skia2/third_party/gif/SkGifImageReader.h

400 lines
14 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: 4; 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 Communicator client 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):
*
* 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 ***** */
#ifndef SkGifImageReader_h
#define 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
// Define ourselves as the clientPtr. Mozilla just hacked their C++ callback class into this old C decoder,
// so we will too.
class SkGifCodec;
#include "SkCodec.h"
#include "SkCodecPriv.h"
#include "SkCodecAnimation.h"
#include "SkColorTable.h"
#include "SkData.h"
#include "SkImageInfo.h"
#include "SkStreamBuffer.h"
#include "../private/SkTArray.h"
#include <memory>
#include <vector>
typedef SkTArray<unsigned char, true> SkGIFRow;
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
#define SK_MAX_DICTIONARY_ENTRY_BITS 12
#define SK_MAX_DICTIONARY_ENTRIES 4096 // 2^SK_MAX_DICTIONARY_ENTRY_BITS
#define SK_MAX_COLORS 256
#define SK_BYTES_PER_COLORMAP_ENTRY 3
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
// List of possible parsing states.
enum SkGIFState {
SkGIFType,
SkGIFGlobalHeader,
SkGIFGlobalColormap,
SkGIFImageStart,
SkGIFImageHeader,
SkGIFImageColormap,
SkGIFImageBody,
SkGIFLZWStart,
SkGIFLZW,
SkGIFSubBlock,
SkGIFExtension,
SkGIFControlExtension,
SkGIFConsumeBlock,
SkGIFSkipBlock,
SkGIFDone,
SkGIFCommentExtension,
SkGIFApplicationExtension,
SkGIFNetscapeExtensionBlock,
SkGIFConsumeNetscapeExtension,
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
};
struct SkGIFFrameContext;
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
// LZW decoder state machine.
class SkGIFLZWContext final : public SkNoncopyable {
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
public:
SkGIFLZWContext(SkGifCodec* client, const SkGIFFrameContext* frameContext)
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(0)
, codemask(0)
, clearCode(0)
, avail(0)
, oldcode(0)
, firstchar(0)
, bits(0)
, datum(0)
, ipass(0)
, irow(0)
, rowsRemaining(0)
, rowIter(0)
, m_client(client)
, m_frameContext(frameContext)
{ }
bool prepareToDecode();
bool outputRow(const unsigned char* rowBegin);
bool doLZW(const unsigned char* block, size_t bytesInBlock);
bool hasRemainingRows() { return SkToBool(rowsRemaining); }
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
private:
// LZW decoding states and output states.
int codesize;
int codemask;
int clearCode; // Codeword used to trigger dictionary reset.
int avail; // Index of next available slot in dictionary.
int oldcode;
unsigned char firstchar;
int bits; // Number of unread bits in "datum".
int datum; // 32-bit input buffer.
int ipass; // Interlace pass; Ranges 1-4 if interlaced.
size_t irow; // Current output row, starting at zero.
size_t rowsRemaining; // Rows remaining to be output.
unsigned short prefix[SK_MAX_DICTIONARY_ENTRIES];
unsigned char suffix[SK_MAX_DICTIONARY_ENTRIES];
unsigned short suffixLength[SK_MAX_DICTIONARY_ENTRIES];
SkGIFRow rowBuffer; // Single scanline temporary buffer.
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
unsigned char* rowIter;
SkGifCodec* const m_client;
const SkGIFFrameContext* m_frameContext;
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
};
class SkGIFColorMap final {
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
public:
SkGIFColorMap()
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_isDefined(false)
, m_colors(0)
, m_packColorProc(nullptr)
{
}
void setNumColors(size_t colors) {
m_colors = colors;
}
size_t numColors() const { return m_colors; }
void setRawData(const char* data, size_t size)
{
// FIXME: Can we avoid this copy?
m_rawData = SkData::MakeWithCopy(data, size);
SkASSERT(m_colors * SK_BYTES_PER_COLORMAP_ENTRY == size);
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_isDefined = true;
}
bool isDefined() const { return m_isDefined; }
// Build RGBA table using the data stream.
