Use 4 value syntax for getComputedStyle of background-position and -webkit-mask-position

Use 4 value syntax for getComputedStyle of background-position and -webkit-mask-position

This patch forces getComputedStyle of background-position to always return
the 4 value syntax (eg. left 10% top 25px) regardless of input format.
Similar behaviour applies to -webkit-mask-position as well.

This format is dictated by the background-position spec:
http://www.w3.org/TR/css3-background/#the-background-position

Existing tests already check this behaviour and have been updated.

BUG=310977
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=161213

[Timothy Loh, 2013-10-24 05:20:41]

What do other browsers return for getComputedStyle on background-position?

[dstockwell, 2013-10-24 05:29:56]

You've linked to the editors draft, but it appears to be specified the same way
in the last CR.

[alancutter (OOO until 2018), 2013-10-24 05:32:55]

On 2013/10/24 05:29:56, dstockwell wrote:
> You've linked to the editors draft, but it appears to be specified the same
way
> in the last CR.

Thanks for spotting that, changed link to CR.

[alancutter (OOO until 2018), 2013-10-24 05:36:38]

On 2013/10/24 05:20:41, Timothy Loh wrote:
> What do other browsers return for getComputedStyle on background-position?

Firefox always returns the 2 value syntax.
"right 10px center" becomes "calc(-10px + 100%) 50%".

[alancutter (OOO until 2018), 2013-10-28 03:20:25]

steveblock: Poke.

[shans, 2013-10-29 23:11:49]

lgtm

[alancutter (OOO until 2018), 2013-10-31 07:17:00]

On 2013/10/29 23:11:49, shans wrote:
> lgtm

steveblock: Poke.

[Steve Block, 2013-11-01 01:08:52]

lgtm

[commit-bot: I haz the power, 2013-11-01 01:29:55]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/alancutter@chromium.org/38573003/1

[commit-bot: I haz the power, 2013-11-01 01:30:06]

Failed to apply patch for
LayoutTests/animations/resources/animation-test-helpers.js:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file LayoutTests/animations/resources/animation-test-helpers.js
  Hunk #1 FAILED at 373.
  1 out of 1 hunk FAILED -- saving rejects to file
LayoutTests/animations/resources/animation-test-helpers.js.rej

Patch:       LayoutTests/animations/resources/animation-test-helpers.js
Index: LayoutTests/animations/resources/animation-test-helpers.js
diff --git a/LayoutTests/animations/resources/animation-test-helpers.js
b/LayoutTests/animations/resources/animation-test-helpers.js
index
49e63e710edc60225558cee8c66386f70ffd582e..00f4e51c769f3672519b6f22f96b10f59571d17a
100644
--- a/LayoutTests/animations/resources/animation-test-helpers.js
+++ b/LayoutTests/animations/resources/animation-test-helpers.js
@@ -373,7 +373,8 @@ function checkExpectedTransitionValue(expected, index)
             for (var i = 0; i < computedStyle.length; ++i) {
                 switch (computedStyle[i].cssValueType) {
                   case CSSValue.CSS_PRIMITIVE_VALUE:
-                   
values.push(computedStyle[i].getFloatValue(CSSPrimitiveValue.CSS_NUMBER));
+                    if (computedStyle[i].primitiveType !==
CSSPrimitiveValue.CSS_IDENT)
+                       
values.push(computedStyle[i].getFloatValue(CSSPrimitiveValue.CSS_NUMBER));
                     break;
                   case CSSValue.CSS_CUSTOM:
                     // arbitrarily pick shadow-x and shadow-y

[commit-bot: I haz the power, 2013-11-01 23:03:46]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/alancutter@chromium.org/38573003/150001

[commit-bot: I haz the power, 2013-11-02 00:46:35]

Change committed as 161213

[darktears, 2013-11-04 11:29:33]

On 2013/11/02 00:46:35, I haz the power (commit-bot) wrote:
> Change committed as 161213

Isn't this going to break various websites and toolkits relying on the 2 values
syntax when querying the computed value? getComputedStyle now outputs values
that old libraries or old code won't be able to parse anymore.

[alancutter (OOO until 2018), 2013-11-04 22:21:50]

On 2013/11/04 11:29:33, darktears wrote:
> On 2013/11/02 00:46:35, I haz the power (commit-bot) wrote:
> > Change committed as 161213
> 
> Isn't this going to break various websites and toolkits relying on the 2
values
> syntax when querying the computed value? getComputedStyle now outputs values
> that old libraries or old code won't be able to parse anymore.

Without metrics for getComputedStyle usage we can only speculate on the real
world dependencies on this output format. The non-stable channels are our
sources of feedback for these changes.

[jsclark, 2013-11-13 23:48:46]

On 2013/11/04 22:21:50, alancutter wrote:
> Without metrics for getComputedStyle usage we can only speculate on the real
> world dependencies on this output format. The non-stable channels are our
> sources of feedback for these changes.

This appears to have gone out in Chrome beta 32.0.1700.6 or 32.0.1700.14

What are the plans for this going into stable?  I work on a web app that this
change will have a significant impact on.  While we can adjust our code to
handle the new format, pushing an update to all our users could be problematic
depending on how quickly this makes it to stable.

Also, is the proper place for this comment, or should have commented on the
ticket or somewhere else?

[alancutter (OOO until 2018), 2013-11-18 08:44:43]

On 2013/11/13 23:48:46, jsclark wrote:
> On 2013/11/04 22:21:50, alancutter wrote:
> > Without metrics for getComputedStyle usage we can only speculate on the real
> > world dependencies on this output format. The non-stable channels are our
> > sources of feedback for these changes.
> 
> This appears to have gone out in Chrome beta 32.0.1700.6 or 32.0.1700.14
> 
> What are the plans for this going into stable?  I work on a web app that this
> change will have a significant impact on.  While we can adjust our code to
> handle the new format, pushing an update to all our users could be problematic
> depending on how quickly this makes it to stable.
> 
> Also, is the proper place for this comment, or should have commented on the
> ticket or somewhere else?

Hi, sorry for the late reply.
The issue tracker is probably the best place to discuss problems with this
change: crbug.com/310977
These code changes aren't searchable in Rietveld while the issue tracker is so
it's more likely to be seen if others are looking for it.
I will reply to you there.