platform/upstream/at-spi2-core.git
10 years agoRelease v2.11.5 AT_SPI2_CORE_2_11_5
Mike Gorse [Mon, 3 Feb 2014 23:54:05 +0000 (17:54 -0600)]
Release v2.11.5

10 years agoAllow caching of data sent with events and requesting such data
Mike Gorse [Sat, 1 Feb 2014 23:39:34 +0000 (17:39 -0600)]
Allow caching of data sent with events and requesting such data

Add atspi_event_listener_register_full, to request that particular data
be sent with an event, and cache the data for the duration of the event
callback.

This also extends the D-Bus protocol, adding an a{sv} to an event to send
this data along, but we will still read events without this array added for
backwards compatibility (and compatibilllity with QT-AT-SPI, which does not
send this data for now).

https://bugzilla.gnome.org/show_bug.cgi?id=708695

10 years agoFix crash when get_process_id called with an app no longer on the bus
Mike Gorse [Thu, 16 Jan 2014 19:46:12 +0000 (13:46 -0600)]
Fix crash when get_process_id called with an app no longer on the bus

10 years agoFix testregistry.py
Mike Gorse [Mon, 23 Dec 2013 20:32:58 +0000 (15:32 -0500)]
Fix testregistry.py

testregistry.py was not kept up-to-date with various AT-SPI changes. Update
to at least run and give output similar to its former output.

10 years agoRelease v2.11.3 AT_SPI2_CORE_2_11_3
Mike Gorse [Mon, 16 Dec 2013 21:34:09 +0000 (16:34 -0500)]
Release v2.11.3

10 years agodocument: add support to current page and page count
Alejandro Piñeiro [Thu, 28 Nov 2013 18:28:09 +0000 (19:28 +0100)]
document: add support to current page and page count

https://bugzilla.gnome.org/show_bug.cgi?id=719508

10 years agoFix atspi_text_get_bounded_ranges
Mike Gorse [Wed, 20 Nov 2013 20:44:45 +0000 (14:44 -0600)]
Fix atspi_text_get_bounded_ranges

10 years agoPost-release version bump
Mike Gorse [Wed, 20 Nov 2013 20:44:38 +0000 (14:44 -0600)]
Post-release version bump

10 years agoRelease v2.11.2 AT_SPI2_CORE_2_11_2
Mike Gorse [Tue, 19 Nov 2013 00:07:01 +0000 (18:07 -0600)]
Release v2.11.2

10 years agoBump version
Mike Gorse [Mon, 18 Nov 2013 22:31:20 +0000 (16:31 -0600)]
Bump version

10 years agostates: add CHECKABLE and HAS_POPUP states
Alejandro Piñeiro [Mon, 18 Nov 2013 16:24:53 +0000 (17:24 +0100)]
states: add CHECKABLE and HAS_POPUP states

https://bugzilla.gnome.org/show_bug.cgi?id=712609

10 years agoroles: new roles for ARIA roles
Alejandro Piñeiro [Mon, 18 Nov 2013 16:20:18 +0000 (17:20 +0100)]
roles: new roles for ARIA roles

https://bugzilla.gnome.org/show_bug.cgi?id=712608

10 years agoroles: add TITLE_BAR, BLOCK_QUOTE, VIDEO and AUDIO roles
Alejandro Piñeiro [Wed, 23 Oct 2013 17:24:31 +0000 (19:24 +0200)]
roles: add TITLE_BAR, BLOCK_QUOTE, VIDEO and AUDIO roles

https://bugzilla.gnome.org/show_bug.cgi?id=710464

10 years agoRelease v2.11.1 AT_SPI2_CORE_2_11_1
Mike Gorse [Mon, 28 Oct 2013 22:18:57 +0000 (17:18 -0500)]
Release v2.11.1

10 years agoFix atspi_component_contains
Mike Gorse [Wed, 23 Oct 2013 20:47:17 +0000 (15:47 -0500)]
Fix atspi_component_contains

