SubtreeStyleChange for :hover/:active/:focus with ::first-letter.

When the ::first-letter pseudo element is added or removed, it causes a
Reattach which means everything will work fine for ::first-letter changes.

However, if there is a change in ::first-letter style, but the element had
::first-letter style both before and after the change, we would not try to
recalc any ::first-letter pseudo for LocalStyleChange.

The introduction of LocalStyleChange for :hover/:active/:focus caused a
regression for the combination of ::first-letter and e.g. :hover.

This fix does a SubtreeStyleChange for :hover/:active/:focus updates when
the existing renderStyle is marked as having ::first-letter pseudo.

::first-line is working fine. Added a layout test to make sure it will
continue to do so.

R=esprehn
BUG=344322
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=168509

[rune, 2014-02-20 23:39:55]

On 2014/02/18 23:05:25, rune - CET wrote:

Another alternative would be to use childrenAffectedByHover/Focus/Active in the
presence of ::first-letter instead.

[rune, 2014-02-21 16:44:41]

On 2014/02/20 23:39:55, rune - CET wrote:
> On 2014/02/18 23:05:25, rune - CET wrote:
> 
> Another alternative would be to use childrenAffectedByHover/Focus/Active in
the
> presence of ::first-letter instead.

Ended up avoiding LocalStyleChange for :hover/:active/:focus updates instead.
Modifying LocalStyleChange behavior for ::first-letter also meant that modifying
the style object/attribute of an element with ::first-letter could have a full
subtree recalc, but changing the style object/attribute can never modify the
::first-letter style directly.

[rune, 2014-03-04 20:08:11]

@esprehn any chance you could take at look at this CL?

[esprehn, 2014-03-04 20:24:10]

On 2014/03/04 20:08:11, rune - CET wrote:
> @esprehn any chance you could take at look at this CL?

lgtm, good catch

[rune, 2014-03-04 20:27:12]

The CQ bit was checked by rune@opera.com

[commit-bot: I haz the power, 2014-03-04 20:28:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/rune@opera.com/169643002/110001

[rune, 2014-03-05 10:35:02]

The CQ bit was unchecked by rune@opera.com

[rune, 2014-03-05 10:35:18]

The CQ bit was checked by rune@opera.com

[commit-bot: I haz the power, 2014-03-05 10:35:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/rune@opera.com/169643002/110001

[commit-bot: I haz the power, 2014-03-05 18:28:46]

Change committed as 168509