]> Pileus Git - ~andy/gtk/log
~andy/gtk
12 years agoAvoid dereferencing NULL pointer in gtk_label_drag_data_get()
Tristan Van Berkom [Fri, 26 Aug 2011 16:58:44 +0000 (18:58 +0200)]
Avoid dereferencing NULL pointer in gtk_label_drag_data_get()

It's possible the GtkLabel receives drag'n'drop related signals
if it was setup to receive them by a container or other external
code, just avoid dereferencing priv->select_info in this case.

12 years agotests: Add test for non-nullterminated gtk_css_provider_load_data()
Michael Terry [Fri, 26 Aug 2011 14:15:36 +0000 (16:15 +0200)]
tests: Add test for non-nullterminated gtk_css_provider_load_data()

https://bugzilla.gnome.org/show_bug.cgi?id=657359

12 years agostyleproperty: Improve error message
Benjamin Otte [Fri, 17 Jun 2011 05:41:37 +0000 (07:41 +0200)]
styleproperty: Improve error message

We do dsupport that feature still and don't want to scare people into
thinking everything broke.

12 years agoAPI: cssprovider: Change parsing-error signal
Benjamin Otte [Fri, 17 Jun 2011 05:39:11 +0000 (07:39 +0200)]
API: cssprovider: Change parsing-error signal

Instead of path, line and position, emit the section the error happened
in. This has a lot more information to consume.

12 years agocssparser: Fix cases where we could overrun the terminating 0
Benjamin Otte [Fri, 17 Jun 2011 05:55:17 +0000 (07:55 +0200)]
cssparser: Fix cases where we could overrun the terminating 0

12 years agocssparser: Make lines and positions 0-indexed
Benjamin Otte [Fri, 17 Jun 2011 05:21:46 +0000 (07:21 +0200)]
cssparser: Make lines and positions 0-indexed

12 years agocssprovider: Store sections with parsed values
Benjamin Otte [Fri, 17 Jun 2011 03:58:04 +0000 (05:58 +0200)]
cssprovider: Store sections with parsed values

Also create a separate section for values to associate the values with.
Finally, use this section information when delayed-parsing a widget
style property.

12 years agocssprovider: Handle non-NULL-terminated data
Benjamin Otte [Fri, 17 Jun 2011 03:15:48 +0000 (05:15 +0200)]
cssprovider: Handle non-NULL-terminated data

Do a dumb copy if that happens.

12 years agocssprovider: Keep only the curent section while parsing
Benjamin Otte [Fri, 17 Jun 2011 02:52:43 +0000 (04:52 +0200)]
cssprovider: Keep only the curent section while parsing

We don't need a list of sections, the section can give us the parent
when we need it (d'oh).

12 years agocssprovider: Remove dead code
Benjamin Otte [Fri, 17 Jun 2011 02:40:17 +0000 (04:40 +0200)]
cssprovider: Remove dead code

The code isn't used anymore since the introduction of rulesets.

12 years agoAPI: css: Add GtkCssSection
Benjamin Otte [Fri, 17 Jun 2011 00:02:24 +0000 (02:02 +0200)]
API: css: Add GtkCssSection

This struct keeps track of an area of text in a CSS file and uses it
when specifying information. Also, the cssprovider keeps track of
sections when parsing a file.

12 years agoSquash two compilation warnings about possibly-uninitialzed
Colin Walters [Thu, 25 Aug 2011 18:47:21 +0000 (14:47 -0400)]
Squash two compilation warnings about possibly-uninitialzed

12 years agoUpdated Russian translation
Yuri Myasoedov [Thu, 25 Aug 2011 18:05:07 +0000 (22:05 +0400)]
Updated Russian translation

12 years agoSignalify (already existing) GtkMenuShell.insert()
Ryan Lortie [Sat, 13 Aug 2011 14:09:28 +0000 (10:09 -0400)]
Signalify (already existing) GtkMenuShell.insert()

gtk_menu_shell_insert() is a virtual function that was being directly
invoked from the class vtable.

