platform/upstream/glib.git
10 years agogio/tests/socket: fix minor memleak
Tim-Philipp Müller [Fri, 13 Jun 2014 18:23:13 +0000 (19:23 +0100)]
gio/tests/socket: fix minor memleak

10 years agogparam: Remove unnecessary conditionals around preconditions
Colin Walters [Thu, 12 Jun 2014 17:37:16 +0000 (13:37 -0400)]
gparam: Remove unnecessary conditionals around preconditions

These cause a static analyzer to think we're trying to actually handle
them being NULL, which is not the case.  They both must not be NULL,
period.

No idea why the code was like this originally.

Reviewed by mclasen on IRC.

10 years agogbookmarkfile: Cleaner error handling code to pacify static analysis
Colin Walters [Thu, 12 Jun 2014 16:53:18 +0000 (12:53 -0400)]
gbookmarkfile: Cleaner error handling code to pacify static analysis

A static analyzer flagged the g_file_get_contents() call as not
checking its return value.  While the code here is actually correct,
it's verbose at best.

I think the "goto out + cleanup" code style is substantially cleaner,
less error prone, and easier to read.  It also will pacify the static
analyzer.

https://bugzilla.gnome.org/show_bug.cgi?id=731584

10 years agogparam: change value of G_PARAM_EXPLICIT_NOTIFY
Ryan Lortie [Fri, 6 Jun 2014 18:35:55 +0000 (14:35 -0400)]
gparam: change value of G_PARAM_EXPLICIT_NOTIFY

GParamSpec has a possibility of user-introduced flags, and we didn't
respect that with the addition of _EXPLICIT_NOTIFY.

Change the documentation for the maximum number of user flags to 10,
just to pick a somewhat random number.  The documentation here was never
correct anyway -- it previously claimed that as many as 38 flags were
possible.

Meanwhile, move G_PARAM_EXPLICIT_NOTIFY next to _DEPRECATED in order to
avoid conflicts with low-numbered user flags (which are in use by at
least evolution-data-server).

https://bugzilla.gnome.org/show_bug.cgi?id=731341

10 years agognextstepsettingsbackend: Use same token for function-call and its availability
Daniel Macks [Mon, 9 Jun 2014 19:17:19 +0000 (15:17 -0400)]
gnextstepsettingsbackend: Use same token for function-call and its availability

Availability of the g_nextstep_settings_backend_get_type() prototype
is controlled by HAVE_COCOA in gsettingsbackendinternal.h and the
actual implemenation by OS_COCOA in Makefile.am. Therefore, the
giomodule.c call to that function should also be protected by a COCOA
token rather than an CARBON token (cocoa and carbon are independent
autoconf tests).

https://bugzilla.gnome.org/show_bug.cgi?id=731425

10 years agoassert-msg-test: Explicitly quit gdb session to avoid zombie
Daniel Macks [Tue, 10 Jun 2014 10:00:33 +0000 (06:00 -0400)]
assert-msg-test: Explicitly quit gdb session to avoid zombie

gdb is run in batch mode, and can leave leave the program being
executed/debugged running when the batchfile is finished. Explicitly
"quit"ing the subprocess prevents it from leaving the stray subprocess
when gdb finishes.

https://bugzilla.gnome.org/show_bug.cgi?id=731366

10 years agogtype: remove interface-after-init exceptions
Ryan Lortie [Fri, 6 Jun 2014 20:51:40 +0000 (16:51 -0400)]
gtype: remove interface-after-init exceptions

A year ago, we tried to remove support for adding interfaces on
already-initialised types.  There were problems with the C++ and C#
bindings at the time, so we added exceptions to give them a bit more
time to catch up.

It's already one cycle after when these exceptions were planned to be
removed, so let's take them out now.

https://bugzilla.gnome.org/show_bug.cgi?id=697229

10 years agoMSVC 2010+ Projects: Update "Installation" Process
Chun-wei Fan [Mon, 9 Jun 2014 02:27:36 +0000 (10:27 +0800)]
MSVC 2010+ Projects: Update "Installation" Process

Currently, due to the way that Visual Studio 2010+ projects are handled,
the "install" project does not re-build upon changes to the sources, as it
does not believe that its dependencies have changed, although the changed
sources are automatically recompiled.  This means that if a part or more
of the solution does not build, or if the sources need some other fixes
or enhancements, the up-to-date build is not copied automatically, which
can be misleading.

Improve on the situation by forcing the "install" project to trigger its
rebuild, so that the updated binaries can be copied.  This does trigger an
MSBuild warning, but having that warning is way better than not having an
up-to-date build, especially during testing and development.

