Remove WebCore::IntRect usage from compositor, except within Region.

Remove WebCore::IntRect usage from compositor, except within Region.

This adds an API to the Region class that is intended to be our final Region
class interface. It uses an iterator to walk the rects in the Region, which is
compatible with the SkRegion API, but can also hide the IntRects exposed by the
WebCore Region API.

Once this is done, there is no need to use cc::IntRect, and we can remove it
entirely.

Covered by existing tests; no change in behaviour.

BUG=147395
R=enne

[danakj, 2012-11-02 19:49:23]

Weee!

[danakj, 2012-11-02 19:50:04]

If you like, I can move this class into cc/ and make it wrap Skia, in the same
CL.

Or, I can do it separately.

[enne (OOO), 2012-11-02 20:30:04]

https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h
File cc/stubs/Region.h (right):

https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h#newcode66
cc/stubs/Region.h:66: int size() { return m_rects.size(); }
Why don't you just add a NumRects() function?

[danakj, 2012-11-02 20:32:30]

https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h
File cc/stubs/Region.h (right):

https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h#newcode66
cc/stubs/Region.h:66: int size() { return m_rects.size(); }
On 2012/11/02 20:30:04, enne wrote:
> Why don't you just add a NumRects() function?

Because it would be super costly with the WebCore implementation, which suggests
that maybe for other Region implementations it would be too. Whereas, there are
a non-zero number of cases where we want the size, and will be iterating
(web_layer_impl.cc - incoming patch version). And constructing and iterator and
taking the size in tests can be slow but we won't care.

[enne (OOO), 2012-11-02 20:40:27]

On 2012/11/02 20:32:30, danakj wrote:
> https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h
> File cc/stubs/Region.h (right):
> 
> https://codereview.chromium.org/11361072/diff/1/cc/stubs/Region.h#newcode66
> cc/stubs/Region.h:66: int size() { return m_rects.size(); }
> On 2012/11/02 20:30:04, enne wrote:
> > Why don't you just add a NumRects() function?
> 
> Because it would be super costly with the WebCore implementation, which
suggests
> that maybe for other Region implementations it would be too. Whereas, there
are
> a non-zero number of cases where we want the size, and will be iterating
> (web_layer_impl.cc - incoming patch version). And constructing and iterator
and
> taking the size in tests can be slow but we won't care.

Ah, right.  Since it returns them all by value.

lgtm