Keep track of content-type change after focus-in

Keep track of content-type change after focus-in

This is an amendment to 6ca5ddb3.  As content-type (input-purpose and hints)
can change after focus-in, we should monitor the changes.

BUG=
R=takao.fujiwara1@gmail.com

Committed: ca128d1

[Daiki Ueno, 2014-04-03 08:44:25]

Tested with:
http://paste.fedoraproject.org/91217/39651452/

[Peng, 2014-04-03 20:44:45]

+takao.fujiwara1

[fujiwara, 2014-04-04 03:40:20]

Looks good while the signal is not called with the current gnome-shell lock
sreen in my test.

https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c
File bus/inputcontext.c (right):

https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c#newcode1165
bus/inputcontext.c:1165: g_signal_emit (context,
I have been thinking if that signal is emitted only when context->has_focus is
TRUE.

https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c
File bus/panelproxy.c (right):

https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c#newcode661
bus/panelproxy.c:661: { "set-content-type",           G_CALLBACK
(_context_set_content_type_cb) },
I've noticed you sometimes trimmed that last comma.
You like to do it here?

[Daiki Ueno, 2014-04-04 06:46:50]

https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c
File bus/inputcontext.c (right):

https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c#newcode1165
bus/inputcontext.c:1165: g_signal_emit (context,
On 2014/04/04 03:40:20, fujiwara wrote:
> I have been thinking if that signal is emitted only when context->has_focus is
> TRUE.

Good point, done.

https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c
File bus/panelproxy.c (right):

https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c#newcode661
bus/panelproxy.c:661: { "set-content-type",           G_CALLBACK
(_context_set_content_type_cb) },
On 2014/04/04 03:40:20, fujiwara wrote:
> I've noticed you sometimes trimmed that last comma.
> You like to do it here?

Not quite sure, but I remember I intentionally did that once in Vala files,
since older Vala compiler didn't support it.  However, in C, it has been
supported since C99, so I followed the original coding style here.

[fujiwara, 2014-04-04 07:22:37]

lgtm

On 2014/04/04 06:46:50, Daiki Ueno wrote:
> https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c
> File bus/inputcontext.c (right):
> 
> https://codereview.appspot.com/83920043/diff/1/bus/inputcontext.c#newcode1165
> bus/inputcontext.c:1165: g_signal_emit (context,
> On 2014/04/04 03:40:20, fujiwara wrote:
> > I have been thinking if that signal is emitted only when context->has_focus
is
> > TRUE.
> 
> Good point, done.
> 
> https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c
> File bus/panelproxy.c (right):
> 
> https://codereview.appspot.com/83920043/diff/1/bus/panelproxy.c#newcode661
> bus/panelproxy.c:661: { "set-content-type",           G_CALLBACK
> (_context_set_content_type_cb) },
> On 2014/04/04 03:40:20, fujiwara wrote:
> > I've noticed you sometimes trimmed that last comma.
> > You like to do it here?
> 
> Not quite sure, but I remember I intentionally did that once in Vala files,
> since older Vala compiler didn't support it.  However, in C, it has been
> supported since C99, so I followed the original coding style here.

[Peng, 2014-04-04 15:43:52]

lgtm

https://codereview.appspot.com/83920043/diff/20001/bus/inputcontext.c
File bus/inputcontext.c (right):

https://codereview.appspot.com/83920043/diff/20001/bus/inputcontext.c#newcode...
bus/inputcontext.c:1165: if (context->has_focus)
Add {} when there is more than one line.

[Daiki Ueno, 2014-04-07 09:06:19]

Committed patchset #2 manually as rca128d1 (presubmit successful).