10 years agotrivial: docs: add some missing commas
Michael Catanzaro [Sat, 7 Jun 2014 00:04:12 +0000 (19:04 -0500)]
trivial: docs: add some missing commas

10 years agodocs: warn about installing properties after init
Ryan Lortie [Fri, 6 Jun 2014 20:42:23 +0000 (16:42 -0400)]
docs: warn about installing properties after init

Leave ourselves a little wiggle room: if people install properties after
initialisation then we reserve the right to handle that in a way that
may not be threadsafe.

https://bugzilla.gnome.org/show_bug.cgi?id=698614

10 years agoRevert "GObject: prevent installing properties after init"
Ryan Lortie [Fri, 6 Jun 2014 20:40:15 +0000 (16:40 -0400)]
Revert "GObject: prevent installing properties after init"

This reverts commit ddb0ce14215cd62c7a2497d6cf9f2ea63c40ebb5.

Conflicts:
gobject/gobject.c

https://bugzilla.gnome.org/show_bug.cgi?id=698614

10 years agogthread: use inline keyword for _get_impl() functions
Tim-Philipp Müller [Sat, 31 May 2014 13:54:08 +0000 (14:54 +0100)]
gthread: use inline keyword for _get_impl() functions

Give compiler a hint that these should be inlined,
which doesn't seem to happen by default with -O2.
Yields 5% speedup in artificial benchmarks, and
1% speedup in a real-world test case doing a lot
of mutex locking and unlocking.

https://bugzilla.gnome.org/show_bug.cgi?id=730807

10 years agogatomic: use GCC C11-style atomics, if available
Ryan Lortie [Tue, 27 May 2014 13:28:08 +0000 (15:28 +0200)]
gatomic: use GCC C11-style atomics, if available

GCC does not yet support ISO C11 atomic operations, but it has
compatible versions available as an extension.  Use these for load and
store if they are available in order to avoid emitting a hard fence
instruction (since in many cases, we do not need it -- on x86, for
example).

For now we use the fully seqentially-consistent memory model, since
these APIs are documented rather explicitly: "This call acts as a full
compiler and hardware memory barrier".

In the future we can consider introducing new APIs for the more relaxed
memory models, if they are available (or fall back to stricter ones
otherwise).

https://bugzilla.gnome.org/show_bug.cgi?id=730807

10 years agogatomic: whitespace fixes
Ryan Lortie [Tue, 27 May 2014 13:27:54 +0000 (15:27 +0200)]
gatomic: whitespace fixes

https://bugzilla.gnome.org/show_bug.cgi?id=730807

10 years agogatomic: fix typo in deprecation attribute
Ryan Lortie [Tue, 27 May 2014 13:25:37 +0000 (15:25 +0200)]
gatomic: fix typo in deprecation attribute

https://bugzilla.gnome.org/show_bug.cgi?id=730807

10 years agogtype: guard uses of new fundamental type check
Ryan Lortie [Fri, 6 Jun 2014 14:45:51 +0000 (10:45 -0400)]
gtype: guard uses of new fundamental type check

g_type_is_fundamentally_a (see bug 730984) is a new API/ABI and is
marked with a version macro.  We should therefore avoid its
unconditional use from G_IS_OBJECT() and G_IS_PARAM_SPEC() which are
APIs that have been around for a long time.

This prevents deprecation warnings from being emitted when these
functions are used with an older GLIB_VERSION_MAX_ALLOWED and also
prevents linking to the new ABI in that case (so that it's possible to
use the resulting binary with an older version of GLib).

https://bugzilla.gnome.org/show_bug.cgi?id=731335

10 years agotest: add test for explicit-notify properties
Ryan Lortie [Fri, 6 Jun 2014 14:25:16 +0000 (10:25 -0400)]
test: add test for explicit-notify properties

https://bugzilla.gnome.org/show_bug.cgi?id=731200

10 years agogobject: add 'explicit notify' GParamSpec flag
Ryan Lortie [Wed, 4 Jun 2014 12:59:50 +0000 (08:59 -0400)]
gobject: add 'explicit notify' GParamSpec flag

Add a flag to prevent the automatic emission of the "notify" signal
during g_object_set_property().