Turn it into a proper signal and emit it in the usual way.

https://bugzilla.gnome.org/show_bug.cgi?id=656565

12 years agogtk+.doap: Add some more info
Javier Jardón [Thu, 25 Aug 2011 11:19:47 +0000 (12:19 +0100)]
gtk+.doap: Add some more info

12 years agoUpdated Galician translations
Fran Dieguez [Wed, 24 Aug 2011 19:03:09 +0000 (21:03 +0200)]
Updated Galician translations

12 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 24 Aug 2011 17:58:42 +0000 (19:58 +0200)]
Updated Spanish translation

12 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 24 Aug 2011 17:58:14 +0000 (19:58 +0200)]
Updated Spanish translation

12 years agoUpdated Norwegian bokmål translation
Kjartan Maraas [Wed, 24 Aug 2011 16:54:00 +0000 (18:54 +0200)]
Updated Norwegian bokmål translation

12 years agoUpdate gtk/Makefile.am
Chun-wei Fan [Fri, 19 Aug 2011 07:39:24 +0000 (15:39 +0800)]
Update gtk/Makefile.am

Distribute fallback implementation source file for rint()/round()

12 years agoUse a weak ref in GtkAccelLabel
Matthias Clasen [Wed, 24 Aug 2011 12:16:03 +0000 (08:16 -0400)]
Use a weak ref in GtkAccelLabel

GtkAccelLabel was holding a strong reference on its parent,
which could, depending on the circumstances, lead to leaks.

https://bugzilla.gnome.org/show_bug.cgi?id=344858

12 years agoTrivial: whitespace fixes
Matthias Clasen [Wed, 24 Aug 2011 11:58:21 +0000 (07:58 -0400)]
Trivial: whitespace fixes

12 years agoTrivial: Silence a compiler warning
Matthias Clasen [Wed, 24 Aug 2011 11:56:13 +0000 (07:56 -0400)]
Trivial: Silence a compiler warning

12 years agoTrivial: Silence a compiler warning
Matthias Clasen [Wed, 24 Aug 2011 11:54:58 +0000 (07:54 -0400)]
Trivial: Silence a compiler warning

12 years agoGtkColorButton: Don't forget to close a cairo context
Matthias Clasen [Wed, 24 Aug 2011 11:46:35 +0000 (07:46 -0400)]
GtkColorButton: Don't forget to close a cairo context

Patch by Jerome Lambourg, bug 655392.

12 years agoUpdated Swedish translation
Daniel Nylander [Wed, 24 Aug 2011 07:22:23 +0000 (09:22 +0200)]
Updated Swedish translation

12 years agoGtkFontChooserDialog: Select when a list item is activated
Matthias Clasen [Wed, 24 Aug 2011 04:49:59 +0000 (00:49 -0400)]
GtkFontChooserDialog: Select when a list item is activated

This is the expected behavior.

12 years agoGtkFontChooser: Add a ::font-activated signal
Matthias Clasen [Wed, 24 Aug 2011 04:49:16 +0000 (00:49 -0400)]
GtkFontChooser: Add a ::font-activated signal

This signal gets emitted when a font in the list is activated.

12 years agoGtkAppChooserDialog: remove an unused variable
Matthias Clasen [Wed, 24 Aug 2011 04:32:20 +0000 (00:32 -0400)]
GtkAppChooserDialog: remove an unused variable

12 years agoGtkAppChooserDialog: Simplify the construction
Matthias Clasen [Wed, 24 Aug 2011 04:11:32 +0000 (00:11 -0400)]
GtkAppChooserDialog: Simplify the construction

We do the same in other places.

12 years agoGtkAppChooserButton: Make activating rows work as expected
Matthias Clasen [Wed, 24 Aug 2011 04:10:34 +0000 (00:10 -0400)]
GtkAppChooserButton: Make activating rows work as expected

When a row is activated in the 'Other...' dialog, we have
to make sure the dialog goes away.

