Bound RasterShapeInterval size to int

Bound RasterShapeInterval size to int

Everything in the code assumes that the interval size will always fit in
an int. However, to make sure there are no negatives, the
RasterShapeInterval constructor takes an unsigned. This clamps that
value to make sure we can't overflow an int later.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=175823

[Bem Jones-Bey (adobe), 2014-06-09 18:05:32]

WDYT?

[leviw_travelin_and_unemployed, 2014-06-09 18:19:06]

On 2014/06/09 at 18:05:32, bjonesbe wrote:
> WDYT?

Can we get a test that tickles this by chance? The change seems fine. It'd just
be neat if we had some bounds tests.

[Bem Jones-Bey (adobe), 2014-06-09 18:30:15]

On 2014/06/09 18:19:06, leviw wrote:
> On 2014/06/09 at 18:05:32, bjonesbe wrote:
> > WDYT?
> 
> Can we get a test that tickles this by chance? The change seems fine. It'd
just
> be neat if we had some bounds tests.

As far as I can tell, the current code only passes in ints, so I don't think
it's possible to write a layout test for this.

I can put together a unit test, though.

[Bem Jones-Bey (adobe), 2014-06-09 21:15:10]

On 2014/06/09 18:30:15, Bem Jones-Bey (adobe) wrote:
> On 2014/06/09 18:19:06, leviw wrote:
> > On 2014/06/09 at 18:05:32, bjonesbe wrote:
> > > WDYT?
> > 
> > Can we get a test that tickles this by chance? The change seems fine. It'd
> just
> > be neat if we had some bounds tests.
> 
> As far as I can tell, the current code only passes in ints, so I don't think
> it's possible to write a layout test for this.
> 
> I can put together a unit test, though.

How badly do you want that test? Unfortunately, the size of the intervals vector
is not part of the public interface of RasterShapeIntervals. While I can come up
with contortions to make it inspectable from the unit tests, I'm not sure it's
worth changing the implementation of RasterShapeIntervals just for this purpose.

[leviw_travelin_and_unemployed, 2014-06-09 21:26:04]

On 2014/06/09 at 21:15:10, bjonesbe wrote:
> How badly do you want that test? Unfortunately, the size of the intervals
vector is not part of the public interface of RasterShapeIntervals. While I can
come up with contortions to make it inspectable from the unit tests, I'm not
sure it's worth changing the implementation of RasterShapeIntervals just for
this purpose.

It's clearly an okay change, so I'll take it. LGTM.

[Bem Jones-Bey (adobe), 2014-06-09 21:37:03]

The CQ bit was checked by bjonesbe@adobe.com

[commit-bot: I haz the power, 2014-06-09 21:37:11]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/bjonesbe@adobe.com/313393004/1

[Bem Jones-Bey (adobe), 2014-06-09 21:37:16]

On 2014/06/09 21:26:04, leviw wrote:
> On 2014/06/09 at 21:15:10, bjonesbe wrote:
> > How badly do you want that test? Unfortunately, the size of the intervals
> vector is not part of the public interface of RasterShapeIntervals. While I
can
> come up with contortions to make it inspectable from the unit tests, I'm not
> sure it's worth changing the implementation of RasterShapeIntervals just for
> this purpose.
> 
> It's clearly an okay change, so I'll take it. LGTM.

Thanks!

[commit-bot: I haz the power, 2014-06-09 21:45:28]

Change committed as 175823