
path.lineTo(1156.6848144531250, 2016.9965820312500);
// Segment 1156.6848182678223 2016.9966052174568 0.0433322464027 -0.0026558844666 -0.0244396787691 -0.0123649140586 1156.7268834554304 2017.0014646778025
path.cubicTo(1156.7281494140625, 2016.9938964843750, 1156.7023925781250, 2016.9891357421875, 1156.7269287109375, 2017.0014648437500);
// split at 0.5581579208374
// path.cubicTo(1156.7089843750000, 2016.9951171875000, 1156.7116699218750, 2016.9929199218750, 1156.7136230468750, 2016.9934082031250);
// path.cubicTo(1156.7150878906250, 2016.9938964843750, 1156.7160644531250, 2016.9959716796875, 1156.7269287109375, 2017.0014648437500);
// Segment 1156.7268834554304 2017.0014646778025 0.0203483811973 0.0039136894418 -0.0070717276929 -0.0122220954353 1156.7093435710913 2016.9886936938033
path.cubicTo(1156.7471923828125, 2017.0053710937500, 1156.7022705078125, 2016.9764404296875, 1156.7093505859375, 2016.9886474609375);
// Segment 1156.7093435710913 2016.9886936938033 0.0120617528380 0.0065934603083 -0.0032637855860 -0.0184860248562 1156.6941780622435 2016.9674742350060
path.cubicTo(1156.7214355468750, 2016.9952392578125, 1156.6909179687500, 2016.9489746093750, 1156.6942138671875, 2016.9675292968750);
// Segment 1156.6941780622435 2016.9674742350060 0.0118595244351 0.0220843520393 0.0028009248351 -0.0403311981485 1156.6974975466728 2017.0000000000000
path.cubicTo(1156.7060546875000, 2016.9895019531250, 1156.7003173828125, 2016.9597167968750, 1156.6975097656250, 2017.0000000000000);
// split at 0.4572408795357
// path.cubicTo(1156.6995849609375, 2016.9775390625000, 1156.7014160156250, 2016.9768066406250, 1156.7014160156250, 2016.9768066406250);
// path.cubicTo(1156.7014160156250, 2016.9769287109375, 1156.6989746093750, 2016.9781494140625, 1156.6975097656250, 2017.0000000000000);
// Segment 1156.6974975466728 2017.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1158.0474975705147 2017.0000000000000
path.lineTo(1158.0474853515625, 2017.0000000000000);
path.close();
// start loop, contour: 13
// Segment 1156.6848182678223 2015.6466051936150 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2015.6466051936150
path.moveTo(1156.6848144531250, 2015.6466064453125);
path.lineTo(1145.3381347656250, 2015.6466064453125);
// Segment 1145.3381097316742 2015.6466051936150 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.3381097316742 2016.9966052174568
path.lineTo(1145.3381347656250, 2016.9965820312500);
// Segment 1145.3381097316742 2016.9966052174568 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2016.9966052174568
path.lineTo(1156.6848144531250, 2016.9965820312500);
// Segment 1156.6848182678223 2016.9966052174568 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1156.6848182678223 2015.6466051936150
path.lineTo(1156.6848144531250, 2015.6466064453125);
path.close();
// start loop, contour: 14
// Segment 1145.8121519375022 2016.8021351246741 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220237907 2014.8347900061515
path.moveTo(1145.8121337890625, 2016.8021240234375);
path.lineTo(1147.8063964843750, 2014.8348388671875);
// Segment 1147.8064220237907 2014.8347900061515 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8583376121346 2013.8737301678750
path.lineTo(1146.8583984375000, 2013.8737792968750);
// Segment 1146.8583376121346 2013.8737301678750 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1144.8640675258462 2015.8410752863977
path.lineTo(1144.8640136718750, 2015.8410644531250);
// Segment 1144.8640675258462 2015.8410752863977 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.8121519375022 2016.8021351246741
path.lineTo(1145.8121337890625, 2016.8021240234375);
path.close();
// start loop, contour: 15
// Segment 1147.8064220516883 2014.8347899786306 0.5430154146087 -0.5356841365729 0.5430154146087 0.5356841365729 1147.8064220516883 2012.9239773430752
path.moveTo(1147.8063964843750, 2014.8348388671875);
path.cubicTo(1148.3494873046875, 2014.2990722656250, 1148.3494873046875, 2013.4597167968750, 1147.8063964843750, 2012.9239501953125);
// Segment 1147.8064220516883 2012.9239773430752 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8583375842370 2013.8850371263100
path.lineTo(1146.8583984375000, 2013.8850097656250);
// Segment 1146.8583375842370 2013.8850371263100 0.0071280060876 0.0070317705240 0.0071280060876 -0.0070317705240 1146.8583375842370 2013.8737301953959
path.cubicTo(1146.8654785156250, 2013.8920898437500, 1146.8654785156250, 2013.8666992187500, 1146.8583984375000, 2013.8737792968750);
// Segment 1146.8583375842370 2013.8737301953959 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220516883 2014.8347899786306
path.lineTo(1147.8063964843750, 2014.8348388671875);
path.close();
// start loop, contour: 16
// Segment 1147.8064220516883 2012.9239773430752 -0.5379138488298 -0.5306514472866 0.5379138488298 -0.5306514472866 1145.8955864341058 2012.9239773430752
path.moveTo(1147.8063964843750, 2012.9239501953125);
path.cubicTo(1147.2685546875000, 2012.3933105468750, 1146.4334716796875, 2012.3933105468750, 1145.8956298828125, 2012.9239501953125);
// Segment 1145.8955864341058 2012.9239773430752 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436709015571 2013.8850371263100
path.lineTo(1146.8436279296875, 2013.8850097656250);
// Segment 1146.8436709015571 2013.8850371263100 0.0122295718664 -0.0120644598103 -0.0122295718664 -0.0120644598103 1146.8583375842370 2013.8850371263100
path.cubicTo(1146.8559570312500, 2013.8729248046875, 1146.8460693359375, 2013.8729248046875, 1146.8583984375000, 2013.8850097656250);
// Segment 1146.8583375842370 2013.8850371263100 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220516883 2012.9239773430752
path.lineTo(1147.8063964843750, 2012.9239501953125);
path.close();
// start loop, contour: 17
// Segment 1145.8955864579798 2012.9239773195236 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615803600 2016.0445936810224
path.moveTo(1145.8956298828125, 2012.9239501953125);
path.lineTo(1142.7322998046875, 2016.0445556640625);
// Segment 1142.7322615803600 2016.0445936810224 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6803460000633 2017.0056535113604
path.lineTo(1143.6802978515625, 2017.0056152343750);
// Segment 1143.6803460000633 2017.0056535113604 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436708776831 2013.8850371498615
path.lineTo(1146.8436279296875, 2013.8850097656250);
// Segment 1146.8436708776831 2013.8850371498615 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.8955864579798 2012.9239773195236
path.lineTo(1145.8956298828125, 2012.9239501953125);
path.close();
// start loop, contour: 18
