skia2/third_party/gif/SkGifImageReader.cpp

1025 lines
40 KiB
C++
Raw Normal View History

Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
/* -*- Mode: C; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* ***** BEGIN LICENSE BLOCK *****
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
*
* The contents of this file are subject to the Mozilla Public License Version
* 1.1 (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
* http://www.mozilla.org/MPL/
*
* Software distributed under the License is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
* for the specific language governing rights and limitations under the
* License.
*
* The Original Code is mozilla.org code.
*
* The Initial Developer of the Original Code is
* Netscape Communications Corporation.
* Portions created by the Initial Developer are Copyright (C) 1998
* the Initial Developer. All Rights Reserved.
*
* Contributor(s):
* Chris Saari <saari@netscape.com>
* Apple Computer
*
* Alternatively, the contents of this file may be used under the terms of
* either the GNU General Public License Version 2 or later (the "GPL"), or
* the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
* in which case the provisions of the GPL or the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of either the GPL or the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the GPL or the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of any one of the MPL, the GPL or the LGPL.
*
* ***** END LICENSE BLOCK ***** */
/*
The Graphics Interchange Format(c) is the copyright property of CompuServe
Incorporated. Only CompuServe Incorporated is authorized to define, redefine,
enhance, alter, modify or change in any way the definition of the format.
CompuServe Incorporated hereby grants a limited, non-exclusive, royalty-free
license for the use of the Graphics Interchange Format(sm) in computer
software; computer software utilizing GIF(sm) must acknowledge ownership of the
Graphics Interchange Format and its Service Mark by CompuServe Incorporated, in
User and Technical Documentation. Computer software utilizing GIF, which is
distributed or may be distributed without User or Technical Documentation must
display to the screen or printer a message acknowledging ownership of the
Graphics Interchange Format and the Service Mark by CompuServe Incorporated; in
this case, the acknowledgement may be displayed in an opening screen or leading
banner, or a closing screen or trailing banner. A message such as the following
may be used:
"The Graphics Interchange Format(c) is the Copyright property of
CompuServe Incorporated. GIF(sm) is a Service Mark property of
CompuServe Incorporated."
For further information, please contact :
CompuServe Incorporated
Graphics Technology Department
5000 Arlington Center Boulevard
Columbus, Ohio 43220
U. S. A.
CompuServe Incorporated maintains a mailing list with all those individuals and
organizations who wish to receive copies of this document when it is corrected
or revised. This service is offered free of charge; please provide us with your
mailing address.
*/
#include "SkGifImageReader.h"
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
#include "SkColorPriv.h"
#include "SkGifCodec.h"
#include <algorithm>
#include <string.h>
// GETN(n, s) requests at least 'n' bytes available from 'q', at start of state 's'.
//
// Note, the hold will never need to be bigger than 256 bytes to gather up in the hold,
// as each GIF block (except colormaps) can never be bigger than 256 bytes.
// Colormaps are directly copied in the resp. global_colormap or dynamically allocated local_colormap.
// So a fixed buffer in SkGifImageReader is good enough.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// This buffer is only needed to copy left-over data from one GifWrite call to the next
#define GETN(n, s) \
do { \
m_bytesToConsume = (n); \
m_state = (s); \
} while (0)
// Get a 16-bit value stored in little-endian format.
#define GETINT16(p) ((p)[1]<<8|(p)[0])
namespace {
bool is_palette_index_valid(int transparentIndex) {
// -1 is a signal that there is no transparent index.
// Otherwise, it is encoded in 8 bits, and all 256 values are considered
// valid since a GIF may use an index outside of the palette to be
// transparent.
return transparentIndex >= 0;
}
} // anonymous namespace
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Send the data to the display front-end.
void SkGIFLZWContext::outputRow(const unsigned char* rowBegin)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
int drowStart = irow;
int drowEnd = irow;
// Haeberli-inspired hack for interlaced GIFs: Replicate lines while
// displaying to diminish the "venetian-blind" effect as the image is
// loaded. Adjust pixel vertical positions to avoid the appearance of the
// image crawling up the screen as successive passes are drawn.
if (m_frameContext->progressiveDisplay() && m_frameContext->interlaced() && ipass < 4) {
unsigned rowDup = 0;
unsigned rowShift = 0;
switch (ipass) {
case 1:
rowDup = 7;
rowShift = 3;
break;
case 2:
rowDup = 3;
rowShift = 1;
break;
case 3:
rowDup = 1;
rowShift = 0;
break;
default:
break;
}
drowStart -= rowShift;
drowEnd = drowStart + rowDup;
// Extend if bottom edge isn't covered because of the shift upward.
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if ((unsigned)((m_frameContext->height() - 1) - drowEnd) <= rowShift)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
drowEnd = m_frameContext->height() - 1;
// Clamp first and last rows to upper and lower edge of image.
if (drowStart < 0)
drowStart = 0;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (drowEnd >= m_frameContext->height())
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
drowEnd = m_frameContext->height() - 1;
}
// Protect against too much image data.
