Add accelerator table entries for hardware back/forward/refresh/favorites keys on linux aura

Add accelerator table entries for VKEY_BROWSER_(BACK|FORWARD) on linux

This registers accelerator table entries for VKEY_BROWSER_(BACK|FORWARD)
to IDC_(BACK|FORWARD) for linux platforms (linux desktop and ChromeOS) so
that hardward back/forward keys navigate the browser.

BUG=338945
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=247853

[jamesr, 2014-01-29 01:54:38]

On windows these buttons generate some other sort of event (I forget which) that
navigates.  If we have both the accelerator table entries and that handling, we
go back/forward twice.

I wonder if this makes this code dead:

https://code.google.com/p/chromium/codesearch#chromium/src/content/shell/brow...

[Elliot Glaysher, 2014-01-29 02:15:18]

On 2014/01/29 01:54:38, jamesr wrote:
> On windows these buttons generate some other sort of event (I forget which)
that
> navigates.  If we have both the accelerator table entries and that handling,
we
> go back/forward twice.
> 
> I wonder if this makes this code dead:
> 
>
https://code.google.com/p/chromium/codesearch#chromium/src/content/shell/brow...

lgtm as long as these aren't the mouse buttons. (Those are handled in the path
from (desktop_)root_window_host_x11.cc to desktop_user_action_handler_aura.cc.

[jamesr, 2014-01-29 03:41:15]

I have a mouse with physical back/forward buttons at home, I can try it out with
this patch tomorrow just to make sure things still behave properly.  I only
expect this patch to change handling of keyboard events.

[sky, 2014-01-29 16:45:39]

https://codereview.chromium.org/149273002/diff/1/chrome/browser/ui/views/acce...
File chrome/browser/ui/views/accelerator_table.cc (right):

https://codereview.chromium.org/149273002/diff/1/chrome/browser/ui/views/acce...
chrome/browser/ui/views/accelerator_table.cc:149: { ui::VKEY_BROWSER_HOME,
ui::EF_NONE, IDC_HOME },
Are any of these other ones useful on non-chromeos?

[jamesr, 2014-01-29 20:56:54]

The home button on my keyboard does something useful in gtk+ and does something
useful in linux-aura only if I move its entry out of the OS_CHROMEOS block.

I'd really like to find a list somewhere of the accelerator keys that did work
in linux-gtk so I can just preserve the existing behavior, but I can't figure
out where such a table might have existed.  Short of that, I can try testing and
mapping all the keys that exist on my keyboard.

[jamesr, 2014-01-29 21:04:45]

This adds an entry for the HOME key, which is the only other key in this table
that I have on my physical keyboard (and thus can test).  The browser search key
on my keyboard behaves the same way in linux-aura as linux-gtk (focuses the
omnibox and replaces the text with a "?"), but I'm not sure by what mechanism
that is.  I don't have physical browser stop or refresh buttons to test.

[sky, 2014-01-29 21:47:02]

LGTM

[jamesr, 2014-01-29 22:35:59]

I found a keyboard with more keys and updated the accelerator table entries for
the 'refresh' button to make it work like it did in linux/gtk and add an entry
for the 'favorites' button to make it open/close the bookmark bar.  This
behavior is different from ChromeOS, where it's mapped to open the bookmark
manager.

The keyboard I have also has a 'stop' button, but it doesn't seem to generate
the expected keycode.  It's spitting out 0xFF69, which according to the X.org
header I could find is 'News'.  I've left it alone.

[sky, 2014-01-29 22:40:19]

SLGTM

[commit-bot: I haz the power, 2014-01-29 22:44:28]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/149273002/40001

[commit-bot: I haz the power, 2014-01-30 03:04:37]

CQ bit was unchecked on CL. Ignoring.

[commit-bot: I haz the power, 2014-01-30 03:13:09]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/149273002/40001

[commit-bot: I haz the power, 2014-01-30 06:25:47]

Change committed as 247853