Compute source/destination rectangles in FEComponentTransfer::applySkia

Compute source/destination rectangles in FEComponentTransfer::applySkia

When applying the filtering operation, care must be taken to use the
proper rectangle from the input efffect/primitive to adjust for a
potential difference in the filter primitive subregion.

TEST=svg/filters/feComponentTransfer-subregion.svg
BUG=327693
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=164019

[Stephen Chennney, 2013-12-17 14:19:51]

LGTM with a couple of nits on the test.

https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
File LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg (right):

https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:2: <text
x="10" y="20">There should be a lime rectangle below, and no red.</text>
Text in ref-tests can sometimes result in flakiness. Please remove it.

https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:3: <rect
x="50" y="50" width="100" height="100" fill="lime"/>
Please use an elicit rgb color spec for the fill. I presume it is #00ff00.

[fs, 2013-12-17 14:37:46]

On 2013/12/17 14:19:51, Stephen Chenney wrote:
> LGTM with a couple of nits on the test.
> 
>
https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
> File LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg
(right):
> 
>
https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
> LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:2: <text
> x="10" y="20">There should be a lime rectangle below, and no red.</text>
> Text in ref-tests can sometimes result in flakiness. Please remove it.

Turned the text into a comment instead.

>
https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
> LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:3: <rect
> x="50" y="50" width="100" height="100" fill="lime"/>
> Please use an elicit rgb color spec for the fill. I presume it is #00ff00.

Fixed (this and the other two). May I ask why this is though?

[commit-bot: I haz the power, 2013-12-17 14:38:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fs@opera.com/112053007/20001

[Stephen Chennney, 2013-12-17 15:44:58]

On 2013/12/17 14:37:46, fs wrote:
> On 2013/12/17 14:19:51, Stephen Chenney wrote:
>
https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
> > LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:3: <rect
> > x="50" y="50" width="100" height="100" fill="lime"/>
> > Please use an elicit rgb color spec for the fill. I presume it is #00ff00.
> 
> Fixed (this and the other two). May I ask why this is though?

For someone like me, who isn't familiar with the exact RGB values that
correspond to a CSS color, it avoids a step in looking up what the value should
be in case of a test failure.

I'm conscious of the issue after looking at some bugs where we were off by 1 in
color calculations.

[fs, 2013-12-17 15:54:48]

On 2013/12/17 15:44:58, Stephen Chenney wrote:
> On 2013/12/17 14:37:46, fs wrote:
> > On 2013/12/17 14:19:51, Stephen Chenney wrote:
> >
>
https://codereview.chromium.org/112053007/diff/1/LayoutTests/svg/filters/feCo...
> > > LayoutTests/svg/filters/feComponentTransfer-subregion-expected.svg:3:
<rect
> > > x="50" y="50" width="100" height="100" fill="lime"/>
> > > Please use an elicit rgb color spec for the fill. I presume it is #00ff00.
> > 
> > Fixed (this and the other two). May I ask why this is though?
> 
> For someone like me, who isn't familiar with the exact RGB values that
> correspond to a CSS color, it avoids a step in looking up what the value
should
> be in case of a test failure.

I think most (CSS) color names -> RGB I know are "testcase colors" (red, lime,
green, blue). I can absolutely see how "lime" is non-obvious though...

[commit-bot: I haz the power, 2013-12-17 16:10:21]

Change committed as 164019