Stef Walter [Thu, 31 Oct 2013 13:39:05 +0000 (14:39 +0100)]
gerror: Don't leak when warning about overwriting an error
While not strictly necessary, this fixes a false positive
leak in the tests.
https://bugzilla.gnome.org/show_bug.cgi?id=627423
Ryan Lortie [Mon, 25 Mar 2013 01:55:08 +0000 (21:55 -0400)]
tests: fix leak in mainloop test
Properly unref a pair of GSources in the unix-fd mainloop test.
valgrind was reporting these as 'still reachable' before (possibly due
to some residual pointers somewhere in memory), but when running with
G_DEBUG=cleanup they were properly reported as leaked.
Ryan Lortie [Mon, 25 Mar 2013 01:25:26 +0000 (21:25 -0400)]
grand: restructure a bit
Instead of having lots of 'if NULL then allocate' code segments for the
global GRand instance, move it to a single getter function that everyone
calls.
Ryan Lortie [Mon, 25 Mar 2013 01:18:30 +0000 (21:18 -0400)]
gslice: don't misuse g_mutex_init()
We were using g_mutex_init() to initialise a pair of mutexes in static
storage, but we should only do that for mutexes that are part of
allocated structures.
Philip Withnall [Tue, 5 Nov 2013 16:43:26 +0000 (16:43 +0000)]
garray: Note lack of bounds checking in g_ptr_array_index() documentation
Chun-wei Fan [Mon, 4 Nov 2013 08:32:01 +0000 (16:32 +0800)]
gio/tests: Clean up inclusion of unistd.h
Include unistd.h only on *NIX and define items as necessary on Windows,
also replace instances of ssize_t with the GLib-equivilant gssize so to fix
the build on platforms that do not have ssize_t, such as Visual C++.
https://bugzilla.gnome.org/show_bug.cgi?id=711047
Chun-wei Fan [Mon, 4 Nov 2013 05:07:52 +0000 (13:07 +0800)]
glib/tests: Clean up inclusion of unistd.h
Include unistd.h only when G_OS_UNIX is defined (or when G_OS_WIN32 is not
defined). This will avoid including unistd.h unconditionally and/or
unecessarily, which may cause problems in certain scenarios, such as when
building the tests on Visual C++, which does not come with a unistd.h and
MinGW, where unistd.h is essentially a wrapper for io.h and process.h.
https://bugzilla.gnome.org/show_bug.cgi?id=711047
Chun-wei Fan [Mon, 4 Nov 2013 04:23:30 +0000 (12:23 +0800)]
glib/tests/fileutils.c: Include unistd.h on *NIX only
...and fix the test on non-English Windows, as gettext on Windows does
not honor LC_ALL = "C" (the default CRT behavior) but requires using
SetThreadLocale() to set the locale as it picks up the user's environment
and the thread's locale. Without doing so the g_format_size_for_display()
et al will display the translated message if the gettext translations have
been installed before, causing the test_format_size_for_display tests to
fail.
https://bugzilla.gnome.org/show_bug.cgi?id=711047
Chun-wei Fan [Tue, 29 Oct 2013 11:12:38 +0000 (19:12 +0800)]
tests/libmoduletestplugins*.c: Update Prototypes
Visual C++ does not like function declarations being different from
their prototypes, so make the prototypes match the declarations by
decorating them with G_MODULE_EXPORT.
https://bugzilla.gnome.org/show_bug.cgi?id=711047
Chun-wei Fan [Tue, 29 Oct 2013 11:17:32 +0000 (19:17 +0800)]
gio/test/resources.c: Fix for Windows
We need to use g_content_type_get_mime_type() to look up the mime type of
the file from the registry on the content type that was acquired on
Windows, as g_file_info_get_content_type() does not acquire the
file mime type (unlike on *NIX).
g_content_type_get_mime_type() on *NIX is more or less an no-op as it
simply returns the g_strdup()-ed version of the passed-in content type.
This will enable the resources test to pass on Windows.
https://bugzilla.gnome.org/show_bug.cgi?id=711047
Ognyan Tonchev [Tue, 29 Oct 2013 15:39:38 +0000 (16:39 +0100)]
gmain: make g_source_add_child_source() thread safe
g_source_add_child_source() releases the context lock before attaching
child_source to context. And this causes trouble if parent source is
blocked and g_main_dispatch() manages to lock the context mutex and call
unblock_source() before child_source gets attached to context.
