MainContext#

class MainContext(**kwargs)#

The GMainContext struct is an opaque data type representing a set of sources to be handled in a main loop.

Constructors#

class MainContext
classmethod new() → MainContext#

Creates a new MainContext structure.

classmethod new_with_flags(flags: MainContextFlags) → MainContext#

Creates a new MainContext structure.

Added in version 2.72.

Parameters:

flags – a bitwise-OR combination of MainContextFlags flags that can only be set at creation time.

Methods#

class MainContext
acquire() → bool#

Tries to become the owner of the specified context. If some other thread is the owner of the context, returns False immediately. Ownership is properly recursive: the owner can require ownership again and will release ownership when release is called as many times as acquire.

You must be the owner of a context before you can call prepare, query, check, dispatch, release.

Since 2.76 context can be None to use the global-default main context.

add_poll(fd: PollFD, priority: int) → None#

Adds a file descriptor to the set of file descriptors polled for this context. This will very seldom be used directly. Instead a typical event source will use g_source_add_unix_fd instead.

Parameters:

  • fd – a PollFD structure holding information about a file descriptor to watch.

  • priority – the priority for this file descriptor which should be the same as the priority used for attach to ensure that the file descriptor is polled whenever the results may be needed.

check(max_priority: int, fds: list[PollFD]) → bool#

Passes the results of polling back to the main loop. You should be careful to pass fds and its length n_fds as received from query, as this functions relies on assumptions on how fds is filled.

You must have successfully acquired the context with acquire before you may call this function.

Since 2.76 context can be None to use the global-default main context.

Parameters:

  • max_priority – the maximum numerical priority of sources to check

  • fds – array of PollFD's that was passed to the last call to query

classmethod default() → MainContext#

Returns the global-default main context. This is the main context used for main loop functions when a main loop is not explicitly specified, and corresponds to the “main” main loop. See also get_thread_default.

dispatch() → None#

Dispatches all pending sources.

You must have successfully acquired the context with acquire before you may call this function.

Since 2.76 context can be None to use the global-default main context.

find_source_by_funcs_user_data(funcs: SourceFuncs, user_data: Any = None) → Source#

Finds a source with the given source functions and user data. If multiple sources exist with the same source function and user data, the first one found will be returned.

Parameters:

  • funcs – the source_funcs passed to new.

  • user_data – the user data from the callback.

find_source_by_id(source_id: int) → Source#

Finds a Source given a pair of context and ID.

It is a programmer error to attempt to look up a non-existent source.

More specifically: source IDs can be reissued after a source has been destroyed and therefore it is never valid to use this function with a source ID which may have already been removed. An example is when scheduling an idle to run in another thread with idle_add: the idle may already have run and been removed by the time this function is called on its (now invalid) source ID. This source ID may have been reissued, leading to the operation being performed against the wrong source.

Parameters:

source_id – the source ID, as returned by get_id.

find_source_by_user_data(user_data: Any = None) → Source#

Finds a source with the given user data for the callback. If multiple sources exist with the same user data, the first one found will be returned.

Parameters:

user_data – the user_data for the callback.

classmethod get_thread_default() → MainContext | None#

Gets the thread-default MainContext for this thread. Asynchronous operations that want to be able to be run in contexts other than the default one should call this method or ref_thread_default to get a MainContext to add their Sources to. (Note that even in single-threaded programs applications may sometimes want to temporarily push a non-default context, so it is not safe to assume that this will always return None if you are running in the default thread.)

If you need to hold a reference on the context, use ref_thread_default instead.

Added in version 2.22.

invoke_full(priority: int, function: Callable[[Any], bool], data: Any = None) → None#

Invokes a function in such a way that context is owned during the invocation of function.

This function is the same as invoke except that it lets you specify the priority in case function ends up being scheduled as an idle and also lets you give a DestroyNotify for data.

notify should not assume that it is called from any particular thread or with any particular context acquired.

Added in version 2.28.

Parameters:

  • priority – the priority at which to run function

  • function – function to call

  • data – data to pass to function

is_owner() → bool#

Determines whether this thread holds the (recursive) ownership of this MainContext. This is useful to know before waiting on another thread that may be blocking to get ownership of context.