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (drowStart >= m_frameContext->height())
return;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// CALLBACK: Let the client know we have decoded a row.
const bool writeTransparentPixels = (SkCodec::kNone == m_frameContext->getRequiredFrame());
m_client->haveDecodedRow(m_frameContext->frameId(), rowBegin,
drowStart, drowEnd - drowStart + 1, writeTransparentPixels);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (!m_frameContext->interlaced())
irow++;
else {
do {
switch (ipass) {
case 1:
irow += 8;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (irow >= (unsigned) m_frameContext->height()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
ipass++;
irow = 4;
}
break;
case 2:
irow += 8;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (irow >= (unsigned) m_frameContext->height()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
ipass++;
irow = 2;
}
break;
case 3:
irow += 4;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (irow >= (unsigned) m_frameContext->height()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
ipass++;
irow = 1;
}
break;
case 4:
irow += 2;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (irow >= (unsigned) m_frameContext->height()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
ipass++;
irow = 0;
}
break;
default:
break;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
} while (irow > (unsigned) (m_frameContext->height() - 1));
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
}
// Perform Lempel-Ziv-Welch decoding.
// Returns true if decoding was successful. In this case the block will have been completely consumed and/or rowsRemaining will be 0.
// Otherwise, decoding failed; returns false in this case, which will always cause the SkGifImageReader to set the "decode failed" flag.
bool SkGIFLZWContext::doLZW(const unsigned char* block, size_t bytesInBlock)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
const int width = m_frameContext->width();
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (rowIter == rowBuffer.end())
return true;
for (const unsigned char* ch = block; bytesInBlock-- > 0; ch++) {
// Feed the next byte into the decoder's 32-bit input buffer.
datum += ((int) *ch) << bits;
bits += 8;
// Check for underflow of decoder's 32-bit input buffer.
while (bits >= codesize) {
// Get the leading variable-length symbol from the data stream.
int code = datum & codemask;
datum >>= codesize;
bits -= codesize;
// Reset the dictionary to its original state, if requested.
if (code == clearCode) {
codesize = m_frameContext->dataSize() + 1;
codemask = (1 << codesize) - 1;
avail = clearCode + 2;
oldcode = -1;
continue;
}
// Check for explicit end-of-stream code.
if (code == (clearCode + 1)) {
// end-of-stream should only appear after all image data.
if (!rowsRemaining)
return true;
return false;
}
const int tempCode = code;
unsigned short codeLength = 0;
if (code < avail) {
// This is a pre-existing code, so we already know what it
// encodes.
codeLength = suffixLength[code];
rowIter += codeLength;
} else if (code == avail && oldcode != -1) {
// This is a new code just being added to the dictionary.
// It must encode the contents of the previous code, plus
// the first character of the previous code again.
codeLength = suffixLength[oldcode] + 1;
rowIter += codeLength;
*--rowIter = firstchar;
code = oldcode;
} else {
// This is an invalid code. The dictionary is just initialized
// and the code is incomplete. We don't know how to handle
// this case.
return false;
}
while (code >= clearCode) {
*--rowIter = suffix[code];
code = prefix[code];
}
*--rowIter = firstchar = suffix[code];
// Define a new codeword in the dictionary as long as we've read
// more than one value from the stream.
if (avail < SK_MAX_DICTIONARY_ENTRIES && oldcode != -1) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
prefix[avail] = oldcode;
suffix[avail] = firstchar;
suffixLength[avail] = suffixLength[oldcode] + 1;
++avail;
// If we've used up all the codewords of a given length
// increase the length of codewords by one bit, but don't
// exceed the specified maximum codeword size.
if (!(avail & codemask) && avail < SK_MAX_DICTIONARY_ENTRIES) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
++codesize;
codemask += avail;
}
}
oldcode = tempCode;
rowIter += codeLength;
// Output as many rows as possible.
unsigned char* rowBegin = rowBuffer.begin();
for (; rowBegin + width <= rowIter; rowBegin += width) {
outputRow(rowBegin);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
rowsRemaining--;
if (!rowsRemaining)
return true;
}
if (rowBegin != rowBuffer.begin()) {
// Move the remaining bytes to the beginning of the buffer.
const size_t bytesToCopy = rowIter - rowBegin;
memcpy(&rowBuffer.front(), rowBegin, bytesToCopy);
rowIter = rowBuffer.begin() + bytesToCopy;
}
}
}
return true;
}
sk_sp<SkColorTable> SkGIFColorMap::buildTable(SkStreamBuffer* streamBuffer, SkColorType colorType,
int transparentPixel) const
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
if (!m_isDefined)
return nullptr;
const PackColorProc proc = choose_pack_color_proc(false, colorType);
if (m_table && proc == m_packColorProc && m_transPixel == transparentPixel) {
SkASSERT(transparentPixel == kNotFound || transparentPixel > m_table->count()
|| m_table->operator[](transparentPixel) == SK_ColorTRANSPARENT);
// This SkColorTable has already been built with the same transparent color and
// packing proc. Reuse it.