// Segment 1142.7322615564860 2016.0445937045740 -0.0343838913237 0.0339196727021 0.0561572931720 -0.0710493024751 1142.5744069596683 2016.2183613784646
path.moveTo(1142.7322998046875, 2016.0445556640625);
path.cubicTo(1142.6978759765625, 2016.0784912109375, 1142.6306152343750, 2016.1473388671875, 1142.5744628906250, 2016.2183837890625);
// Segment 1142.5744069596683 2016.2183613784646 -0.0547779032556 0.0720510806539 0.0000000000000 -0.2570904015602 1142.3937679156661 2016.7286419868469
path.cubicTo(1142.5196533203125, 2016.2904052734375, 1142.3937988281250, 2016.4715576171875, 1142.3937988281250, 2016.7286376953125);
// Segment 1142.3937679156661 2016.7286419868469 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.7437679395080 2016.7286419868469
path.lineTo(1143.7437744140625, 2016.7286376953125);
// Segment 1143.7437679395080 2016.7286419868469 -0.0051909534315 0.0665915567290 0.0133980913650 -0.0361675066532 1143.6976291086639 2016.9514128270803
path.cubicTo(1143.7385253906250, 2016.7952880859375, 1143.7110595703125, 2016.9152832031250, 1143.6976318359375, 2016.9514160156250);
// Segment 1143.6976291086639 2016.9514128270803 -0.0142876819622 0.0277028472317 0.0040377216094 -0.0063254385208 1143.6490888124401 2017.0354042045738
path.cubicTo(1143.6833496093750, 2016.9791259765625, 1143.6530761718750, 2017.0290527343750, 1143.6490478515625, 2017.0354003906250);
// Segment 1143.6490888124401 2017.0354042045738 -0.0045813437564 0.0032098513409 -0.0343840362634 0.0339198156850 1143.6803460239373 2017.0056534878088
path.cubicTo(1143.6445312500000, 2017.0385742187500, 1143.6459960937500, 2017.0395507812500, 1143.6802978515625, 2017.0056152343750);
// Segment 1143.6803460239373 2017.0056534878088 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615564860 2016.0445937045740
path.lineTo(1142.7322998046875, 2016.0445556640625);
path.close();
// start loop, contour: 19
// Segment 1142.5947256938614 2016.2481120952295 -0.1857487117715 0.1832409092043 0.0167379373694 -0.0990717748979 1142.3430278987244 2016.7518748698508
path.moveTo(1142.5947265625000, 2016.2481689453125);
path.cubicTo(1142.4089355468750, 2016.4313964843750, 1142.3597412109375, 2016.6528320312500, 1142.3430175781250, 2016.7518310546875);
// Segment 1142.3430278987244 2016.7518748698508 -0.0156657977007 0.1069052535795 0.0000000000000 -0.0339197441936 1142.3249999880791 2017.0000000000000
path.cubicTo(1142.3273925781250, 2016.8587646484375, 1142.3249511718750, 2016.9660644531250, 1142.3249511718750, 2017.0000000000000);
// Segment 1142.3249999880791 2017.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6750000119209 2017.0000000000000
path.lineTo(1143.6750488281250, 2017.0000000000000);
// Segment 1143.6750000119209 2017.0000000000000 0.0000000000000 -0.0339197441936 -0.0015261841961 -0.0051459911965 1143.6741640831724 2016.9767671169961
path.cubicTo(1143.6750488281250, 2016.9660644531250, 1143.6726074218750, 2016.9716796875000, 1143.6741943359375, 2016.9768066406250);
// Segment 1143.6741640831724 2016.9767671169961 -0.0007886982052 0.0013596649622 0.0074114058388 -0.0224954551713 1143.6525251830094 2017.0486861571169
path.cubicTo(1143.6733398437500, 2016.9781494140625, 1143.6599121093750, 2017.0262451171875, 1143.6524658203125, 2017.0487060546875);
// split at 0.4203657805920
// path.cubicTo(1143.6738281250000, 2016.9774169921875, 1143.6712646484375, 2016.9862060546875, 1143.6678466796875, 2016.9979248046875);
// path.cubicTo(1143.6630859375000, 2017.0140380859375, 1143.6567382812500, 2017.0356445312500, 1143.6524658203125, 2017.0487060546875);
// Segment 1143.6525251830094 2017.0486861571169 -0.0119644334077 0.0236755853369 0.0381324473830 -0.0447670202574 1143.5428101613127 2017.2091718784643
path.cubicTo(1143.6405029296875, 2017.0723876953125, 1143.5809326171875, 2017.1644287109375, 1143.5428466796875, 2017.2092285156250);
// Segment 1143.5428101613127 2017.2091718784643 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.5947256938614 2016.2481120952295
path.lineTo(1142.5947265625000, 2016.2481689453125);
path.close();
// start loop, contour: 20
// Segment 1142.3249999880791 2017.0000000000000 0.0000000000000 0.0339197441936 -0.0156657977007 -0.1069052535795 1142.3430278987244 2017.2481251301492
path.moveTo(1142.3249511718750, 2017.0000000000000);
path.cubicTo(1142.3249511718750, 2017.0339355468750, 1142.3273925781250, 2017.1412353515625, 1142.3430175781250, 2017.2481689453125);
// Segment 1142.3430278987244 2017.2481251301492 0.0167379373694 0.0990717748979 -0.1857487117715 -0.1832409092043 1142.5947256938614 2017.7518879047705
path.cubicTo(1142.3597412109375, 2017.3471679687500, 1142.4089355468750, 2017.5686035156250, 1142.5947265625000, 2017.7518310546875);
// split at 0.4008532166481
// path.cubicTo(1142.3497314453125, 2017.2878417968750, 1142.3616943359375, 2017.3471679687500, 1142.3854980468750, 2017.4158935546875);
// path.cubicTo(1142.4211425781250, 2017.5185546875000, 1142.4833984375000, 2017.6420898437500, 1142.5947265625000, 2017.7518310546875);
// Segment 1142.5947256938614 2017.7518879047705 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.5428101613127 2016.7908281215357
path.lineTo(1143.5428466796875, 2016.7907714843750);
// Segment 1143.5428101613127 2016.7908281215357 0.0381324473830 0.0447670202574 -0.0119644334077 -0.0236755853369 1143.6525251830094 2016.9513138428831
path.cubicTo(1143.5809326171875, 2016.8355712890625, 1143.6405029296875, 2016.9276123046875, 1143.6524658203125, 2016.9512939453125);
// Segment 1143.6525251830094 2016.9513138428831 0.0074114058388 0.0224954551713 -0.0007886982052 -0.0013596649622 1143.6741640831724 2017.0232328830039
path.cubicTo(1143.6599121093750, 2016.9737548828125, 1143.6733398437500, 2017.0218505859375, 1143.6741943359375, 2017.0231933593750);
// Segment 1143.6741640831724 2017.0232328830039 -0.0015261841961 0.0051459911965 0.0000000000000 0.0339197441936 1143.6750000119209 2017.0000000000000
path.cubicTo(1143.6726074218750, 2017.0283203125000, 1143.6750488281250, 2017.0339355468750, 1143.6750488281250, 2017.0000000000000);
// Segment 1143.6750000119209 2017.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.3249999880791 2017.0000000000000
path.lineTo(1142.3249511718750, 2017.0000000000000);
path.close();
// start loop, contour: 21
// Segment 1142.5947256938614 2017.7518879047705 -0.0799271403989 -0.1522613934208 -0.2174629955730 -0.2879403701950 1142.7322615564860 2017.9554062954260
path.moveTo(1142.5947265625000, 2017.7518310546875);
path.cubicTo(1142.5147705078125, 2017.5996093750000, 1142.5147705078125, 2017.6674804687500, 1142.7322998046875, 2017.9554443359375);