To fix this we call g_source_attach_unlocked() before releasing the
context mutex.
https://bugzilla.gnome.org/show_bug.cgi?id=711064
Ross Lagerwall [Tue, 29 Oct 2013 04:02:59 +0000 (06:02 +0200)]
gio: Don't allow skipping past the end of GLocalFileInputStream
The overridden implementation of the skip method for
GLocalFileInputStream allows skipping past the end of the file which is
inconsistent with the documentation. Prevent this by first seeking to
the end of the file and then seeking backwards from there as much as
is necessary.
https://bugzilla.gnome.org/show_bug.cgi?id=711048
Dan Winship [Sat, 2 Nov 2013 16:18:45 +0000 (12:18 -0400)]
gmacros.h: fix C99 check
G_STRFUNC was checking __STDC_VERSION__ against the wrong value
(though it didn't actually matter, since __STDC_VERSION__ wasn't
defined in C90, so the check still only matched C99 and above anyway).
Fran Diéguez [Fri, 1 Nov 2013 11:52:43 +0000 (12:52 +0100)]
Updated Galician translations
Fran Diéguez [Mon, 28 Oct 2013 00:13:41 +0000 (01:13 +0100)]
Updated Galician translations
Daniel Mustieles [Thu, 31 Oct 2013 13:36:50 +0000 (14:36 +0100)]
Updated Spanish translation
Murray Cumming [Thu, 31 Oct 2013 12:15:01 +0000 (13:15 +0100)]
docs: GDataTime: *_add_*(): Mention that you can subtract.
Because this was not obvious to at least one person on irc.
Ross Lagerwall [Wed, 30 Oct 2013 18:34:14 +0000 (20:34 +0200)]
gio: Clean up trashinfo file if trashing fails
https://bugzilla.gnome.org/show_bug.cgi?id=687202
Xavier Claessens [Mon, 28 Oct 2013 21:02:44 +0000 (17:02 -0400)]
Tests: It is useless to unset "DBUS_SESSION_BUS_ADDRESS" manually
If the goal is to make sure we don't have a dbus connection, it has
to call g_test_dbus_unset() instead which is much more complete.
In this case, g_test_dbus_unset() is called already, so it should be
fine.
https://bugzilla.gnome.org/show_bug.cgi?id=697348
Xavier Claessens [Mon, 28 Oct 2013 18:54:22 +0000 (14:54 -0400)]
GTestDBus: Make sure only DBUS_SESSION_BUS_ADDRESS is set by default
g_test_dbus_unset() now also unset DBUS_STARTER_ADDRESS and
DBUS_STARTER_BUS_TYPE.
https://bugzilla.gnome.org/show_bug.cgi?id=697348
Xavier Claessens [Mon, 28 Oct 2013 21:18:54 +0000 (17:18 -0400)]
Tests: add session_bus_run() and use it where possible
This is to avoid having again the subtil bug in dbus-appinfo.c:
session_bus_down() was called before g_test_run() so the test was
running on the user's dbus session.
https://bugzilla.gnome.org/show_bug.cgi?id=697348
Colin Walters [Tue, 29 Oct 2013 13:47:34 +0000 (09:47 -0400)]
tests/gdbus-introspection: Add a less trivial test case for _IGNORE_QUALIFIED
This is distilled from PackageKit which currently aborts on startup.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Ryan Lortie [Tue, 29 Oct 2013 16:01:32 +0000 (09:01 -0700)]
GMarkup: clear attributes on ignorned tags
Make sure that if we ignore a tag then we also clear the attributes that
we already collected so that they don't end up on the next unignored tag
opening.
Also add some extra brackets for clarity (it doesn't make any difference
-- I just think it reads nicer this way).
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Lars Uebernickel [Tue, 29 Oct 2013 06:55:19 +0000 (23:55 -0700)]
gtk-notification-backend: remove unused variable
A reference to the session bus is now stored in GNotificationBackend.
Remove the extraneous one in the gtk backend and stop using it in
withdraw_notification.
https://bugzilla.gnome.org/show_bug.cgi?id=711051
Ross Lagerwall [Tue, 29 Oct 2013 12:53:38 +0000 (14:53 +0200)]
gio: Clear error properly to prevent crash
https://bugzilla.gnome.org/show_bug.cgi?id=711070
Ryan Lortie [Mon, 28 Oct 2013 18:44:16 +0000 (11:44 -0700)]
GCancellable: drop lock for callback during connect()
Don't hold the lock when calling the user's callback during
g_cancellable_connect() for the case that the cancellable has already
fired.
Taken from a patch by Alex Larsson.
