Initialize ExtensionAccessibilityEventRouter earlier.

Initialize ExtensionAccessibilityEventRouter earlier.

On Chrome OS it needs to be initialized during the OOBE, and on other
platforms there's no harm in allowing it to initialize itself in its
constructor rather than waiting until a profile is loaded.

BUG=97614
TEST=manual testing

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=102422

[Yoyo Zhou, 2011-09-22 20:24:25]

LGTM

The C++ style guide advises against doing work in constructors, but this is
definitely not the only place I've seen them registering for notifications.

[dmazzoni, 2011-09-22 20:27:28]

On Thu, Sep 22, 2011 at 1:24 PM, <yoz@chromium.org> wrote:

> The C++ style guide advises against doing work in constructors, but this is
> definitely not the only place I've seen them registering for notifications.
>

One alternative would be if ProfileImpl::InitExtensions called an init
method on ExtensionAccessibilityEventRouter explicitly - would you prefer
that?

Otherwise, I think doing work in this constructor is harmless.

- Dominic

[Yoyo Zhou, 2011-09-22 20:29:14]

On 2011/09/22 20:27:28, Dominic Mazzoni wrote:
> On Thu, Sep 22, 2011 at 1:24 PM, <mailto:yoz@chromium.org> wrote:
> 
> > The C++ style guide advises against doing work in constructors, but this is
> > definitely not the only place I've seen them registering for notifications.
> >
> 
> One alternative would be if ProfileImpl::InitExtensions called an init
> method on ExtensionAccessibilityEventRouter explicitly - would you prefer
> that?
> 
> Otherwise, I think doing work in this constructor is harmless.
> 
> - Dominic

I think it's fine as you have it.

[commit-bot: I haz the power, 2011-09-22 23:23:49]

CQ is trying tha patch. Follow status at
https://chromium-status.appspot.com/cq/dmazzoni@chromium.org/8002001/1

[commit-bot: I haz the power, 2011-09-23 01:54:15]

Change committed as 102422