skia2/include/core/SkDrawLooper.h
commit-bot@chromium.org 8b0e8ac5f5 Refactor read and write buffers.
Eliminates SkFlattenable{Read,Write}Buffer, promoting SkOrdered{Read,Write}Buffer
a step each in the hierarchy.

What used to be this:

SkFlattenableWriteBuffer -> SkOrderedWriteBuffer
SkFlattenableReadBuffer  -> SkOrderedReadBuffer
SkFlattenableReadBuffer  -> SkValidatingReadBuffer

is now

SkWriteBuffer
SkReadBuffer -> SkValidatingReadBuffer

Benefits:
  - code is simpler, names are less wordy
  - the generic SkFlattenableFooBuffer code in SkPaint was incorrect; removed
  - write buffers are completely devirtualized, important for record speed

This refactoring was mostly mechanical.  You aren't going to find anything
interesting in files with less than 10 lines changed.

BUG=skia:
R=reed@google.com, scroggo@google.com, djsollen@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/134163010

git-svn-id: http://skia.googlecode.com/svn/trunk@13245 2bbb7eff-a529-9590-31e7-b0007b416f81
2014-01-30 18:58:24 +00:00

78 lines
2.7 KiB
C++

/*
* Copyright 2011 The Android Open Source Project
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#ifndef SkDrawLooper_DEFINED
#define SkDrawLooper_DEFINED
#include "SkFlattenable.h"
class SkCanvas;
class SkPaint;
struct SkRect;
class SkString;
/** \class SkDrawLooper
Subclasses of SkDrawLooper can be attached to a SkPaint. Where they are,
and something is drawn to a canvas with that paint, the looper subclass will
be called, allowing it to modify the canvas and/or paint for that draw call.
More than that, via the next() method, the looper can modify the draw to be
invoked multiple times (hence the name loop-er), allow it to perform effects
like shadows or frame/fills, that require more than one pass.
*/
class SK_API SkDrawLooper : public SkFlattenable {
public:
SK_DECLARE_INST_COUNT(SkDrawLooper)
/**
* Called right before something is being drawn. This will be followed by
* calls to next() until next() returns false.
*/
virtual void init(SkCanvas*) = 0;
/**
* Called in a loop (after init()). Each time true is returned, the object
* is drawn (possibly with a modified canvas and/or paint). When false is
* finally returned, drawing for the object stops.
*
* On each call, the paint will be in its original state, but the canvas
* will be as it was following the previous call to next() or init().
*
* The implementation must ensure that, when next() finally returns false,
* that the canvas has been restored to the state it was initially, before
* init() was first called.
*/
virtual bool next(SkCanvas*, SkPaint* paint) = 0;
/**
* The fast bounds functions are used to enable the paint to be culled early
* in the drawing pipeline. If a subclass can support this feature it must
* return true for the canComputeFastBounds() function. If that function
* returns false then computeFastBounds behavior is undefined otherwise it
* is expected to have the following behavior. Given the parent paint and
* the parent's bounding rect the subclass must fill in and return the
* storage rect, where the storage rect is with the union of the src rect
* and the looper's bounding rect.
*/
virtual bool canComputeFastBounds(const SkPaint& paint);
virtual void computeFastBounds(const SkPaint& paint,
const SkRect& src, SkRect* dst);
SkDEVCODE(virtual void toString(SkString* str) const = 0;)
SK_DEFINE_FLATTENABLE_TYPE(SkDrawLooper)
protected:
SkDrawLooper() {}
SkDrawLooper(SkReadBuffer& buffer) : INHERITED(buffer) {}
private:
typedef SkFlattenable INHERITED;
};
#endif