Bug 50797 – canvas/philip/tests/2d.fillStyle.parse.invalid.rgba-6.html fails

• Status: RESOLVED
• Resolution: FIXED
• Component: CSS

[Mihai Parparita]

This tests checks that we don't allow the invalid color value "rgba(255, 0, 0, 1.)" to be used as fillStyle. We actually allow this, even though 1. is not a valid CSS <alphavalue> (which is a <number>, which is defined as "A <number> can either be an <integer>, or it can be zero or more digits followed by a dot (.) followed by one or more digits." at http://www.w3.org/TR/css3-values/#numbers).

Since this would involve making our CSS color parser stricter, I'm not sure that this is worth fixing.

[Sudarsana Nagineni (babu)]

*** Bug 93493 has been marked as a duplicate of this bug. ***

[Sudarsana Nagineni (babu)]

Created attachment 157462 [details]
Patch

[Andreas Kling]

(In reply to comment #0)
> Since this would involve making our CSS color parser stricter, I'm not sure that this is worth fixing.

Indeed, I wonder the same thing. What are other browsers doing?

[Sudarsana Nagineni (babu)]

(In reply to comment #3)
> Indeed, I wonder the same thing. What are other browsers doing?

Looks like other browsers (FireFox and Opera) are treating 1. is an invalid CSS alpha value, so the test is passing. Full parser doing the same in our case also. The issue is only with the fast-path color parser.

[Sudarsana Nagineni (babu)]

Kling, could you please have a look at this bug again?

[Andreas Kling]

Comment on attachment 157462 [details]
Patch

Okay.

[WebKit Review Bot]

Comment on attachment 157462 [details]
Patch

Clearing flags on attachment: 157462

Committed r126192: <http://trac.webkit.org/changeset/126192>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.