Let CSS Color handle CSS <color> value parsing

[annevk]

Our definition had issues such as not resolving currentcolor correctly when used as part of CSS color functions.

Fixes #9940.

• At least two implementers are interested (and none opposed):
  • Chromium
  • WebKit
• Tests are written and can be reviewed and commented upon at:
  • https://wpt.fyi/results/html/canvas/element/fill-and-stroke-styles/2d.fillStyle.colormix.currentcolor.html
• Implementation bugs are filed:
  • Chromium: N/A
  • Gecko: https://bugzilla.mozilla.org/show_bug.cgi?id=1883586
  • WebKit: https://bugs.webkit.org/show_bug.cgi?id=270509
• MDN issue is filed: N/A
• The top of this comment includes a clear commit message to use.

(See WHATWG Working Mode: Changes for more details.)

---

/canvas.html ( diff )
/infrastructure.html ( diff )
/semantics.html ( diff )

[annevk]

@Loirooriol do you know if this is tested? Is any browser non-compliant?

[Loirooriol]

There is web-platform-tests/wpt@cf3bea4, e.g. https://wpt.fyi/results/html/canvas/element/fill-and-stroke-styles/2d.fillStyle.colormix.currentcolor.html?label=master&label=experimental&product=chrome&product=edge&product=firefox&product=safari&product=servo

WebKit rejects color-mix(), while Gecko treats currentcolor as #000000. Blink passes the test, Servo is also right but fails due to #8917.

[annevk]

Thank you! @domenic any objection to treating this as an obvious bug fix?

[Kaiido]

I notice that shadowColor uses some old attribute syntax which makes the use of the parsing algo a bit odd (potentially setting the color to failure, before pretending nothing did happen).
Though I suppose it would be best to handle that in a separate PR that would update the 4 CanvasShadowStyles attributes to use getter & setter steps instead?

[domenic]

@domenic any objection to treating this as an obvious bug fix?

I guess it doesn't seem necessary. If Blink already implements this and you're happy to speak for WebKit, and we have tests... might as well just check all the boxes.

[annevk]

@Kaiido I meant to reply before closing this. It seems worthwhile updating those getter/setters, yes.