Fix 'border' serialization to fail in cssText when it is invalid

Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363
Committed: https://crrev.com/5cf5e5234ad01ef8e22f10f25f674232a2ae7d2e
Cr-Commit-Position: refs/heads/master@{#394409}

[Timothy Loh, 2016-05-17 01:50:44]

Description was changed from

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

This patch makes us match Firefox in behaviour.
==========

to

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

[Timothy Loh, 2016-05-17 01:50:44]

timloh@chromium.org changed reviewers:
+ alancutter@chromium.org, nainar@chromium.org

[Timothy Loh, 2016-05-17 01:52:28]

Description was changed from

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

to

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

[Timothy Loh, 2016-05-17 01:53:53]

https://codereview.chromium.org/1982903002/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/fast/dom/css-shorthand-common-value.html
(right):

https://codereview.chromium.org/1982903002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/fast/dom/css-shorthand-common-value.html:8:
<p>getPropertyValue('border') should not return a value for any property that
doesn't have the same value for top, left, right and bottom, even if the values
that differ are implicitly set by a shorthand.
Interestingly, this description was here since this test's inception but the
test itself was always incorrect.

[nainar, 2016-05-17 23:55:17]

lgtm

[alancutter (OOO until 2018), 2016-05-18 01:58:51]

lgtm

[Timothy Loh, 2016-05-18 02:17:41]

The CQ bit was checked by timloh@chromium.org

[commit-bot: I haz the power, 2016-05-18 02:18:15]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1982903002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1982903002/1

[commit-bot: I haz the power, 2016-05-18 04:24:56]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-05-18 04:24:57]

Try jobs failed on following builders:
  win_chromium_compile_dbg_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_comp...)

[Timothy Loh, 2016-05-18 04:26:14]

The CQ bit was checked by timloh@chromium.org

[commit-bot: I haz the power, 2016-05-18 04:26:28]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1982903002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1982903002/1

[commit-bot: I haz the power, 2016-05-18 05:27:00]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-05-18 05:27:01]

Try jobs failed on following builders:
  win_chromium_compile_dbg_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_comp...)

[Timothy Loh, 2016-05-18 07:28:52]

The CQ bit was checked by timloh@chromium.org

[commit-bot: I haz the power, 2016-05-18 07:29:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1982903002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1982903002/1

[commit-bot: I haz the power, 2016-05-18 08:22:45]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-05-18 08:22:46]

Try jobs failed on following builders:
  win_chromium_compile_dbg_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_comp...)

[Timothy Loh, 2016-05-18 12:55:22]

The CQ bit was checked by timloh@chromium.org

[commit-bot: I haz the power, 2016-05-18 12:55:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1982903002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1982903002/1

[commit-bot: I haz the power, 2016-05-18 13:58:10]

Description was changed from

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

to

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

[commit-bot: I haz the power, 2016-05-18 13:58:11]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-05-18 13:59:28]

Description was changed from

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363
==========

to

==========
Fix 'border' serialization to fail in cssText when it is invalid

Currently we serialize 'border' correctly in e.style.border, but when
reading cssText (which serializes an entire declaration) we sometimes
get this wrong. Since the 'border' shorthand always sets the four
(top bottom left right) longhands for each type (color style width)
to the same value, if any of the types have unequal longhands then
seialization should fail.

The goal here is to make cssText serialization do the same thing as
the other serialization, so we can make it just call getPropertyValue
and have generic checks (e.g. for 'initial') in that function.

This patch makes us match Firefox in behaviour.

BUG=612363

Committed: https://crrev.com/5cf5e5234ad01ef8e22f10f25f674232a2ae7d2e
Cr-Commit-Position: refs/heads/master@{#394409}
==========

[commit-bot: I haz the power, 2016-05-18 13:59:29]

Patchset 1 (id:??) landed as
https://crrev.com/5cf5e5234ad01ef8e22f10f25f674232a2ae7d2e
Cr-Commit-Position: refs/heads/master@{#394409}