Per the spec, org.a11y.Atspi.Component.Contains takes a uint32 for the coord
type, not an int16

https://bugzilla.gnome.org/show_bug.cgi?id=710730

10 years agoCheck returned error before using when registering with session manager
Mike Gorse [Wed, 23 Oct 2013 15:14:23 +0000 (10:14 -0500)]
Check returned error before using when registering with session manager

Apparently, the call to register with gnome-session-manager can fail without
setting a GError, and, in any case, it doesn't hurt to check that the error
is set before using it.

https://bugzilla.gnome.org/show_bug.cgi?id=710689

10 years agoCast to correct type when calling g_setenv
Mike Gorse [Tue, 15 Oct 2013 22:42:26 +0000 (17:42 -0500)]
Cast to correct type when calling g_setenv

10 years agoUnref process_deferred_messages_source when clearing it
Mike Gorse [Wed, 9 Oct 2013 21:22:17 +0000 (16:22 -0500)]
Unref process_deferred_messages_source when clearing it

Since g_source_attach adds a ref to a source and we keep the source stored
in a global variable and, therefore, do not unref it after attaching it, we
thus need to unref it when it is being destroyed and we clear the global
variable.

https://bugzilla.gnome.org/show_bug.cgi?id=709625

10 years agoFix a few more warnings
Mike Gorse [Thu, 10 Oct 2013 20:22:31 +0000 (15:22 -0500)]
Fix a few more warnings

10 years agoSkip some non-introspectable functions
Mike Gorse [Thu, 10 Oct 2013 19:44:47 +0000 (14:44 -0500)]
Skip some non-introspectable functions

libdbus is not introspectable, so functions taking or returning a
DBusConnection cannot be bound anyhow.

10 years agoRemove calls to g_type_init()
Mike Gorse [Thu, 10 Oct 2013 19:36:56 +0000 (14:36 -0500)]
Remove calls to g_type_init()

g_type_init() has been deprecated since 2.36.

10 years agoDefine ATSPI_CACHE_DEFAULT on one line
Mike Gorse [Thu, 10 Oct 2013 19:33:46 +0000 (14:33 -0500)]
Define ATSPI_CACHE_DEFAULT on one line

glib-mkenums seems to be happier this way.

10 years agoFix some warnings and documentation issues
Mike Gorse [Thu, 10 Oct 2013 18:48:04 +0000 (13:48 -0500)]
Fix some warnings and documentation issues

10 years agoMake AtspiEventListenerCb not call for a const datum
Mike Gorse [Thu, 10 Oct 2013 18:31:31 +0000 (13:31 -0500)]
Make AtspiEventListenerCb not call for a const datum

AtspiEventListenerCb is currently defined as (transfer
full), so the caller owns the struct and needs to free it, so it should not be const.

10 years agoAdd prototype for atspi_device_event_get_type
Mike Gorse [Tue, 8 Oct 2013 22:59:57 +0000 (17:59 -0500)]
Add prototype for atspi_device_event_get_type

10 years agoRelease as 2.10.0 AT_SPI2_CORE_2_10_0
Mike Gorse [Mon, 23 Sep 2013 21:13:54 +0000 (16:13 -0500)]
Release as 2.10.0

10 years agoRelease v2.9.92 AT_SPI2_CORE_2_9_92
Mike Gorse [Mon, 16 Sep 2013 21:33:46 +0000 (16:33 -0500)]
Release v2.9.92

10 years agoDocument that multiple attribute values can now be listed in match rules.
Mike Gorse [Sun, 15 Sep 2013 17:04:22 +0000 (12:04 -0500)]
Document that multiple attribute values can now be listed in match rules.

https://bugzilla.gnome.org/show_bug.cgi?id=700865

10 years agodbind: remove call to atspi_dbus_connection_setup_with_g_main
Mike Gorse [Thu, 5 Sep 2013 23:58:44 +0000 (18:58 -0500)]
dbind: remove call to atspi_dbus_connection_setup_with_g_main

dbind isn't currently running a GMainLoop, so it shouldn't call
atspi_dbus_connection_setup_with_g_main, and calling this is harmful if
it is being called from a context other than the default context, since
D-Bus messages will no longer be delivered.

https://bugzilla.gnome.org/show_bug.cgi?id=707218

10 years agomisc: Simplify code flow in process_deferred_messages
Jasper St. Pierre [Thu, 5 Sep 2013 13:57:39 +0000 (09:57 -0400)]
misc: Simplify code flow in process_deferred_messages

Rather than use complicated gboolean-cast-to-gpointer to make
decisions, simply use two different callbacks.