If this flag is set then the class must explicitly emit the notify
for themselves.  This is already standard practice on most classes, but
we cannot simply remove the existing behaviour because there are surely
many cases where it is needed.

https://bugzilla.gnome.org/show_bug.cgi?id=731200

10 years agoAdd a test for GMenuAttributeIter
Matthias Clasen [Thu, 5 Jun 2014 12:04:19 +0000 (08:04 -0400)]
Add a test for GMenuAttributeIter

10 years agovalgrind.h: Update to latest upstream version
David King [Fri, 16 May 2014 11:32:49 +0000 (12:32 +0100)]
valgrind.h: Update to latest upstream version

Avoid crashes in g_type_free_instance() on mingw64.

https://bugzilla.gnome.org/show_bug.cgi?id=730198

10 years agoSupport Building Using Visual Studio 2013
Chun-wei Fan [Tue, 3 Jun 2014 07:50:37 +0000 (15:50 +0800)]
Support Building Using Visual Studio 2013

Like the Visual Studio 2012 project files, the Visual Studio 2013 files are
largely the same as the Visual Studio 2010 project files, so support
Visual Studio 2013 by updating the autotools scripts that is used for
Visual Studio 2012.  This means that project files for Visual Studio 2012
and Visual Studio 2013 can be maintained by simply maintaining the Visual
Studio 2010 project files, adding minimal maintenance overhead.

10 years agodocs: Add missing language annotations
Matthias Clasen [Sun, 1 Jun 2014 13:38:49 +0000 (09:38 -0400)]
docs: Add missing language annotations

10 years agodocs: GType documentation cleanup
Matthias Clasen [Sun, 1 Jun 2014 00:55:00 +0000 (20:55 -0400)]
docs: GType documentation cleanup

Still some markup left.

10 years agodocs: Replace <function> XML tag with back ticks
Emmanuele Bassi [Sat, 31 May 2014 18:21:23 +0000 (19:21 +0100)]
docs: Replace <function> XML tag with back ticks

https://bugzilla.gnome.org/show_bug.cgi?id=731050

10 years agodocs: Replace <structname> tags with back ticks
Emmanuele Bassi [Sat, 31 May 2014 18:20:03 +0000 (19:20 +0100)]
docs: Replace <structname> tags with back ticks

https://bugzilla.gnome.org/show_bug.cgi?id=731050

10 years agodocs: Add an index for 2.40 api
Matthias Clasen [Sat, 31 May 2014 15:32:55 +0000 (11:32 -0400)]
docs: Add an index for 2.40 api

10 years agoMinor documentation additions and corrections
Matthias Clasen [Sat, 31 May 2014 13:04:11 +0000 (09:04 -0400)]
Minor documentation additions and corrections

Going for 100%.

10 years agodocs: Fill in the gspawn long description
Matthias Clasen [Sat, 31 May 2014 14:38:47 +0000 (10:38 -0400)]
docs: Fill in the gspawn long description

10 years agodocs: Fill in gshell long description
Matthias Clasen [Sat, 31 May 2014 14:18:07 +0000 (10:18 -0400)]
docs: Fill in gshell long description

10 years agoDocument g_signal_handlers_destroy
Matthias Clasen [Sat, 31 May 2014 13:02:27 +0000 (09:02 -0400)]
Document g_signal_handlers_destroy

10 years agodocs: Fix up varargs documentation
Matthias Clasen [Sat, 31 May 2014 12:55:21 +0000 (08:55 -0400)]
docs: Fix up varargs documentation

gtk-doc expects varargs to be documented in a specific way,
otherwise it complains.

10 years agodocs: Add some missing apis
Matthias Clasen [Sat, 31 May 2014 12:54:52 +0000 (08:54 -0400)]
docs: Add some missing apis

10 years agodocs: Add a 2.42 api index
Matthias Clasen [Sat, 31 May 2014 12:54:16 +0000 (08:54 -0400)]
docs: Add a 2.42 api index

10 years agoGParamSpec: Use new fundamental instance check
Edward Hervey [Fri, 30 May 2014 10:25:27 +0000 (12:25 +0200)]
GParamSpec: Use new fundamental instance check

https://bugzilla.gnome.org/show_bug.cgi?id=730984

10 years agogobject: Use fast fundamental instance type check
Edward Hervey [Wed, 28 May 2014 08:59:14 +0000 (10:59 +0200)]
gobject: Use fast fundamental instance type check

Speeds up g_object_ref/_unref by 50%-65% (i.e. takes 60-65% of the time
it used to take).

https://bugzilla.gnome.org/show_bug.cgi?id=730984

10 years agogtype: Add check for fundamental instance type
Edward Hervey [Wed, 28 May 2014 08:57:28 +0000 (10:57 +0200)]
gtype: Add check for fundamental instance type

When checking whether an instance is of a given fundamental type (such
as G_TYPE_OBJECT), we can avoid over 60%+ of the cost of checking types.

https://bugzilla.gnome.org/show_bug.cgi?id=730984

10 years agoFix the closure test in continuous
Matthias Clasen [Fri, 30 May 2014 14:21:43 +0000 (10:21 -0400)]
Fix the closure test in continuous

This test has the same problem as the mapping-test - it uses
SIGUSR1 without checking the signal mask. Apply the same fix
here.