return m_table;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
m_packColorProc = proc;
m_transPixel = transparentPixel;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const size_t bytes = m_colors * SK_BYTES_PER_COLORMAP_ENTRY;
sk_sp<SkData> rawData(streamBuffer->getDataAtPosition(m_position, bytes));
if (!rawData) {
return nullptr;
}
SkASSERT(m_colors <= SK_MAX_COLORS);
const uint8_t* srcColormap = rawData->bytes();
SkPMColor colorStorage[SK_MAX_COLORS];
for (int i = 0; i < m_colors; i++) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (i == transparentPixel) {
colorStorage[i] = SK_ColorTRANSPARENT;
} else {
colorStorage[i] = proc(255, srcColormap[0], srcColormap[1], srcColormap[2]);
}
srcColormap += SK_BYTES_PER_COLORMAP_ENTRY;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
for (int i = m_colors; i < SK_MAX_COLORS; i++) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
colorStorage[i] = SK_ColorTRANSPARENT;
}
m_table = sk_sp<SkColorTable>(new SkColorTable(colorStorage, SK_MAX_COLORS));
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
return m_table;
}
sk_sp<SkColorTable> SkGifImageReader::getColorTable(SkColorType colorType, int index) {
if (index < 0 || index >= m_frames.count()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
return nullptr;
}
const SkGIFFrameContext* frameContext = m_frames[index].get();
const SkGIFColorMap& localColorMap = frameContext->localColorMap();
const int transPix = frameContext->transparentPixel();
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (localColorMap.isDefined()) {
return localColorMap.buildTable(&m_streamBuffer, colorType, transPix);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
if (m_globalColorMap.isDefined()) {
return m_globalColorMap.buildTable(&m_streamBuffer, colorType, transPix);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
return nullptr;
}
// Perform decoding for this frame. frameComplete will be true if the entire frame is decoded.
// Returns false if a decoding error occurred. This is a fatal error and causes the SkGifImageReader to set the "decode failed" flag.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Otherwise, either not enough data is available to decode further than before, or the new data has been decoded successfully; returns true in this case.
bool SkGIFFrameContext::decode(SkStreamBuffer* streamBuffer, SkGifCodec* client,
bool* frameComplete)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
*frameComplete = false;
if (!m_lzwContext) {
// Wait for more data to properly initialize SkGIFLZWContext.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (!isDataSizeDefined() || !isHeaderDefined())
return true;
m_lzwContext.reset(new SkGIFLZWContext(client, this));
if (!m_lzwContext->prepareToDecode()) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
m_lzwContext.reset();
return false;
}
m_currentLzwBlock = 0;
}
// Some bad GIFs have extra blocks beyond the last row, which we don't want to decode.
while (m_currentLzwBlock < m_lzwBlocks.count() && m_lzwContext->hasRemainingRows()) {
const auto& block = m_lzwBlocks[m_currentLzwBlock];
const size_t len = block.blockSize;
sk_sp<SkData> data(streamBuffer->getDataAtPosition(block.blockPosition, len));
if (!data) {
return false;
}
if (!m_lzwContext->doLZW(reinterpret_cast<const unsigned char*>(data->data()), len)) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
return false;
}
++m_currentLzwBlock;
}
// If this frame is data complete then the previous loop must have completely decoded all LZW blocks.
// There will be no more decoding for this frame so it's time to cleanup.
if (isComplete()) {
*frameComplete = true;
m_lzwContext.reset();
}
return true;
}
// Decode a frame.
// This method uses SkGIFFrameContext:decode() to decode the frame; decoding error is reported to client as a critical failure.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Return true if decoding has progressed. Return false if an error has occurred.
bool SkGifImageReader::decode(int frameIndex, bool* frameComplete)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
SkGIFFrameContext* currentFrame = m_frames[frameIndex].get();
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
return currentFrame->decode(&m_streamBuffer, m_client, frameComplete);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
// Parse incoming GIF data stream into internal data structures.
SkCodec::Result SkGifImageReader::parse(SkGifImageReader::SkGIFParseQuery query)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
if (m_parseCompleted) {
return SkCodec::kSuccess;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
if (SkGIFLoopCountQuery == query && m_loopCount != cLoopCountNotSeen) {
// Loop count has already been parsed.
return SkCodec::kSuccess;
}
// SkGIFSizeQuery and SkGIFFrameCountQuery are negative, so this is only meaningful when >= 0.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const int lastFrameToParse = (int) query;
if (lastFrameToParse >= 0 && m_frames.count() > lastFrameToParse
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
&& m_frames[lastFrameToParse]->isComplete()) {
// We have already parsed this frame.
return SkCodec::kSuccess;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
while (true) {
if (!m_streamBuffer.buffer(m_bytesToConsume)) {
// The stream does not yet have enough data.
return SkCodec::kIncompleteInput;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
switch (m_state) {
case SkGIFLZW: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
SkASSERT(!m_frames.empty());
auto* frame = m_frames.back().get();
frame->addLzwBlock(m_streamBuffer.markPosition(), m_bytesToConsume);
GETN(1, SkGIFSubBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFLZWStart: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
SkASSERT(!m_frames.empty());
auto* currentFrame = m_frames.back().get();
currentFrame->setDataSize(this->getOneByte());
GETN(1, SkGIFSubBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFType: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const char* currentComponent = m_streamBuffer.get();
// All GIF files begin with "GIF87a" or "GIF89a".
if (!memcmp(currentComponent, "GIF89a", 6))
m_version = 89;
else if (!memcmp(currentComponent, "GIF87a", 6))
m_version = 87;
else {
// This prevents attempting to continue reading this invalid stream.
GETN(0, SkGIFDone);
return SkCodec::kInvalidInput;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
GETN(7, SkGIFGlobalHeader);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFGlobalHeader: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
// This is the height and width of the "screen" or frame into which
// images are rendered. The individual images can be smaller than
// the screen size and located with an origin anywhere within the
// screen.
// Note that we don't inform the client of the size yet, as it might
// change after we read the first frame's image header.
