Get rid of Reset and ResetStyleResolverAndFontSelector.

Get rid of Reset and ResetStyleResolverAndFontSelector.

Now that we store RuleSets on StyleSheetContents, recreating
the StyleResolver is much cheaper. By getting rid of the Reset
codepaths we can simplify a lot of code and get rid of a lot
of duplication.

Also, this is generally less error prone since we don't have to worry
about some random subset of StyleResolver's member variables not
getting updated properly.

If we find recreating the StyleResolver on profiles, we should
look into moving the more expensive bits to better locations
(e.g. maybe fontSelector shouldn't live on StyleResolver).

[ojan, 2013-11-26 01:54:50]

https://codereview.chromium.org/85693009/diff/1/LayoutTests/fast/css/add-remo...
File LayoutTests/fast/css/add-remove-stylesheets-minimal-recalc-style.html
(right):

https://codereview.chromium.org/85693009/diff/1/LayoutTests/fast/css/add-remo...
LayoutTests/fast/css/add-remove-stylesheets-minimal-recalc-style.html:60: //
Recalc 3 elements + documentElement.
Don't really understand why this causes the documentElement to get recalced.
Different timing I suppose. In either case, it's not a big deal. We're just
recalcing 1 extra element.

[tasak, 2013-11-26 04:19:52]

lgtm.

[commit-bot: I haz the power, 2013-11-26 17:07:28]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ojan@chromium.org/85693009/1

[ojan, 2013-11-28 00:41:13]

Since the RuleSet on StyleSheetContents patch got reverted, I'm going to wait
until it relands before trying to land this.

[ojan, 2013-12-03 20:36:09]

It looks like resetting the fontSelector is expensive still. So, until we
divorce fontSelector from StyleResolver and make it so that the Reconstruct
codepath doesn't reset the font selector, this patch will be a regression. See
https://code.google.com/p/chromium/issues/detail?id=324909 for an example.