GitLab

Implement all border-radisu properties as specified by
http://www.w3.org/TR/css3-background/#the-border-radius
The end goal here is to get joined buttons to really look joined.

We will want to print doubles more often, so add a little helper
function for it. We cannot use printf() here, because we want English
locale output.

We don't want to use it for GPtrArray.

It's not used yet, but it's now possible to specify a custom read or
write func.

Rename them to GtkStyleParseFunc and GtkStylePrintFunc, so we can export
them.

Just shuffles code around for parsing properties.

... and take an optional style property as argument. This way, we can
allow custom parse functions for properties. The style property needs to
be optional so that we can use it for widget style properties, too.

I want to add parse_func as something different soon.

Name it _gtk_style_property_print_value() and actually pass it the style
property. This way, we can later change it to use custom print functions
for different style properties.

It's preferrable to print to an existing GString instead of returning a
char* everywhere that needs to be freed later.

Include the child widget path in the returned path now. This allows
customizing the path of the current widgets - like adding flags to child
widgets (and maybe siblings in the future).

Includes a bunch of header shuffling so we can use GtkWidget* in
gtkwidgetpath.h

This code should be more obvious: The check only needs to check for
scrollable widgets, and only needs to compare to tab_width.

As a side note, for vis_pages == 1, tab_width will be smaller than
tab_max + stuff, so this code will not behave differently.

We don't want tabs to resize when a 2nd page is added, just because the
scroll arrows are now visible. And we do want the arrow size to be
included.

Note: Previosuly this code was never run, as the check
  requisition->width < tab_width
was only run when tab_width wasn't computed, so was still equal to 0
(same for the height). So what this patch essentially does it add the
functionality of the if switch.

For shift-clicks inside the selection, shrink the selection.
https://bugzilla.gnome.org/show_bug.cgi?id=353712

To avoid quadratic behaviour when building up
paths by repeated appending indices. Bug 634491.

Patch by Vincent Untz,
https://bugzilla.gnome.org/show_bug.cgi?id=642768

https://bugzilla.gnome.org/show_bug.cgi?id=646859

John Lindgren points out in bug 650202 that using judicious
use of g_key_file_has_key() can save some cycles in application
startup.

As pointed out by John Lindgren in bug 650114, GtkEntry forgets
to disconnect an idle source when the completion is set to NULL.

https://bugzilla.gnome.org/show_bug.cgi?id=650167

Patch by John Lindgren,
https://bugzilla.gnome.org/show_bug.cgi?id=650110

Replacing the function gtk_misc_set_alignment () with
gtk_widget_set_halign () and gtk_widget_set_valign () in
gtk/gtkcolorsel.c

https://bugzilla.gnome.org/show_bug.cgi?id=650762

Replacing the function gtk_misc_set_alignment () with
gtk_widget_set_halign () and gtk_widget_set_valign () in
gtk/gtknotebook.c

https://bugzilla.gnome.org/show_bug.cgi?id=650762

Replacing the function gtk_misc_set_alignment () with
gtk_widget_set_halign () and gtk_widget_set_valign () in
gtk/gtkcustompaperunixdialog.c

https://bugzilla.gnome.org/show_bug.cgi?id=650762

Replacing the function gtk_misc_set_alignment () with
gtk_widget_set_halign () and gtk_widget_set_valign () in
gtk/gtkfilechooserbutton.c

https://bugzilla.gnome.org/show_bug.cgi?id=650762

Add a cross-reference to GContentType.
https://bugzilla.gnome.org/show_bug.cgi?id=650012

This is useful to let unrelated parts of an application (or
plugins) learn about the windows of an application. Based
on a patch by Matt Barnes,
https://bugzilla.gnome.org/show_bug.cgi?id=641087

Based on a patch by Matt Barnes,
https://bugzilla.gnome.org/show_bug.cgi?id=624539

Use the GtkStyleContext accessors for boxed properties where possible,
to reduce allocations.

Use the GtkStyleContext accessors for boxed properties where possible,
to reduce allocations.

The call to gtk_border_free() within unpack_border() felt completely
in the wrong place, as the border actually pertains to the GValue
being unpacked. Plus, the GValue itself was also being leaked.