Only default to transition shorthand if all 4 values are set

Only default to transition shorthand if all 4 values are set

When setting any individual transition property the
transition shorthand was used for cssText. This patch
changes the behavior so that only when all 4 individual
properties have a value set, the shorthand is used. A
test is added to verify this.

Matches Firefox behavior.

BUG=310403
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=183024

[rwlbuis, 2014-09-29 21:46:28]

rob.buis@samsung.com changed reviewers:
+ timloh@google.com

[rwlbuis, 2014-09-29 21:46:33]

PTAL

[Timothy Loh, 2014-09-30 11:37:05]

On 2014/09/29 21:46:33, rwlbuis wrote:
> PTAL

Is this something we also want to do for other callers to
getLayeredShorthandValue? Looks like the same behaviour occurs with animations.

[rwlbuis, 2014-09-30 11:57:22]

On 2014/09/30 11:37:05, Timothy Loh wrote:
> On 2014/09/29 21:46:33, rwlbuis wrote:
> > PTAL
> 
> Is this something we also want to do for other callers to
> getLayeredShorthandValue? Looks like the same behaviour occurs with
animations.

Good point. I'll try that and see how (much) that changes the test results. Oh,
and I'll actually
add the expected file :)

[rwlbuis, 2014-09-30 15:34:07]

On 2014/09/30 11:57:22, rwlbuis wrote:
> On 2014/09/30 11:37:05, Timothy Loh wrote:
> > On 2014/09/29 21:46:33, rwlbuis wrote:
> > > PTAL
> > 
> > Is this something we also want to do for other callers to
> > getLayeredShorthandValue? Looks like the same behaviour occurs with
> animations.
> 
> Good point. I'll try that and see how (much) that changes the test results.
Oh,
> and I'll actually
> add the expected file :)

As expected, when including the check for all getLayeredShorthandValue callers I
get a lot of FAILS for
things like -webkit-mask. Since I don't know the other properties to well I'd
prefer to use the new code path for
animation/transition.

[Timothy Loh, 2014-10-01 00:29:31]

On 2014/09/30 15:34:07, rwlbuis wrote:
> On 2014/09/30 11:57:22, rwlbuis wrote:
> > On 2014/09/30 11:37:05, Timothy Loh wrote:
> > > On 2014/09/29 21:46:33, rwlbuis wrote:
> > > > PTAL
> > > 
> > > Is this something we also want to do for other callers to
> > > getLayeredShorthandValue? Looks like the same behaviour occurs with
> > animations.
> > 
> > Good point. I'll try that and see how (much) that changes the test results.
> Oh,
> > and I'll actually
> > add the expected file :)
> 
> As expected, when including the check for all getLayeredShorthandValue callers
I
> get a lot of FAILS for
> things like -webkit-mask. Since I don't know the other properties to well I'd
> prefer to use the new code path for
> animation/transition.

Seems reasonable to me. LGTM

[rwlbuis, 2014-10-01 01:33:37]

The CQ bit was checked by rob.buis@samsung.com

[commit-bot: I haz the power, 2014-10-01 01:34:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/617493002/60001

[commit-bot: I haz the power, 2014-10-01 02:53:02]

Committed patchset #4 (id:60001) as 183024