// Segment 1142.7322615564860 2017.9554062954260 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6803460239373 2016.9943465121912
path.lineTo(1143.6802978515625, 2016.9943847656250);
// Segment 1143.6803460239373 2016.9943465121912 0.0799271403989 0.1522613934208 0.2174629955730 0.2879403701950 1143.5428101613127 2016.7908281215357
path.cubicTo(1143.7602539062500, 2017.1466064453125, 1143.7602539062500, 2017.0787353515625, 1143.5428466796875, 2016.7907714843750);
// Segment 1143.5428101613127 2016.7908281215357 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.5947256938614 2017.7518879047705
path.lineTo(1142.5947265625000, 2017.7518310546875);
path.close();
// start loop, contour: 22
// Segment 1142.7322615746241 2017.9554063133189 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.8955864522438 2021.0760227493236
path.moveTo(1142.7322998046875, 2017.9554443359375);
path.lineTo(1145.8956298828125, 2021.0760498046875);
// Segment 1145.8955864522438 2021.0760227493236 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8436708834190 2020.1149629303029
path.lineTo(1146.8436279296875, 2020.1149902343750);
// Segment 1146.8436708834190 2020.1149629303029 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1143.6803460057993 2016.9943464942983
path.lineTo(1143.6802978515625, 2016.9943847656250);
// Segment 1143.6803460057993 2016.9943464942983 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1142.7322615746241 2017.9554063133189
path.lineTo(1142.7322998046875, 2017.9554443359375);
path.close();
// start loop, contour: 23
// Segment 1145.8955864341058 2021.0760227314306 0.2730164534637 0.2693304447891 -0.3016608168437 0.0000000000000 1146.8510041236877 2021.4740112423897
path.moveTo(1145.8956298828125, 2021.0760498046875);
path.cubicTo(1146.1685791015625, 2021.3453369140625, 1146.5493164062500, 2021.4739990234375, 1146.8509521484375, 2021.4739990234375);
// Segment 1146.8510041236877 2021.4740112423897 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8510041236877 2020.1240112185478
path.lineTo(1146.8509521484375, 2020.1240234375000);
// Segment 1146.8510041236877 2020.1240112185478 -0.0031276099109 0.0031991747760 0.0281856144058 0.0140930868099 1146.8580791488898 2020.1202473991566
path.cubicTo(1146.8479003906250, 2020.1271972656250, 1146.8862304687500, 2020.1343994140625, 1146.8580322265625, 2020.1202392578125);
// split at 0.3845077157021
// path.cubicTo(1146.8497314453125, 2020.1252441406250, 1146.8547363281250, 2020.1270751953125, 1146.8596191406250, 2020.1280517578125);
// path.cubicTo(1146.8675537109375, 2020.1296386718750, 1146.8753662109375, 2020.1289062500000, 1146.8580322265625, 2020.1202392578125);
// Segment 1146.8580791488898 2020.1202473991566 -0.0369995545027 -0.0123195805663 0.0067223483810 0.0136883790721 1146.8436709015571 2020.1149629481959
path.cubicTo(1146.8210449218750, 2020.1079101562500, 1146.8503417968750, 2020.1286621093750, 1146.8436279296875, 2020.1149902343750);
// Segment 1146.8436709015571 2020.1149629481959 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.8955864341058 2021.0760227314306
path.lineTo(1145.8956298828125, 2021.0760498046875);
path.close();
// start loop, contour: 24
// Segment 1146.8510041236877 2021.4740112423897 0.3016605789999 0.0000000000000 -0.2730166120260 0.2693306012106 1147.8064220516883 2021.0760227314306
path.moveTo(1146.8509521484375, 2021.4739990234375);
path.cubicTo(1147.1527099609375, 2021.4739990234375, 1147.5334472656250, 2021.3453369140625, 1147.8063964843750, 2021.0760498046875);
// Segment 1147.8064220516883 2021.0760227314306 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8583375842370 2020.1149629481959
path.lineTo(1146.8583984375000, 2020.1149902343750);
// Segment 1146.8583375842370 2020.1149629481959 -0.0067222671256 0.0136883164611 0.0369996293611 -0.0123196021258 1146.8439293663473 2020.1202473404985
path.cubicTo(1146.8515625000000, 2020.1286621093750, 1146.8809814453125, 2020.1079101562500, 1146.8438720703125, 2020.1202392578125);
// Segment 1146.8439293663473 2020.1202473404985 -0.0281857033438 0.0140931104690 0.0031276541428 0.0031991704542 1146.8510041236877 2020.1240112185478
path.cubicTo(1146.8157958984375, 2020.1343994140625, 1146.8541259765625, 2020.1271972656250, 1146.8509521484375, 2020.1240234375000);
// Segment 1146.8510041236877 2020.1240112185478 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8510041236877 2021.4740112423897
path.lineTo(1146.8509521484375, 2021.4739990234375);
path.close();
// start loop, contour: 25
// Segment 1147.8064220516883 2021.0760227314306 0.5430154146087 -0.5356841365729 0.5430154146087 0.5356841365729 1147.8064220516883 2019.1652101405787
path.moveTo(1147.8063964843750, 2021.0760498046875);
path.cubicTo(1148.3494873046875, 2020.5402832031250, 1148.3494873046875, 2019.7009277343750, 1147.8063964843750, 2019.1651611328125);
// Segment 1147.8064220516883 2019.1652101405787 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8583375842370 2020.1262699238134
path.lineTo(1146.8583984375000, 2020.1262207031250);
// Segment 1146.8583375842370 2020.1262699238134 0.0071280060876 0.0070317705240 0.0071280060876 -0.0070317705240 1146.8583375842370 2020.1149629481959
path.cubicTo(1146.8654785156250, 2020.1333007812500, 1146.8654785156250, 2020.1079101562500, 1146.8583984375000, 2020.1149902343750);
// Segment 1146.8583375842370 2020.1149629481959 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220516883 2021.0760227314306
path.lineTo(1147.8063964843750, 2021.0760498046875);
path.close();
// start loop, contour: 26
// Segment 1147.8064220383478 2019.1652101274185 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1145.8121519520594 2017.1978648896866
path.moveTo(1147.8063964843750, 2019.1651611328125);
path.lineTo(1145.8121337890625, 2017.1978759765625);
// Segment 1145.8121519520594 2017.1978648896866 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1144.8640675112890 2018.1589246992417
path.lineTo(1144.8640136718750, 2018.1589355468750);
// Segment 1144.8640675112890 2018.1589246992417 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1146.8583375975775 2020.1262699369736
path.lineTo(1146.8583984375000, 2020.1262207031250);
// Segment 1146.8583375975775 2020.1262699369736 0.0000000000000 0.0000000000000 0.0000000000000 0.0000000000000 1147.8064220383478 2019.1652101274185
path.lineTo(1147.8063964843750, 2019.1651611328125);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads66(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(2, 0);
path.quadTo(3, 1, 2, 2);
path.lineTo(2, 3);
path.close();
path.moveTo(2, 1);
path.lineTo(2, 1);
path.quadTo(1, 2, 2, 2);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads67(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(3, 2);
