Bug 84006 – support canvas.resetTransform()

• Component: Canvas
• Status: RESOLVED CONFIGURATION CHANGED

[Lu Guanqun]

Support canvas.resetTransform() interface.

[Lu Guanqun]

Created attachment 137269 [details]
Patch

[WebKit Review Bot]

Comment on attachment 137269 [details]
Patch

Attachment 137269 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/12408859

New failing tests:
platform/chromium/virtual/gpu/fast/canvas/canvas-resetTransform.html
fast/canvas/canvas-resetTransform.html

[WebKit Review Bot]

Created attachment 137276 [details]
Archive of layout-test-results from ec2-cr-linux-03

The attached test failures were seen while running run-webkit-tests on the chromium-ews.
Bot: ec2-cr-linux-03  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

[Lu Guanqun]

Created attachment 137278 [details]
Patch

[Alexey Proskuryakov]

Comment on attachment 137278 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=137278&action=review

> Source/WebCore/html/canvas/CanvasRenderingContext2D.cpp:670
> +    state().m_transform = AffineTransform();

Shouldn't this be makeIdentity()?

[Dean Jackson]

Comment on attachment 137278 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=137278&action=review

> LayoutTests/fast/canvas/script-tests/canvas-resetTransform.js:15
> +ctx.beginPath();
> +ctx.scale(0.5, 0.5);
> +ctx.resetTransform();
> +ctx.fillStyle = 'green';
> +ctx.fillRect(0, 0, 100, 100);
> +
> +var imageData = ctx.getImageData(1, 1, 98, 98);
> +var imgdata = imageData.data;
> +shouldBe("imgdata[4]", "0");
> +shouldBe("imgdata[5]", "128");
> +shouldBe("imgdata[6]", "0");

I would expect this test to draw two identical rects with different colours, one with a scale and one after the resetTransform, and test both corners of the imagedata. Am I missing something?

And maybe have some debug() statements to explain what's going on.

> Source/WebCore/html/canvas/CanvasRenderingContext2D.cpp:668
> +    if (!ctm.isInvertible())
> +        return;

Is this behaviour explained anywhere? I can understand why the other transformation functions have it, but in this case we're trying to reset the transform to identity, so I don't know why it is should fail if there is a currently un-invertable state.

>> Source/WebCore/html/canvas/CanvasRenderingContext2D.cpp:670
>> +    state().m_transform = AffineTransform();
> 
> Shouldn't this be makeIdentity()?

Agreed.

[Adele Peterson]

Comment on attachment 137278 [details]
Patch

r- until Dean and Alexey's comments are addressed.

[Joseph Pecoraro]

It seems we already support resetTransform. Lets close this out.