A tricky fix for Issue 1845 (Take 2)....

A tricky fix for Issue 1845 (Take 2).

This is almost the same change as <http://codereview.chromium.org/39252/show>;, which caused a build break on a Linux buildbot while compiling my new template function in "chrome/common/render_messages.h".
Even though I was not able to reproduce the build errors on my Linux box, I removed this function and use the int type in my IPC message 'ViewMsg_SetTextDirection'.

BUG=1845
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=12434

[jeremy, 2009-03-23 16:20:44]

(Adding deanm as a reviewer since this CL adds an ipc message type)

http://codereview.chromium.org/42495/diff/44/58
File chrome/renderer/render_widget.cc (right):

http://codereview.chromium.org/42495/diff/44/58#newcode642
Line 642: return;
Since this is coming over IPC, could you verify the input and return if
direction isn't a known value.

[Hironori Bono, 2009-03-24 09:48:38]

Thank you for your review and comment.
I updated this change to apply a comment from jeremy.

http://codereview.chromium.org/42495/diff/44/58
File chrome/renderer/render_widget.cc (right):

http://codereview.chromium.org/42495/diff/44/58#newcode642
Line 642: return;
On 2009/03/23 16:20:44, jeremy wrote:
> Since this is coming over IPC, could you verify the input and return if
> direction isn't a known value.

Done. Thank you for noticing this. You are definitely right. I forgot adding
code which verifies its input values when changing this IPC parameter to int.

[jeremy, 2009-03-24 15:48:09]

LGTM with 1 comment

http://codereview.chromium.org/42495/diff/2001/2015
File chrome/renderer/render_widget.cc (right):

http://codereview.chromium.org/42495/diff/2001/2015#newcode649
Line 649: }
How about an else ASSERT(FALSE) here so that if we unintentionally pass a bad
value here we can get a hang of it quickly.

[Dean McNamee, 2009-03-24 15:50:39]

http://codereview.chromium.org/42495/diff/2001/2015
File chrome/renderer/render_widget.cc (right):

http://codereview.chromium.org/42495/diff/2001/2015#newcode649
Line 649: }
You mean
} else {
  NOTREACHED();
}

On 2009/03/24 15:48:10, jeremy wrote:
> How about an else ASSERT(FALSE) here so that if we unintentionally pass a bad
> value here we can get a hang of it quickly.

[brettw, 2009-03-24 16:31:35]

LGTM with the other comments fixed.

http://codereview.chromium.org/42495/diff/2001/2009
File chrome/browser/renderer_host/render_widget_host.h (right):

http://codereview.chromium.org/42495/diff/2001/2009#newcode225
Line 225: //     UpdateTextDirectioN(direction);
Here and once below: lower-case the N :)