Track actual offsets of image dimensions in UMA |
||
Issue descriptionImageDecoders.EffectiveDimensionsLocation tracks how much of an image file Chrome needed to read to get the dimensions information, but Chrome not optimized to stop reading at the byte location where the dimensions actually reside. The code below provides one way to track that actual location, for JPEG. As appropriate, we should track for other formats as well. https://codereview.chromium.org/1647703006/diff/20001/third_party/WebKit/Source/platform/image-decoders/jpeg/JPEGImageDecoder.cpp
,
May 17 2016
Since bengr@ already coded up a rolled image decoder for blink, I will try to land that. For PNG, we know it is 32 bytes, and for GIF the canvas size is in the first 20 bytes. I'll look to see if any other formats need this as well, but I suspect that JPEG is the only one that is necessary.
,
Oct 28 2016
|
||
►
Sign in to add a comment |
||
Comment 1 by bengr@chromium.org
, May 17 2016