10 years agoFix the mapping test in continuous
Matthias Clasen [Fri, 30 May 2014 14:18:36 +0000 (10:18 -0400)]
Fix the mapping test in continuous

It turns out that due to a recent gdm change, the inherited
signal mask has SIGUSR1 blocked - which is bad news for
tests using SIGUSR1. Fix the test by explicitly checking the
signal mask before using SIGUSR1.

10 years agogconvert: mention that the g_convert len should be in bytes
Thiago Santos [Thu, 29 May 2014 22:13:37 +0000 (19:13 -0300)]
gconvert: mention that the g_convert len should be in bytes

Some charsets have each char with more than one byte, make it clear that
the length should be in bytes

https://bugzilla.gnome.org/show_bug.cgi?id=730963

10 years agoAdd more debug spew to the mapping-test
Matthias Clasen [Wed, 28 May 2014 10:14:56 +0000 (06:14 -0400)]
Add more debug spew to the mapping-test

10 years agogvariant: Clarify type and format strings in the docs
Philip Withnall [Tue, 27 May 2014 14:50:00 +0000 (15:50 +0100)]
gvariant: Clarify type and format strings in the docs

'@' and '&' are only used in format strings as prefixes to type
strings and not to full format strings.

https://bugzilla.gnome.org/show_bug.cgi?id=729269

10 years agomapping-test: Add debug spew
Matthias Clasen [Wed, 28 May 2014 02:13:33 +0000 (22:13 -0400)]
mapping-test: Add debug spew

This might help figuring out why the test is failing in continuous.

10 years agogio/tests/network-address: fix when no network is available
Dan Winship [Tue, 27 May 2014 13:39:46 +0000 (09:39 -0400)]
gio/tests/network-address: fix when no network is available

GResolver doesn't do full validation of its inputs, so in some of
these tests, the fact that we were getting back
G_RESOLVER_ERROR_NOT_FOUND is because the junk string was getting
passed to an upstream DNS resolver, which returned NXDOMAIN. But if
there's no network on the machine then we'd get
G_RESOLVER_ERROR_INTERNAL instead in that case.

10 years agoGLib 2.41.0
Ryan Lortie [Mon, 26 May 2014 15:34:48 +0000 (17:34 +0200)]
GLib 2.41.0

10 years agogio tests: rm xdgconfighome on 'make clean'
Ryan Lortie [Mon, 26 May 2014 21:01:25 +0000 (23:01 +0200)]
gio tests: rm xdgconfighome on 'make clean'

10 years agoRevert "gvariant: Fix confusion between type and format strings in the docs"
Ryan Lortie [Mon, 26 May 2014 15:33:57 +0000 (17:33 +0200)]
Revert "gvariant: Fix confusion between type and format strings in the docs"

This reverts commit fca1b83be74c9840037cfee84015fd8460cc33f9.

The original documentation was correct.

https://bugzilla.gnome.org/show_bug.cgi?id=729269

10 years agogtestutils: Error out if /dev/null stdin redirection fails
Philip Withnall [Fri, 16 May 2014 22:08:10 +0000 (23:08 +0100)]
gtestutils: Error out if /dev/null stdin redirection fails

https://bugzilla.gnome.org/show_bug.cgi?id=730189

10 years agogio: port annotations from the Vala metadata.
Evan Nemerson [Wed, 21 May 2014 07:27:36 +0000 (00:27 -0700)]
gio: port annotations from the Vala metadata.

https://bugzilla.gnome.org/show_bug.cgi?id=730493

10 years agoconfig.h.win32.in: Define _WIN32_WINNT Conditionally
Chun-wei Fan [Fri, 23 May 2014 02:14:16 +0000 (10:14 +0800)]
config.h.win32.in: Define _WIN32_WINNT Conditionally

This is done so that _WIN32_WINNT may be overridden in the project files,
if needed, so that one can access the Vista+ (or so) Windows APIs easier
by using "preprocessor defines" (or so) in the Visual C++ project files.

10 years agoTry to fix mapping-test
Matthias Clasen [Thu, 22 May 2014 21:22:56 +0000 (17:22 -0400)]
Try to fix mapping-test

The mapping-test is failing under gnome-continuous. I suspect this
is simply due to running many tests in parallel, and mapping-test
being racy. Replace the blind sleep by signals, to avoid the
races.

