Treat inert nodes as aria-hidden instead of aria-disabled.

Treat inert nodes as aria-hidden instead of aria-disabled.

As per the spec change: http://html5.org/r/8209

BUG=304779
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=161015

[falken, 2013-10-30 09:31:24]

Hi Dominic, could you review this patch?

It's unclear to me if there are caching problems. I have to test end-to-end how
AT reacts when a modal dialog opens. I was using ChromeVox and
chrome://accessibility to test. But it turns out ChromeVox just looks at DOM, so
unless we mutate 'aria-hidden' directly it won't understand inertness (I think
we shouldn't do that, and hope ChromeVox can learn about inertness some other
way). And in chrome://accessibility, I must click "Refresh accessibility tree"
to refresh.

[dmazzoni, 2013-10-30 23:32:22]

lgtm with nits

There is still a potential caching problem, but let's address that in a separate
change. I'll file a bug with details.

https://codereview.chromium.org/51743003/diff/30001/LayoutTests/accessibility...
File LayoutTests/accessibility/inert-node-is-hidden.html (right):

https://codereview.chromium.org/51743003/diff/30001/LayoutTests/accessibility...
LayoutTests/accessibility/inert-node-is-hidden.html:31:
shouldBeTrue("treeContainsText(webArea, 'This content should not be exposed to
AT when a modal dialog is open')");
A simpler way to test this is accessibilityController.accessibleElementById(id)

https://codereview.chromium.org/51743003/diff/30001/Source/core/accessibility...
File Source/core/accessibility/AXObject.h (right):

https://codereview.chromium.org/51743003/diff/30001/Source/core/accessibility...
Source/core/accessibility/AXObject.h:558: bool ariaIsHidden() const;
With the name ariaIsHidden, this should be grouped under "ARIA attributes",
above.

But I'm thinking that it's misleading to call it ariaIsHidden.

The Blink style is to be clear, even if it's slightly more verbose. How about
isInertOrAriaHidden, and group it under the "Whether objects are ignored"
section.

[falken, 2013-10-31 02:27:40]

Thanks for the review.

https://codereview.chromium.org/51743003/diff/30001/LayoutTests/accessibility...
File LayoutTests/accessibility/inert-node-is-hidden.html (right):

https://codereview.chromium.org/51743003/diff/30001/LayoutTests/accessibility...
LayoutTests/accessibility/inert-node-is-hidden.html:31:
shouldBeTrue("treeContainsText(webArea, 'This content should not be exposed to
AT when a modal dialog is open')");
On 2013/10/30 23:32:22, Dominic Mazzoni wrote:
> A simpler way to test this is
accessibilityController.accessibleElementById(id)

I wanted to ensure it works for an anon child of body, and it looked a bit
tricky to navigate exactly to the text using childAtIndex (it's two levels deep,
and its position would change when the dialog is opened). So I just did a brute
force search of the tree.

https://codereview.chromium.org/51743003/diff/30001/Source/core/accessibility...
File Source/core/accessibility/AXObject.h (right):

https://codereview.chromium.org/51743003/diff/30001/Source/core/accessibility...
Source/core/accessibility/AXObject.h:558: bool ariaIsHidden() const;
On 2013/10/30 23:32:22, Dominic Mazzoni wrote:
> With the name ariaIsHidden, this should be grouped under "ARIA attributes",
> above.
> 
> But I'm thinking that it's misleading to call it ariaIsHidden.
> 
> The Blink style is to be clear, even if it's slightly more verbose. How about
> isInertOrAriaHidden, and group it under the "Whether objects are ignored"
> section.

Done.

[commit-bot: I haz the power, 2013-10-31 02:28:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/falken@chromium.org/51743003/150001

[commit-bot: I haz the power, 2013-10-31 03:56:35]

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, 2013-10-31 04:16:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/falken@chromium.org/51743003/150001

[commit-bot: I haz the power, 2013-10-31 04:17:33]

Change committed as 161015