Methods

Asynchronously connects to the message bus specified by busType.

When the operation is finished, callback will be invoked. You can then call busGetFinish to get the result of the operation.

This is an asynchronous failable function. See busGetSync for the synchronous version.

Since: 2.26

Finishes an operation started with busGet.

The returned object is a singleton, that is, shared with other callers of busGet and busGetSync for busType. In the event that you need a private message bus connection, use dbusAddressGetForBusSync and dBusConnectionNewForAddress.

Note that the returned DBusConnection object will (usually) have the DBusConnection:exit-on-close property set to True.

Since: 2.26

Synchronously connects to the message bus specified by busType. Note that the returned object may shared with other callers, e.g. if two separate parts of a process calls this function with the same busType, they will share the same object.

This is a synchronous failable function. See busGet and busGetFinish for the asynchronous version.

The returned object is a singleton, that is, shared with other callers of busGet and busGetSync for busType. In the event that you need a private message bus connection, use dbusAddressGetForBusSync and dBusConnectionNewForAddress.

Note that the returned DBusConnection object will (usually) have the DBusConnection:exit-on-close property set to True.

Since: 2.26

Version of g_bus_own_name() using closures instead of callbacks for easier binding in other languages.

Since: 2.26

Version of g_bus_own_name_on_connection() using closures instead of callbacks for easier binding in other languages.

Since: 2.26

Stops owning a name.

Note that there may still be D-Bus traffic to process (relating to owning and unowning the name) in the current thread-default MainContext after this function has returned. You should continue to iterate the MainContext until the DestroyNotify function passed to g_bus_own_name() is called, in order to avoid memory leaks through callbacks queued on the MainContext after it’s stopped being iterated.

Since: 2.26

Stops watching a name.

Note that there may still be D-Bus traffic to process (relating to watching and unwatching the name) in the current thread-default MainContext after this function has returned. You should continue to iterate the MainContext until the DestroyNotify function passed to g_bus_watch_name() is called, in order to avoid memory leaks through callbacks queued on the MainContext after it’s stopped being iterated.

Since: 2.26

Version of g_bus_watch_name() using closures instead of callbacks for easier binding in other languages.

Since: 2.26

Version of g_bus_watch_name_on_connection() using closures instead of callbacks for easier binding in other languages.

Since: 2.26

Checks if a content type can be executable. Note that for instance things like text files can be executables (i.e. scripts and batch files).

Compares two content types for equality.

Tries to find a content type based on the mime type name.

Since: 2.18

Gets the human readable description of the content type.

Gets the generic icon name for a content type.

See the shared-mime-info specification for more on the generic icon name.

Since: 2.34

Gets the icon for a content type.

Get the list of directories which MIME data is loaded from. See contentTypeSetMimeDirs for details.

Since: 2.60

Gets the mime type for the content type, if one is registered.

Gets the symbolic icon for a content type.

Since: 2.34

Guesses the content type based on example data. If the function is uncertain, resultUncertain will be set to True. Either filename or data may be Nothing, in which case the guess will be based solely on the other argument.

Tries to guess the type of the tree with root root, by looking at the files it contains. The result is an array of content types, with the best guess coming first.

The types returned all have the form x-content/foo, e.g. x-content/audio-cdda (for audio CDs) or x-content/image-dcf (for a camera memory card). See the shared-mime-info specification for more on x-content types.

This function is useful in the implementation of mountGuessContentType.

Since: 2.18

Determines if type is a subset of supertype.

Determines if type is a subset of mimeType. Convenience wrapper around contentTypeIsA.

Since: 2.52

Checks if the content type is the generic "unknown" type. On UNIX this is the "application/octet-stream" mimetype, while on win32 it is "*" and on OSX it is a dynamic type or octet-stream.

Set the list of directories used by GIO to load the MIME database. If dirs is Nothing, the directories used are the default:

• the mime subdirectory of the directory in $XDG_DATA_HOME
• the mime subdirectory of every directory in $XDG_DATA_DIRS

This function is intended to be used when writing tests that depend on information stored in the MIME database, in order to control the data.

Typically, in case your tests use TEST_OPTION_ISOLATE_DIRS, but they depend on the system’s MIME database, you should call this function with dirs set to Nothing before calling g_test_init(), for instance:

C code

 // Load MIME data from the system
 g_content_type_set_mime_dirs (NULL);
 // Isolate the environment
 g_test_init (&argc, &argv, G_TEST_OPTION_ISOLATE_DIRS, NULL);

 …

 return g_test_run ();

Since: 2.60

Gets a list of strings containing all the registered content types known to the system. The list and its data should be freed using g_list_free_full (list, g_free).