sk_sp<SkColorTable> buildTable(SkColorType dstColorType, size_t transparentPixel) const;
private:
bool m_isDefined;
size_t m_colors;
sk_sp<SkData> m_rawData;
mutable PackColorProc m_packColorProc;
mutable sk_sp<SkColorTable> m_table;
};
// LocalFrame output state machine.
struct SkGIFFrameContext : SkNoncopyable {
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
public:
SkGIFFrameContext(int id)
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_frameId(id)
, m_xOffset(0)
, m_yOffset(0)
, m_width(0)
, m_height(0)
, m_transparentPixel(kNotFound)
, m_disposalMethod(SkCodecAnimation::Keep_DisposalMethod)
, m_requiredFrame(SkCodec::kNone)
, m_dataSize(0)
, m_progressiveDisplay(false)
, m_interlaced(false)
, m_delayTime(0)
, m_currentLzwBlock(0)
, m_isComplete(false)
, m_isHeaderDefined(false)
, m_isDataSizeDefined(false)
{
}
static constexpr size_t kNotFound = static_cast<size_t>(-1);
~SkGIFFrameContext()
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
{
}
void addLzwBlock(const void* data, size_t size)
{
m_lzwBlocks.push_back(SkData::MakeWithCopy(data, size));
}
bool decode(SkGifCodec* client, bool* frameDecoded);
int frameId() const { return m_frameId; }
void setRect(unsigned x, unsigned y, unsigned width, unsigned height)
{
m_xOffset = x;
m_yOffset = y;
m_width = width;
m_height = height;
}
SkIRect frameRect() const { return SkIRect::MakeXYWH(m_xOffset, m_yOffset, m_width, m_height); }
unsigned xOffset() const { return m_xOffset; }
unsigned yOffset() const { return m_yOffset; }
unsigned width() const { return m_width; }
unsigned height() const { return m_height; }
size_t transparentPixel() const { return m_transparentPixel; }
void setTransparentPixel(size_t pixel) { m_transparentPixel = pixel; }
SkCodecAnimation::DisposalMethod getDisposalMethod() const { return m_disposalMethod; }
void setDisposalMethod(SkCodecAnimation::DisposalMethod disposalMethod) { m_disposalMethod = disposalMethod; }
size_t getRequiredFrame() const { return m_requiredFrame; }
void setRequiredFrame(size_t req) { m_requiredFrame = req; }
unsigned delayTime() const { return m_delayTime; }
void setDelayTime(unsigned delay) { m_delayTime = delay; }
bool isComplete() const { return m_isComplete; }
void setComplete() { m_isComplete = true; }
bool isHeaderDefined() const { return m_isHeaderDefined; }
void setHeaderDefined() { m_isHeaderDefined = true; }
bool isDataSizeDefined() const { return m_isDataSizeDefined; }
int dataSize() const { return m_dataSize; }
void setDataSize(int size)
{
m_dataSize = size;
m_isDataSizeDefined = true;
}
bool progressiveDisplay() const { return m_progressiveDisplay; }
void setProgressiveDisplay(bool progressiveDisplay) { m_progressiveDisplay = progressiveDisplay; }
bool interlaced() const { return m_interlaced; }
void setInterlaced(bool interlaced) { m_interlaced = interlaced; }
void clearDecodeState() { m_lzwContext.reset(); }
const SkGIFColorMap& localColorMap() const { return m_localColorMap; }
SkGIFColorMap& localColorMap() { return m_localColorMap; }
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
private:
int m_frameId;
unsigned m_xOffset;
unsigned m_yOffset; // With respect to "screen" origin.
unsigned m_width;
unsigned m_height;
size_t m_transparentPixel; // Index of transparent pixel. Value is kNotFound if there is no transparent pixel.
SkCodecAnimation::DisposalMethod m_disposalMethod; // Restore to background, leave in place, etc.
size_t m_requiredFrame;
int m_dataSize;
bool m_progressiveDisplay; // If true, do Haeberli interlace hack.
bool m_interlaced; // True, if scanlines arrive interlaced order.
unsigned m_delayTime; // Display time, in milliseconds, for this image in a multi-image GIF.
std::unique_ptr<SkGIFLZWContext> m_lzwContext;
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
std::vector<sk_sp<SkData>> m_lzwBlocks; // LZW blocks for this frame.
