Only match style from element's TreeScope for normal rules.

Only match style from element's TreeScope for normal rules.

We collected all scoped resolvers up to the document resolver while
collecting normal style rules. That was necessary for <style scoped>, but
now that all scoping is done through shadow roots, only boundary crossing
rules can match from other author style scopes. Those rules are handled
through the global TreeBoundaryCrossingRules.

Instead, collect using the ScopedStyleResolver for the element's tree
scope if present. The need for CascadeScope disappears, so it's been
removed.

There is an exception for custom pseudo elements and VTT elements (styled
through ::cue) since those rules are crossing a shadow boundary, yet not
handled as part of boundary crossing rules. I chose to do the special
handling instead of start adding them to the boundary crossing rules
since the boundary crossing rule handling is about to change as part of
401359 anyway, and the special handling is safer since it's closer to the
previous behavior.

R=hayato@chromium.org
BUG=401359
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=188292

[rune, 2015-01-08 13:44:10]

ptal

[rune, 2015-01-08 14:16:58]

On 2015/01/08 at 13:44:10, rune wrote:
> ptal

Regressed:

media/track/css-cue-for-video-in-shadow-2.html

[rune, 2015-01-08 23:53:43]

On 2015/01/08 at 14:16:58, rune wrote:
> On 2015/01/08 at 13:44:10, rune wrote:
> > ptal
> 
> Regressed:
> 
> media/track/css-cue-for-video-in-shadow-2.html

This fails in general for elements with shadowPseudoId. Things like:

"::-webkit-meter-bar"

is represented internally as "* <ShadowPseudo> ::-webkit-meter-bar" effectively
making it a tree-boundary-crossing selector. Yet, the rule is not added to
treeBoundaryCrossingRules because the internal ShadowPseudo combinator is not
considered to be boundary crossing even though it is.

[hayato, 2015-01-09 08:36:52]

This looks similar to https://codereview.chromium.org/451983002/, as you might
be already aware of that.

On 2015/01/08 23:53:43, rune wrote:
> On 2015/01/08 at 14:16:58, rune wrote:
> > On 2015/01/08 at 13:44:10, rune wrote:
> > > ptal
> > 
> > Regressed:
> > 
> > media/track/css-cue-for-video-in-shadow-2.html
> 
> This fails in general for elements with shadowPseudoId. Things like:
> 
> "::-webkit-meter-bar"
> 
> is represented internally as "* <ShadowPseudo> ::-webkit-meter-bar"
effectively
> making it a tree-boundary-crossing selector. Yet, the rule is not added to
> treeBoundaryCrossingRules because the internal ShadowPseudo combinator is not
> considered to be boundary crossing even though it is.

[rune, 2015-01-09 10:21:10]

On 2015/01/09 at 08:36:52, hayato wrote:
> This looks similar to https://codereview.chromium.org/451983002/, as you might
be already aware of that.

I wasn't. It's that and similar to https://codereview.chromium.org/585683002/

I have a fix for custom pseudo elements and VTT now. I'll update the
description.

[rune, 2015-01-09 11:30:28]

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

[rune, 2015-01-09 11:30:28]

Adding esprehn@ since he did https://codereview.chromium.org/585683002/

That CL does a little bit more refactoring than this.

[rune, 2015-01-09 15:55:33]

On 2015/01/09 at 11:30:28, rune wrote:
> Adding esprehn@ since he did https://codereview.chromium.org/585683002/
> 
> That CL does a little bit more refactoring than this.

Follow-up refactoring here:

https://codereview.chromium.org/844963002

[rune, 2015-01-13 07:46:31]

hayato@ could you PTAL?

[hayato, 2015-01-13 08:46:53]

Yeah, let me take a look soon.
Sorry for the late response. Monday was a national local holiday in Tokyo and
today I am a blink gardener :)

[rune, 2015-01-13 08:51:56]

On 2015/01/13 at 08:46:53, hayato wrote:
> Yeah, let me take a look soon.
> Sorry for the late response. Monday was a national local holiday in Tokyo and
today I am a blink gardener :)

np.

[hayato, 2015-01-13 10:40:57]

LGTM

[rune, 2015-01-13 10:44:48]

On 2015/01/13 at 10:40:57, hayato wrote:
> LGTM

thanks!

[rune, 2015-01-13 10:44:53]

The CQ bit was checked by rune@opera.com

[commit-bot: I haz the power, 2015-01-13 10:45:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/843773002/20001

[commit-bot: I haz the power, 2015-01-13 12:23:13]

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

[commit-bot: I haz the power, 2015-01-13 12:23:14]

Try jobs failed on following builders:
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/45908)

[rune, 2015-01-13 12:32:33]

The CQ bit was checked by rune@opera.com

[commit-bot: I haz the power, 2015-01-13 12:33:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/843773002/20001

[commit-bot: I haz the power, 2015-01-13 13:18:09]

Committed patchset #2 (id:20001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=188292