10 years agoGDBusError enum: Remove trailing comma.
Murray Cumming [Tue, 29 Apr 2014 11:21:41 +0000 (13:21 +0200)]
GDBusError enum: Remove trailing comma.

This is helpful to people using the g++ --pedantic option.

10 years agogobject: unref unused class
Marek Kasik [Thu, 22 May 2014 09:00:14 +0000 (11:00 +0200)]
gobject: unref unused class

If g_type_class_ref() returns a class which is not
a GObjectClass we need to unref it before return in
object_interface_check_properties().

https://bugzilla.gnome.org/show_bug.cgi?id=706983

10 years agoTrivial: fix a guard comment
Matthias Clasen [Wed, 21 May 2014 11:40:09 +0000 (07:40 -0400)]
Trivial: fix a guard comment

The conventional comment for the #endif at the bottom of the
header was not matching the #ifdef at the top in glib-private.h.
Fix that.

10 years agogdbus-tool: Remove dead variables
Philip Withnall [Sat, 17 May 2014 11:08:58 +0000 (12:08 +0100)]
gdbus-tool: Remove dead variables

Coverity issues: #1159447, #1159446

https://bugzilla.gnome.org/show_bug.cgi?id=730295

10 years agogthread: Fix use of a local variable after it went out of scope
Philip Withnall [Fri, 16 May 2014 20:47:24 +0000 (21:47 +0100)]
gthread: Fix use of a local variable after it went out of scope

This could theoretically cause problems, although in practice we would
have seen them by now (the bug was introduced in 2012).

Coverity issue: #1159486

https://bugzilla.gnome.org/show_bug.cgi?id=730277

10 years agogsubprocess: Add a missing va_end() call
Philip Withnall [Fri, 16 May 2014 21:09:29 +0000 (22:09 +0100)]
gsubprocess: Add a missing va_end() call

Coverity issues: #1214070, #1214069

https://bugzilla.gnome.org/show_bug.cgi?id=730278

10 years agoUpdate config.h.win32.in for Newer Windows
Chun-wei Fan [Mon, 19 May 2014 06:49:39 +0000 (14:49 +0800)]
Update config.h.win32.in for Newer Windows

Make use of if_indextoname() and if_nametoindex() when building against
Window Vista/Server 2008 or later, as these are provided by the system.

This is not turned on by default as we still want to support XP and
Server 2003-turn this on by changing _WIN32_WINNT to 0x600 or later prior
to compiling GLib.

https://bugzilla.gnome.org/show_bug.cgi?id=730352

10 years agoUpdated Czech translation
Marek Černocký [Sun, 18 May 2014 06:38:05 +0000 (08:38 +0200)]
Updated Czech translation

10 years agoUpdated Hebrew translation
Yosef Or Boczko [Fri, 16 May 2014 01:37:21 +0000 (04:37 +0300)]
Updated Hebrew translation

10 years agogsocket: Add missing preconditions to g_socket_send_message()
Philip Withnall [Thu, 15 May 2014 11:16:00 +0000 (12:16 +0100)]
gsocket: Add missing preconditions to g_socket_send_message()

https://bugzilla.gnome.org/show_bug.cgi?id=730190

10 years agogresource-tool: Fix an FD leak in an error path
Philip Withnall [Thu, 15 May 2014 09:18:38 +0000 (10:18 +0100)]
gresource-tool: Fix an FD leak in an error path

Coverity issues: #1214054, #1214055, #1214056

10 years agogtestutils: Don’t free unassigned variables in an error path
Philip Withnall [Thu, 15 May 2014 09:15:26 +0000 (10:15 +0100)]
gtestutils: Don’t free unassigned variables in an error path

If the stream is invalid, msg.nums and msg.strings have not been
assigned to, so don’t free them.

Coverity issue: #1159505

10 years agoAvoid overeager warning about deprecated properties
Matthias Clasen [Tue, 13 May 2014 03:05:41 +0000 (23:05 -0400)]
Avoid overeager warning about deprecated properties

Construct properties are always set during construction.
It makes no sense to warn about this even if the property
is marked as deprecated; the deprecation warning should
only be issues for explicit uses of the property after
construction.

https://bugzilla.gnome.org/show_bug.cgi?id=730045

10 years agoRemove some markup from docs
Matthias Clasen [Thu, 8 May 2014 03:08:50 +0000 (23:08 -0400)]
Remove some markup from docs

10 years agogio/gtestdbus.c: Fix write_config_file() for Windows
Chun-wei Fan [Tue, 18 Mar 2014 04:54:10 +0000 (12:54 +0800)]
gio/gtestdbus.c: Fix write_config_file() for Windows

Windows does not like g_unlink() to be called on files whose file
descriptor is still open, so doing that would cause a permission
denied error.  Since the fd is not used in that function after
acquiring the temp file, close it earlier before
g_file_set_contents(), so that it can complete successfully.