10 years agomisc: Fix bad type
Jasper St. Pierre [Thu, 5 Sep 2013 13:53:57 +0000 (09:53 -0400)]
misc: Fix bad type

10 years agoMove the process_deferred_messages idle cb when switching main contexts
Mike Gorse [Thu, 5 Sep 2013 02:51:05 +0000 (21:51 -0500)]
Move the process_deferred_messages idle cb when switching main contexts

If at-spi2-atk enters a new main context and calls
atspi_set_main_context and process_deferred_messages has already been
scheduled for the original context, then it would not be called until
the original main loop is re-entered, meaning that Orca will sometimes
deadlock in its preferences dialog, since its key listener won't get the
notification for its own key event, as it will be deferred, while
at-spi2-atk waits for the listener to be processed.

https://bugzilla.gnome.org/show_bug.cgi?id=707218

10 years agoRelease v2.9.91 AT_SPI2_CORE_2_9_91
Mike Gorse [Mon, 2 Sep 2013 19:07:39 +0000 (14:07 -0500)]
Release v2.9.91

10 years agoInclude config.h in private header
Mike Gorse [Mon, 2 Sep 2013 17:32:15 +0000 (12:32 -0500)]
Include config.h in private header

atspi-misc.c was not including config.h anywhere, resulting in HAVE_X11
not being defined, so X would not be queried when fetching the
accessibility bus address.

https://bugzilla.gnome.org/show_bug.cgi?id=707084

10 years agoDon't dispatch while calling get_accessibility_bus_address
Mike Gorse [Mon, 2 Sep 2013 17:31:38 +0000 (12:31 -0500)]
Don't dispatch while calling get_accessibility_bus_address

Calling dbus_connection_dispatch isn't really needed when fetching the
bus address--this is only done once.

10 years agoDon't call dbus_connection_dispatch recursively
Mike Gorse [Mon, 2 Sep 2013 17:28:27 +0000 (12:28 -0500)]
Don't call dbus_connection_dispatch recursively

dbus_connection_dispatch is trying to lock the connection when it is
already locked, resulting in a deadlock, so just don't call it
recursively for now.

10 years agoUpdated Irish translation
Seán de Búrca [Sat, 31 Aug 2013 08:40:20 +0000 (02:40 -0600)]
Updated Irish translation

10 years agoskipping atspi_event_listener_[register/deregister]_no_data
Alejandro Piñeiro [Thu, 29 Aug 2013 13:27:12 +0000 (15:27 +0200)]
skipping atspi_event_listener_[register/deregister]_no_data

https://bugzilla.gnome.org/show_bug.cgi?id=707010

10 years agoaccessible: remove an unneeded unref
Alejandro Piñeiro [Tue, 27 Aug 2013 14:00:00 +0000 (16:00 +0200)]
accessible: remove an unneeded unref

That unref is done inside an if that starts with a !parent condition, so
unrefing it doesn't makes sense

https://bugzilla.gnome.org/show_bug.cgi?id=706893

10 years agoAdded Irish translation
Seán de Búrca [Mon, 26 Aug 2013 09:20:17 +0000 (03:20 -0600)]
Added Irish translation

10 years agoDispatch dbus messages after method calls when no main loop running
Mike Gorse [Wed, 21 Aug 2013 19:33:18 +0000 (14:33 -0500)]
Dispatch dbus messages after method calls when no main loop running

