[pygtk] insert_at_cursor signal - cannot connect?

[Doug Quale]

Christian Robottom Reis <kiko at async.com.br> writes:

> I've grepped through the GTK+ source and I don't see where the
> insert_at_cursor signal could be emitted, but I don't claim expert
> status there either. delete_from_cursor itself is emitted by GDK.Delete,
> visibly in gtktextview.c:
> 
>       /* Deleting text */
>         gtk_binding_entry_add_signal (binding_set, GDK_Delete, 0,
>                                       "delete_from_cursor", 2,
>                                       G_TYPE_ENUM, GTK_DELETE_CHARS,
>                                       G_TYPE_INT, 1);
> 
> with similar lines for KP_Delete and Backspace. So the question is: who,
> if anybody, emits insert_at_cursor? Or is this simply meant to be
> emitted from userspace when convenient?

Hmmm.  Maybe that's right.  It looks like "delete_from_cursor" is only
emitted by a few keybindings.  Perhaps "insert_at_cursor" is available
for symmetry and for use by a keybinding, but none of the default
keybindings for TextView use it.

I was experimenting with deleting text directly from the TextView
TextBuffer and the "delete_at_cursor" signal is not emitted when the
TextBuffer.delete() method is used.  This means it's possible to
delete text from the TextView without triggering the
"delete_at_cursor" signal, although probably not from the keyboard.  I
think the "insert_at_cursor" and "delete_from_cursor" TextView signals
aren't really suitable for what the original poster wants to do.  It's
probably better to look at the underlying TextBuffer directly.  In
fact, I think he hinted that all he needs to know is if the buffer is
empty.  The TextBuffer "changed" signal would be the easiest one to
use for that.