This fixes a number of GTK+ tests on Windows.

https://bugzilla.gnome.org/show_bug.cgi?id=719344

10 years agodoc: better ordering of GSubprocessLauncher functions
Sébastien Wilmet [Sun, 11 May 2014 21:41:23 +0000 (23:41 +0200)]
doc: better ordering of GSubprocessLauncher functions

Take the same order as the header.

10 years agoUpdated Spanish translation
Daniel Mustieles [Sat, 10 May 2014 17:43:59 +0000 (19:43 +0200)]
Updated Spanish translation

10 years agogobject.py: Port to gdb 7.7 frame filter API
Tom Tromey [Fri, 21 Feb 2014 19:16:00 +0000 (19:16 +0000)]
gobject.py: Port to gdb 7.7 frame filter API

https://bugzilla.gnome.org/show_bug.cgi?id=623552

10 years agodoc: various improvements
Sébastien Wilmet [Mon, 21 Apr 2014 12:50:53 +0000 (14:50 +0200)]
doc: various improvements

- GSubprocessLauncher exists since 2.40, not 2.36
- more logical order for g_markup functions
- fix short description of GMarkup
- GMarkupParser: specify that some parameters are NULL-terminated.
- g_string_new (NULL); is possible.
- other trivial fixes.

https://bugzilla.gnome.org/show_bug.cgi?id=728983

10 years agogio: cleanup gdbusmessage.c file
Lukasz Skalski [Fri, 9 May 2014 13:06:52 +0000 (15:06 +0200)]
gio: cleanup gdbusmessage.c file

* removed passing GError to ensure_input_padding() function
  - it was necessary before commit 3e5214c15c2a19b64d74e714
  when we used GData*Streams and GMemoryInputStream with
  g_seekable_seek() - now it's useless,

* removed checking return value of ensure_input_padding()
  function - in previous implementation (like above)
  g_seekable_seek() could return FALSE - now it's always TRUE,

* removed passing GError to g_memory_buffer_read_*() functions
  and checking returned value - it also has been inherited after
  old implementation with g_data_input_stream_read_*() functions
  - now it's also useless

* cleaned up code formatting,

https://bugzilla.gnome.org/show_bug.cgi?id=729875

10 years agotests: add testcase for {Only,Not}ShowIn
Ryan Lortie [Thu, 8 May 2014 20:16:35 +0000 (16:16 -0400)]
tests: add testcase for {Only,Not}ShowIn

https://bugzilla.gnome.org/show_bug.cgi?id=729813

10 years agoAppInfo: use XDG_CURRENT_DESKTOP for OnlyShowIn
Ryan Lortie [Thu, 8 May 2014 12:57:50 +0000 (08:57 -0400)]
AppInfo: use XDG_CURRENT_DESKTOP for OnlyShowIn

Expand the functionality of g_desktop_app_info_set_desktop_env() to
include the possibility of passing strings containing ':' characters (as
some apps, such as gnome-session, are directly passing the value of
XDG_CURRENT_DESKTOP).  At the same time, deprecate it, since now we get
the list from the environment variable for ourselves.

Modify the checks in g_desktop_app_info_get_show_in() to deal with
multiple items listed in XDG_CURRENT_DESKTOP.  For example, if we find
that we have

  XDG_CURRENT_DESKTOP=GNOME-Classic:GNOME

and a desktop file contains:

  OnlyShowIn=GNOME

then we will show this file because of the fallback to GNOME.  If the
file _also_ contains the line:

  NotShowIn=GNOME-Classic

Then we will not show it, because GNOME-Classic comes before GNOME in
XDG_CURRENT_DESKTOP.

https://bugzilla.gnome.org/show_bug.cgi?id=729813

10 years agotests: add tests for GDesktopAppInfo Implements=
Ryan Lortie [Fri, 15 Nov 2013 19:44:33 +0000 (14:44 -0500)]
tests: add tests for GDesktopAppInfo Implements=

https://bugzilla.gnome.org/show_bug.cgi?id=712391

10 years agodocs: add two missing symbols on GDesktopAppInfo
Ryan Lortie [Fri, 15 Nov 2013 19:44:11 +0000 (14:44 -0500)]
docs: add two missing symbols on GDesktopAppInfo

https://bugzilla.gnome.org/show_bug.cgi?id=712391

10 years agoAdd g_desktop_app_info_get_implementations()
Ryan Lortie [Fri, 15 Nov 2013 19:40:37 +0000 (14:40 -0500)]
Add g_desktop_app_info_get_implementations()

