Simplify CSSSelector::isAttributeSelector() by not checking each Match type

Simplify CSSSelector::isAttributeSelector() by not checking each Match type

Simplify CSSSelector::isAttributeSelector() by not checking each Match type
explicitly and reordering the Match enum instead so that the attribute
selector match type are contiguous and at the end of the enum. This way, the
check can become (m_match >= FirstAttributeSelectorMatch).

This optimization is similar to the one we already use in CollectionType.h
for collection types. isAttributeSelector() is called from
SelectorChecker::checkOne() which is hot code.

I see a ~2% progression on Dromaeo's cssquery-jquery:
http://dromaeo.com/?id=222514,222518

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=175632

[esprehn, 2014-06-04 22:38:43]

lgtm, fix nit before landing.

https://codereview.chromium.org/315043005/diff/1/Source/core/css/CSSSelector.h
File Source/core/css/CSSSelector.h (right):

https://codereview.chromium.org/315043005/diff/1/Source/core/css/CSSSelector....
Source/core/css/CSSSelector.h:124: static const Match
FirstAttributeSelectorMatch = Exact;
Just do this inside the enum itself. You don't need a separate static.

[Inactive, 2014-06-04 23:49:51]

https://codereview.chromium.org/315043005/diff/20001/Source/core/css/CSSSelec...
File Source/core/css/CSSSelector.h (right):

https://codereview.chromium.org/315043005/diff/20001/Source/core/css/CSSSelec...
Source/core/css/CSSSelector.h:117: FirstAttributeSelectorMatch = Exact,
esprehn@, is this what you had in mind? I'd like not to increase the size of
this enum unnecessarily since we are storing this in a bitfield (there is still
some room for now, but we don't want to use those bits if we don't need to).

[Inactive, 2014-06-05 02:00:22]

https://codereview.chromium.org/315043005/diff/20001/Source/core/css/CSSSelec...
File Source/core/css/CSSSelector.h (right):

https://codereview.chromium.org/315043005/diff/20001/Source/core/css/CSSSelec...
Source/core/css/CSSSelector.h:117: FirstAttributeSelectorMatch = Exact,
On 2014/06/04 23:49:52, Chris Dumez wrote:
> esprehn@, is this what you had in mind? I'd like not to increase the size of
> this enum unnecessarily since we are storing this in a bitfield (there is
still
> some room for now, but we don't want to use those bits if we don't need to).

Maybe I should move this one to the end of the enum to make it clearer it is not
a normal enum value?

[Inactive, 2014-06-06 01:30:11]

The CQ bit was checked by ch.dumez@samsung.com

[commit-bot: I haz the power, 2014-06-06 01:30:50]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/ch.dumez@samsung.com/315043005/30001

[commit-bot: I haz the power, 2014-06-06 03:30:06]

Change committed as 175632