path.quadTo(1, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(1, 0);
path.quadTo(2, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads68(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(1, 2);
path.quadTo(0, 3, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(0, 1);
path.quadTo(1, 3, 2, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads69(skiatest::Reporter* reporter,const char* filename) {
SkPath path;
path.moveTo(1, 0);
path.quadTo(2, 2, 2, 3);
path.lineTo(2, 3);
path.close();
path.moveTo(1, 0);
path.lineTo(1, 0);
path.quadTo(3, 0, 1, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads70(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 1);
path.quadTo(2, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(2, 0);
path.lineTo(2, 2);
path.quadTo(1, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads71(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 1);
path.quadTo(2, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(3, 0);
path.lineTo(2, 2);
path.quadTo(1, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
}
static void testQuads72(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(1, 1);
path.quadTo(2, 3, 3, 3);
path.lineTo(3, 3);
path.close();
path.moveTo(0, 1);
path.lineTo(2, 2);
path.quadTo(1, 3, 3, 3);
path.close();
testSimplify(reporter, path, filename);
fix security bug This fix is a tradeoff. It changes intersection to treat a case where one coincident run is intersected at one point and the other edge is not as continuing to be a span. The old code tried to treat this as a single point. The old code is probably right, but this change alone made the data structures inconsistent. Later, extending the coincident runs would fail by incorrectly discarding the single point intersection. As a result, this fixes the security test and one other, but makes a different test fail. Isolating the failure uncovered a reduced case that fails with and without the change, so there are more serious problems here. Those problems are addressed in a separate CL. Many of the test edits below remove ill-thought out debugging messaging that fire off global state, which isn't usable in a multi-threaded test environment. In the end, with this fix, all existing tests (modulo one new failure and one new non-failure) pass in debug and in the extended release test suites. TBR=reed@google.com BUG=614248 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2018513003 Review-Url: https://codereview.chromium.org/2018513003
2016-05-26 16:01:47 +00:00
}
quad and conic do not intersect If a quad a conic intersect only where the end of one is contained by the convex hull of the other, and the curve contained by the hull is nearly a straight line, treating it as a line may move the end point to the other side of the curve. Detect this by checking to see if the end point is in the hull, and if so, continue to subdivide the curve rather than treating it as a line. This fixes several existing tests that were disabled earlier this year. A typo in SkDCurve::nearPoint() prevented detecting when the end of a line was nearly touching a curve. Also fixed concidence a bit to get the second half of tiger further along. All existing tests, including extended testing in Release and the first half of tiger, work. TBR=reed@google.com BUG=skia:5131 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2338323002 Review-Url: https://codereview.chromium.org/2338323002
2016-09-15 14:48:18 +00:00
static void testQuads73(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(0, 0);
path.quadTo(0, 1, 1, 2);
path.lineTo(0, 3);
path.close();
path.moveTo(0, 0);
path.lineTo(0, 0);
path.quadTo(0, 1, 1, 1);
path.close();
testSimplify(reporter, path, filename);
}
fix cubic linear test Check to see if the line between end points is degenerate before measuring control points. Also, add test case for a bug to see if it shows up on any platform. TBR=reed@google.com BUG=skia:5169, skia:5240 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2375053002 Review-Url: https://codereview.chromium.org/2375053002
2016-09-28 16:22:17 +00:00
static void bug5169(skiatest::Reporter* reporter, const char* filename) {
SkPath path;
path.moveTo(SkBits2Float(0x00000000), SkBits2Float(0x4281c71c)); // 0, 64.8889f
path.cubicTo(SkBits2Float(0x434e0000), SkBits2Float(0x4281c71c), SkBits2Float(0x00000000), SkBits2Float(0xc2a238e4), SkBits2Float(0x00000000), SkBits2Float(0x4281c71c)); // 206, 64.8889f, 0, -81.1111f, 0, 64.8889f
path.moveTo(SkBits2Float(0x43300000), SkBits2Float(0x41971c72)); // 176, 18.8889f
path.cubicTo(SkBits2Float(0xc29e0000), SkBits2Float(0xc25c71c7), SkBits2Float(0x42b20000), SkBits2Float(0x42fbc71c), SkBits2Float(0x43300000), SkBits2Float(0x41971c72)); // -79, -55.1111f, 89, 125.889f, 176, 18.8889f
testSimplify(reporter, path, filename);
}
cumulative pathops patch Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer. Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering. Unify t and point values so that small intervals can be eliminated in one pass. Break cubics up front to eliminate loops and cusps. Make the Simplify and Op code more regular and eliminate arbitrary differences. Add a builder that takes an array of paths and operators. Delete unused code. BUG=skia:3588 R=reed@google.com Review URL: https://codereview.chromium.org/1037573004
2015-03-26 14:52:43 +00:00
static void (*skipTest)(skiatest::Reporter* , const char* filename) = 0;
add tiger tests Add the tiger test suite. In extended mode (-x) this adds 700K new pathop tests. Normally, it adds about 3500 tests. TBR=reed@google.com BUG=skia:5131 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2373533002 Review-Url: https://codereview.chromium.org/2373533002
2016-09-26 17:25:26 +00:00
static void (*firstTest)(skiatest::Reporter* , const char* filename) = 0;
cumulative pathops patch Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer. Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering. Unify t and point values so that small intervals can be eliminated in one pass. Break cubics up front to eliminate loops and cusps. Make the Simplify and Op code more regular and eliminate arbitrary differences. Add a builder that takes an array of paths and operators. Delete unused code. BUG=skia:3588 R=reed@google.com Review URL: https://codereview.chromium.org/1037573004
2015-03-26 14:52:43 +00:00
static void (*stopTest)(skiatest::Reporter* , const char* filename) = 0;
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static TestDesc tests[] = {
