Added an automation IPC to set the parent of a window.

Updated the AutomationMsg_TabReposition IPC to be able to set the parent of a window.

Bug=1751431
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=13253

[amit, 2009-04-05 04:52:06]

http://codereview.chromium.org/62042/diff/1/6
File chrome/browser/automation/automation_provider.cc (right):

http://codereview.chromium.org/62042/diff/1/6#newcode2799
Line 2799: void AutomationProvider::OnTabSetParentWindow(int tab_handle,
This IPC and its processing is quite identical to OnTabReposition. So I am
wondering if it makes sense to handle it there by addibg the following to
IPC::Reposition_Params:

bool set_parent;
HWND parent_window;

Also, in theory we shouldn't need to pass window handle to the tab window from
outside, it should be easily available inside, but that's a different story.

[ananta, 2009-04-05 21:42:57]

http://codereview.chromium.org/62042/diff/1/6
File chrome/browser/automation/automation_provider.cc (right):

http://codereview.chromium.org/62042/diff/1/6#newcode2799
Line 2799: void AutomationProvider::OnTabSetParentWindow(int tab_handle,
On 2009/04/05 04:52:06, amit wrote:
> This IPC and its processing is quite identical to OnTabReposition. So I am
> wondering if it makes sense to handle it there by addibg the following to
> IPC::Reposition_Params:
> 
> bool set_parent;
> HWND parent_window;
> 
> Also, in theory we shouldn't need to pass window handle to the tab window from
> outside, it should be easily available inside, but that's a different story.

Done.

[amit, 2009-04-05 22:23:53]

lgtm, with a nit. 

Changes to automation_messages_internal.h are not part of this?

http://codereview.chromium.org/62042/diff/17/2003
File chrome/browser/automation/automation_provider.cc (right):

http://codereview.chromium.org/62042/diff/17/2003#newcode2795
Line 2795: if (!IsWindow(params.parent_window))
Instead of aborting the whole operation, maybe just
DCHECK(IsWindow(params.parent.window) and then log last error if SetParent
fails.

[ananta, 2009-04-05 22:31:21]

On 2009/04/05 22:23:53, amit wrote:
> lgtm, with a nit. 
> 
> Changes to automation_messages_internal.h are not part of this?
> 
> http://codereview.chromium.org/62042/diff/17/2003
> File chrome/browser/automation/automation_provider.cc (right):
> 
> http://codereview.chromium.org/62042/diff/17/2003#newcode2795
> Line 2795: if (!IsWindow(params.parent_window))
> Instead of aborting the whole operation, maybe just
> DCHECK(IsWindow(params.parent.window) and then log last error if SetParent
> fails.

Done. The change is in automation_messages.h. Added to this CB.

[amit, 2009-04-06 16:14:27]

lgtm++