If no main loop is running, then the dbus connection's incoming message
queue was never emptied, so with, eg, dogtail,
dbus_connection_send_with_reply_and_block would take longer and longer
to look through the message queue to find a reply. This also effectively
led to a memory leak.

https://bugzilla.gnome.org/show_bug.cgi?id=701145

10 years agoRelease v2.9.90 AT_SPI2_CORE_2_9_90
Mike Gorse [Mon, 19 Aug 2013 21:54:29 +0000 (16:54 -0500)]
Release v2.9.90

10 years agoAdd atspi_text_get_string_at_offset
Mike Gorse [Mon, 19 Aug 2013 21:43:02 +0000 (16:43 -0500)]
Add atspi_text_get_string_at_offset

10 years agoDo not query X unless DISPLAY or AT_SPI_DISPLAY are set
Mike Gorse [Thu, 8 Aug 2013 16:21:15 +0000 (11:21 -0500)]
Do not query X unless DISPLAY or AT_SPI_DISPLAY are set

Previously, AT-SPI would hard-code a default display name of ":0" if
neither DISPLAY nor AT_SPI_DISPLAY were set. This seems unnecessary--the
environment variable ought to be set--and causes a warning to be
generated whenever a Gtk app is started under Wayland.

https://bugzilla.gnome.org/show_bug.cgi?id=701378

10 years agoAllow building without X11
Mike Gorse [Wed, 7 Aug 2013 14:58:14 +0000 (09:58 -0500)]
Allow building without X11

Separate out X11-specific code in preparation for Wayland support and to
avoid calling X under Mac OS. It should now be possible to build without
X, although keyboard and mouse-related functionality will likely not
work.

https://bugzilla.gnome.org/show_bug.cgi?id=682199

10 years agoBump version to 2.9.5 (should've done this before)
Mike Gorse [Mon, 29 Jul 2013 22:42:43 +0000 (17:42 -0500)]
Bump version to 2.9.5 (should've done this before)

10 years agoRelease v2.9.5 AT_SPI2_CORE_2_9_5
Mike Gorse [Mon, 29 Jul 2013 22:03:28 +0000 (17:03 -0500)]
Release v2.9.5

10 years agoSkip atspi_(event|device)_listener_new_simple
Mike Gorse [Mon, 15 Jul 2013 18:17:18 +0000 (13:17 -0500)]
Skip atspi_(event|device)_listener_new_simple

These are intended as C convenience functions. Gjs doesn't support
having a GDestroyNotify without a user_data and prints out a warning, so
might as well skip for purposes of introspection.

10 years agoCancel pending calls when giving up on them
Mike Gorse [Wed, 10 Jul 2013 20:20:53 +0000 (15:20 -0500)]
Cancel pending calls when giving up on them

When timing out without having received a reply from a pending call,
then
cancel the call so that it will be removed from the DBusConnection's
queue. This will hopefully prevent the call and the
SpiReentrantCallClosure from being leaked.

Note that this code is still behaving oddly (why isn't a notification
being sent by libdbus when the specified amount of time has ellapsed?)

10 years agoFix memory leak in atspi_accessible_get_role_name
Mike Gorse [Wed, 10 Jul 2013 18:47:55 +0000 (13:47 -0500)]
Fix memory leak in atspi_accessible_get_role_name

Only dup the string once.

10 years agoRelease v2.9.4 AT_SPI2_CORE_2_9_4
Mike Gorse [Mon, 8 Jul 2013 21:31:29 +0000 (16:31 -0500)]
Release v2.9.4

10 years agoAdd atspi_set_main_context
Mike Gorse [Mon, 1 Jul 2013 19:32:23 +0000 (14:32 -0500)]
Add atspi_set_main_context

Add a function to allow at-spi2-atk to set the main context that is used
when adding idle handlers (ie, process_deferred_messages, which is used
to prevent new events from being sent to the AT while other events are
being processed).

Also bump version to 2.9.4.

