Re: Partial patch for bug 55767

[Owen Taylor <otaylor redhat com>]

Kristian Rietveld <kristian planet nl> writes:

> Hi all,
> 
> I worked on a patch for bug 55767 ("Getters for all setters"). More than
> 50% of the missings getters in GTK+ have been added. Attached are:
> 
> 1. 'added-getters-Jun15'. This is a list with all missing getters. Every
> function with 'DONE' before it has been added. Every function with 'XXXX'
> before it won't be added (with a short explanation behind the function).
> And all functions between square brackets aren't implemented yet, because
> I don't know what to do with them. For some I don't know if we want them
> and/or I don't know on which way we want them implemented (some functions
> require new signals and stuff).
> 
> 2. 'missing-getters-Jun15.diff.gz', the actual patch. I created it with
> cvs diff (my first cvs diff, so I hope I did it the right way ;). It
> compiles on my machine without problems. As you'll understand, I didn't
> test every added function seperately ;).
> 
> I hope somebody is able to comment on those functions between square
> brackets and of course on the patch itself.

Here's comments on square brackets (and a few others) in your list.
Comments on the patch will follow.

> DONE gtk_cell_renderer_text_get_fixed_size

API here isn't (I think) finalized, Jonathan should check this.

> DONE gtk_check_menu_item_get_active

> [gtk_clipboard_get_with_data]
> [gtk_clipboard_get_with_owner]
> [gtk_clipboard_get_text]

These don't make sense.

> [gtk_dialog_get_response_sensitive]

Just a convenience function, not needed. (And doesn't really
make sense.)

> [gtk_dialog_get_default_response]

Doesn't make sense, again a convenience function.

> [gtk_drag_dest_get_proxy]

If we wanted to add something for this, it would need to
be gtk_drag_dest_get(), that handled both gtk_drag_dest_set(),
and gtk_drag_dest_set_proxy().

But I'd just skip it - gtk_drag_dest_set_proxy() is really
specialized.

> [gtk_drag_source_get_icon]
> [gtk_drag_get_icon_widget]
> [gtk_drag_get_icon_pixmap]
> [gtk_drag_get_default_icon]

I have some API changes oustanding here - I'll look at adding
getters at the same time. (See #56263 that I just filed)

> DONE gtk_entry_get_editable

Deprecated in favor of gtk_editable_get_editable()

> gtk_entry_get_position [this one is static in gtkentry.c, should we make it
> 			public?]

gtk_entry_set_position() is deprecated in favor of 
gtk_editable_set_position()

> [gtk_im_context_get_client_window]
> [gtk_im_context_get_cursor_location]
> [gtk_im_context_get_use_preedit]

I think we can avoid adding these - I can't see where they would
be useful and would require adding extra stuff to the GtkIMContext
vtable.

> DONE gtk_label_get_attributes
> [gtk_label_get_markup]
> [gtk_label_get_markup_with_mnemonic]
> [gtk_label_get_text_with_mnemonic]
> [gtk_label_get_pattern]

There is a basic problem here that GtkLabel conceptually works the way
that the properties are set up:

 label: The text of the label
 use_markup: Whether the label contains markup
 use_underline: Whether the label contains an underline