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
fScreenWidth = GETINT16(currentComponent);
fScreenHeight = GETINT16(currentComponent + 2);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const int globalColorMapColors = 2 << (currentComponent[4] & 0x07);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if ((currentComponent[4] & 0x80) && globalColorMapColors > 0) { /* global map */
m_globalColorMap.setNumColors(globalColorMapColors);
GETN(SK_BYTES_PER_COLORMAP_ENTRY * globalColorMapColors, SkGIFGlobalColormap);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFGlobalColormap: {
m_globalColorMap.setTablePosition(m_streamBuffer.markPosition());
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFImageStart: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const char currentComponent = m_streamBuffer.get()[0];
if (currentComponent == '!') { // extension.
GETN(2, SkGIFExtension);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
if (currentComponent == ',') { // image separator.
GETN(9, SkGIFImageHeader);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
// If we get anything other than ',' (image separator), '!'
// (extension), or ';' (trailer), there is extraneous data
// between blocks. The GIF87a spec tells us to keep reading
// until we find an image separator, but GIF89a says such
// a file is corrupt. We follow Mozilla's implementation and
// proceed as if the file were correctly terminated, so the
// GIF will display.
GETN(0, SkGIFDone);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFExtension: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
size_t bytesInBlock = currentComponent[1];
SkGIFState exceptionState = SkGIFSkipBlock;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
switch (*currentComponent) {
case 0xf9:
// The GIF spec mandates that the GIFControlExtension header block length is 4 bytes,
exceptionState = SkGIFControlExtension;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// and the parser for this block reads 4 bytes, so we must enforce that the buffer
// contains at least this many bytes. If the GIF specifies a different length, we
// allow that, so long as it's larger; the additional data will simply be ignored.
bytesInBlock = std::max(bytesInBlock, static_cast<size_t>(4));
break;
// The GIF spec also specifies the lengths of the following two extensions' headers
// (as 12 and 11 bytes, respectively). Because we ignore the plain text extension entirely
// and sanity-check the actual length of the application extension header before reading it,
// we allow GIFs to deviate from these values in either direction. This is important for
// real-world compatibility, as GIFs in the wild exist with application extension headers
// that are both shorter and longer than 11 bytes.
case 0x01:
// ignoring plain text extension
break;
case 0xff:
exceptionState = SkGIFApplicationExtension;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
case 0xfe:
exceptionState = SkGIFConsumeComment;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
if (bytesInBlock)
GETN(bytesInBlock, exceptionState);
else
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFConsumeBlock: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char currentComponent = this->getOneByte();
if (!currentComponent)
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
else
GETN(currentComponent, SkGIFSkipBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFSkipBlock: {
GETN(1, SkGIFConsumeBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFControlExtension: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
addFrameIfNecessary();
SkGIFFrameContext* currentFrame = m_frames.back().get();
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (*currentComponent & 0x1)
currentFrame->setTransparentPixel(currentComponent[3]);
// We ignore the "user input" bit.
// NOTE: This relies on the values in the FrameDisposalMethod enum
// matching those in the GIF spec!
int rawDisposalMethod = ((*currentComponent) >> 2) & 0x7;
switch (rawDisposalMethod) {
case 1:
case 2:
case 3:
currentFrame->setDisposalMethod((SkCodecAnimation::DisposalMethod) rawDisposalMethod);
break;
case 4:
// Some specs say that disposal method 3 is "overwrite previous", others that setting
// the third bit of the field (i.e. method 4) is. We map both to the same value.
currentFrame->setDisposalMethod(SkCodecAnimation::DisposalMethod::kRestorePrevious);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
default:
// Other values use the default.
currentFrame->setDisposalMethod(SkCodecAnimation::DisposalMethod::kKeep);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
currentFrame->setDuration(GETINT16(currentComponent + 1) * 10);
GETN(1, SkGIFConsumeBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFCommentExtension: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char currentComponent = this->getOneByte();
if (currentComponent)
GETN(currentComponent, SkGIFConsumeComment);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
else
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFConsumeComment: {
GETN(1, SkGIFCommentExtension);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFApplicationExtension: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Check for netscape application extension.
if (m_bytesToConsume == 11) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
if (!memcmp(currentComponent, "NETSCAPE2.0", 11) || !memcmp(currentComponent, "ANIMEXTS1.0", 11))
GETN(1, SkGIFNetscapeExtensionBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
if (m_state != SkGIFNetscapeExtensionBlock)
GETN(1, SkGIFConsumeBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
// Netscape-specific GIF extension: animation looping.
case SkGIFNetscapeExtensionBlock: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const int currentComponent = this->getOneByte();
// SkGIFConsumeNetscapeExtension always reads 3 bytes from the stream; we should at least wait for this amount.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (currentComponent)
GETN(std::max(3, currentComponent), SkGIFConsumeNetscapeExtension);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
else
GETN(1, SkGIFImageStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
// Parse netscape-specific application extensions
case SkGIFConsumeNetscapeExtension: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
int netscapeExtension = currentComponent[0] & 7;
// Loop entire animation specified # of times. Only read the loop count during the first iteration.
if (netscapeExtension == 1) {
m_loopCount = GETINT16(currentComponent + 1);
// Zero loop count is infinite animation loop request.
if (!m_loopCount)
m_loopCount = SkCodec::kRepetitionCountInfinite;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
GETN(1, SkGIFNetscapeExtensionBlock);
if (SkGIFLoopCountQuery == query) {
m_streamBuffer.flush();
return SkCodec::kSuccess;
}
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
} else if (netscapeExtension == 2) {
// Wait for specified # of bytes to enter buffer.