12 years agoGtkSwitch: Return FALSE from the button press handler
Matthias Clasen [Wed, 24 Aug 2011 03:29:19 +0000 (23:29 -0400)]
GtkSwitch: Return FALSE from the button press handler

This makes the behavior identical to other widgets which
handle button presses, and it avoids problems when placing
switches into a windows main toolbar.

https://bugzilla.gnome.org/show_bug.cgi?id=656986

12 years ago[filechooser] bgo#349502: Context menu option to copy selected files to the clipboard
Arx Cruz [Thu, 4 Aug 2011 18:34:16 +0000 (15:34 -0300)]
[filechooser] bgo#349502: Context menu option to copy selected files to the clipboard

12 years agoUpdated POTFILES.in
Piotr Drąg [Tue, 23 Aug 2011 20:28:43 +0000 (22:28 +0200)]
Updated POTFILES.in

12 years agoAdd tests/*ui to POTFILES.skip
Gabor Kelemen [Tue, 23 Aug 2011 20:18:37 +0000 (22:18 +0200)]
Add tests/*ui to POTFILES.skip

12 years agoThere is no need for [type: ...] tags in POTFILES.skip
Gabor Kelemen [Tue, 23 Aug 2011 20:16:09 +0000 (22:16 +0200)]
There is no need for [type: ...] tags in POTFILES.skip

12 years agoFixed POTFILES.in (error in D-L)
Daniel Mustieles [Tue, 23 Aug 2011 20:04:04 +0000 (22:04 +0200)]
Fixed POTFILES.in (error in D-L)

12 years agoFixed POTFILES.in (error in D-L)
Daniel Mustieles [Tue, 23 Aug 2011 19:56:15 +0000 (21:56 +0200)]
Fixed POTFILES.in (error in D-L)

12 years agoFixed POTFILES.in (error in D-L)
Daniel Mustieles [Tue, 23 Aug 2011 19:55:04 +0000 (21:55 +0200)]
Fixed POTFILES.in (error in D-L)

12 years agoFixed bug #657146
Daniel Mustieles [Tue, 23 Aug 2011 19:52:16 +0000 (21:52 +0200)]
Fixed bug #657146

12 years agoUpdated POTFILES.in
Piotr Drąg [Tue, 23 Aug 2011 17:54:06 +0000 (19:54 +0200)]
Updated POTFILES.in

12 years agoUpdated Serbian translation
Мирослав Николић [Tue, 23 Aug 2011 17:15:48 +0000 (19:15 +0200)]
Updated Serbian translation

12 years agoChange 'Ok' to 'Select' in the color chooser dialog
Matthias Clasen [Tue, 23 Aug 2011 14:16:34 +0000 (10:16 -0400)]
Change 'Ok' to 'Select' in the color chooser dialog

We have meaningful labels for the Ok buttons in all other choosers,
so just use the same string we use in the app chooser and the
font chooser.

Also move the mnemonic from s to a in 'Saturation', to avoid a
mnemonic conflict. String change !

https://bugzilla.gnome.org/show_bug.cgi?id=657006

12 years agoAdd a mnemonic to the 'Select' button
Matthias Clasen [Tue, 23 Aug 2011 14:08:33 +0000 (10:08 -0400)]
Add a mnemonic to the 'Select' button

This makes the string the same as in the app chooser dialog.

12 years agoUpdated Gujarati Translations
Sweta Kothari [Tue, 23 Aug 2011 09:16:57 +0000 (14:46 +0530)]
Updated Gujarati Translations

12 years agoUpdated Indonesian translation
Andika Triwidada [Mon, 22 Aug 2011 22:32:59 +0000 (05:32 +0700)]
Updated Indonesian translation

12 years agowayland: Acquire cairo device during operations that need the GL context
Daniel Nicoara [Mon, 22 Aug 2011 20:31:27 +0000 (16:31 -0400)]
wayland: Acquire cairo device during operations that need the GL context

12 years agowayland: We need to require egl as well
Kristian Høgsberg [Mon, 22 Aug 2011 18:54:12 +0000 (14:54 -0400)]
wayland: We need to require egl as well

