RuleSet invalidations on single feature only, including tags.

RuleSet invalidations on single feature only, including tags.

If we add a stylesheet with selector "#x span", we would
invalidate both #x and span elements regardless of whether the span
elements were descendants of #x. Instead introduce a bit on
invalidation sets saying if it should be scheduled for ruleset
invalidations or not, and mark only a single invalidation set per
selector.

Consider ".x #y". We will walk from right to left and remember one
simple selector for which we use invalidation sets. We choose more
specific selectors over less specific and choose the leftmost simple
selector for selectors with the same specificity. So in this case, we
will mark the invalidation set for "#y" to be used for ruleset
invalidation. For ".x .y", we will mark the invalidation set for ".x".

Additionally we create a singleton invalidation set to be scheduled on
the root node for element type invalidations. That is, for selectors
which contain a type selector in the rightmost compound, but do not
contain id, class, or attribute, we add that type selector to the
type invalidation set and schedule it on the root node for ruleset
invalidations.

BUG=680549

[rune, 2017-02-10 15:57:22]

The CQ bit was checked by rune@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2017-02-10 15:57:52]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-02-10 17:20:57]

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

[commit-bot: I haz the power, 2017-02-10 17:20:58]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[rune, 2017-02-14 23:32:55]

The CQ bit was checked by rune@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2017-02-14 23:34:01]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[rune, 2017-02-15 00:21:44]

https://codereview.chromium.org/2683373003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/inspector/tracing/timeline-style/parse-author-style-sheet-expected.txt
(right):

https://codereview.chromium.org/2683373003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector/tracing/timeline-style/parse-author-style-sheet-expected.txt:3:
error: Failed to lookup the event that scheduled a style invalidator
invalidation.
This is because the introduced type rule invalidation set is not traced. It's
easy to trigger this error for existing invalidations as well, even if we
generate a trace event for the invalidation set scheduling. Asked for the status
of the invalidation tracing and reported issue 692283.

[commit-bot: I haz the power, 2017-02-15 01:34:50]

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

[commit-bot: I haz the power, 2017-02-15 01:34:50]

Dry run: This issue passed the CQ dry run.

[rune, 2017-02-15 16:56:47]

Description was changed from

==========
RuleSet invalidations on single feature only, including tags.

If we have add a stylesheet with selector "#x span", we would
invalidate both #x and span elements regardless of whether the span
elements were descendants of #x. Instead introduce a bit on
invalidation sets saying if it should be scheduled for ruleset
invalidations or not, and mark only a single invalidation set per
selector.

Consider ".x #y", we will walk from right to left and remember one
simple selector for which we use invalidation sets. We choose more
specific selectors over less specific and choose the leftmost simple
selector for selectors with the same specificity. So in this case, we
will mark the invalidation set for "#y" to be used for ruleset
invalidation. For ".x .y", we will mark the invalidation set for ".x".

Additionally we create a singleton invalidation set to be scheduled on
the root node for element type invalidations. That is, for selectors
which contain a type selector in the rightmost compound, but do not
contain id, class, or attribute, we add that type selector to the
type invalidation set and schedule it on the root node for ruleset
invalidations.

BUG=680549
==========

to

==========
RuleSet invalidations on single feature only, including tags.

If we add a stylesheet with selector "#x span", we would
invalidate both #x and span elements regardless of whether the span
elements were descendants of #x. Instead introduce a bit on
invalidation sets saying if it should be scheduled for ruleset
invalidations or not, and mark only a single invalidation set per
selector.

Consider ".x #y". We will walk from right to left and remember one
simple selector for which we use invalidation sets. We choose more
specific selectors over less specific and choose the leftmost simple
selector for selectors with the same specificity. So in this case, we
will mark the invalidation set for "#y" to be used for ruleset
invalidation. For ".x .y", we will mark the invalidation set for ".x".

Additionally we create a singleton invalidation set to be scheduled on
the root node for element type invalidations. That is, for selectors
which contain a type selector in the rightmost compound, but do not
contain id, class, or attribute, we add that type selector to the
type invalidation set and schedule it on the root node for ruleset
invalidations.

BUG=680549
==========