// Don't do this, this extension doesn't exist (isn't used at all)
// and doesn't do anything, as our streaming/buffering takes care of it all...
// See: http://semmix.pl/color/exgraf/eeg24.htm
GETN(1, SkGIFNetscapeExtensionBlock);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
} else {
// 0,3-7 are yet to be defined netscape extension codes
// This prevents attempting to continue reading this invalid stream.
GETN(0, SkGIFDone);
return SkCodec::kInvalidInput;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
break;
}
case SkGIFImageHeader: {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
int height, width, xOffset, yOffset;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const unsigned char* currentComponent =
reinterpret_cast<const unsigned char*>(m_streamBuffer.get());
/* Get image offsets, with respect to the screen origin */
xOffset = GETINT16(currentComponent);
yOffset = GETINT16(currentComponent + 2);
/* Get image width and height. */
width = GETINT16(currentComponent + 4);
height = GETINT16(currentComponent + 6);
// Some GIF files have frames that don't fit in the specified
// overall image size. For the first frame, we can simply enlarge
// the image size to allow the frame to be visible. We can't do
// this on subsequent frames because the rest of the decoding
// infrastructure assumes the image size won't change as we
// continue decoding, so any subsequent frames that are even
// larger will be cropped.
// Luckily, handling just the first frame is sufficient to deal
// with most cases, e.g. ones where the image size is erroneously
// set to zero, since usually the first frame completely fills
// the image.
if (currentFrameIsFirstFrame()) {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
fScreenHeight = std::max(fScreenHeight, yOffset + height);
fScreenWidth = std::max(fScreenWidth, xOffset + width);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
// NOTE: Chromium placed this block after setHeaderDefined, down
// below we returned true when asked for the size. So Chromium
// created an image which would fail. Is this the correct behavior?
// We choose to return false early, so we will not create an
// SkCodec.
// Work around more broken GIF files that have zero image width or
// height.
if (!height || !width) {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
height = fScreenHeight;
width = fScreenWidth;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (!height || !width) {
// This prevents attempting to continue reading this invalid stream.
GETN(0, SkGIFDone);
return SkCodec::kInvalidInput;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
}
const bool isLocalColormapDefined = SkToBool(currentComponent[8] & 0x80);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// The three low-order bits of currentComponent[8] specify the bits per pixel.
const int numColors = 2 << (currentComponent[8] & 0x7);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
if (currentFrameIsFirstFrame()) {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
const int transPix = m_frames.empty() ? SkGIFColorMap::kNotFound
: m_frames[0]->transparentPixel();
if (is_palette_index_valid(transPix)) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
m_firstFrameHasAlpha = true;
} else {
const bool frameIsSubset = xOffset > 0 || yOffset > 0
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
|| width < fScreenWidth
|| height < fScreenHeight;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
m_firstFrameHasAlpha = frameIsSubset;
}
}
Do not create SkGifCodec if true size is not known If there is enough data in the stream to read the reported canvas size, but not enough to read the first image's header, we do not know the true canvas size, since we may expand it to fit the first frame. In that case, return nullptr from NewFromStream. Add a test. SkGifCodec.cpp: Correct a comment - parse returns false if there is a fatal error. parse() returning true does not guarantee that the size was found. Instead of checking the width and height, check to see whether the first frame exists and has its header defined. If not, we do not yet know the true canvas size. Assert that the canvas size is non-zero, which is a fatal error from parse. SkGifImageReader.cpp: Move the code to set the header defined before the SkGIFSizeQuery exit condition. This allows SkGifCodec to check the first frame's header to determine whether the size is known. GifTest.cpp: Add a test which truncates the file just before the image header (and after the global header). Prior to the other changes, this would create an SkCodec. For an image that needs its canvas size expanded, the SkCodec would have an incorrect size. CodecPartialTest.cpp: randPixels.gif now needs more than half of its data to create an SkCodec, so set a minimum for test_partial. Change-Id: I40482f524128b2f1fe59b8f27dd64c7cbe793079 Reviewed-on: https://skia-review.googlesource.com/5701 Reviewed-by: Derek Sollenberger <djsollen@google.com> Commit-Queue: Leon Scroggins <scroggo@google.com>
2016-12-08 16:54:04 +00:00
addFrameIfNecessary();
SkGIFFrameContext* currentFrame = m_frames.back().get();
currentFrame->setHeaderDefined();
if (query == SkGIFSizeQuery) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// The decoder needs to stop, so we return here, before
// flushing the buffer. Next time through, we'll be in the same
// state, requiring the same amount in the buffer.
return SkCodec::kSuccess;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
currentFrame->setXYWH(xOffset, yOffset, width, height);
currentFrame->setInterlaced(SkToBool(currentComponent[8] & 0x40));
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Overlaying interlaced, transparent GIFs over
// existing image data using the Haeberli display hack
// requires saving the underlying image in order to
// avoid jaggies at the transparency edges. We are
// unprepared to deal with that, so don't display such
// images progressively. Which means only the first
// frame can be progressively displayed.