fix cubic linear test Check to see if the line between end points is degenerate before measuring control points. Also, add test case for a bug to see if it shows up on any platform. TBR=reed@google.com BUG=skia:5169, skia:5240 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2375053002 Review-Url: https://codereview.chromium.org/2375053002
2016-09-28 16:22:17 +00:00
TEST(bug5169),
split tight quads and conics Tight quads and conics may nearly fold over on themselves, confusing coincidence against other curves. Split them at their max curvature early on to avoid complicating later logic. TBR=reed@google.com BUG=skia:5131 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2357353002 Review-Url: https://codereview.chromium.org/2357353002
2016-09-23 12:47:20 +00:00
TEST(testQuads73),
pathops coincidence and security rewrite Most changes stem from working on an examples bracketed by #if DEBUG_UNDER_DEVELOPMENT // tiger These exposed many problems with coincident curves, as well as errors throughout the code. Fixing these errors also fixed a number of fuzzer-inspired bug reports. * Line/Curve Intersections Check to see if the end of the line nearly intersects the curve. This was a FIXME in the old code. * Performance Use a central chunk allocator. Plumb the allocator into the global variable state so that it can be shared. (Note that 'SkGlobalState' is allocated on the stack and is visible to children functions but not other threads.) * Refactor Let SkOpAngle grow up from a structure to a class. Let SkCoincidentSpans grow up from a structure to a class. Rename enum Alias to AliasMatch. * Coincidence Rewrite Add more debugging to coincidence detection. Parallel debugging routines have read-only logic to report the current coincidence state so that steps through the logic can expose whether things got better or worse. More functions can error-out and cause the pathops engine to non-destructively exit. * Accuracy Remove code that adjusted point locations. Instead, offset the curve part so that sorted curves all use the same origin. Reduce the size (and influence) of magic numbers. * Testing The debug suite with verify and the full release suite ./out/Debug/pathops_unittest -v -V ./out/Release/pathops_unittest -v -V -x expose one error. That error is captured as cubics_d3. This error exists in the checked in code as well. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2128633003 Review-Url: https://codereview.chromium.org/2128633003
2016-07-18 17:01:36 +00:00
TEST(testQuads72),
TEST(testQuads71),
TEST(testQuads70),
TEST(testQuads69),
TEST(testQuads68),
TEST(testQuads67),
TEST(testQuads66),
TEST(dean4),
fix security bug This fix is a tradeoff. It changes intersection to treat a case where one coincident run is intersected at one point and the other edge is not as continuing to be a span. The old code tried to treat this as a single point. The old code is probably right, but this change alone made the data structures inconsistent. Later, extending the coincident runs would fail by incorrectly discarding the single point intersection. As a result, this fixes the security test and one other, but makes a different test fail. Isolating the failure uncovered a reduced case that fails with and without the change, so there are more serious problems here. Those problems are addressed in a separate CL. Many of the test edits below remove ill-thought out debugging messaging that fire off global state, which isn't usable in a multi-threaded test environment. In the end, with this fix, all existing tests (modulo one new failure and one new non-failure) pass in debug and in the extended release test suites. TBR=reed@google.com BUG=614248 GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2018513003 Review-Url: https://codereview.chromium.org/2018513003
2016-05-26 16:01:47 +00:00
TEST(fuzz763_4713_b),
fix pathops coincidence fuzz bug Simplifying a series of rects with very large bounds triggers a coincidence bug where, after one of the intersection points that marks a coincident range has been deleted, it is referenced. Both the deletion and reference is (probably) happening in the SkOpCoincidence::AddExpanded() phase of HandleCoincidence(), and may signify a bug that could happen with usable input data, but I haven't been able to determine that. For now, abort the Simplify() when the erroneous condition is detected. TBR=reed@google.com BUG=558281 Review URL: https://codereview.chromium.org/1463923002
2015-11-20 22:06:28 +00:00
TEST(fuzz_twister2),
Enabling clip stack flattening exercises path ops. Iterating through the 903K skps that represent the imagable 1M top web pages triggers a number of bugs, some of which are addressed here. Some web pages trigger intersecting cubic representations of arc with their conic counterparts. This exposed a flaw in coincident detection that caused an infinite loop. The loop alternatively extended the coincident section and, determining the that the bounds of the curve pairs did not overlap, deleted the extension. Track the number of times the coincident detection is called, and if it exceeds an empirically found limit, assume that the curves are coincident and force it to be so. The loop count limit can be determined by enabling DEBUG_T_SECT_LOOP_COUNT and running all tests. The largest count is reported on completion. Another class of bugs was caused by concident detection duplicating nearly identical points that had been merged earlier. To track these bugs, the 'handle coincidence' code was duplicated as a const debug variety that reported if one of a dozen or so irregularities are present; then it is easier to see when a block of code that fixes one irregularity regresses another. Creating the debug const code version exposed some non-debug code that could be const, and some that was experimental and could be removed. Set DEBUG_COINCIDENCE to track coincidence health and handling. For running on Chrome, DEBUG_VERIFY checks the result of pathops against the same operation using SkRegion to verify that the results are nearly the same. When visualizing the pathops work using tools/pathops_visualizer.htm, set DEBUG_DUMP_ALIGNMENT to see the curves after they've been aligned for coincidence. Other bugs fixed include detecting when a section of a pair of curves have devolved into lines and are coincident. TBR=reed@google.com Review URL: https://codereview.chromium.org/1394503003
2015-10-16 16:03:38 +00:00
TEST(fuzz_twister),
looking for fuzz failure These tests were isolated from their respective minimized test cases. The tests work fine and pass path ops internal validation; hopefully some more intensive x-san or valgrind test will help isolate the bug. Sheriff, please revert if it fails and I don't get to it first. TBR=reed@google.com,halcanary@google.com BUG=535127,535151 Review URL: https://codereview.chromium.org/1359263003