10 years agoRelease v2.9.3 AT_SPI2_CORE_2_9_3
Mike Gorse [Mon, 17 Jun 2013 21:25:10 +0000 (16:25 -0500)]
Release v2.9.3

10 years agoAdd atspi_is_initialized
Mike Gorse [Sun, 16 Jun 2013 19:02:22 +0000 (14:02 -0500)]
Add atspi_is_initialized

Added a function to query whether libatspi has been initialized. Used by
at-spi2-atk. Also, bump version.

10 years agoRelease v2.9.2 AT_SPI2_CORE_2_9_2
Mike Gorse [Tue, 28 May 2013 15:32:08 +0000 (10:32 -0500)]
Release v2.9.2

10 years agoAdd template files to EXTRA_DIST
Mike Gorse [Tue, 28 May 2013 15:30:59 +0000 (10:30 -0500)]
Add template files to EXTRA_DIST

11 years agoReinstate atspi_role_get_name
Mike Gorse [Wed, 22 May 2013 20:38:15 +0000 (15:38 -0500)]
Reinstate atspi_role_get_name

Atspi_role_get_name is used (by Orca, for instance) and was
inadvertently removed with the enum refactor.

11 years agoRename various functions to avoid name collisions
Mike Gorse [Tue, 14 May 2013 15:53:45 +0000 (10:53 -0500)]
Rename various functions to avoid name collisions

Having functions with common suffixes (ie, atspi_text_get_attributes and
atspi_document_get_attributes) results in problems for language
bindings. Deprecating these functions in favor of functions with new
names to avoid these collisions.

https://bugzilla.gnome.org/show_bug.cgi?id=700243

11 years agoAnnotate filter to allow none for device event listeners
Mike Gorse [Tue, 14 May 2013 18:00:54 +0000 (13:00 -0500)]
Annotate filter to allow none for device event listeners

Atspi_register_device_event_listener and
atspi_deregister_device_event_listener take a filter (currently unused).
It can be NULL, so it should be annotated as such.

https://bugzilla.gnome.org/show_bug.cgi?id=700237

11 years agoneed to dup a role name when fetching it from a value
Mike Gorse [Tue, 14 May 2013 15:51:45 +0000 (10:51 -0500)]
need to dup a role name when fetching it from a value

11 years agobuild: Fix srcdir != builddir
Colin Walters [Sat, 11 May 2013 00:17:23 +0000 (20:17 -0400)]
build: Fix srcdir != builddir

11 years agoMerge branch 'enum-types'
Mike Gorse [Fri, 10 May 2013 23:56:32 +0000 (18:56 -0500)]
Merge branch 'enum-types'

Conflicts:
atspi/Makefile.am
atspi/atspi-accessible.c

11 years agoversioned-introspection.py: Define prototypes as extern
Mike Gorse [Wed, 20 Mar 2013 21:14:36 +0000 (16:14 -0500)]
versioned-introspection.py: Define prototypes as extern

11 years agoFix introspection for GetActions
Mike Gorse [Wed, 20 Mar 2013 21:07:30 +0000 (16:07 -0500)]
Fix introspection for GetActions

11 years agoAdded Friulian translation
Fabio Tomat [Tue, 26 Mar 2013 12:09:12 +0000 (13:09 +0100)]
Added Friulian translation

11 years agoAdded Friulian translation
Fabio Tomat [Tue, 26 Mar 2013 12:09:09 +0000 (13:09 +0100)]
Added Friulian translation

11 years agoRelease v2.8.0 AT_SPI2_CORE_2_8_0
Mike Gorse [Mon, 25 Mar 2013 22:08:09 +0000 (17:08 -0500)]
Release v2.8.0

11 years ago[l10n] Added Tadjik translation
Victor Ibragimov [Tue, 19 Mar 2013 21:26:32 +0000 (22:26 +0100)]
[l10n] Added Tadjik translation

11 years agoRelease v2.7.91 AT_SPI2_CORE_2_7_91
Mike Gorse [Mon, 4 Mar 2013 21:41:46 +0000 (15:41 -0600)]
Release v2.7.91