Escape string so it can appear in a D-Bus address as the value part of a key-value pair.

For instance, if string is /run/bus-for-:0, this function would return /run/bus-for-%3A0, which could be used in a D-Bus address like unix:nonce-tcp:host=127.0.0.1,port=42,noncefile=/run/bus-for-%3A0.

Since: 2.36

Synchronously looks up the D-Bus address for the well-known message bus instance specified by busType. This may involve using various platform specific mechanisms.

The returned address will be in the D-Bus address format.

Since: 2.26

Asynchronously connects to an endpoint specified by address and sets up the connection so it is in a state to run the client-side of the D-Bus authentication conversation. address must be in the D-Bus address format.

When the operation is finished, callback will be invoked. You can then call dbusAddressGetStreamFinish to get the result of the operation.

This is an asynchronous failable function. See dbusAddressGetStreamSync for the synchronous version.

Since: 2.26

Finishes an operation started with dbusAddressGetStream.

Since: 2.26

Synchronously connects to an endpoint specified by address and sets up the connection so it is in a state to run the client-side of the D-Bus authentication conversation. address must be in the D-Bus address format.

This is a synchronous failable function. See dbusAddressGetStream for the asynchronous version.

Since: 2.26

Generate a D-Bus GUID that can be used with e.g. dBusConnectionNew.

See the D-Bus specification regarding what strings are valid D-Bus GUID (for example, D-Bus GUIDs are not RFC-4122 compliant).

Since: 2.26

Converts a Value to a GVariant of the type indicated by the type parameter.

The conversion is using the following rules:

• G_TYPE_STRING: 's', 'o', 'g' or 'ay'
• G_TYPE_STRV: 'as', 'ao' or 'aay'
• G_TYPE_BOOLEAN: 'b'
• G_TYPE_UCHAR: 'y'
• G_TYPE_INT: 'i', 'n'
• G_TYPE_UINT: 'u', 'q'
• G_TYPE_INT64 'x'
• G_TYPE_UINT64: 't'
• G_TYPE_DOUBLE: 'd'
• G_TYPE_VARIANT: Any VariantType

This can fail if e.g. gvalue is of type G_TYPE_STRING and type is ['i'][G-VARIANT-TYPE-INT32:CAPS]. It will also fail for any GType (including e.g. G_TYPE_OBJECT and G_TYPE_BOXED derived-types) not in the table above.

Note that if gvalue is of type G_TYPE_VARIANT and its value is Nothing, the empty GVariant instance (never Nothing) for type is returned (e.g. 0 for scalar types, the empty string for string types, '/' for object path types, the empty array for any array type and so on).

See the dbusGvariantToGvalue function for how to convert a GVariant to a Value.

Since: 2.30

Converts a GVariant to a Value. If value is floating, it is consumed.

The rules specified in the dbusGvalueToGvariant function are used - this function is essentially its reverse form. So, a GVariant containing any basic or string array type will be converted to a Value containing a basic value or string array. Any other GVariant (handle, variant, tuple, dict entry) will be converted to a Value containing that GVariant.

The conversion never fails - a valid Value is always returned in outGvalue.

Since: 2.30

Checks if string is a D-Bus address.

This doesn't check if string is actually supported by DBusServer or DBusConnection - use dbusIsSupportedAddress to do more checks.

Since: 2.26

Checks if string is a D-Bus GUID.

See the D-Bus specification regarding what strings are valid D-Bus GUID (for example, D-Bus GUIDs are not RFC-4122 compliant).

Since: 2.26

Checks if string is a valid D-Bus interface name.

Since: 2.26

Checks if string is a valid D-Bus member (e.g. signal or method) name.

Since: 2.26

Checks if string is a valid D-Bus bus name (either unique or well-known).

Since: 2.26

Like dbusIsAddress but also checks if the library supports the transports in string and that key/value pairs for each transport are valid. See the specification of the D-Bus address format.

Since: 2.26

Checks if string is a valid D-Bus unique bus name.

Since: 2.26

Converts errno.h error codes into GIO error codes. The fallback value IOErrorEnumFailed is returned for error codes not currently handled (but note that future GLib releases may return a more specific value instead).

As errno is global and may be modified by intermediate function calls, you should save its value as soon as the call which sets it

Gets the GIO Error Quark.

Scans all the modules in the specified directory, ensuring that any extension point implemented by a module is registered.

This may not actually load and initialize all the types in each module, some modules may be lazily loaded and initialized when an extension point it implementes is used with e.g. iOExtensionPointGetExtensions or iOExtensionPointGetExtensionByName.

If you need to guarantee that all types are loaded in all the modules, use g_io_modules_load_all_in_directory().

