Summary: | Implement "bitmaprenderer" CanvasRenderingContext | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Dean Jackson <dino> | ||||
Component: | Canvas | Assignee: | Dean Jackson <dino> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | buildbot, cdumez, dino, esprehn+autocc, graouts, gyuyoung.kim, kondapallykalyan, sam, webkit-bug-importer | ||||
Priority: | P2 | Keywords: | InRadar | ||||
Version: | WebKit Nightly Build | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Dean Jackson
2017-10-29 14:04:04 PDT
Created attachment 325292 [details]
Patch
Comment on attachment 325292 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=325292&action=review > Source/WebCore/html/HTMLCanvasElement.cpp:507 > +// FIXME: Needs to accept ImageBitmapRenderingContext::Settings. Might as well file a bug for this, even if you fix it soon. > Source/WebCore/html/canvas/ImageBitmapRenderingContext.cpp:55 > + // 1. If a bitmap argument was not provided, then: I'd prefer a simple link to the spec than copying the entire copy as-is which may end up being stale. Comment on attachment 325292 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=325292&action=review > Source/WebCore/html/ImageBitmap.cpp:563 > + m_detached = true; Not related to this change, but is the m_detached useful? I would think it always corresponds to whether m_bitmapData is null or not. Comment on attachment 325292 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=325292&action=review >> Source/WebCore/html/ImageBitmap.cpp:563 >> + m_detached = true; > > Not related to this change, but is the m_detached useful? I would think it always corresponds to whether m_bitmapData is null or not. It might be possible to transfer a null ImageBitmap, although I'm not sure how you'd ever create one. I think you're right though, so I'll make the change. >> Source/WebCore/html/canvas/ImageBitmapRenderingContext.cpp:55 >> + // 1. If a bitmap argument was not provided, then: > > I'd prefer a simple link to the spec than copying the entire copy as-is which may end up being stale. I prefer the specification as comments because: 1. It makes it easier to implement. 2. It makes it easier to review - you don't have to be reading the specification and linking the logic up. 3. It makes it more clear when our implementation gets out of date compared to the specification. Committed r224195: <https://trac.webkit.org/changeset/224195> |