11 years agoRevert "Do not run at-spi-dbus-bus at initialization phase"
Mike Gorse [Mon, 4 Mar 2013 21:32:10 +0000 (15:32 -0600)]
Revert "Do not run at-spi-dbus-bus at initialization phase"

The bus launcher may be started after applications that use AT-SPI, such
as caribou, if it is started in the application phase. I'm unsure of the
rationale for the original commit. I'm also unsure if the .desktop file
is still needed, since dbus-daemon should automatically launch the bus
launcher on request, but would rather not eliminate it this close to
3.8.

https://bugzilla.gnome.org/show_bug.cgi?id=694984

This reverts commit 1089e7c3687a48bb7810c40d6c9ff37e7acd10a1.

11 years agoAdd atspi_accessible_get_object_locale
Mike Gorse [Tue, 26 Feb 2013 00:36:12 +0000 (18:36 -0600)]
Add atspi_accessible_get_object_locale

https://bugzilla.gnome.org/show_bug.cgi?id=694368

11 years agoRelease v2.7.5 AT_SPI2_CORE_2_7_90
Mike Gorse [Mon, 18 Feb 2013 22:41:54 +0000 (16:41 -0600)]
Release v2.7.5

11 years agoRemove DEBUG_REF_COUNTS define
Mike Gorse [Thu, 7 Feb 2013 15:06:05 +0000 (09:06 -0600)]
Remove DEBUG_REF_COUNTS define

DEBUG_REF_COUNTS is intended for debugging and should not be defined by
default.

11 years agoRelease v2.7.5 AT_SPI2_CORE_2_7_5
Mike Gorse [Mon, 4 Feb 2013 22:29:14 +0000 (16:29 -0600)]
Release v2.7.5

11 years agoMore README updates
Mike Gorse [Mon, 4 Feb 2013 19:48:56 +0000 (13:48 -0600)]
More README updates

The directory structure was not completely documented.

11 years agoUpdate README
Mike Gorse [Mon, 4 Feb 2013 19:27:10 +0000 (13:27 -0600)]
Update README

The README was very outdated.

11 years agoFix some warnings
Mike Gorse [Mon, 4 Feb 2013 18:53:46 +0000 (12:53 -0600)]
Fix some warnings

11 years agoBump version
Mike Gorse [Thu, 31 Jan 2013 22:20:24 +0000 (16:20 -0600)]
Bump version

11 years agoAdd ATSPI_ROLE_LEVEL_BAR
Mike Gorse [Thu, 31 Jan 2013 21:40:37 +0000 (15:40 -0600)]
Add ATSPI_ROLE_LEVEL_BAR

Add "level bar" role corresponding to the new atk role.

11 years agoFix various warnings
Mike Gorse [Wed, 16 Jan 2013 10:45:30 +0000 (04:45 -0600)]
Fix various warnings

11 years agoRelease v2.7.4.1
Mike Gorse [Tue, 15 Jan 2013 17:01:47 +0000 (11:01 -0600)]
Release v2.7.4.1

11 years agoCheck that message is non-NULL before calling libdbus
Mike Gorse [Tue, 15 Jan 2013 16:07:12 +0000 (10:07 -0600)]
Check that message is non-NULL before calling libdbus

_atspi_dbus_call_partial now checks for an error message, but it should
first check that the message is not NULL, or libdbus will abort.

11 years agoRelease v2.7.4 AT_SPI2_CORE_2_7_4
Mike Gorse [Mon, 14 Jan 2013 23:22:19 +0000 (17:22 -0600)]
Release v2.7.4

11 years agoFix ref count leaks in atspi_accessible_get_application
Mike Gorse [Mon, 14 Jan 2013 23:16:38 +0000 (17:16 -0600)]
Fix ref count leaks in atspi_accessible_get_application

11 years ago_atspi-dbus_call_partial: Check for error.
Mike Gorse [Mon, 14 Jan 2013 19:03:52 +0000 (13:03 -0600)]
_atspi-dbus_call_partial: Check for error.

