Update the accessibility tree when a modal dialog is opened/closed.

Update the accessibility tree when a modal dialog is opened/closed.

This patch is an improved approach from r161595. It turns out it's not
sufficient to traverse the new AT tree calling
notifyIfIgnoreValueChanged on each node, since the browser-side can hold
on to descendents that have been pruned out of the tree and never get a
child changed notification for them.

This patch blows away the entire tree and forces it to rebuild when a
modal dialog is opened/closed. It's a big hammer but seems to be the
most foolproof approach. We considered approaches like traversing
m_objects calling notifyIfIgnoreValueChanged on each, but it's
conceivable that m_objects isn't up-to-date. Since a modal dialog likely
causes a massive change to the tree, starting over entirely isn't so
inefficient.

The test for this patch is Chrome-side:
https://codereview.chromium.org/64273003/

This patch causes inert-node-is-hidden.html to fail, because
AccessibilityController doesn't expect the AX cache to get destroyed
and recreated. Since the DumpAccessibilityTree tests on Chrome-side
cover the behavior and the plan is to have only DumpAccessibilityTree
tests in the future, I'm removing inert-node-is-hidden.html.

BUG=304779
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=161880

[falken, 2013-11-12 12:32:06]

Dominic, please take a look.

[dmazzoni, 2013-11-12 21:26:25]

lgtm

I like it! Should be much more foolproof.

[falken, 2013-11-13 03:42:24]

On 2013/11/12 21:26:25, Dominic Mazzoni wrote:
> lgtm
> 
> I like it! Should be much more foolproof.

Thanks! This patch causes the LayoutTest inert-node-is-hidden.html to fail,
because AccessibilityController doesn't expect the AX cache to get destroyed and
recreated. Since the DumpAccessibilityTree tests on Chrome-side cover the
behavior and (as I understood from our offline convo yesterday) the plan is to
have only DumpAccessibilityTree tests in the future, I'm removing
inert-node-is-hidden.html.

[commit-bot: I haz the power, 2013-11-13 03:56:16]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/falken@chromium.org/70213002/50001

[commit-bot: I haz the power, 2013-11-13 05:22:17]

Retried try job too often on linux_blink_rel for step(s) webkit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_blin...

[commit-bot: I haz the power, 2013-11-13 05:27:03]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/falken@chromium.org/70213002/50001

[commit-bot: I haz the power, 2013-11-13 07:05:34]

Change committed as 161880