
Gnome::Gio::R-AsyncResult
Description
Gnome::Gio::R-AsyncResult provides a base class for implementing asynchronous function results.
Asynchronous operations are broken up into two separate operations which are chained together by a Gnome::Gio::T-iotypes. To begin an asynchronous operation, provide a Gnome::Gio::T-iotypes to the asynchronous function. This callback will be triggered when the operation has completed, and must be run in a later iteration of the thread-default main context (see .push-thread-default()
in class Gnome::Gio::AsyncResult
) from where the operation was initiated. It will be passed a Gnome::Gio::R-AsyncResult instance filled with the details of the operation's success or failure, the object the asynchronous function was started for and any error codes returned. The asynchronous callback function is then expected to call the corresponding _finish()` function, passing the object the function was called for, the Gnome::Gio::R-AsyncResult instance, and (optionally) an $error
to grab any error conditions that may have occurred.
The _finish()` function for an operation takes the generic result (of type Gnome::Gio::R-AsyncResult) and returns the specific result that the operation in question yields (e.g. a Gnome::Gio::AsyncResult for a "enumerate children" operation). If the result or error status of the operation is not needed, there is no need to call the _finish()` function; GIO will take care of cleaning up the result and error information after the Gnome::Gio::T-iotypes returns. You can pass undefined for the Gnome::Gio::T-iotypes if you don't need to take any action at all after the operation completes. Applications may also take a reference to the Gnome::Gio::R-AsyncResult and call _finish()` later; however, the _finish()` function may be called at most once.
Example of a typical asynchronous operation flow:
The callback for an asynchronous operation is called only once, and is always called, even in the case of a cancelled operation. On cancellation the result is a G_IO_ERROR_CANCELLED error.
I/O Priority
Many I/O-related asynchronous operations have a priority parameter, which is used in certain cases to determine the order in which operations are executed. They are not used to determine system-wide I/O scheduling. Priorities are integers, with lower numbers indicating higher priority. It is recommended to choose priorities between Gnome::Glib::T-main and Gnome::Glib::T-main, with Gnome::Glib::T-main as a default.
Methods
get-source-object
Gets the source object from a Gnome::Gio::R-AsyncResult.
method get-source-object (--> N-Object )
Return value; a new reference to the source object for the $res
, or undefined if there is none..
get-user-data
Gets the user data from a Gnome::Gio::R-AsyncResult.
method get-user-data (--> gpointer )
Return value; the user data for $res
..
is-tagged
Checks if $res
has the given $source-tag
(generally a function pointer indicating the function $res
was created by).
method is-tagged ( gpointer $source-tag --> Bool )
$source-tag; an application-defined tag.
Return value; True
if $res
has the indicated $source-tag
, False
if not..
legacy-propagate-error
If $res
is a Gnome::Gio::SimpleAsyncResult, this is equivalent to g_simple_async_result_propagate_error(). Otherwise it returns False
.
This can be used for legacy error handling in async *_finish() wrapper functions that traditionally handled Gnome::Gio::SimpleAsyncResult error returns themselves rather than calling into the virtual method. This should not be used in new code; Gnome::Gio::R-AsyncResult errors that are set by virtual methods should also be extracted by virtual methods, to enable subclasses to chain up correctly.
method legacy-propagate-error ( CArray[N-Error] $err --> Bool )
$err; Error object. When defined, an error can be returned when there is one. Use
Pointer
when you want to ignore the error. .
Return value; True
if $error
is has been filled in with an error from $res
, False
if not..