Dbind may return a message containing an error, so check for this in
_atspi_dbus_call_partial. I an error is set, then do not return the
message, but set the GError. This fixes some warnings where the error
message was being passed to _atspi_dbus_return_accessible_from_message.

11 years agoFix build with picky linkers
Behdad Esfahbod [Mon, 31 Dec 2012 02:07:20 +0000 (20:07 -0600)]
Fix build with picky linkers

11 years agobuild: really fix srcdir != builddir
Giovanni Campagna [Fri, 28 Dec 2012 00:11:22 +0000 (01:11 +0100)]
build: really fix srcdir != builddir

The file we want is in srcdir, not builddir.

http://ostree.gnome.org/work/tasks/gnomeos-3.8-build/2012.2854/log

11 years agobuild: Fix srcdir != builddir
Colin Walters [Fri, 28 Dec 2012 00:05:01 +0000 (19:05 -0500)]
build: Fix srcdir != builddir

11 years agoFix some ref count issues
Mike Gorse [Thu, 27 Dec 2012 21:50:04 +0000 (16:50 -0500)]
Fix some ref count issues

Add refs in some places where they were previously not present, and be more
thorough about removing refs when disposing. This should fix some leaks,
though it is likely that many still remain.
Also add a test program that can be used to test for some leaks.

11 years agoRelease v2.7.3 AT_SPI2_CORE_2_7_3
Mike Gorse [Mon, 17 Dec 2012 21:58:45 +0000 (16:58 -0500)]
Release v2.7.3

11 years agoA few documentation fixes
Mike Gorse [Mon, 17 Dec 2012 21:18:23 +0000 (16:18 -0500)]
A few documentation fixes

11 years agoMake documentation for atspi_event_listener_new() appear.
Patrick Welche [Sat, 15 Dec 2012 19:48:27 +0000 (19:48 +0000)]
Make documentation for atspi_event_listener_new() appear.

11 years agoDocumentation fix
Patrick Welche [Thu, 13 Dec 2012 21:21:10 +0000 (21:21 +0000)]
Documentation fix

This change makes public that a return value of 1 from atspi_init()
means that atspi was already initialized, so testing for atspi_init()>1
makes sense as a check for errors.

11 years agoFix deregistering of applications
Mike Gorse [Fri, 7 Dec 2012 17:21:19 +0000 (11:21 -0600)]
Fix deregistering of applications

11 years agoFix another crash when removing hung processes
Mike Gorse [Fri, 7 Dec 2012 16:27:58 +0000 (10:27 -0600)]
Fix another crash when removing hung processes

There was another instance where we were freeing the wrong data.

11 years agoRemove the correct datum when clearing a hung process
Mike Gorse [Thu, 6 Dec 2012 15:50:18 +0000 (09:50 -0600)]
Remove the correct datum when clearing a hung process

When clearing hung processes, we were passing the wrong variable to
g_list_remove, meaning that the datum would never be removed from the
list, even though it had been freed, resulting in a crash.

11 years agoFix a crash when parsing events
Mike Gorse [Wed, 5 Dec 2012 05:54:55 +0000 (23:54 -0600)]
Fix a crash when parsing events

If an event listener specifies a detail for an event, but an event
arrives with an empty detail, then libatspi might crash.

11 years agobuild: Use gobject-introspection's Makefile instead of rolling our own
Rui Matos [Mon, 26 Nov 2012 21:45:25 +0000 (22:45 +0100)]
build: Use gobject-introspection's Makefile instead of rolling our own

Invoking the GI compiler through libtool would cause us to embed an
RPATH in our binaries.

https://bugzilla.gnome.org/show_bug.cgi?id=640303

11 years agoFix atspi_device_listener_new_simple and some other warnings
Mike Gorse [Mon, 26 Nov 2012 19:51:28 +0000 (13:51 -0600)]
Fix atspi_device_listener_new_simple and some other warnings

11 years agoRe-register keystroke listeners if the registry goes away and returns
Mike Gorse [Tue, 20 Nov 2012 22:29:50 +0000 (16:29 -0600)]
Re-register keystroke listeners if the registry goes away and returns