This provides support for the draft addition of 'Implements=' to the Desktop
Entry specification.

https://bugzilla.gnome.org/show_bug.cgi?id=712391

10 years agoAdd a testcase for the previous fix
Matthias Clasen [Tue, 6 May 2014 00:30:35 +0000 (20:30 -0400)]
Add a testcase for the previous fix

This testcase tests that short option arguments are
not erroneously added to the remaining argument array
when g_option_context_set_ignore_unknown_options is
called.
https://bugzilla.gnome.org/show_bug.cgi?id=729563

10 years agogoption: Don't include parsed option values in G_OPTION_REMAINING
Kjell Ahlstedt [Mon, 5 May 2014 12:40:31 +0000 (14:40 +0200)]
goption: Don't include parsed option values in G_OPTION_REMAINING

After a call to g_option_context_set_ignore_unknown_options(context, TRUE),
the values of short options were included in the array returned by a
G_OPTION_REMAINING option.

https://bugzilla.gnome.org/show_bug.cgi?id=729563

10 years agogio: iostream: use GTask instead of internal pointer
Lionel Landwerlin [Tue, 21 Jan 2014 19:12:08 +0000 (19:12 +0000)]
gio: iostream: use GTask instead of internal pointer

This fixes an infinite recursion loop when calling
g_io_stream_close_async() on a SoupIOStream.

https://bugzilla.gnome.org/show_bug.cgi?id=722723

10 years agoFix some typos in documentation
Mathieu Bridon [Mon, 5 May 2014 10:30:56 +0000 (18:30 +0800)]
Fix some typos in documentation

10 years agodocs: Add a README.rationale documenting major design decisions
Philip Withnall [Fri, 2 May 2014 07:42:06 +0000 (09:42 +0200)]
docs: Add a README.rationale documenting major design decisions

It will be useful to document the major decisions which affect the whole
of GLib in one centralised, easily-greppable file, otherwise they will
get lost forever in Bugzilla.

This file should contain a brief explanation of the decision and its
rationale, plus a link to further discussion (e.g. on a mailing list or
bug report).

This contains an initial discussion about use of compiler attributes in
GLib.

https://bugzilla.gnome.org/show_bug.cgi?id=113075

10 years agogstrfuncs: Add missing preconditions to g_str_match_string()
Philip Withnall [Thu, 28 Nov 2013 20:39:56 +0000 (20:39 +0000)]
gstrfuncs: Add missing preconditions to g_str_match_string()

https://bugzilla.gnome.org/show_bug.cgi?id=113075

10 years agogio: add a missing property to GDummyTlsConnection
Dan Winship [Sat, 3 May 2014 21:24:48 +0000 (17:24 -0400)]
gio: add a missing property to GDummyTlsConnection

GDummyTlsConnection didn't implement the "interaction" property,
meaning you'd get warnings if you tried to set it while creating a
GTlsConnection when using the dummy backend. (Of course, trying to
create the GTlsConnection will fail anyway, but it ought to fail
without hitting any g_warnings.)

10 years ago[l10n] Update Catalan translation
Pau Iranzo [Fri, 2 May 2014 10:41:30 +0000 (12:41 +0200)]
[l10n] Update Catalan translation

10 years agodocs: Fix some minor typos in the GVariant documentation
Philip Withnall [Wed, 30 Apr 2014 14:58:21 +0000 (16:58 +0200)]
docs: Fix some minor typos in the GVariant documentation

https://bugzilla.gnome.org/show_bug.cgi?id=729269

10 years agogvariant: Fix confusion between type and format strings in the docs
Philip Withnall [Wed, 30 Apr 2014 14:37:30 +0000 (16:37 +0200)]
gvariant: Fix confusion between type and format strings in the docs

‘@’ and ‘&’ are only used in format strings, not type strings.

https://bugzilla.gnome.org/show_bug.cgi?id=729269

10 years agogobject: Document that classes/objects/interfaces are zero-filled
Philip Withnall [Tue, 29 Apr 2014 07:47:14 +0000 (08:47 +0100)]
gobject: Document that classes/objects/interfaces are zero-filled