Doc updates from Colin Walters.
https://bugzilla.gnome.org/show_bug.cgi?id=705395
Chun-wei Fan [Tue, 29 Oct 2013 06:49:13 +0000 (14:49 +0800)]
gio/gsubprocess.c: Use g_snprintf()
snprintf() is unfortunately still not universally available, so use
g_snprintf() to ensure the build works on all platforms.
https://bugzilla.gnome.org/show_bug.cgi?id=711049
Chun-wei Fan [Tue, 29 Oct 2013 04:46:32 +0000 (12:46 +0800)]
gobject/gvaluetransform.c: Cleanup #ifndef _MSC_VER
The two casts that were unsupported with Visual Studio is now properly
supported, so build this code like how the other compilers build the code.
Ryan Lortie [Tue, 29 Oct 2013 03:48:53 +0000 (20:48 -0700)]
GSettingsSchemaSource: fix parameter name in header
list_schemas() had the parameters called 'recursive' and 'non-recursive'
in the header instead of 'relocatable'. Fix that.
Ryan Lortie [Mon, 28 Oct 2013 21:49:14 +0000 (14:49 -0700)]
exporter: give error on Describe of missing action
If someone calls org.gtk.Actions.Describe on a non-existent action then
return an exception instead of a trivial description (disabled, no
state, etc.).
https://bugzilla.gnome.org/show_bug.cgi?id=687185
Ryan Lortie [Mon, 28 Oct 2013 21:49:14 +0000 (14:49 -0700)]
exporter: give error on Describe of missing action
If someone calls org.gtk.Actions.Describe on a non-existent action then
return an exception instead of a trivial description (disabled, no
state, etc.).
https://bugzilla.gnome.org/show_bug.cgi?id=687185
Ryan Lortie [Mon, 28 Oct 2013 21:59:26 +0000 (14:59 -0700)]
GSimpleAction: don't allow changing state type
g_object_set() allowed us to bypass the usual checks that the state
doesn't change type and also leaked.
Fix that up by turning the state into a construct property (so that it
always gets set once during construction, even if only to NULL) and
then route the further sets through the C API so that they are subject
to the same checks.
https://bugzilla.gnome.org/show_bug.cgi?id=696424
Ryan Lortie [Mon, 28 Oct 2013 22:30:25 +0000 (15:30 -0700)]
glib-compile-schemas: ignore qualified tags/attrs
Ignore qualified (in the XML namespace sense) tags and attributes when
compiling GSettings schemas.
This will allow people to add custom tags and attributes to their schemas
without tripping up the compiler.
https://bugzilla.gnome.org/show_bug.cgi?id=635641
Ryan Lortie [Mon, 28 Oct 2013 22:27:29 +0000 (15:27 -0700)]
GDBus: ignore qualified XML tags and attributes
Ignore qualified (in the XML namespace sense) tags and attributes when
parsing D-Bus introspection XML.
This will allow people to add custom tags and attributes to their D-Bus
interfaces without tripping up GDBus.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Ryan Lortie [Mon, 28 Oct 2013 21:11:05 +0000 (14:11 -0700)]
tests: add a ignore-qualified markup-collect case
Add a case to markup-collect that exercises the new IGNORE_QUALIFIED
flag.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Ryan Lortie [Mon, 28 Oct 2013 21:05:51 +0000 (14:05 -0700)]
GMarkup: add G_MARKUP_IGNORE_QUALIFIED
Add a flag to GMarkupParserFlags to ignore qualified tags (along with
their contents) and attributes.
This will provide a nice way for some of our parsers (GDBus
introspection, GSettings schema, etc) to ignore additional tags that
users have added to their files, under a different namespace.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Ryan Lortie [Mon, 28 Oct 2013 20:56:26 +0000 (13:56 -0700)]
GMarkup: share some common code for closing tags
The code for dealing with </foo> and the second half of <foo/> was
largely duplicated. We can share a lot of it by using a common
function.
This slightly changes the behaviour of the parser under error
circumstances: previously the parser would deal with '<foo/}' by first
issuing the end_element callback and then flagging the error due to the
unexpected character. Now we will flag the unexpected character error
first, skipping the callback.
This behaviour change required modifying the testsuite.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Stef Walter [Mon, 28 Oct 2013 09:12:29 +0000 (10:12 +0100)]
test: g_debug messages shouldn't affect g_assert_expected_messages
Debug messages are meant to give insight into how a process is
proceeding, and are unpredictable in nature. They also often have
line numbers in them.
This patch ignores debug messages in g_test_assert_expected_messages().
https://bugzilla.gnome.org/show_bug.cgi?id=710991
Ryan Lortie [Mon, 28 Oct 2013 19:10:13 +0000 (12:10 -0700)]
GSettings: fix for loop in 'extends' functionality
Obvious mistake :(
Ryan Lortie [Mon, 28 Oct 2013 17:54:08 +0000 (10:54 -0700)]
gsettings(1): stop using deprecated APIs
Stop using the recently-deprecated GSettings APIs.