// FIXME: It is possible that a non-transparent frame
// can be interlaced and progressively displayed.
currentFrame->setProgressiveDisplay(currentFrameIsFirstFrame());
if (isLocalColormapDefined) {
currentFrame->localColorMap().setNumColors(numColors);
GETN(SK_BYTES_PER_COLORMAP_ENTRY * numColors, SkGIFImageColormap);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
setAlphaAndRequiredFrame(currentFrame);
GETN(1, SkGIFLZWStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFImageColormap: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
SkASSERT(!m_frames.empty());
auto* currentFrame = m_frames.back().get();
auto& cmap = currentFrame->localColorMap();
cmap.setTablePosition(m_streamBuffer.markPosition());
setAlphaAndRequiredFrame(currentFrame);
GETN(1, SkGIFLZWStart);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
}
case SkGIFSubBlock: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
const size_t bytesInBlock = this->getOneByte();
if (bytesInBlock)
GETN(bytesInBlock, SkGIFLZW);
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
else {
// Finished parsing one frame; Process next frame.
SkASSERT(!m_frames.empty());
// Note that some broken GIF files do not have enough LZW blocks to fully
// decode all rows but we treat it as frame complete.
m_frames.back()->setComplete();
GETN(1, SkGIFImageStart);
if (lastFrameToParse >= 0 && m_frames.count() > lastFrameToParse) {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
m_streamBuffer.flush();
return SkCodec::kSuccess;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
}
break;
}
case SkGIFDone: {
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
m_parseCompleted = true;
return SkCodec::kSuccess;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
default:
// We shouldn't ever get here.
// This prevents attempting to continue reading this invalid stream.
GETN(0, SkGIFDone);
return SkCodec::kInvalidInput;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
break;
} // switch
m_streamBuffer.flush();
}
}
void SkGifImageReader::addFrameIfNecessary()
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
if (m_frames.empty() || m_frames.back()->isComplete()) {
const int i = m_frames.count();
m_frames.emplace_back(new SkGIFFrameContext(i));
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
}
}
static SkIRect frame_rect_on_screen(SkIRect frameRect,
const SkIRect& screenRect) {
if (!frameRect.intersect(screenRect)) {
return SkIRect::MakeEmpty();
}
return frameRect;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
static bool independent(const SkFrame& frame) {
return frame.getRequiredFrame() == SkCodec::kNone;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
static bool restore_bg(const SkFrame& frame) {
return frame.getDisposalMethod() == SkCodecAnimation::DisposalMethod::kRestoreBGColor;
}
SkEncodedInfo::Alpha SkGIFFrameContext::onReportedAlpha() const {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
// Note: We could correct these after decoding - i.e. some frames may turn out to be
// independent and opaque if they do not use the transparent pixel, but that would require
// checking whether each pixel used the transparent index.
return is_palette_index_valid(this->transparentPixel()) ? SkEncodedInfo::kBinary_Alpha
: SkEncodedInfo::kOpaque_Alpha;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
}
void SkFrameHolder::setAlphaAndRequiredFrame(SkFrame* frame) {
const bool reportsAlpha = frame->reportedAlpha() != SkEncodedInfo::kOpaque_Alpha;
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
const auto screenRect = SkIRect::MakeWH(fScreenWidth, fScreenHeight);
const auto frameRect = frame_rect_on_screen(frame->frameRect(), screenRect);
const int i = frame->frameId();
if (0 == i) {
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
frame->setHasAlpha(reportsAlpha || frameRect != screenRect);
frame->setRequiredFrame(SkCodec::kNone);
return;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
const bool blendWithPrevFrame = frame->getBlend() == SkCodecAnimation::Blend::kPriorFrame;
if ((!reportsAlpha || !blendWithPrevFrame) && frameRect == screenRect) {
frame->setHasAlpha(reportsAlpha);
frame->setRequiredFrame(SkCodec::kNone);
return;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
const SkFrame* prevFrame = this->getFrame(i-1);
while (prevFrame->getDisposalMethod() == SkCodecAnimation::DisposalMethod::kRestorePrevious) {
const int prevId = prevFrame->frameId();
if (0 == prevId) {
frame->setHasAlpha(true);
frame->setRequiredFrame(SkCodec::kNone);
return;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
prevFrame = this->getFrame(prevId - 1);
}
const bool clearPrevFrame = restore_bg(*prevFrame);
auto prevFrameRect = frame_rect_on_screen(prevFrame->frameRect(), screenRect);
if (clearPrevFrame) {
if (prevFrameRect == screenRect || independent(*prevFrame)) {
frame->setHasAlpha(true);
frame->setRequiredFrame(SkCodec::kNone);
return;
}
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
if (reportsAlpha && blendWithPrevFrame) {
// Note: We could be more aggressive here. If prevFrame clears
// to background color and covers its required frame (and that
// frame is independent), prevFrame could be marked independent.
