Support unit-less lengths for white-list of properties only.

Support unit-less lengths for white-list of properties only.

The Quirks Mode spec[1] has a white-list of properties for which we should
accept unitless lengths in quirks mode. This CL limits the quirk to apply
to those properties/short-hands.

[1] https://quirks.spec.whatwg.org/#the-unitless-length-quirk

BUG=436145
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=187959

[Timothy Loh, 2014-11-27 00:13:02]

FYI SVG lengths can be unitless unconditionally

http://www.w3.org/TR/SVG/types.html#BasicDataTypes

"For SVG-specific properties, the length unit identifier is optional. If a unit
is not provided, the length value represents a distance in the current user
coordinate system."

[rune, 2014-11-27 14:29:15]

On 2014/11/27 at 00:13:02, timloh wrote:
> FYI SVG lengths can be unitless unconditionally
> 
> http://www.w3.org/TR/SVG/types.html#BasicDataTypes
> 
> "For SVG-specific properties, the length unit identifier is optional. If a
unit is not provided, the length value represents a distance in the current user
coordinate system."

Yes, the current patch fails big time on svg.

[rune, 2014-12-01 11:29:43]

rune@opera.com changed reviewers:
+ timloh@chromium.org

[rune, 2014-12-01 11:29:43]

Ready for review.

[rune, 2015-01-05 14:22:03]

More properties (from CSS2.1) were added to the white-list in the spec. Patch
updated.

[Timothy Loh, 2015-01-05 23:12:20]

On 2015/01/05 14:22:03, rune wrote:
> More properties (from CSS2.1) were added to the white-list in the spec. Patch
> updated.

Sorry for the delay, I didn't notice this the first time around. Will get to
this shortly.

[Timothy Loh, 2015-01-06 23:21:00]

lgtm

I don't really like the unitless variable that spans the scope of the entire
switch statement, but I couldn't think of a better way to do it.

I think you might need a separate OWNER for Source/web test changes

https://codereview.chromium.org/764703002/diff/100001/LayoutTests/TestExpecta...
File LayoutTests/TestExpectations (right):

https://codereview.chromium.org/764703002/diff/100001/LayoutTests/TestExpecta...
LayoutTests/TestExpectations:358: crbug.com/436145
animations/keyframe-multiple-timing-functions-transform.html [ NeedsRebaseline ]
What needs to be rebaselined here?

[rune, 2015-01-07 08:40:23]

On 2015/01/06 at 23:21:00, timloh wrote:

> I think you might need a separate OWNER for Source/web test changes

Looks like I can revert those test changes since they change
min/max-width/height which were still going to support the quirk after all.

[rune, 2015-01-07 08:40:37]

https://codereview.chromium.org/764703002/diff/100001/LayoutTests/TestExpecta...
File LayoutTests/TestExpectations (right):

https://codereview.chromium.org/764703002/diff/100001/LayoutTests/TestExpecta...
LayoutTests/TestExpectations:358: crbug.com/436145
animations/keyframe-multiple-timing-functions-transform.html [ NeedsRebaseline ]
On 2015/01/06 at 23:21:00, Timothy Loh wrote:
> What needs to be rebaselined here?

The integer 1 in this short-hand in the test were interpreted as a unitless
delay of 1ms:

"horizontal 2s linear 1 alternate"

I assumed the intention of the test was to use 1 as the
animation-iteration-count, so instead of adding "ms", I wanted to do a
rebaseline. The rebaseline is necessary because the removed 1ms animation-delay
causes a rendering difference.

[rune, 2015-01-07 09:32:08]

The CQ bit was checked by rune@opera.com

[commit-bot: I haz the power, 2015-01-07 09:32:42]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/764703002/120001

[commit-bot: I haz the power, 2015-01-07 10:46:20]

Committed patchset #7 (id:120001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=187959