2015-09-24 01:07:51 +00:00
TEST(fuzz994s_3414),
TEST(fuzz994s_11),
compute split conic endpoints exactly The divide by w can generate slightly erroneous results even for t == 0 or t == 1. The error in turn defeats detecting a point in common for a pair of curves that travel in opposite directions. Instead, special case endpoints when the t is 0 or 1. TBR=reed@google.com BUG=514118 Review URL: https://codereview.chromium.org/1259513004
2015-07-28 12:12:19 +00:00
TEST(cr514118),
fix path ops fuzz buster Mark collapsed segments as done and remove collapsed segment references from the coincidence array. Also add test names to global debugging. R=fmalita@chromium.org BUG=512592 Review URL: https://codereview.chromium.org/1250293002
2015-07-23 19:40:22 +00:00
TEST(fuzz864a),
When three or more edges are coincident, the logic needs to compute the overlapping ranges and combine the winding into a single destination. This computes coincidence more rigorously, fixing the edge cases exposed by this bug. Also, add the ability to debug and dump pathop structures from the coincident context. TBR=reed@google.com BUG=skia:3651 Review URL: https://codereview.chromium.org/1182493015
2015-07-06 18:38:33 +00:00
TEST(testQuads65),
fix builder winding again Record the nesting level when finding the edge winding contribution so that inner edges can be reversed as needed. R=fmalita@chromium.org BUG=skia:3838 Review URL: https://codereview.chromium.org/1140383002
2015-05-18 19:56:57 +00:00
TEST(testIssue3838_3),
fix path op builder The rewrite of path ops caused the inner contour direction to be reversed. This exposed an existing bug in path ops builder, namely that the implicit winding of the internal sum path could hide inner contours if they ended up in the wrong direction. Setting the sum path's fill type to even-odd ensures that the inner contours aren't discarded. R=fmalita@chromium.org BUG=skia:3838 Review URL: https://codereview.chromium.org/1126193004
2015-05-14 21:18:13 +00:00
TEST(testIssue3838),
deal more consistently with unsortable edges Improve line/curve coincident detection and resolution. This fixed the remaining simple failures. When an edge is unsortable, use the ray intersection to determine the angles' winding. Deal with degenerate segments. TBR=reed@google.com BUG=skia:3588,skia:3762 Review URL: https://codereview.chromium.org/1140813002
2015-05-13 15:23:48 +00:00
TEST(testArc),
Path ops formerly found the topmost unprocessed edge and determined its angle sort order to initialize the winding. This never worked correctly with cubics and was flaky with paths consisting mostly of vertical edges. This replacement shoots axis-aligned rays through all intersecting edges to find the outermost one either horizontally or vertically. The resulting code is smaller and twice as fast. To support this, most of the horizontal / vertical intersection code was rewritten and standardized, and old code supporting the top-directed winding was deleted. Contours were pointed to by an SkTDArray. Instead, put them in a linked list, and designate the list head with its own class to ensure that methods that take lists of contours start at the top. This change removed a large percentage of memory allocations used by path ops. TBR=reed@google.com BUG=skia:3588 Review URL: https://codereview.chromium.org/1111333002
2015-05-11 14:21:27 +00:00
TEST(testTriangle2),
TEST(testTriangle1),
Now, path ops natively intersect conics, quads, and cubics in any combination. There are still a class of cubic tests that fail and a handful of undiagnosed failures from skps and fuzz tests, but things are much better overall. Extended tests (150M+) run to completion in release in about 6 minutes; the standard test suite exceeds 100K and finishes in a few seconds on desktops. TBR=reed BUG=skia:3588 Review URL: https://codereview.chromium.org/1037953004
2015-04-20 15:31:59 +00:00
TEST(testQuads64),
TEST(testQuads63),
TEST(testQuads62),
TEST(testRect4),
update pathops core and tests split out skpclip (the test of 1M pictures) into its own project TBR=reed Author: caryclark@google.com Review URL: https://codereview.chromium.org/400033002
2014-07-18 12:08:14 +00:00
TEST(testRect3),
Enabling the canvas bit to turn the clip stack into a flat replace exposed around 100 failures when testing the 800K skp set generated from the top 1M web sites. This fixes all but one of those failures. Major changes include: - Replace angle indices with angle pointers. This was motivated by the need to add angles later but not renumber existing angles. - Aggressive segment chase. When the winding is known on a segment, more aggressively passing that winding to adjacent segments allows fragmented data sets to succeed. - Line segments with ends nearly the same are treated as coincident first. - Transfer partial coincidence by observing that if segment A is partially coincident to B and C then B and C may be partially coincident. TBR=reed Author: caryclark@google.com Review URL: https://codereview.chromium.org/272153002
2014-06-17 12:15:38 +00:00
TEST(testQuadralateral10),
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
TEST(testQuads61),
TEST(testQuads60),
TEST(testQuads59),
TEST(testQuads58),
TEST(testQuads57),
TEST(testQuads56),
TEST(testQuads54),
TEST(testQuads53),
TEST(testQuads52),
TEST(testQuads51),
TEST(testQuads50),
TEST(testQuads49),
TEST(testQuads48),
TEST(testQuads47),
TEST(testQuads46x),
TEST(testQuads45),
TEST(testQuads44),
TEST(testQuads43),
TEST(testQuads42),
TEST(testQuads41),
TEST(testQuads36),
TEST(testQuads37),
TEST(testQuads38),
TEST(testQuads39),
TEST(testQuads40),
TEST(testQuads16),
TEST(testQuads17),
TEST(testQuads18),
TEST(testQuads19),
TEST(testQuads20),
TEST(testQuads21),
TEST(testQuads22),
TEST(testQuads23),
TEST(testQuads24),
TEST(testQuads25),
TEST(testQuads26),
TEST(testQuads27),
TEST(testQuads28),
TEST(testQuads29),
TEST(testQuads30),
TEST(testQuads31),
TEST(testQuads32),
TEST(testQuads33),
TEST(testQuads34),
TEST(testQuads35),
TEST(testDegenerates1),
TEST(testQuad13),
TEST(testQuad14),
TEST(testQuad15),
TEST(testQuadratic56),
TEST(testQuadralateral4),
TEST(testQuadralateral3),
TEST(testDegenerate5),
TEST(testQuad12),
cumulative pathops patch Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer. Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering. Unify t and point values so that small intervals can be eliminated in one pass. Break cubics up front to eliminate loops and cusps. Make the Simplify and Op code more regular and eliminate arbitrary differences. Add a builder that takes an array of paths and operators. Delete unused code. BUG=skia:3588 R=reed@google.com Review URL: https://codereview.chromium.org/1037573004