Added in version 2.10.

iteration(may_block=True)#

Runs a single iteration for the given main loop. This involves checking to see if any event sources are ready to be processed, then if no events sources are ready and may_block is True, waiting for a source to become ready, then dispatching the highest priority events sources that are ready. Otherwise, if may_block is False sources are not waited to become ready, only those highest priority events sources will be dispatched (if any), that are ready at this given moment without further waiting.

Note that even when may_block is True, it is still possible for iteration to return False, since the wait may be interrupted for other reasons than an event source becoming ready.

Parameters:

may_block – whether the call may block.

pending() → bool#

Checks if any sources have pending events for the given context.

pop_thread_default() → None#

Pops context off the thread-default context stack (verifying that it was on the top of the stack).

Added in version 2.22.

prepare() → tuple[bool, int]#

Prepares to poll sources within a main loop. The resulting information for polling is determined by calling query.

You must have successfully acquired the context with acquire before you may call this function.

push_thread_default() → None#

Acquires context and sets it as the thread-default context for the current thread. This will cause certain asynchronous operations (such as most [gio][gio]-based I/O) which are started in this thread to run under context and deliver their results to its main loop, rather than running under the global default main context in the main thread. Note that calling this function changes the context returned by get_thread_default, not the one returned by default, so it does not affect the context used by functions like idle_add.

Normally you would call this function shortly after creating a new thread, passing it a MainContext which will be run by a MainLoop in that thread, to set a new default context for all async operations in that thread. In this case you may not need to ever call pop_thread_default, assuming you want the new MainContext to be the default for the whole lifecycle of the thread.

If you don’t have control over how the new thread was created (e.g. in the new thread isn’t newly created, or if the thread life cycle is managed by a ThreadPool), it is always suggested to wrap the logic that needs to use the new MainContext inside a push_thread_default / pop_thread_default pair, otherwise threads that are re-used will end up never explicitly releasing the MainContext reference they hold.

In some cases you may want to schedule a single operation in a non-default context, or temporarily use a non-default context in the main thread. In that case, you can wrap the call to the asynchronous operation inside a push_thread_default / pop_thread_default pair, but it is up to you to ensure that no other asynchronous operations accidentally get started while the non-default context is active.

Beware that libraries that predate this function may not correctly handle being used from a thread with a thread-default context. Eg, see g_file_supports_thread_contexts().

Added in version 2.22.

query(max_priority: int) → tuple[int, int, list[PollFD]]#

Determines information necessary to poll this main loop. You should be careful to pass the resulting fds array and its length n_fds as is when calling check, as this function relies on assumptions made when the array is filled.

You must have successfully acquired the context with acquire before you may call this function.

Parameters:

max_priority – maximum priority source to check

release() → None#

Releases ownership of a context previously acquired by this thread with acquire. If the context was acquired multiple times, the ownership will be released only when release is called as many times as it was acquired.

You must have successfully acquired the context with acquire before you may call this function.

remove_poll(fd: PollFD) → None#

Removes file descriptor from the set of file descriptors to be polled for a particular context.

Parameters:

fd – a PollFD descriptor previously added with add_poll

wait(cond: Cond, mutex: Mutex) → bool#

Tries to become the owner of the specified context, as with acquire. But if another thread is the owner, atomically drop mutex and wait on cond until that owner releases ownership or until cond is signaled, then try again (once) to become the owner.

Deprecated since version 2.58:

Use is_owner and separate

locking instead.

Parameters:

  • cond – a condition variable

  • mutex – a mutex, currently held

wakeup() → None#

If context is currently blocking in iteration waiting for a source to become ready, cause it to stop blocking and return. Otherwise, cause the next invocation of iteration to return without blocking.

This API is useful for low-level control over MainContext; for example, integrating it with main loop implementations such as MainLoop.

Another related use for this function is when implementing a main loop with a termination condition, computed from multiple threads:

#define NUM_TASKS 10
static gint tasks_remaining = NUM_TASKS;  // (atomic)
...

while (g_atomic_int_get (&tasks_remaining) != 0)
  g_main_context_iteration (NULL, TRUE);

Then in a thread:

perform_work();

if (g_atomic_int_dec_and_test (&tasks_remaining))
  g_main_context_wakeup (NULL);