Use the GSettingsSchema-based APIs instead.
This fixes a number of bugs and also a net reduction of code. In
particular, list-schemas will now work in context of a given --schemadir
argument.
https://bugzilla.gnome.org/show_bug.cgi?id=695558
Ryan Lortie [Mon, 28 Oct 2013 17:23:29 +0000 (10:23 -0700)]
gsettings(1): use global variables
The number of arguments passed to each function is about to increase, so
just use global variables instead.
This is a commandline tool, after all...
https://bugzilla.gnome.org/show_bug.cgi?id=695558
Ryan Lortie [Mon, 28 Oct 2013 00:03:04 +0000 (17:03 -0700)]
GSettingsSchemaKey: add introspection APIs
Add g_settings_schema_has_key() and _get_range(), _range_check(),
_get_value_type(), _get_default_value() methods on GSettingsSchemaKey.
Deprecate the equivalent APIs on GSettings.
https://bugzilla.gnome.org/show_bug.cgi?id=683017
Ryan Lortie [Mon, 28 Oct 2013 17:23:29 +0000 (10:23 -0700)]
bump version
dconf needs the new version with read_user_value() on the backend API.
Ryan Lortie [Sun, 27 Oct 2013 17:34:01 +0000 (10:34 -0700)]
GSettings: add getters for user/default value
Add two new APIs: g_settings_get_user_value() and
g_settings_get_default_value(). Together, these should allow the
inspection of all interesting cases of "is this key set?" and "what
would happen if I reset this key?"
https://bugzilla.gnome.org/show_bug.cgi?id=668233
Ryan Lortie [Sun, 27 Oct 2013 17:33:08 +0000 (10:33 -0700)]
GSettings: small internal refactor
Add two boolean parameters to our internal getter utility function in
anticipation of the coming addition of g_settings_get_user_value() and
g_settings_get_default_value() APIs.
https://bugzilla.gnome.org/show_bug.cgi?id=668233
Ryan Lortie [Sun, 27 Oct 2013 16:42:32 +0000 (09:42 -0700)]
GSettingsBackend: add read_user_value() API
This will get the 'user' value from the database (ie: the one that the user has
control over).
Provide a default implementation that chains to ->read(). That will work for
all of our internal backends which don't have a concept of layering or
lockdown.
The delayed backend implments "user value" by returning anything that's
in the changeset (incuding an explicit NULL) or chaining up otherwise.
We will use this for g_settings_get_user_value().
https://bugzilla.gnome.org/show_bug.cgi?id=668233
Philip Withnall [Mon, 28 Oct 2013 17:07:49 +0000 (17:07 +0000)]
gio: Remove an unused variable from Makefile.am
RESOLVER_LIBADD was not defined anywhere (its functionality has
migrated to NETWORK_LIBS instead).
Emmanuele Bassi [Mon, 14 Oct 2013 18:41:54 +0000 (19:41 +0100)]
docs: Mention that generated code should not be kept
Code generated by gdbus-codegen should neither be checked in into
revision control, nor should be distributed.
https://bugzilla.gnome.org/show_bug.cgi?id=710133
Emmanuele Bassi [Mon, 14 Oct 2013 18:40:53 +0000 (19:40 +0100)]
gdbus-codegen maintains ABI for type structures
Make it explicit, to avoid changes that could potentially lead to
breakage in user code.
https://bugzilla.gnome.org/show_bug.cgi?id=710133
Emmanuele Bassi [Mon, 14 Oct 2013 18:30:21 +0000 (19:30 +0100)]
gdbus-codegen: Take into consideration MAX_ALLOWED for private data
The G_ADD_PRIVATE() macro, and the auto-generated get_instance_private()
internal function, should be used conditionally depending on the maximum
allowed version of GLib, as defined by the GLIB_VERSION_MAX_ALLOWED
pre-processor symbol.
This allows generating code that can be compiled in projects that wish
to use an older API version of GLib through the use of the
GLIB_VERSION_MAX_ALLOWED symbol.
https://bugzilla.gnome.org/show_bug.cgi?id=710133
Ryan Lortie [Mon, 28 Oct 2013 16:29:15 +0000 (09:29 -0700)]
g_settings_schema_list: some fixes
Prevent a crash in the case that gvdb_table_list() returns NULL (ie:
because a schema has no keys).
Stop a memory leak caused by pointlessly stealing keys from a hashtable
(after we quarked them already).