Since: 2.24

Scans all the modules in the specified directory, ensuring that any extension point implemented by a module is registered.

This may not actually load and initialize all the types in each module, some modules may be lazily loaded and initialized when an extension point it implementes is used with e.g. iOExtensionPointGetExtensions or iOExtensionPointGetExtensionByName.

If you need to guarantee that all types are loaded in all the modules, use g_io_modules_load_all_in_directory().

Since: 2.30

Cancels all cancellable I/O jobs.

A job is cancellable if a Cancellable was passed into ioSchedulerPushJob.

Schedules the I/O job to run in another thread.

notify will be called on userData after jobFunc has returned, regardless whether the job was cancelled or has run to completion.

If cancellable is not Nothing, it can be used to cancel the I/O job by calling cancellableCancel or by calling ioSchedulerCancelAllJobs.

Creates a keyfile-backed SettingsBackend.

The filename of the keyfile to use is given by filename.

All settings read to or written from the backend must fall under the path given in rootPath (which must start and end with a slash and not contain two consecutive slashes). rootPath may be "/".

If rootGroup is non-Nothing then it specifies the name of the keyfile group used for keys that are written directly below rootPath. For example, if rootPath is "/apps/example/" and rootGroup is "toplevel", then settings the key "/apps/example/enabled" to a value of True will cause the following to appear in the keyfile:

 [toplevel]
 enabled=true

If rootGroup is Nothing then it is not permitted to store keys directly below the rootPath.

For keys not stored directly below rootPath (ie: in a sub-path), the name of the subpath (with the final slash stripped) is used as the name of the keyfile group. To continue the example, if "/apps/example/profiles/default/font-size" were set to 12 then the following would appear in the keyfile:

 [profiles/default]
 font-size=12

The backend will refuse writes (and return writability as being False) for keys outside of rootPath and, in the event that rootGroup is Nothing, also for keys directly under rootPath. Writes will also be refused if the backend detects that it has the inability to rewrite the keyfile (ie: the containing directory is not writable).

There is no checking done for your key namespace clashing with the syntax of the key file format. For example, if you have '[' or ']' characters in your path names or '=' in your key names you may be in trouble.

The backend reads default values from a keyfile called defaults in the directory specified by the GKeyfileSettingsBackend:defaults-dir property, and a list of locked keys from a text file with the name locks in the same location.

Creates a memory-backed SettingsBackend.

This backend allows changes to settings, but does not write them to any backing storage, so the next time you run your application, the memory backend will start out with the default values again.

Since: 2.28

Initializes the platform networking libraries (eg, on Windows, this calls WSAStartup()). GLib will call this itself if it is needed, so you only need to call it if you directly call system networking functions (without calling any GLib networking functions first).

Since: 2.36

Creates a readonly SettingsBackend.

This backend does not allow changes to settings, so all settings will always have their default values.

Since: 2.28

Utility method for PollableInputStream and PollableOutputStream implementations. Creates a new Source that expects a callback of type PollableSourceFunc. The new source does not actually do anything on its own; use sourceAddChildSource to add other sources to it to cause it to trigger.

Since: 2.28

Utility method for PollableInputStream and PollableOutputStream implementations. Creates a new Source, as with pollableSourceNew, but also attaching childSource (with a dummy callback), and cancellable, if they are non-Nothing.

Since: 2.34

Tries to read from stream, as with inputStreamRead (if blocking is True) or pollableInputStreamReadNonblocking (if blocking is False). This can be used to more easily share code between blocking and non-blocking implementations of a method.

If blocking is False, then stream must be a PollableInputStream for which pollableInputStreamCanPoll returns True, or else the behavior is undefined. If blocking is True, then stream does not need to be a PollableInputStream.

Since: 2.34

Tries to write to stream, as with outputStreamWrite (if blocking is True) or pollableOutputStreamWriteNonblocking (if blocking is False). This can be used to more easily share code between blocking and non-blocking implementations of a method.

If blocking is False, then stream must be a PollableOutputStream for which pollableOutputStreamCanPoll returns True or else the behavior is undefined. If blocking is True, then stream does not need to be a PollableOutputStream.

Since: 2.34

Tries to write count bytes to stream, as with outputStreamWriteAll, but using pollableStreamWrite rather than outputStreamWrite.

On a successful write of count bytes, True is returned, and bytesWritten is set to count.

If there is an error during the operation (including IOErrorEnumWouldBlock in the non-blocking case), False is returned and error is set to indicate the error status, bytesWritten is updated to contain the number of bytes written into the stream before the error occurred.