SkGIFColorMap m_localColorMap;
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
size_t m_currentLzwBlock;
bool m_isComplete;
bool m_isHeaderDefined;
bool m_isDataSizeDefined;
};
class SkGifImageReader final : public SkNoncopyable {
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
public:
// This takes ownership of stream.
SkGifImageReader(SkStream* stream)
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_client(nullptr)
, m_state(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
, m_bytesToConsume(6) // Number of bytes for GIF type, either "GIF87a" or "GIF89a".
, m_version(0)
, m_screenWidth(0)
, m_screenHeight(0)
, m_loopCount(cLoopCountNotSeen)
, m_streamBuffer(stream)
, m_parseCompleted(false)
, m_firstFrameHasAlpha(false)
, m_firstFrameSupportsIndex8(false)
{
}
static constexpr int cLoopCountNotSeen = -2;
~SkGifImageReader()
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
{
}
void setClient(SkGifCodec* client) { m_client = client; }
unsigned screenWidth() const { return m_screenWidth; }
unsigned screenHeight() const { return m_screenHeight; }
// Option to pass to parse(). All enums are negative, because a non-negative value is used to
// indicate that the Reader should parse up to and including the frame indicated.
enum SkGIFParseQuery {
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 enough to determine the size. Note that this parses the first frame's header,
// since we may decide to expand based on the frame's dimensions.
SkGIFSizeQuery = -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
// Parse to the end, so we know about all frames.
SkGIFFrameCountQuery = -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
};
// Parse incoming GIF data stream into internal data structures.
// Non-negative values are used to indicate to parse through that frame.
// Return true if parsing has progressed or there is not enough data.
// Return false if a fatal error is encountered.
bool parse(SkGIFParseQuery);
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
// Decode the frame indicated by frameIndex.
// frameComplete will be set to true if the frame is completely decoded.
// The method returns false if there is an error.
bool decode(size_t frameIndex, bool* frameComplete);
size_t imagesCount() const
{
if (m_frames.empty())
return 0;
// This avoids counting an empty frame when the file is truncated right after
// SkGIFControlExtension but before 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
// FIXME: This extra complexity is not necessary and we should just report m_frames.size().
return m_frames.back()->isHeaderDefined() ? m_frames.size() : m_frames.size() - 1;
}
int loopCount() const { return m_loopCount; }
const SkGIFColorMap& globalColorMap() 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
{
return m_globalColorMap;
}
const SkGIFFrameContext* frameContext(size_t index) 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
{
return index < m_frames.size() ? m_frames[index].get() : 0;
}
void clearDecodeState() {
for (size_t index = 0; index < m_frames.size(); index++) {
m_frames[index]->clearDecodeState();
}
}
// Return the color table for frame index (which may be the global color table).
sk_sp<SkColorTable> getColorTable(SkColorType dstColorType, size_t index) const;
bool firstFrameHasAlpha() const { return m_firstFrameHasAlpha; }
bool firstFrameSupportsIndex8() const { return m_firstFrameSupportsIndex8; }
private:
// Requires that one byte has been buffered into m_streamBuffer.
unsigned char getOneByte() const {
return reinterpret_cast<const unsigned char*>(m_streamBuffer.get())[0];
}
void addFrameIfNecessary();
bool currentFrameIsFirstFrame() const
{
return m_frames.empty() || (m_frames.size() == 1u && !m_frames[0]->isComplete());
}
// Unowned pointer
SkGifCodec* m_client;
// Parsing state machine.
SkGIFState m_state; // Current decoder master state.
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
size_t m_bytesToConsume; // Number of bytes to consume for next stage of parsing.
// Global (multi-image) state.
int m_version; // Either 89 for GIF89 or 87 for GIF87.
unsigned m_screenWidth; // Logical screen width & height.
unsigned m_screenHeight;
SkGIFColorMap m_globalColorMap;
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 m_loopCount; // Netscape specific extension block to control the number of animation loops a GIF renders.
std::vector<std::unique_ptr<SkGIFFrameContext>> m_frames;
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
SkStreamBuffer m_streamBuffer;
bool m_parseCompleted;
// These values can be computed before we create a SkGIFFrameContext, so we
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
// store them here instead of on m_frames[0].
bool m_firstFrameHasAlpha;
bool m_firstFrameSupportsIndex8;
};
#endif