Re-enable WebView ChromeVox manual script injection when accessibility status changes.

Re-enable WebView ChromeVox  manual script injection when accessibility status changes.

This is a partial revert of
https://codereview.chromium.org/1140173003/

TEST=in certain contexts involving WebView, toggling on/off ChromeVox repeatedly does not break ChromeVox.
BUG=496664
Committed: https://crrev.com/d35baea6341e533124003af62a255868b1beb703
Cr-Commit-Position: refs/heads/master@{#333587}

[David Tseng, 2015-06-09 16:51:52]

dtseng@chromium.org changed reviewers:
+ dmazzoni@chromium.org

[dmazzoni, 2015-06-09 17:07:27]

Does this double-inject if ChromeVox is already running when you open a webview,
like in the new oobe? Should we limit this to whitelisted urls that are known to
be broken?

[David Tseng, 2015-06-09 18:38:18]

It shouldn't double inject in the scenario since the partial revert only
injects on toggling on ChromeVox. I am, however, seeing a crash in OOBE
when toggling on/off repeatedly.

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[David Tseng, 2015-06-09 18:45:47]

Patchset #3 (id:40001) has been deleted

[David Tseng, 2015-06-09 18:45:54]

I fixed the crash with the cl but it also uncovered an existing issue.
accessibility_manager.cc::113 was crashing because TabHelper is a nullptr.
This code was being triggered as part of dynamic injection via
ChromeWebViewGutestDelegate.

A preexisting issue is if you toggle off/on ChromeVox, this webview remains
silent. This is similar to the issue being fixed here, except in that
context, TabHelper is non-null.

On Tue, Jun 9, 2015 at 11:07 AM, David Tseng <dtseng@chromium.org> wrote:

> It shouldn't double inject in the scenario since the partial revert only
> injects on toggling on ChromeVox. I am, however, seeing a crash in OOBE
> when toggling on/off repeatedly.
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[dmazzoni, 2015-06-09 19:16:09]

ok, lgtm

Note to other reviewers: we hope to remove this code once again in M45, but I
don't see any other option for M44.

[David Tseng, 2015-06-09 19:21:21]

dtseng@chromium.org changed reviewers:
+ fsamuel@chromium.org

[David Tseng, 2015-06-09 19:21:22]

+ fsamuel for OWNERS.

[Fady Samuel, 2015-06-09 19:48:45]

lgtm

[David Tseng, 2015-06-09 20:51:17]

The CQ bit was checked by dtseng@chromium.org

[commit-bot: I haz the power, 2015-06-09 20:52:53]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1173523006/60001

[commit-bot: I haz the power, 2015-06-09 22:11:20]

Committed patchset #3 (id:60001)

[commit-bot: I haz the power, 2015-06-09 22:12:16]

Patchset 3 (id:??) landed as
https://crrev.com/d35baea6341e533124003af62a255868b1beb703
Cr-Commit-Position: refs/heads/master@{#333587}