2015-03-26 14:52:43 +00:00
TEST(testQuadratic51),
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
TEST(testQuad8),
TEST(testQuad11),
TEST(testQuad10),
TEST(testQuad9),
TEST(testTriangles4x),
TEST(testTriangles3x),
TEST(testRect2),
TEST(testRect1),
TEST(tooCloseTest),
TEST(skphealth_com76),
TEST(testQuadLineIntersect1),
TEST(testQuadLineIntersect2),
TEST(testQuadLineIntersect3),
TEST(testQuad7),
TEST(testQuad6),
TEST(testQuad5),
TEST(testQuad4),
TEST(testQuad3),
TEST(testQuad2),
TEST(testAddTCoincident2),
TEST(testAddTCoincident1),
TEST(testTriangles2),
TEST(testTriangles1),
TEST(testQuadratic97),
TEST(testQuadratic96),
TEST(testQuadratic95),
TEST(testQuadratic94),
TEST(testQuadralateral2),
TEST(testQuad1),
TEST(testCubic2),
TEST(testCubic1),
TEST(testQuadralateral1),
TEST(testLine85),
TEST(testLine84),
TEST(testLine84x),
TEST(testLine83),
TEST(testLine82h),
TEST(testLine82g),
TEST(testLine82f),
TEST(testLine82e),
TEST(testLine82d),
TEST(testLine82c),
TEST(testLine82b),
TEST(testLine82a),
TEST(testLine82),
TEST(testQuadratic93),
TEST(testQuadratic92x),
TEST(testQuadratic91),
TEST(testQuadratic90x),
TEST(testQuadratic89x),
TEST(testQuadratic88),
TEST(testQuadratic87),
TEST(testQuadratic86),
TEST(testQuadratic85),
TEST(testQuadratic84),
TEST(testQuadratic83),
TEST(testQuadratic82),
TEST(testQuadratic81),
TEST(testQuadratic80),
TEST(testEight1),
TEST(testEight2),
TEST(testEight3),
TEST(testEight4),
TEST(testEight5),
TEST(testEight6),
TEST(testEight7),
TEST(testEight8),
TEST(testEight9),
TEST(testEight10),
TEST(testQuadratic79),
TEST(testQuadratic78),
TEST(testQuadratic77),
TEST(testQuadratic76),
TEST(testQuadratic75),
TEST(testQuadratic74),
TEST(testQuadratic73),
TEST(testQuadratic72),
TEST(testQuadratic71),
TEST(testQuadratic70x),
TEST(testQuadratic69),
TEST(testQuadratic68),
TEST(testQuadratic67x),
TEST(testQuadratic65),
TEST(testQuadratic64),
TEST(testQuadratic63),
TEST(testLine1a),
TEST(testLine1ax),
TEST(testQuadratic59),
TEST(testQuadratic59x),
TEST(testQuadratic58),
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
TEST(testQuadratic55),
TEST(testQuadratic53),
TEST(testQuadratic38),
TEST(testQuadratic37),
TEST(testQuadratic36),
TEST(testQuadratic35),
TEST(testQuadratic34),
TEST(testQuadratic33),
TEST(testQuadratic32),
TEST(testQuadratic31),
TEST(testQuadratic30),
TEST(testQuadratic29),
TEST(testQuadratic28),
TEST(testQuadratic27),
TEST(testQuadratic26),
TEST(testQuadratic25),
TEST(testQuadratic24),
TEST(testQuadratic23),
TEST(testQuadratic22),
TEST(testQuadratic21),
TEST(testQuadratic20),
TEST(testQuadratic19),
TEST(testQuadratic18),
TEST(testQuadratic17x),
TEST(testQuadratic15),
TEST(testQuadratic14),
TEST(testQuadratic9),
TEST(testQuadratic8),
TEST(testQuadratic7),
TEST(testQuadratic6),
TEST(testQuadratic5),
TEST(testQuadratic4x),
TEST(testQuadratic3x),
TEST(testQuadratic2x),
TEST(testQuadratic1x),
TEST(testQuadratic4),
TEST(testQuadratic3),
TEST(testQuadratic2),
TEST(testQuadratic1),
TEST(testLine4ax),
TEST(testLine3aax),
TEST(testLine2ax),
TEST(testLine1ax),
TEST(testQuadralateral9x),
TEST(testQuadralateral8x),
TEST(testQuadralateral7x),
TEST(testQuadralateral6x),
TEST(testQuadralateral6ax),
TEST(testQuadralateral9),
TEST(testQuadralateral8),
TEST(testQuadralateral7),
TEST(testQuadralateral6),
TEST(testQuadralateral6a),
TEST(testFauxQuadralateral6dx),
TEST(testFauxQuadralateral6cx),
TEST(testFauxQuadralateral6bx),
TEST(testFauxQuadralateral6ax),
TEST(testFauxQuadralateral6x),
TEST(testFauxQuadralateral6d),
TEST(testFauxQuadralateral6c),
TEST(testFauxQuadralateral6b),
TEST(testFauxQuadralateral6a),
TEST(testFauxQuadralateral6),
TEST(testQuadralateral5x),
TEST(testQuadralateral5),
TEST(testNondegenerate4x),
TEST(testNondegenerate3x),
TEST(testNondegenerate2x),
TEST(testNondegenerate1x),
TEST(testNondegenerate4),
TEST(testNondegenerate3),
TEST(testNondegenerate2),
TEST(testNondegenerate1),
TEST(testDegenerate4x),
TEST(testDegenerate3x),
TEST(testDegenerate2x),
TEST(testDegenerate1x),
TEST(testDegenerate4),
TEST(testDegenerate3),
TEST(testDegenerate2),
TEST(testDegenerate1),
TEST(testLine79x),
TEST(testLine78x),
TEST(testLine77x),
TEST(testLine76x),
TEST(testLine75x),
TEST(testLine74x),
TEST(testLine73x),
TEST(testLine72x),
TEST(testLine71x),
TEST(testLine70x),
TEST(testLine69x),
TEST(testLine68hx),
TEST(testLine68gx),
TEST(testLine68fx),
TEST(testLine68ex),
TEST(testLine68dx),
TEST(testLine68cx),
TEST(testLine68bx),
TEST(testLine68ax),
TEST(testLine67x),
TEST(testLine66x),
TEST(testLine65x),
TEST(testLine64x),
TEST(testLine63x),
TEST(testLine62x),
TEST(testLine61x),
TEST(testLine60x),
TEST(testLine59x),
TEST(testLine58x),
TEST(testLine57x),
TEST(testLine56x),
TEST(testLine55x),
TEST(testLine54x),
TEST(testLine53x),
TEST(testLine52x),
TEST(testLine51x),
TEST(testLine50x),
TEST(testLine49x),
TEST(testLine48x),
TEST(testLine47x),
TEST(testLine46x),
TEST(testLine45x),
TEST(testLine44x),
TEST(testLine43x),
TEST(testLine42x),
TEST(testLine41x),
TEST(testLine40x),
TEST(testLine38x),
TEST(testLine37x),
TEST(testLine36x),
TEST(testLine35x),
TEST(testLine34x),
TEST(testLine33x),
TEST(testLine32x),
TEST(testLine31x),
TEST(testLine30x),
TEST(testLine29x),
TEST(testLine28x),
TEST(testLine27x),
TEST(testLine26x),
TEST(testLine25x),
TEST(testLine24ax),
TEST(testLine24x),
TEST(testLine23x),
TEST(testLine22x),
TEST(testLine21x),
TEST(testLine20x),
TEST(testLine19x),
TEST(testLine18x),
TEST(testLine17x),
TEST(testLine16x),
TEST(testLine15x),
TEST(testLine14x),
TEST(testLine13x),
TEST(testLine12x),
TEST(testLine11x),
TEST(testLine10ax),
TEST(testLine10x),
TEST(testLine9x),
TEST(testLine8x),
TEST(testLine7bx),
TEST(testLine7ax),
TEST(testLine7x),
TEST(testLine6x),
TEST(testLine5x),
TEST(testLine4x),
TEST(testLine3bx),
TEST(testLine3ax),
TEST(testLine3x),
TEST(testLine2x),
TEST(testLine1x),
TEST(testLine81),
TEST(testLine80),
TEST(testLine79),
TEST(testLine78),
TEST(testLine77),
TEST(testLine76),
TEST(testLine75),
TEST(testLine74),
TEST(testLine73),
TEST(testLine72),
TEST(testLine71),
TEST(testLine70),
TEST(testLine69),
TEST(testLine68h),
TEST(testLine68g),
TEST(testLine68f),
TEST(testLine68e),
TEST(testLine68d),
TEST(testLine68c),
TEST(testLine68b),
TEST(testLine68a),
TEST(testLine67),
TEST(testLine66),
TEST(testLine65),
TEST(testLine64),
TEST(testLine63),
TEST(testLine62),
TEST(testLine61),
TEST(testLine60),
TEST(testLine59),
TEST(testLine58),
TEST(testLine57),
TEST(testLine56),
TEST(testLine55),