As with pollableStreamWrite, if blocking is False, then stream must be a PollableOutputStream for which pollableOutputStreamCanPoll returns True or else the behavior is undefined. If blocking is True, then stream does not need to be a PollableOutputStream.

Since: 2.34

Returns all the names of children at the specified path in the set of globally registered resources. The return result is a Nothing terminated list of strings which should be released with strfreev.

lookupFlags controls the behaviour of the lookup.

Since: 2.32

Looks for a file at the specified path in the set of globally registered resources and if found returns information about it.

lookupFlags controls the behaviour of the lookup.

Since: 2.32

Looks for a file at the specified path in the set of globally registered resources and returns a Bytes that lets you directly access the data in memory.

The data is always followed by a zero byte, so you can safely use the data as a C string. However, that byte is not included in the size of the GBytes.

For uncompressed resource files this is a pointer directly into the resource bundle, which is typically in some readonly data section in the program binary. For compressed files we allocate memory on the heap and automatically uncompress the data.

lookupFlags controls the behaviour of the lookup.

Since: 2.32

Looks for a file at the specified path in the set of globally registered resources and returns a InputStream that lets you read the data.

lookupFlags controls the behaviour of the lookup.

Since: 2.32

Registers the resource with the process-global set of resources. Once a resource is registered the files in it can be accessed with the global resource lookup functions like resourcesLookupData.

Since: 2.32

Unregisters the resource from the process-global set of resources.

Since: 2.32

Reports an error in an idle function. Similar to g_simple_async_report_error_in_idle(), but takes a GError rather than building a new one.

Determines if mountPath is considered an implementation of the OS. This is primarily used for hiding mountable and mounted volumes that only are used in the OS and has little to no relevance to the casual user.

Determines if devicePath is considered a block device path which is only used in implementation of the OS. This is primarily used for hiding mounted volumes that are intended as APIs for programs to read, and system administrators at a shell; rather than something that should, for example, appear in a GUI. For example, the Linux /proc filesystem.

The list of device paths considered ‘system’ ones may change over time.

Since: 2.56

Determines if fsType is considered a type of file system which is only used in implementation of the OS. This is primarily used for hiding mounted volumes that are intended as APIs for programs to read, and system administrators at a shell; rather than something that should, for example, appear in a GUI. For example, the Linux /proc filesystem.

The list of file system types considered ‘system’ ones may change over time.

Since: 2.56

Gets a UnixMountEntry for a given mount path. If timeRead is set, it will be filled with a unix timestamp for checking if the mounts have changed since with unixMountsChangedSince.

If more mounts have the same mount path, the last matching mount is returned.

Compares two unix mounts.

Makes a copy of mountEntry.

Since: 2.54

Gets a UnixMountEntry for a given file path. If timeRead is set, it will be filled with a unix timestamp for checking if the mounts have changed since with unixMountsChangedSince.

If more mounts have the same mount path, the last matching mount is returned.

Since: 2.52

Frees a unix mount.

Gets the device path for a unix mount.

Gets the filesystem type for the unix mount.

Gets the mount path for a unix mount.

Gets a comma-separated list of mount options for the unix mount. For example, rw,relatime,seclabel,data=ordered.

This is similar to unixMountPointGetOptions, but it takes a UnixMountEntry as an argument.

Since: 2.58

Gets the root of the mount within the filesystem. This is useful e.g. for mounts created by bind operation, or btrfs subvolumes.

For example, the root path is equal to "/" for mount created by "mount /dev/sda1 /mnt/foo" and "/bar" for "mount --bind /mnt/foo/bar /mnt/bar".

Since: 2.60

Guesses whether a Unix mount can be ejected.

Guesses the icon of a Unix mount.

Guesses the name of a Unix mount. The result is a translated string.

Guesses whether a Unix mount should be displayed in the UI.

Guesses the symbolic icon of a Unix mount.

Since: 2.34

Checks if a unix mount is mounted read only.

Checks if a Unix mount is a system mount. This is the Boolean OR of unixIsSystemFsType, unixIsSystemDevicePath and unixIsMountPathSystemInternal on mountEntry’s properties.

The definition of what a ‘system’ mount entry is may change over time as new file system types and device paths are ignored.

Checks if the unix mount points have changed since a given unix time.

Gets a List of UnixMountPoint containing the unix mount points. If timeRead is set, it will be filled with the mount timestamp, allowing for checking if the mounts have changed with unixMountPointsChangedSince.

Checks if the unix mounts have changed since a given unix time.

Gets a List of UnixMountEntry containing the unix mounts. If timeRead is set, it will be filled with the mount timestamp, allowing for checking if the mounts have changed with unixMountsChangedSince.