On initialisation, GObject guarantees to zero-fill
class/object/interface structures. Document this so people don’t spend
forever writing:
    my_object->priv->some_member = NULL;
    my_object->priv->some_other_member = NULL;

https://bugzilla.gnome.org/show_bug.cgi?id=729167

10 years agoUpdated Slovenian translation
Matej Urbančič [Mon, 28 Apr 2014 19:47:39 +0000 (21:47 +0200)]
Updated Slovenian translation

10 years agoUpdated Slovenian translation
Matej Urbančič [Mon, 28 Apr 2014 19:45:20 +0000 (21:45 +0200)]
Updated Slovenian translation

10 years agoUpdated Slovenian translation
Matej Urbančič [Mon, 28 Apr 2014 19:28:16 +0000 (21:28 +0200)]
Updated Slovenian translation

10 years agogmenuexporter: fix protocol documentation link
Lars Uebernickel [Mon, 28 Apr 2014 07:26:32 +0000 (09:26 +0200)]
gmenuexporter: fix protocol documentation link

10 years agoUpdated Brazilian Portuguese translation
Enrico Nicoletto [Sun, 27 Apr 2014 00:19:44 +0000 (00:19 +0000)]
Updated Brazilian Portuguese translation

10 years agoAdded type check to GPermission public functions.
Juan Pablo Ugarte [Fri, 25 Apr 2014 22:35:58 +0000 (19:35 -0300)]
Added type check to GPermission public functions.

10 years agogresolver: Ensure GThreadedResolver always sets an error if resolution fails
Philip Withnall [Thu, 24 Apr 2014 08:26:12 +0000 (09:26 +0100)]
gresolver: Ensure GThreadedResolver always sets an error if resolution fails

It was previously possible for GThreadedResolver to return an empty list
and no error in response to a g_resolver_lookup_by_name() call, if it
happened that all the addresses returned by getaddrinfo() could not be
converted from native addresses to GSocketAddresses.

Fix that by setting a G_RESOLVER_ERROR_NOT_FOUND if the returned list is
empty.

https://bugzilla.gnome.org/show_bug.cgi?id=728776

10 years agogresolver: Document that GResolver lists are non-empty on success
Philip Withnall [Thu, 24 Apr 2014 08:24:48 +0000 (09:24 +0100)]
gresolver: Document that GResolver lists are non-empty on success

The documentation previously wasn’t clear about whether the GResolver
methods could return an empty list and no error. On balance, this seems
like a bad idea, and GResolver should commit to always return a
non-empty list, or an error (which should be G_RESOLVER_ERROR_NOT_FOUND
if the list would otherwise be empty).

https://bugzilla.gnome.org/show_bug.cgi?id=728776

10 years agogio: remove an unneeded pragma
Dan Winship [Thu, 24 Apr 2014 16:15:55 +0000 (12:15 -0400)]
gio: remove an unneeded pragma

ddf82a25 removed the use of non-literal format strings from
gthreadedresolver.c, but left the "#pragma GCC diagnostic ignored
-Wformat-nonliteral" behind.

10 years agodocs: Remove <!-- --> comment before plural s
Volker Sobek [Wed, 16 Apr 2014 21:56:02 +0000 (23:56 +0200)]
docs: Remove <!-- --> comment before plural s

These did show up in the html. Since symbol names are checked for a
trailing plural s when generating the docs, the links stay functional
after removing these comments.

https://bugzilla.gnome.org/show_bug.cgi?id=728380

10 years agogaction: Minor clarifications in the GAction documentation
Philip Withnall [Wed, 6 Nov 2013 10:54:06 +0000 (10:54 +0000)]
gaction: Minor clarifications in the GAction documentation

https://bugzilla.gnome.org/show_bug.cgi?id=728350

10 years agoplatform_get_argv0: drop unneeded headers for OpenBSD
Antoine Jacoutot [Wed, 16 Apr 2014 08:27:09 +0000 (10:27 +0200)]
platform_get_argv0: drop unneeded headers for OpenBSD

And properly set the size of len.
There is also no need for realloc(), g_malloc0 will do just fine.

https://bugzilla.gnome.org/show_bug.cgi?id=728280

10 years agogfile: More explicitly document the context for GFileProgressCallback
Philip Withnall [Sat, 19 Apr 2014 17:22:24 +0000 (18:22 +0100)]
gfile: More explicitly document the context for GFileProgressCallback

Be more explicit in the documentation for g_file_copy_async() about
which GMainContext its progress callback is executed in.

https://bugzilla.gnome.org/show_bug.cgi?id=728565

10 years agoUpdated Greek translation
marablack3 [Fri, 18 Apr 2014 17:55:02 +0000 (17:55 +0000)]
Updated Greek translation