12 years agoFix typo in doc comment
Kristian Rietveld [Mon, 22 Aug 2011 19:26:44 +0000 (21:26 +0200)]
Fix typo in doc comment

12 years agoAdd forgotten calls to signal_monitor_assert_is_empty()
Kristian Rietveld [Sat, 20 Aug 2011 19:51:49 +0000 (21:51 +0200)]
Add forgotten calls to signal_monitor_assert_is_empty()

12 years agoFix some memory leaks in treemodel unit tests
Kristian Rietveld [Sat, 20 Aug 2011 19:51:14 +0000 (21:51 +0200)]
Fix some memory leaks in treemodel unit tests

12 years agotreemodelsort: eliminate unread variable assigments
Kristian Rietveld [Sat, 20 Aug 2011 09:16:29 +0000 (11:16 +0200)]
treemodelsort: eliminate unread variable assigments

Courtesy of scan-build/clang.

12 years agotreemodelfilter: eliminate unread variable assigments
Kristian Rietveld [Sat, 20 Aug 2011 09:13:54 +0000 (11:13 +0200)]
treemodelfilter: eliminate unread variable assigments

Courtesy of scan-build/clang.

12 years agoAdd some notes on how filter model testing can be made (even) more thorough
Kristian Rietveld [Sat, 20 Aug 2011 08:59:36 +0000 (10:59 +0200)]
Add some notes on how filter model testing can be made (even) more thorough

12 years agoUpdate filtermodel ref count unit tests to new cache depth
Kristian Rietveld [Sat, 20 Aug 2011 08:55:27 +0000 (10:55 +0200)]
Update filtermodel ref count unit tests to new cache depth

12 years agoFix bad typo in gtk_tree_model_filter_build_level() ...
Kristian Rietveld [Sat, 20 Aug 2011 08:54:49 +0000 (10:54 +0200)]
Fix bad typo in gtk_tree_model_filter_build_level() ...

12 years agogtktreemodelfilter: only maintain cache at 1 level depth instead of 2
Kristian Rietveld [Sat, 20 Aug 2011 08:54:00 +0000 (10:54 +0200)]
gtktreemodelfilter: only maintain cache at 1 level depth instead of 2

12 years agoExtend public and internal documentation about GtkTreeModelFilter
Kristian Rietveld [Sat, 20 Aug 2011 07:36:35 +0000 (09:36 +0200)]
Extend public and internal documentation about GtkTreeModelFilter

12 years agoExtend documentation about GtkTreeModelSort internals
Kristian Rietveld [Thu, 11 Aug 2011 10:23:52 +0000 (12:23 +0200)]
Extend documentation about GtkTreeModelSort internals

12 years agoGtkTreeModel: add more detailed documentation on reference counting
Kristian Rietveld [Thu, 11 Aug 2011 08:42:29 +0000 (10:42 +0200)]
GtkTreeModel: add more detailed documentation on reference counting

12 years agoReplace GArray with GSequence in GtkTreeModelSort
Kristian Rietveld [Wed, 10 Aug 2011 21:24:58 +0000 (23:24 +0200)]
Replace GArray with GSequence in GtkTreeModelSort

This is done in the same way as GArray was replaced with GSequence in
GtkTreeModelFilter, in a patch done by Xavier Claessens.

All sorting code has been reworked to use the GSequence sorting
and insert sort functions, instead of doing a lot on our own.

12 years agoAdd more unit tests for GtkTreeModelSort
Kristian Rietveld [Wed, 10 Aug 2011 21:12:08 +0000 (23:12 +0200)]
Add more unit tests for GtkTreeModelSort

To test ref counting during rows-reordered and the new_order arrays supplied
with rows-reordered signals.

12 years agoMove SignalMonitor to treemodel.[ch] so it can be used by other tests
Kristian Rietveld [Wed, 10 Aug 2011 20:51:48 +0000 (22:51 +0200)]
Move SignalMonitor to treemodel.[ch] so it can be used by other tests

12 years agoRevert usage of parent_elt_index back to parent_elt pointers
Kristian Rietveld [Sun, 7 Aug 2011 15:58:37 +0000 (17:58 +0200)]
Revert usage of parent_elt_index back to parent_elt pointers

Based on earlier patch for GtkTreeModelFilter by Xavier Claessens.