Stop allocating an extra entry at the end of an array for a terminator
(that we never wrote anyway) when all functions using this API refer to
the out-parameter length array.
https://bugzilla.gnome.org/show_bug.cgi?id=711016
Matthias Clasen [Mon, 28 Oct 2013 16:31:31 +0000 (12:31 -0400)]
Fix handling of icons in GNotification
The fdo backend was sending a uri where a file path
was expected.
Stef Walter [Wed, 28 Nov 2012 21:01:21 +0000 (22:01 +0100)]
Add a request_certificate virtual method to GTlsInteraction
This allows GTlsConnection implementations to request a certificate
from the user.
Fix ups by Dan Winship <danw@gnome.org>
https://bugzilla.gnome.org/show_bug.cgi?id=637257
Matthew Barnes [Mon, 28 Oct 2013 00:26:33 +0000 (17:26 -0700)]
GKeyFileSettingsBackend: ignore file deletions
Ignore the keyfile being deleted (ie: by the user) instead of reporting
it to the application as all values being reverted back to their
originals.
https://bugzilla.gnome.org/show_bug.cgi?id=637956
Christian Persch [Mon, 28 Oct 2013 00:22:38 +0000 (17:22 -0700)]
Add test for extended schemas
The test checks that the keys of the base schema are all present in
the extended schema.
https://bugzilla.gnome.org/show_bug.cgi?id=645453
Ryan Lortie [Mon, 28 Oct 2013 00:18:10 +0000 (17:18 -0700)]
GSettings: properly support 'extends'
Support the 'extends' attribute that has been supported by the compiler
for a long time by doing three things:
- when creating a schema that extends another schema, lookup that other
schema
- when looking up keys and we can't find them in the schema, check
(recursively) in the 'extends' schema
- when listing all keys in a schema, also visit the extends schemas,
but take care to avoid duplicates caused by overrides
Extend the testsuite to verify that it works.
https://bugzilla.gnome.org/show_bug.cgi?id=645453
Ross Lagerwall [Fri, 25 Oct 2013 08:05:02 +0000 (10:05 +0200)]
gio: Fix typo in documentation
https://bugzilla.gnome.org/show_bug.cgi?id=710859
Ryan Lortie [Sun, 27 Oct 2013 02:12:46 +0000 (19:12 -0700)]
Add g_settings_schema_source_list_schemas()
g_settings_list_schemas() and g_settings_list_relocatable_schemas() are
now deprecated.
This will allow listing off schemas on non-default sources and is a
better fit with the new direction the API is going.
https://bugzilla.gnome.org/show_bug.cgi?id=680838
Colin Walters [Sun, 27 Oct 2013 14:34:08 +0000 (10:34 -0400)]
gsubprocess: Add documentation about default input directions
Colin Walters [Sun, 27 Oct 2013 14:21:13 +0000 (10:21 -0400)]
gsubprocess: Fall back to plain F_DUPFD+fcntl for OS X <= Snow Lion
Since it apparently doesn't have F_DUPFD_CLOEXEC.
https://bugzilla.gnome.org/show_bug.cgi?id=710962
Ryan Lortie [Sat, 26 Oct 2013 22:58:06 +0000 (18:58 -0400)]
Add a testcase for summary/description
https://bugzilla.gnome.org/show_bug.cgi?id=668232
Ryan Lortie [Sat, 26 Oct 2013 22:57:55 +0000 (18:57 -0400)]
Add g_settings_schema_key_get_summary/description
Add an API to read the summary and description from the .xml schema
files.
This will be used by dconf-editor and gnome-tweak-tool.
This API is a bit heavy -- it parses the XML and builds a table. It
also loads gettext domains for translation. It only does these things
if it is used, however, so it will not impact normal applications.
We store the summary/description in a pair of hash tables on the schema
source (which we have a backref to as of a few commits ago). We can't
use a global table because people might want to request summary and
description from non-default sources. We don't want to use per-schema
tables because we'd have to reparse the directory every time (since we
cannot guess which file a schema may have been in).
https://bugzilla.gnome.org/show_bug.cgi?id=668232
Ryan Lortie [Sat, 26 Oct 2013 22:57:24 +0000 (18:57 -0400)]
Make GSettingsSchemaKey public
Take this private API and make it public along with a boxed type and
ref/unref functions.
Future commits will add accessors with new functionality and some that
allow us to deprecate functions on GSettings itself (such as
g_settings_get_range).
https://bugzilla.gnome.org/show_bug.cgi?id=668232
Ryan Lortie [Sat, 26 Oct 2013 22:50:39 +0000 (18:50 -0400)]
GSettingsSchema: store a backref to the source
Record in each GSettingsSchema object the source from which it came.
