Allow CanvasSkia to bind to an existing SkCanvas.

Allow CanvasSkia to bind to an existing SkCanvas.

BUG=None
TEST=None


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=104146

[piman, 2011-10-05 16:58:04]

This is needed for the webkit compositor, which provides the canvas when
painting.
Most of the meat of this change is in ui/gfx/canvas* and skia/ext/canvas_paint*
The rest is mostly mechanical changes.

[Ben Goodger (Google), 2011-10-05 17:11:45]

http://codereview.chromium.org/8122013/diff/6092/ui/gfx/canvas.h
File ui/gfx/canvas.h (right):

http://codereview.chromium.org/8122013/diff/6092/ui/gfx/canvas.h#newcode237
ui/gfx/canvas.h:237: virtual const CanvasSkia* AsCanvasSkia() const;
Did anyone use this for anything other than to get a SkCanvas? Can we just
remove this?

[piman, 2011-10-05 17:31:28]

http://codereview.chromium.org/8122013/diff/6092/ui/gfx/canvas.h
File ui/gfx/canvas.h (right):

http://codereview.chromium.org/8122013/diff/6092/ui/gfx/canvas.h#newcode237
ui/gfx/canvas.h:237: virtual const CanvasSkia* AsCanvasSkia() const;
On 2011/10/05 17:11:45, Ben Goodger (Google) wrote:
> Did anyone use this for anything other than to get a SkCanvas? Can we just
> remove this?

I wanted to get rid of it, but there's a dozen or so calls that need this to
access a method of CanvasSkia that isn't an override of Canvas. Some of them
probably could be made that way, but others probably don't (like ExtractBitmap,
that returns a SkBitmap&), so I didn't push it. We could.

[Ben Goodger (Google), 2011-10-05 17:33:15]

OK. I figure when we nix CanvasDirect2D and consolidate Canvas/CanvasSkia the
need for the method will disappear. This change LGTM though.

[commit-bot: I haz the power, 2011-10-05 17:34:25]

CQ is trying the patch. Follow status at
https://chromium-status.appspot.com/cq/piman@chromium.org/8122013/6092

[commit-bot: I haz the power, 2011-10-05 19:05:10]

Change committed as 104146