12 years agoReplace GArray with GSequence in GtkTreeModelFilter
Xavier Claessens [Sun, 7 Aug 2011 15:11:13 +0000 (17:11 +0200)]
Replace GArray with GSequence in GtkTreeModelFilter

Significantly improves performance when e.g. removing (filtering) a lot
of rows from the filter model.  Fixes bug 616871.

This commit includes changes by Kristian Rietveld to make the patch apply
on top of the treemodel-fix branch and pass all newly written unit tests.

12 years agoRevert usage of parent_elt_index back to parent_elt pointers
Xavier Claessens [Fri, 5 Aug 2011 15:40:04 +0000 (17:40 +0200)]
Revert usage of parent_elt_index back to parent_elt pointers

This was introduced in commit a59c39f3703e81f560aa946c25145413192d795b.
The revert is part of bug 616871.  Adaption to recent GtkTreeModelFilter
changes by Kristian Rietveld.

12 years agoDo not emit rows-reordered for levels with no visible rows
Kristian Rietveld [Sun, 24 Jul 2011 16:52:59 +0000 (18:52 +0200)]
Do not emit rows-reordered for levels with no visible rows

12 years agoRemove left-over dump call
Kristian Rietveld [Sun, 24 Jul 2011 16:49:27 +0000 (18:49 +0200)]
Remove left-over dump call

12 years agoCorrect ref transfer in GtkTreeModelFilter rows-reordered handling
Kristian Rietveld [Sun, 24 Jul 2011 16:46:46 +0000 (18:46 +0200)]
Correct ref transfer in GtkTreeModelFilter rows-reordered handling

12 years agoTest GtkTreeModelFilter's reference transfers
Kristian Rietveld [Sun, 24 Jul 2011 16:44:25 +0000 (18:44 +0200)]
Test GtkTreeModelFilter's reference transfers

12 years agoUnit test GtkTreeModelFilter's rows-reordered emissions
Kristian Rietveld [Sun, 24 Jul 2011 16:43:10 +0000 (18:43 +0200)]
Unit test GtkTreeModelFilter's rows-reordered emissions

12 years agoUnit test rows reordered ref count handling in GtkTreeView
Kristian Rietveld [Mon, 18 Jul 2011 08:03:03 +0000 (10:03 +0200)]
Unit test rows reordered ref count handling in GtkTreeView

12 years agoShould validate TreeModelRefCount before the sort and filter model
Kristian Rietveld [Mon, 18 Jul 2011 07:41:34 +0000 (09:41 +0200)]
Should validate TreeModelRefCount before the sort and filter model

12 years agoAdd reference counting unit tests for GtkTreeModelFilter
Kristian Rietveld [Sun, 10 Jul 2011 19:56:43 +0000 (21:56 +0200)]
Add reference counting unit tests for GtkTreeModelFilter

12 years agoAdd more "has child" unit tests
Kristian Rietveld [Sun, 10 Jul 2011 19:53:25 +0000 (21:53 +0200)]
Add more "has child" unit tests

This allows for more thorough testing of "has child" filter functions.
We also test a has child filter function with a sort model as
child model, to verify that we receive enough signals to function
properly.

12 years agoAdd cleaned up and easier version of unit test for bug 311955
Kristian Rietveld [Sun, 10 Jul 2011 19:48:23 +0000 (21:48 +0200)]
Add cleaned up and easier version of unit test for bug 311955

12 years agoCheck level length in unit test for bug 311955
Kristian Rietveld [Sun, 10 Jul 2011 19:46:02 +0000 (21:46 +0200)]
Check level length in unit test for bug 311955

12 years agoRemove checks for superfluous signals which are not longer emitted
Kristian Rietveld [Sun, 10 Jul 2011 19:44:06 +0000 (21:44 +0200)]
Remove checks for superfluous signals which are not longer emitted

12 years agoCorrect/expand base tests to check emitted signals
Kristian Rietveld [Sun, 10 Jul 2011 19:34:38 +0000 (21:34 +0200)]
Correct/expand base tests to check emitted signals

The filter model is now more strict about the signals which are emitted,
so the base tests have been expanded to test with both the root level
nodes collapsed and expanded.