This will be useful in future commits.
https://bugzilla.gnome.org/show_bug.cgi?id=668232
Ryan Lortie [Sat, 26 Oct 2013 22:49:58 +0000 (18:49 -0400)]
Clean up GSettingsSchema logic
The way we created the global schema list predates
g_settings_schema_source_new_from_directory() and therefore doesn't use
it.
Update it to use that function, removing some code.
https://bugzilla.gnome.org/show_bug.cgi?id=668232
Ryan Lortie [Sun, 27 Oct 2013 16:24:40 +0000 (09:24 -0700)]
tests: test g_hash_table_get_keys_as_array()
https://bugzilla.gnome.org/show_bug.cgi?id=710964
Ryan Lortie [Sun, 27 Oct 2013 01:48:42 +0000 (18:48 -0700)]
Add g_hash_table_get_keys_as_array()
Returns a %NULL-terminated array of the keys of a hashtable.
In the case that the hash table has strings for keys, this is actually a
gchar**.
https://bugzilla.gnome.org/show_bug.cgi?id=710964
Giovanni Campagna [Fri, 25 Oct 2013 16:02:16 +0000 (18:02 +0200)]
gapplication: don't emit warnings for DBus calls
If a broken (or malicious) remote client calls Open or CommandLine
on a GApplication that does not implement those, return a DBus
error instead of going through and then emitting a warning.
https://bugzilla.gnome.org/show_bug.cgi?id=710885
Giovanni Campagna [Fri, 25 Oct 2013 16:01:51 +0000 (18:01 +0200)]
gapplication: fix memory leak
g_variant_get_child_value() returns a new GVariant, which we must
free.
https://bugzilla.gnome.org/show_bug.cgi?id=710885
Ryan Lortie [Fri, 25 Oct 2013 16:22:42 +0000 (12:22 -0400)]
win32: use real random data for seed on win32
We can get cryptographically secure data from rand_s().
https://bugzilla.gnome.org/show_bug.cgi?id=710738
Nilamdyuti Goswami [Fri, 25 Oct 2013 07:14:48 +0000 (12:44 +0530)]
Assamese translation updated
Ryan Lortie [Wed, 23 Oct 2013 19:09:07 +0000 (15:09 -0400)]
GLib 2.39.0
Ryan Lortie [Thu, 24 Oct 2013 18:24:04 +0000 (14:24 -0400)]
docs: dist manpage gapplication.xml
Philip Withnall [Thu, 24 Oct 2013 11:09:27 +0000 (12:09 +0100)]
tests: Add tests for the thumbnail verification code in GIO
This code was added for use by the G_FILE_ATTRIBUTE_THUMBNAIL_IS_VALID
file attribute, but may end up being used elsewhere (e.g. in GVfs) as well.
As it’s dealing with untrusted external files, and the non-trivial PNG file
format, this commit adds several test cases to cover valid and invalid PNG
files.
The security model for the thumbnail verification code is that the user’s
cache directory is untrusted, and potentially any PNG file which is passed
to the verifier has been manipulated arbitrarily by an attacker.
This is a follow-up to commit
fe7069749fe39a006985ec266260a3c02ee8c855.
https://bugzilla.gnome.org/show_bug.cgi?id=709898
Colin Walters [Thu, 17 Oct 2013 20:57:10 +0000 (16:57 -0400)]
Revert "Use g_return_val_if_fail() for developer-only messages"
This reverts commit
00f0795a84d23f2e2654a86f8bd3a233c8af3771.
https://bugzilla.gnome.org/show_bug.cgi?id=569017
Jiro Matsuzawa [Wed, 28 Aug 2013 16:05:29 +0000 (01:05 +0900)]
gsettings-tool: Add --version into general usage
https://bugzilla.gnome.org/show_bug.cgi?id=707002
Ryan Lortie [Wed, 23 Oct 2013 20:42:12 +0000 (16:42 -0400)]
GSettings: verify path validity on constructors
Don't allow constructing GSettings objects with invalid paths.
https://bugzilla.gnome.org/show_bug.cgi?id=704802
Olivier Brunel [Tue, 8 Oct 2013 14:11:44 +0000 (16:11 +0200)]
gthemedicon: Fallback to non-symbolic icons
When doing fallback for symbolic icons, we first shorten the name at dashes
while preserving the -symbolic suffix. But after exhausting that, we should also
try stripping the suffix.
