Merge Element::offsetParentForBindings() into offsetParent()

Merge Element::offsetParentForBindings() into offsetParent()

This split was introduced with a fix for "offsetParent should never
return nodes in user agent Shadow DOM to script" in 2013:
https://trac.webkit.org/changeset/146037

SpatialNavigations's rectToAbsoluteCoordinates was the only internal use
of Element::offsetParent(). Rewrite that in a way that preserves the
existing behavior.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=191902

[philipj_slow, 2015-03-10 10:31:10]

philipj@opera.com changed reviewers:
+ dominicc@chromium.org

[philipj_slow, 2015-03-10 10:31:11]

PTAL

[dominicc (has gone to gerrit), 2015-03-10 13:14:15]

On 2015/03/10 at 10:31:11, philipj wrote:
> PTAL

What do you make of the test failures on eg Mac?

I think the if (!element ...) part is wrong; you want something that boils down
to:

if (!element)
  return nullptr;

if (!element->isInShadowTree())
  return element;

if (element->containingShadowRoot()->shouldExposeToBindings())
  return element;

return nullptr;

You can wrap that up into compound expressions and helpers any way you think
helps readability.

[philipj_slow, 2015-03-10 16:21:28]

oops

[philipj_slow, 2015-03-10 16:24:32]

On 2015/03/10 13:14:15, dominicc wrote:
> On 2015/03/10 at 10:31:11, philipj wrote:
> > PTAL
> 
> What do you make of the test failures on eg Mac?
> 
> I think the if (!element ...) part is wrong; you want something that boils
down
> to:
> 
> if (!element)
>   return nullptr;
> 
> if (!element->isInShadowTree())
>   return element;
> 
> if (element->containingShadowRoot()->shouldExposeToBindings())
>   return element;
> 
> return nullptr;
> 
> You can wrap that up into compound expressions and helpers any way you think
> helps readability.

Ugh, good thing you wrote a proper test case! I had missed to invert the
shouldExposeToBindings bit. Let's see if this works better.

There are so many ways of grouping the expressions and I've gone back and forth.
If you like any particular form, just say the word and I will make it so.

[philipj_slow, 2015-03-11 18:31:32]

Ping, the bots are happy now.

[philipj_slow, 2015-03-15 03:52:11]

rebase

[dominicc (has gone to gerrit), 2015-03-16 09:12:25]

On 2015/03/15 at 03:52:11, philipj wrote:
> rebase

Very neat. LGTM.

[dominicc (has gone to gerrit), 2015-03-16 09:12:29]

The CQ bit was checked by dominicc@chromium.org

[dominicc (has gone to gerrit), 2015-03-16 09:12:30]

lgtm

[commit-bot: I haz the power, 2015-03-16 09:12:38]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/994903003/40001

[philipj_slow, 2015-03-16 09:14:04]

On 2015/03/16 09:12:25, dominicc wrote:
> On 2015/03/15 at 03:52:11, philipj wrote:
> > rebase
> 
> Very neat. LGTM.

Thanks for the review!

[commit-bot: I haz the power, 2015-03-16 10:37:33]

Committed patchset #3 (id:40001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=191902