12 years agotreemodelrefcount: check_iter(): check if iter is valid
Kristian Rietveld [Sun, 10 Jul 2011 15:37:05 +0000 (17:37 +0200)]
treemodelrefcount: check_iter(): check if iter is valid

12 years agotreemodelrefcount: make it possible to assert within check functions
Kristian Rietveld [Sun, 10 Jul 2011 15:35:49 +0000 (17:35 +0200)]
treemodelrefcount: make it possible to assert within check functions

12 years agotreemodelfilter: rework reference counting
Kristian Rietveld [Sun, 10 Jul 2011 14:35:38 +0000 (16:35 +0200)]
treemodelfilter: rework reference counting

 - Before we kept a reference on all nodes in non-root levels.  This has
   been changed, now we keep a reference on the first node of each level.
   If, due to changes in the model, another node becomes the first node in
   the level, the reference is transferred to this new first node.
 - All non-root levels keep a reference on their parent.
 - By making use of the external ref count, the filter model now emits less
   unnecessary signals.
 - GtkTreeModelFilter does support filter functions which decide visibility
   of a given node based on the number of or visibility of children.
   To accomplish this, a child level of a node is cached when its
   parent has an external ref count > 0, because changes to the node might
   affect this parent.
 - An optimization for not building the root level in case the inserted
   node is not visible in gtk_tree_model_filter_row_inserted() has been
   removed.  In this case, we still need to build the root level and
   possibly a child level to monitor for signals which might make
   this row visible.

12 years agotreemodelfilter: handle another case in check_ancestors
Kristian Rietveld [Sun, 10 Jul 2011 14:24:05 +0000 (16:24 +0200)]
treemodelfilter: handle another case in check_ancestors

In gtk_tree_model_filter_check_ancestors(), also handle the case when
a node is already in the cache, but invisible, in the root level.
With the upcoming changes to GtkTreeModelFilter's ref counting this
case can occur.

12 years agotreemodelfilter: guard for building already existing levels
Kristian Rietveld [Sun, 10 Jul 2011 14:21:08 +0000 (16:21 +0200)]
treemodelfilter: guard for building already existing levels

12 years agotreemodelfilter: Introduce notion of external ref count
Kristian Rietveld [Sun, 26 Jun 2011 20:04:05 +0000 (22:04 +0200)]
treemodelfilter: Introduce notion of external ref count

We need to distinguish between the ref count objects have on us versus
the ref count we have on our child model.  To keep track of the former,
we introduce the "external ref count" in this commit.  The zero_ref_count
needs to be determined from the external ref count, because objects that
have a ref count on us have say in which levels must be cached and which
can be released.

Before the caching in GtkTreeModelFilter was essentially broken and
levels were never released.  This was caused because the zero_ref_count
was connected to the ref count the filter model had on its child model.
Now that this depends on the external ref count, this is working fine and
as to be expected.

12 years agotreemodelfilter: Move more missions of row-deleted down
Kristian Rietveld [Sun, 26 Jun 2011 19:44:13 +0000 (21:44 +0200)]
treemodelfilter: Move more missions of row-deleted down

The row-deleted signal should be emitted after the internal data
structures have been updated.  In gtk_tree_model_filter_remove_elt_from_level
and gtk_tree_model_filter_virtual_root_deleted the signal was still being
emitted before the updates were carried out.