See https://bugzilla.gnome.org/show_bug.cgi?id=708163
Signed-off-by: Olivier Brunel <jjk@jjacky.com>
https://bugzilla.gnome.org/show_bug.cgi?id=710254
Colin Walters [Wed, 23 Oct 2013 20:41:09 +0000 (16:41 -0400)]
grand: Document this is not for cryptographic purposes
While this is obvious to seasoned experts upon inspection, it is
really worth documenting explicitly.
https://bugzilla.gnome.org/show_bug.cgi?id=710736
Philip Withnall [Wed, 23 Oct 2013 14:45:15 +0000 (15:45 +0100)]
gdbus: Ensure message matching always succeeds against path_namespace='/'
This copies the fix from upstream D-Bus bug
https://bugs.freedesktop.org/show_bug.cgi?id=70799 to the GDBusDaemon
implementation, ensuring that matching against path_namespace='/' succeeds
for all keys (i.e. it’s a no-op).
https://bugzilla.gnome.org/show_bug.cgi?id=710726
Ryan Lortie [Wed, 23 Oct 2013 18:17:48 +0000 (14:17 -0400)]
gtester: only remove source if not already dead
Don't attempt to g_source_remove() a source for which we already returned FALSE
from the handler.
https://bugzilla.gnome.org/show_bug.cgi?id=710724
Ryan Lortie [Wed, 23 Oct 2013 15:10:00 +0000 (11:10 -0400)]
tests: expect critical on failure to remove source
We've added a g_critical() on failure to remove sources, so make sure we
expect to see that (instead of failing the test due to the unexpected
message).
https://bugzilla.gnome.org/show_bug.cgi?id=710724
Ryan Lortie [Wed, 23 Oct 2013 15:04:01 +0000 (11:04 -0400)]
gmain: test g_source_remove() with invalid ID
Make sure we get the proper critical displayed.
https://bugzilla.gnome.org/show_bug.cgi?id=710724
Bastien Nocera [Wed, 23 Oct 2013 13:38:58 +0000 (15:38 +0200)]
gmain: Warn when g_source_remove() fails
Trying to remove a non-existent source should really be
a programming error, as the programmer could be trying to
use the wrong function to remove a callback, as seen when
GtkScrolledWindow tried to remove ID from another function
using g_source_remove().
See https://bugzilla.gnome.org/show_bug.cgi?id=710666#c12
https://bugzilla.gnome.org/show_bug.cgi?id=710724
Ryan Lortie [Fri, 11 Oct 2013 10:22:31 +0000 (11:22 +0100)]
file-info: Add a G_FILE_ATTRIBUTE_THUMBNAIL_IS_VALID attribute
This indicates whether the thumbnail (given by G_FILE_ATTRIBUTE_THUMBNAIL_PATH)
is valid — i.e. to represent the file in its current state. If
G_FILE_ATTRIBUTE_THUMBNAIL_IS_VALID is FALSE (for a normal _or_ failed
thumbnail) it means the file has changed since the thumbnail was generated, and
the thumbnail is out of date.
Part of checking thumbnail validity (by the spec) involves parsing
headers out of the thumbnail .png so we include some (small) code to do
that in a separate file. We will likely want to copy this code to gvfs
to do the same for GVfsFile.
Heavily based on a patch from Philip Withnall <philip.withnall@collabora.co.uk>
who suggested the feature and designed the API.
https://bugzilla.gnome.org/show_bug.cgi?id=709898
Ryan Lortie [Tue, 22 Oct 2013 19:51:15 +0000 (15:51 -0400)]
GMemoryOutputStream: docs and whitespace fixes
Document the difference between resizable and fixed-sized streams,
particularly with regards to sizing and seeking.
https://bugzilla.gnome.org/show_bug.cgi?id=684842
Ryan Lortie [Tue, 22 Oct 2013 20:15:29 +0000 (16:15 -0400)]
GMemoryOutputStream: improve seek tests
Improve test coverage for testing seeking on fixed vs. resizable
GMemoryOutputStream.
https://bugzilla.gnome.org/show_bug.cgi?id=684842
Ryan Lortie [Wed, 23 Oct 2013 15:25:56 +0000 (11:25 -0400)]
Change semantics of seek on memory output stream
It is our intention that memory output streams should operate in two
distinct modes, depending on if a realloc function was provided or not.
In the case that we have a realloc function (resizable mode), we want
the stream to behave as if it were a file that started out empty. In
the case that we don't have a realloc function (fixed-sized mode), we
want the stream to behave as a block device would.
To this end, we introduce two changes in functionality:
- seeking to SEEK_END on a resizable stream will now seek to the end of
the valid data region, not to the end of the allocated memory (which
is really just an implementation detail)
- seeks past the end of the allocated memory size are now permitted,
but only on resizable streams. The next write will grow the buffer
(inserting zeros between).