TEST(testLine54),
TEST(testLine53),
TEST(testLine52),
TEST(testLine51),
TEST(testLine50),
TEST(testLine49),
TEST(testLine48),
TEST(testLine47),
TEST(testLine46),
TEST(testLine45),
TEST(testLine44),
TEST(testLine43),
TEST(testLine42),
TEST(testLine41),
TEST(testLine40),
TEST(testLine38),
TEST(testLine37),
TEST(testLine36),
TEST(testLine35),
TEST(testLine34),
TEST(testLine33),
TEST(testLine32),
TEST(testLine31),
TEST(testLine30),
TEST(testLine29),
TEST(testLine28),
TEST(testLine27),
TEST(testLine26),
TEST(testLine25),
TEST(testLine24a),
TEST(testLine24),
TEST(testLine23),
TEST(testLine22),
TEST(testLine21),
TEST(testLine20),
TEST(testLine19),
TEST(testLine18),
TEST(testLine17),
TEST(testLine16),
TEST(testLine15),
TEST(testLine14),
TEST(testLine13),
TEST(testLine12),
TEST(testLine11),
TEST(testLine10a),
TEST(testLine10),
TEST(testLine9),
TEST(testLine8),
TEST(testLine7b),
TEST(testLine7a),
TEST(testLine7),
TEST(testLine6),
TEST(testLine5),
TEST(testLine4),
TEST(testLine3b),
TEST(testLine3a),
TEST(testLine3),
TEST(testLine2),
TEST(testLine1),
};
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
static const size_t testCount = SK_ARRAY_COUNT(tests);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
static TestDesc subTests[] = {
looking for fuzz failure These tests were isolated from their respective minimized test cases. The tests work fine and pass path ops internal validation; hopefully some more intensive x-san or valgrind test will help isolate the bug. Sheriff, please revert if it fails and I don't get to it first. TBR=reed@google.com,halcanary@google.com BUG=535127,535151 Review URL: https://codereview.chromium.org/1359263003
2015-09-24 01:07:51 +00:00
TEST(fuzz994s_3414),
TEST(fuzz994s_11),
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
};
path ops work in progress standardize tests use SK_ARRAY_COUNT everywhere debug why x87 differs from SIMD 64 various platform specific fixes git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-15 19:13:59 +00:00
static const size_t subTestCount = SK_ARRAY_COUNT(subTests);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static void (*firstSubTest)(skiatest::Reporter* , const char* filename) = 0;
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
static bool runSubTests = false;
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
static bool runSubTestsFirst = false;
turn off debugging printfs fix pathops issues 1417, 1418 be more rigorous about pulling intersections of lines to end points rewrite cubic/line and quad/line intersections to share style BUG= Review URL: https://codereview.chromium.org/19543005 git-svn-id: http://skia.googlecode.com/svn/trunk@10270 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-07-23 15:27:41 +00:00
static bool runReverse = false;
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
Get rid of DEFINE_TESTCLASS_SHORT() macro. Instead tests should be written using DEF_TEST() macro, which is much nicer and simplifies the process of setting up an unit test. BUG=None TEST=skpskgr_test, pathops_unittest R=mtklein@google.com Review URL: https://codereview.chromium.org/117863005 git-svn-id: http://skia.googlecode.com/svn/trunk@12870 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-01-02 21:45:03 +00:00
DEF_TEST(PathOpsSimplify, reporter) {
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
if (runSubTests && runSubTestsFirst) {
Style Change: NULL->nullptr DOCS_PREVIEW= https://skia.org/?cl=1316233002 Review URL: https://codereview.chromium.org/1316233002
2015-08-27 14:41:13 +00:00
RunTestSet(reporter, subTests, subTestCount, firstSubTest, nullptr, stopTest, runReverse);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
cumulative pathops patch Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer. Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering. Unify t and point values so that small intervals can be eliminated in one pass. Break cubics up front to eliminate loops and cusps. Make the Simplify and Op code more regular and eliminate arbitrary differences. Add a builder that takes an array of paths and operators. Delete unused code. BUG=skia:3588 R=reed@google.com Review URL: https://codereview.chromium.org/1037573004
2015-03-26 14:52:43 +00:00
RunTestSet(reporter, tests, testCount, firstTest, skipTest, stopTest, runReverse);
Mike R: please sanity check SkPostConfig.h Mike K: please sanity check Test.cpp and skia_test.cpp Feel free to look at the rest, but I don't expect any in depth review of path ops innards. Path Ops first iteration used QuickSort to order segments radiating from an intersection to compute the winding rule. This revision uses a circular sort instead. Breaking out the circular sort into its own long-lived structure (SkOpAngle) allows doing less work and provides a home for caching additional sorting data. The circle sort is more stable than the former sort, has a robust ordering and fewer exceptions. It finds unsortable ordering less often. It is less reliant on the initial curve tangent, using convex hulls instead whenever it can. Additional debug validation makes sure that the computed structures are self-consistent. A new visualization tool helps verify that the angle ordering is correct. The 70+M tests pass with this change on Windows, Mac, Linux 32 and Linux 64 in debug and release. R=mtklein@google.com, reed@google.com Author: caryclark@google.com Review URL: https://codereview.chromium.org/131103009 git-svn-id: http://skia.googlecode.com/svn/trunk@14183 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-04-14 17:08:59 +00:00
if (runSubTests && !runSubTestsFirst) {
Style Change: NULL->nullptr DOCS_PREVIEW= https://skia.org/?cl=1316233002 Review URL: https://codereview.chromium.org/1316233002
2015-08-27 14:41:13 +00:00
RunTestSet(reporter, subTests, subTestCount, firstSubTest, nullptr, stopTest, runReverse);
Add implementation of path ops This CL depends on https://codereview.chromium.org/12880016/ "Add intersections for path ops" Given a path, iterate through its contour, and construct an array of segments containing its curves. Intersect each curve with every other curve, and for cubics, with itself. Given the set of intersections, find one with the smallest y and sort the curves eminating from the intersection. Assign each curve a winding value. Operate on the curves, keeping and discarding them according to the current operation and the sum of the winding values. Assemble the kept curves into an output path. Review URL: https://codereview.chromium.org/13094010 git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
2013-04-08 11:50:46 +00:00
}
}
Reference in New Issue Copy Permalink