// Would this extra complexity be worth it?
frame->setRequiredFrame(prevFrame->frameId());
frame->setHasAlpha(prevFrame->hasAlpha() || clearPrevFrame);
return;
}
while (frameRect.contains(prevFrameRect)) {
const int prevRequiredFrame = prevFrame->getRequiredFrame();
if (prevRequiredFrame == SkCodec::kNone) {
frame->setRequiredFrame(SkCodec::kNone);
frame->setHasAlpha(true);
return;
}
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
prevFrame = this->getFrame(prevRequiredFrame);
prevFrameRect = frame_rect_on_screen(prevFrame->frameRect(), screenRect);
}
if (restore_bg(*prevFrame)) {
frame->setHasAlpha(true);
if (prevFrameRect == screenRect || independent(*prevFrame)) {
frame->setRequiredFrame(SkCodec::kNone);
} else {
// Note: As above, frame could still be independent, e.g. if
// prevFrame covers its required frame and that frame is
// independent.
frame->setRequiredFrame(prevFrame->frameId());
}
return;
}
SkASSERT(prevFrame->getDisposalMethod() == SkCodecAnimation::DisposalMethod::kKeep);
frame->setRequiredFrame(prevFrame->frameId());
Add animation support to SkWebpCodec TBR=reed@google.com (No change to the public API, but changed a header file) SkWebpCodec: - Implement onGetFrameCount, onGetFrameInfo, and onGetRepetitionCount - Respect the alpha reported by libwebp. Although the spec states that it is only a hint, the libwebp encoder uses it properly. Respecting allows us to draw opaque images faster and decode them to 565. This also matches other SkCodecs (and Chromium). - onGetPixels: - Decode the frame requested, recursively decoding required frame if necessary - When blending with a prior frame, use SkRasterPipeline SkCodec: - Move check for negative index to getFrameInfo - Reset the colorXform if one is not needed SkCodecAnimation: - Add new blend enum, for WebP's (and APNG's) non-blending option SkFrameHolder: - New base classes for frames and the owner of the frames, allowing code sharing between SkWebpCodec and SkGifCodec (particularly for determining whether a frame has alpha and what frame it depends on) - When moving items from SkGIFFrameContext, use Skia conventions (i.e. int instead of unsigned) - Rename "delay time" to "duration", to match e.g. SkFrameInfo:: fDuration SkGifImageReader: - Move pieces to SkFrameHolder, and adapt to changes made in the process - Make setAlphaAndRequiredFrame (now on the base class SkFrameHolder) more general to support webp, and add support for frames that do not blend - Change SkGIFFrameContext from a struct to a class, to match how we use the distinction elsewhere (i.e. struct is a small object with public fields) - Rework hasTransparentPixel (now hasTransparency, since it returns true in some cases where there is not a transparent pixel) to better fit with the modified setAlphaAndRequiredFrame. Also be more consistent when there is no transparent pixel but no color map. - Simplify an if condition that was previously simplified in 2d61e717 but accidentally got reverted in a4db9be6 CodecAnimTest: - Test new animated webp files - Rearrange the test to more cleanly print alpha type mismatches for the first frame resources: - webp-animated.webp - animated webp from Chromium - blendBG.webp - new webp file using bits of webp-animated-semitransparent4.webp from Chromium - tests required frame and alpha when using the non-blending mode - frames have the following properties: - Frame 0: no alpha, fills screen - Frame 1: alpha, fills screen - Frame 2: no alpha, fills screen - Frame 3: alpha, fills screen, blendBG - Frame 4: no alpha, fills screen, blendBG - Frame 5: alpha, blendBG - Frame 6: covers 4, has alpha, blendBG - also used to test decoding to 565 if the new frame data has alpha but blends onto an opaque frame DM.cpp: - Test animated images to non-native 8888 and unpremul DMSrcSink.cpp: - Do not test non-native 8888 decodes to f16 dst - Test unpremul decodes to f16 - Copy a frame of an animated image prior to drawing, since in unpremul mode, the DM code will premultiply first. Bug: skia: 3315 Change-Id: I4e55ae2ee5bc095b37a743bdcfac644be603b980 Reviewed-on: https://skia-review.googlesource.com/16707 Commit-Queue: Mike Reed <reed@google.com> Reviewed-by: Mike Reed <reed@google.com> Reviewed-by: Leon Scroggins <scroggo@google.com> Reviewed-by: Matt Sarett <msarett@google.com>
2017-05-23 13:37:21 +00:00
frame->setHasAlpha(prevFrame->hasAlpha() || (reportsAlpha && !blendWithPrevFrame));
}
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// FIXME: Move this method to close to doLZW().
bool SkGIFLZWContext::prepareToDecode()
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
{
SkASSERT(m_frameContext->isDataSizeDefined() && m_frameContext->isHeaderDefined());
// Since we use a codesize of 1 more than the datasize, we need to ensure
// that our datasize is strictly less than the SK_MAX_DICTIONARY_ENTRY_BITS.
if (m_frameContext->dataSize() >= SK_MAX_DICTIONARY_ENTRY_BITS)
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
return false;
clearCode = 1 << m_frameContext->dataSize();
avail = clearCode + 2;
oldcode = -1;
codesize = m_frameContext->dataSize() + 1;
codemask = (1 << codesize) - 1;
datum = bits = 0;
ipass = m_frameContext->interlaced() ? 1 : 0;
irow = 0;
// We want to know the longest sequence encodable by a dictionary with
// SK_MAX_DICTIONARY_ENTRIES entries. If we ignore the need to encode the base
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// values themselves at the beginning of the dictionary, as well as the need
// for a clear code or a termination code, we could use every entry to
// encode a series of multiple values. If the input value stream looked
// like "AAAAA..." (a long string of just one value), the first dictionary
// entry would encode AA, the next AAA, the next AAAA, and so forth. Thus
// the longest sequence would be SK_MAX_DICTIONARY_ENTRIES + 1 values.
