Tor Lillqvist [Wed, 8 Sep 2010 17:33:30 +0000 (20:33 +0300)]
Get rid of the rest of variables exported from libgtk
Especially the gtk_*_type ones in gtktexttypes.h were mentioned in
gtk.symbols presumably by accident. That header isn't even installed,
so no way can they be supposed to be public.
gtk_text_attr_appearance_type is from the installed but "semi-private"
gtktextlayout.h, so drop that one too from gtk.symbols for now.
The use of gtk_text_unknown_char_utf8 is bit of a mess. Code in a few
files knew implicitly that it is three bytes. Define a symbolic name
for the length of it instead. Add an exported function
gtk_text_unknown_char_utf8_gtk_tests_only() that returns a pointer to
it just for the sake of gtk/tests/textbuffer.c. Prefix the variable
with an underscore.
I doubt the usefulness of the test_utf8() in textbuffer.c. If it could
be dropped, gtk_text_unknown_char_utf8_gtk_tests_only() could be
dropped, too.
Tor Lillqvist [Wed, 8 Sep 2010 15:17:57 +0000 (18:17 +0300)]
Replace gtk_debug_flags with getter and setter functions
Preferrably should be made just into a local variable for libgtk like
_gdk_debug_flags for libgdk. But for now used by
gtk/tests/textbuffer.c and modules/printbackends/cups/gtkprintbackendcups.c.
Tor Lillqvist [Wed, 8 Sep 2010 14:36:10 +0000 (17:36 +0300)]
Turn the gtk version and age variables into functions
Having variables exported from a DLL is slightly painful and
potentially error-prone on Windows, so let's try get rid of them now
when we can. Starting with these.
Completely removed requisition cache from GtkWidget instance structure.
Since we have a new mechanism for requesting sizes: GtkSizeRequestIface;
it makes no sense to maintain this cache on the GtkWidget structure...
removing the requisition cache however does not break the old "size-request"
signal which is there for backwards compatability reasons.
In any case widget->requisition should not have been accessed,
gtk_widget_get_child_requisition() would have been the correct way
to consult the cache.
This commit also deprecates the newly added gtk_widget_get_requisition()
API and makes it fallback on gtk_size_request_get_size().
Tor Lillqvist [Wed, 8 Sep 2010 12:55:45 +0000 (15:55 +0300)]
Remove variables from the platform-independent GDK API
gdk_threads_mutex, gdk_threads_lock and gdk_threads_unlock are removed
from public API. gdk_threads_mutex was deprecated already. Instead of
using gdk_threads_lock and _unlock one was presumably supposed to use
the GDK_THREADS_ENTER and _LEAVE macros, which now simply call the
corresponding gdk_threads_enter() and _leave() functions.
Remove he dllimport/dllexport ugliness for GDK for Windows.
There is still a gdk_display variable being exported by the X11
backend.
add gtk_radio_button_join_group method for bindings
* this mirrors the committed change for gtk_radio_action_join_group in
commit 85b53969b24d318b219663841e678943516f443a
* Due to object ownership issues it is impossible to correctly use
get_group/set_group from a GI binding
* join_group is safer because at the binding level it works with individual
GtkRadioButton objects and not with the list of objects that gets
modified in the library
These members of GtkWidgetAuxData can no longer be set after
the removal of deprecated gtk_widget_set_uposition(), now removed
these members completely and removed the access to them from
gtk_widget_size_allocate().
_gtk_window_reposition() was only some extra confusing code, internally
it only calls the public API gtk_window_move()... replaced calls to
_gtk_window_reposition() with direct calls to gtk_window_move() and
removed the redundant internal API (probably there only for some historical
reason).
Tor Lillqvist [Thu, 2 Sep 2010 21:13:36 +0000 (00:13 +0300)]
Just use TrackMouseEvent directly
TrackMouseEvent is present in user32.dll in all Windows versions we
support. No need to look it up dynamically. No need to fallback to
_TrackMouseEvent from comctrl32.dll.
Enhanced/Simplified GtkWrapBox api as per Havoc's comments.
Made an enum GtkWrapBoxPacking for the expand/fill horizontal/vertical
boolean options... changed xpadding/ypadding to be horizontal-padding
and vertical-padding for a more consistent api and better readablility.
Fixed GtkNotebook to not consult GtkWidget->requisition directly
GtkNotebook was filling widget->requisition directly at "size-request"
time instead of filling in the *requsition argument, also (more importantly)
at size_allocate time GtkNotebook was consulting the action widget's
widget->requisition directly instead of safely calling
gtk_widget_get_child_requisition(). This commit closes bgo #628068.
Rename the ::orientation property of GtkIconView to item-orientation,
which is both clearer and avoids the conflict with the orientation
property of GtkOrientable implementations.
This was a style property to let theme engines 'opt-in' to more
correct behaviour while maintaining compatibility with existing
themes. GTK+ 3 engines are expected to handle the more correct
behaviour.
This was a style property to let theme engines 'opt-in' to more
correct behaviour while maintaining compatibility with existing
themes. GTK+ 3 engines are expected to handle the more correct
behaviour.
This was a style property to let theme engines 'opt-in' to more
correct behaviour while maintaining compatibility with existing
themes. GTK+ 3 engines are expected to handle the more correct
behaviour.
This was a style property to let theme engines 'opt-in' to more
correct behaviour while maintaining compatibility with existing
themes. GTK+ 3 engines are expected to handle the more correct
behaviour.
This was a style property to let theme engines 'opt-in' to more
correct behaviour while maintaining compatibility with existing
themes. GTK+ 3 engines are expected to handle the more correct
behaviour.