12 years agoUnit test reference counting in GtkTreeModelSort
Kristian Rietveld [Thu, 2 Jun 2011 20:03:44 +0000 (22:03 +0200)]
Unit test reference counting in GtkTreeModelSort

12 years agoAvoid unreferencing deleted nodes
Kristian Rietveld [Thu, 2 Jun 2011 20:01:45 +0000 (22:01 +0200)]
Avoid unreferencing deleted nodes

Now that we call unref_node in free_level, we have to take care that
free_level may only unref (parent) nodes when these still exist in the
child model.  After row-deleted has been received for a node, its
children may no longer unref this node.

12 years agoBug 611922 - gtk_tree_model_sort_ref_node() is too slow
Kristian Rietveld [Thu, 2 Jun 2011 19:51:51 +0000 (21:51 +0200)]
Bug 611922 - gtk_tree_model_sort_ref_node() is too slow

Referencing a parent node for each referenced node is overdone.  Instead,
we now reference the parent from build_level and unreference in free_level.
Each level keeps a single reference on its immediate parent.  This both
alleviates the performence problems and should perfectly serve the purpose.

12 years agoMove asserts to gtktreemodelrefcount.h for other unit tests to use
Kristian Rietveld [Thu, 2 Jun 2011 14:20:08 +0000 (16:20 +0200)]
Move asserts to gtktreemodelrefcount.h for other unit tests to use

12 years agoAdd generic GtkTreeModel reference counting unit tests
Kristian Rietveld [Sat, 28 May 2011 14:48:15 +0000 (16:48 +0200)]
Add generic GtkTreeModel reference counting unit tests

12 years agoAdd GtkTreeModelRefCount, only to be used in unit tests
Kristian Rietveld [Sat, 28 May 2011 14:41:49 +0000 (16:41 +0200)]
Add GtkTreeModelRefCount, only to be used in unit tests

12 years agotreemodelsort: Remove unnecessary comment
Kristian Rietveld [Tue, 24 May 2011 20:19:38 +0000 (22:19 +0200)]
treemodelsort: Remove unnecessary comment

(I am no longer 17 years old and have grown up by now).

12 years agoAdd some comments to gtk_tree_model_filter_remove_elt_from_level
Kristian Rietveld [Sun, 22 May 2011 16:40:21 +0000 (18:40 +0200)]
Add some comments to gtk_tree_model_filter_remove_elt_from_level

12 years agoFix a couple of memleaks in the filter model tests
Kristian Rietveld [Sun, 22 May 2011 16:26:37 +0000 (18:26 +0200)]
Fix a couple of memleaks in the filter model tests

(Not all of them yet, need to do the remainder on Linux since valgrind
on Mac OS X is a bit too noisy).

12 years agoBug 621076 - GtkTreeModelFilter does not emit all signals in some ...
Kristian Rietveld [Sat, 21 May 2011 09:45:08 +0000 (11:45 +0200)]
Bug 621076 - GtkTreeModelFilter does not emit all signals in some ...

The bulk of the fix is to walk the chain of ancestors, starting at the
root level, and check if the visibility of any of the ancestors has
changed.  If yes, the necessary signals are emitted so that this change
is propagated properly.  This walk is done after a node has been
inserted, changed or deleted, see function
gtk_tree_model_filter_check_ancestors().

Bug reported, and initial debugging and analysis, by Xavier Claessens.

12 years agoDecrease code duplication in _remove_elt_from_level take two
Kristian Rietveld [Fri, 20 May 2011 18:43:48 +0000 (20:43 +0200)]
Decrease code duplication in _remove_elt_from_level take two

12 years agoDecrease amount of code duplication in _remove_elt_from_level
Kristian Rietveld [Fri, 20 May 2011 18:41:50 +0000 (20:41 +0200)]
Decrease amount of code duplication in _remove_elt_from_level

12 years agoRename gtk_tree_model_filter_remove_node for internal consistency
Kristian Rietveld [Fri, 20 May 2011 18:39:12 +0000 (20:39 +0200)]
Rename gtk_tree_model_filter_remove_node for internal consistency

12 years agoChange signature of gtk_tree_model_filter_remove_node
Kristian Rietveld [Fri, 20 May 2011 18:33:30 +0000 (20:33 +0200)]
Change signature of gtk_tree_model_filter_remove_node