Some tweaks to testcases were required in order not to break the build,
which indicates that this is an API break, but it seems unlikely that
anyone will be effected by these changes 'in the real world'.
Updates to documentation and further testcases are in following commits.
Based on a patch from Maciej Piechotka <uzytkownik2@gmail.com>.
https://bugzilla.gnome.org/show_bug.cgi?id=684842
Ryan Lortie [Tue, 22 Oct 2013 19:01:16 +0000 (15:01 -0400)]
GSeekable: document seek-past-end semantics
Introduce the concept of "fixed" vs. "resizable" streams and document
how g_seekable_seek() works for each case.
We don't include g_seekable_is_fixed_size() at this point because we
don't know if anyone would require it. This may appear in the future if
someone asks for it, however.
https://bugzilla.gnome.org/show_bug.cgi?id=684842
Philip Withnall [Wed, 23 Oct 2013 14:07:46 +0000 (15:07 +0100)]
gdbus: Work around a D-Bus bug with path_namespace='/' match rules
D-Bus versions < 1.6.18 (i.e. all current versions) have a bug with the
path_namespace='/' match rule key. It should conceptually match everything,
but actually matches nothing. This results in no property change (or other)
signals being forwarded to the D-Bus client.
The work-around implemented in GDBusObjectManagerClient is to remove the
path_namespace match key if its value is ‘/’.
For the upstream D-Bus bug, see:
https://bugs.freedesktop.org/show_bug.cgi?id=70799
https://bugzilla.gnome.org/show_bug.cgi?id=710726
Bastien Nocera [Fri, 18 Oct 2013 12:37:31 +0000 (14:37 +0200)]
g_file_error_from_errno: Remove unneeded breaks
Just wasting space
https://bugzilla.gnome.org/show_bug.cgi?id=710625
Ryan Lortie [Tue, 22 Oct 2013 20:24:41 +0000 (16:24 -0400)]
GKeyfile: don't leak on failed get_(u)int64
In the case that g_key_file_get_(u)int64 fails to parse the integer,
make sure we free the string before returning.
Reported by Andrew Stone <astonecc@gmail.com>
https://bugzilla.gnome.org/show_bug.cgi?id=710313
Dan Winship [Sun, 20 Oct 2013 13:27:11 +0000 (09:27 -0400)]
gobject: simplify object-in-construction handling
Rather than keeping a global list of objects that are being
constructed, use qdata on the object itself like we do with several
other properties now.
https://bugzilla.gnome.org/show_bug.cgi?id=661576
Dan Winship [Mon, 29 Apr 2013 17:04:11 +0000 (13:04 -0400)]
gobject: forbid finalization-during-construction
If a constructor() implementation created an object but then unreffed
it rather than returning it, that object would get left on the
construction_objects list, which would cause problems later when that
memory location got reused by another object.
"Fix" this by making it fail intentionally, and add a test for it (and
for the normal, working singleton case).
https://bugzilla.gnome.org/show_bug.cgi?id=661576
Matthias Clasen [Mon, 21 Oct 2013 22:46:48 +0000 (18:46 -0400)]
Clarify the g_application_withdraw_notification docs
Mention that notifications are dismissed when activated.
Ryan Lortie [Mon, 21 Oct 2013 18:55:52 +0000 (14:55 -0400)]
Don't build gapplication(1) on win32
It's pretty closely tied to desktop files...
Ryan Lortie [Mon, 21 Oct 2013 18:55:21 +0000 (14:55 -0400)]
GSubprocess win32 fixups
Note: we go out of our way not to pass a child setup function on win32
(even if it does nothing) because we get a g_warning() from gspawn if we
do so.
Colin Walters [Mon, 21 Oct 2013 19:05:45 +0000 (15:05 -0400)]
GNotification: Don't install private headers, avoid duplicated prototype
These both break gobject-introspection's build, and moreover are
things we don't want to do anyways.
Ryan Lortie [Sat, 19 Oct 2013 00:40:57 +0000 (20:40 -0400)]
Don't g_test_assert_expected_messages for g_error
Don't allow the user to assert for expected g_error(). They need to use
subprocess for this.
https://bugzilla.gnome.org/show_bug.cgi?id=709615
Ryan Lortie [Mon, 21 Oct 2013 18:29:44 +0000 (14:29 -0400)]
GNotification: finish documentation
Add the missing documentation and add the section to the GIO reference
docs. Fix up a few small documentation issues.
https://bugzilla.gnome.org/show_bug.cgi?id=688492