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
//
// However, we have to account for reserved entries. The first |datasize|
// bits are reserved for the base values, and the next two entries are
// reserved for the clear code and termination code. In theory a GIF can
// set the datasize to 0, meaning we have just two reserved entries, making
// the longest sequence (SK_MAX_DICTIONARY_ENTIRES + 1) - 2 values long. Since
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// each value is a byte, this is also the number of bytes in the longest
// encodable sequence.
const size_t maxBytes = SK_MAX_DICTIONARY_ENTRIES - 1;
Add support for multiple frames in SkCodec Add an interface to decode frames beyond the first in SkCodec, and add an implementation for SkGifCodec. Add getFrameData to SkCodec. This method reads ahead in the stream to return a vector containing meta data about each frame in the image. This is not required in order to decode frames beyond the first, but it allows a client to learn extra information: - how long the frame should be displayed - whether a frame should be blended with a prior frame, allowing the client to provide the prior frame to speed up decoding Add a new fields to SkCodec::Options: - fFrameIndex - fHasPriorFrame The API is designed so that SkCodec never caches frames. If a client wants a frame beyond the first, they specify the frame in Options.fFrameIndex. If the client does not have the frame's required frame (the frame that this frame must be blended on top of) cached, they pass false for Options.fHasPriorFrame. Unless the frame is independent, the codec will then recursively decode all frames necessary to decode fFrameIndex. If the client has the required frame cached, they can put it in the dst they pass to the codec, and the codec will only draw fFrameIndex onto it. Replace SkGifCodec's scanline decoding support with progressive decoding, and update the tests accordingly. Implement new APIs in SkGifCodec. Instead of using gif_lib, use GIFImageReader, imported from Chromium (along with its copyright headers) with the following changes: - SkGifCodec is now the client - Replace blink types - Combine GIFColorMap::buildTable and ::getTable into a method that creates and returns an SkColorTable - Input comes from an SkStream, instead of a SegmentReader. Add SkStreamBuffer, which buffers the (potentially partial) stream in order to decode progressively. (FIXME: This requires copying data that previously was read directly from the SegmentReader. Does this hurt performance? If so, can we fix it?) - Remove UMA code - Instead of reporting screen width and height to the client, allow the client to query for it - Fail earlier if the first frame AND screen have size of zero - Compute required previous frame when adding a new one - Move GIFParseQuery from GIFImageDecoder to GIFImageReader - Allow parsing up to a specific frame (to skip parsing the rest of the stream if a client only wants the first frame) - Compute whether the first frame has alpha and supports index 8, to create the SkImageInfo. This happens before reporting that the size has been decoded. Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from Chromium (along with its copyright header), with the following changes: - Add support for sampling - Use the swizzler - Keep track of the rows decoded - Do *not* keep track of whether we've seen alpha Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF scanline decoding. Call onRewind even if there is no stream (SkGifCodec needs to clear its decoded state so it will decode from the beginning). Add a method to SkSwizzler to access the offset into the dst, taking subsetting into account. Add a GM that animates a GIF. Add tests for the new APIs. *** Behavior changes: * Previously, we reported that an image with a subset frame and no transparent index was opaque and used the background index (if present) to fill the background. This is necessary in order to support index 8, but it does not match viewers/browsers I have seen. Examples: - Chromium and Gimp render the background transparent - Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for a single frame image) This CL matches Chromium's behavior and renders the background transparent. This allows us to have consistent behavior across products and simplifies the code (relative to what we would have to do to continue the old behavior on Android). It also means that we will no longer support index 8 for some GIFs. * Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a. This matches Chromium. I suspect that bugs would have been reported if valid GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode in Chromium). *** Future work not included in this CL: * Move some checks out of haveDecodedRow, since they are the same for the entire frame e.g. - intersecting the frameRect with the full image size - whether there is a color table * Change when we write transparent pixels - In some cases, Chromium deemed this unnecessary, but I suspect it is slower than the fallback case. There will continue to be cases where we should *not* write them, but for e.g. the first pass where we have already cleared to transparent (which we may also be able to skip) writing the transparent pixels will not make anything incorrect. * Report color type and alpha type per frame - Depending on alpha values, disposal methods, frame rects, etc, subsequent frames may have different properties than the first. * Skip copies of the encoded data - We copy the encoded data in case the stream is one that cannot be rewound, so we can parse and then decode (possibly not immediately). For some input streams, this is unnecessary. - I was concerned this cause a performance regression, but on average the new code is faster than the old for the images I tested [1]. - It may cause a performance regression for Chromium, though, where we can always move back in the stream, so this should be addressed. Design doc: https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/ [1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/ GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002 Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 16:03:26 +00:00
// Now allocate the output buffer. We decode directly into this buffer
// until we have at least one row worth of data, then call outputRow().
// This means worst case we may have (row width - 1) bytes in the buffer
// and then decode a sequence |maxBytes| long to append.
rowBuffer.reset(m_frameContext->width() - 1 + maxBytes);
rowIter = rowBuffer.begin();
rowsRemaining = m_frameContext->height();
// Clearing the whole suffix table lets us be more tolerant of bad data.
for (int i = 0; i < clearCode; ++i) {
suffix[i] = i;
suffixLength[i] = 1;
}
return true;
}