Re: [gtk-list] Re: Patches for gtkwidget.[ch]

On Thu, 2 Mar 2000, Kent Schumacher wrote:

> Tim Janik wrote:

> > second, we don't need to integrate this into gtk proper, just create
> > your signal with gtk_object_class_user_signal_new() (omitting the
> > default hanlder portion) without modifying gtkwidget internals, to get
> > the same effect.
> I'm not sure I understand how gtk_object_class_user_signal_new() works.
> My guess is that an object needs to be created (i.e. gtk_button_new())
> before a gtk_object_class_user_signal_new() can be used.  If this is the
> case, I feel this is too unwieldy.  My application has several hundred
> ..._new() calls for focusable widget.  The only commonality is that they
> all inherit from GtkWidget.

nope, the prefix here is gtk_object_class

> I know it is possible for me to go and chase down every ..._new() call and
> add the signal, but I have a hard time understanding how this is more
> desirable than adding an additional
> Note that I feel there needs to be some necessary glue in gtk to make
> it easy to ubiquitously trigger events on a widget regardless of whether
> the source is menu-item, toolbar-button, keypress, voice-activation,
> bar-code reader, dart-gun, whatever.

there is glue code in place to create user defined signals:
gtk_object_class_user_signal_new ().
that should be enough to allow you to emit self defined signals
and thus trigger your callbacks (rewriting your patch):

static guint signal_user_action = 0;

signal_user_action =
  gtk_object_class_user_signal_new (gtk_type_class (GTK_TYPE_WIDGET),
                    GTK_RUN_LAST | GTK_RUN_ACTION,
                    GTK_TYPE_NONE, 1,
gtk_widget_user_action (GtkWidget *widget,
                        gpointer   act_data)
  g_return_if_fail (GTK_IS_WIDGET (widget));

  gtk_signal_emit (GTK_OBJECT (widget), signal_user_action, act_data);

> Thanks for the quick response,
> Kent


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]