1.7.3. Executors

Executors control, if a connector is executed sequentially, in a parallel thread or even a separate process. Use the connectors.executor() function to instantiate an executor.

1.7.3.1. Base class

class connectors._common._executors.Executor

a base class for managing the event loop and the execution in threads or processes.

get_event_loop()

Returns the currently active event loop. This can be None, if no coroutine, task or future is currently being processed.

Returns:the event loop or None

run_coroutine(coro)

Takes a coroutine and runs it in a newly created event loop.

Parameters:coro – the coroutine Returns:the return value of the coroutine

run_coroutines(coros)

Takes multiple coroutines and runs them in a newly created event loop.

Parameters:coros – a sequence of coroutines

run_method(parallelization, method, instance, *args, **kwargs)

Abstract method, whose overrides shall execute the given method. The parallelization shall be implemented in this method.

Parameters:

• parallelization – a flag of connectors.Parallelization, that specifies how the given method can be parallelized
• method – the unbound method, that shall be executed
• instance – the instance of which the method shall be executed
• *args,**kwargs – arguments for the method

Returns:

the return value of the method

run_until_complete(future)

Takes a future or a task and runs it in a newly created event loop. This is a wrapper for the event loop’s run_until_complete() method.

Parameters:future – the future or the task Returns:the return value of the execution

1.7.3.2. Executor classes

class connectors._common._executors.SequentialExecutor

An executor class, that executes everything sequentially.

get_event_loop()

Returns the currently active event loop. This can be None, if no coroutine, task or future is currently being processed.

Returns:the event loop or None

run_coroutine(coro)

Takes a coroutine and runs it in a newly created event loop.

Parameters:coro – the coroutine Returns:the return value of the coroutine

run_coroutines(coros)

Takes multiple coroutines and runs them in a newly created event loop.

Parameters:coros – a sequence of coroutines

run_method(parallelization, method, instance, *args, **kwargs)

Executes the given method sequentially.

Parameters:

• parallelization – a flag of connectors.Parallelization, that specifies how the given method can be parallelized
• method – the unbound method, that shall be executed
• instance – the instance of which the method shall be executed
• *args,**kwargs – arguments for the method

Returns:

the return value of the method

run_until_complete(future)

Takes a future or a task and runs it in a newly created event loop. This is a wrapper for the event loop’s run_until_complete() method.

Parameters:future – the future or the task Returns:the return value of the execution

class connectors._common._executors.ThreadingExecutor(number_of_threads)

An executor class, that can parallelize computations with threads.

Parameters:number_of_threads – the maximum number of threads, that shall be created, or None to determine this number automatically.

get_event_loop()

Returns the currently active event loop. This can be None, if no coroutine, task or future is currently being processed.

Returns:the event loop or None

run_coroutine(coro)

Takes a coroutine and runs it in a newly created event loop.

Parameters:coro – the coroutine Returns:the return value of the coroutine

run_coroutines(coros)

Takes multiple coroutines and runs them in a newly created event loop.

Parameters:coros – a sequence of coroutines

run_method(parallelization, method, instance, *args, **kwargs)

Executes the given method in a thread if possible and falls back to sequential execution if not.

Parameters:

• parallelization – a flag of connectors.Parallelization, that specifies how the given method can be parallelized
• method – the unbound method, that shall be executed
• instance – the instance of which the method shall be executed
• *args,**kwargs – arguments for the method

Returns:

the return value of the method

run_until_complete(future)

Takes a future or a task and runs it in a newly created event loop. This is a wrapper for the event loop’s run_until_complete() method.

Parameters:future – the future or the task Returns:the return value of the execution

class connectors._common._executors.MultiprocessingExecutor(number_of_processes)

An executor class, that can parallelize computations with processes.

Parameters:number_of_processes – the maximum number of processes, that shall be created, or None to determine this number automatically (in this case, the number of CPU cores will be taken).

get_event_loop()

Returns the currently active event loop. This can be None, if no coroutine, task or future is currently being processed.

Returns:the event loop or None

run_coroutine(coro)

Takes a coroutine and runs it in a newly created event loop.

Parameters:coro – the coroutine Returns:the return value of the coroutine

run_coroutines(coros)

Takes multiple coroutines and runs them in a newly created event loop.

Parameters:coros – a sequence of coroutines

run_method(parallelization, method, instance, *args, **kwargs)

Executes the given method in a process if possible and falls back to sequential execution if not.

Parameters:

• parallelization – a flag of connectors.Parallelization, that specifies how the given method can be parallelized
• method – the unbound method, that shall be executed
• instance – the instance of which the method shall be executed
• *args,**kwargs – arguments for the method

Returns:

the return value of the method

run_until_complete(future)

Takes a future or a task and runs it in a newly created event loop. This is a wrapper for the event loop’s run_until_complete() method.

Parameters:future – the future or the task Returns:the return value of the execution

class connectors._common._executors.ThreadingMultiprocessingExecutor(number_of_threads, number_of_processes)

An executor class, that can parallelize computations with both threads and processes.

Parameters:

• number_of_threads – the maximum number of threads, that shall be created, or None to determine this number automatically.
• number_of_processes – the maximum number of processes, that shall be created, or None to determine this number automatically (in this case, the number of CPU cores will be taken).

get_event_loop()

Returns the currently active event loop. This can be None, if no coroutine, task or future is currently being processed.

Returns:the event loop or None

run_coroutine(coro)

Takes a coroutine and runs it in a newly created event loop.

Parameters:coro – the coroutine Returns:the return value of the coroutine

run_coroutines(coros)

Takes multiple coroutines and runs them in a newly created event loop.

Parameters:coros – a sequence of coroutines

run_method(parallelization, method, instance, *args, **kwargs)

Executes the given method in a process if possible and falls back to threaded and then sequential execution if not.

Parameters:

• parallelization – a flag of connectors.Parallelization, that specifies how the given method can be parallelized
• method – the unbound method, that shall be executed
• instance – the instance of which the method shall be executed
• *args,**kwargs – arguments for the method

Returns:

the return value of the method

run_until_complete(future)

Takes a future or a task and runs it in a newly created event loop. This is a wrapper for the event loop’s run_until_complete() method.

Parameters:future – the future or the task Returns:the return value of the execution