pytams.taskrunner
=================

.. py:module:: pytams.taskrunner


Exceptions
----------

.. autoapisummary::

   pytams.taskrunner.RunnerError


Classes
-------

.. autoapisummary::

   pytams.taskrunner.WorkerLoggerPlugin
   pytams.taskrunner.BaseRunner
   pytams.taskrunner.AsIORunner
   pytams.taskrunner.DaskRunner


Functions
---------

.. autoapisummary::

   pytams.taskrunner.get_runner_type


Module Contents
---------------

.. py:class:: WorkerLoggerPlugin(params: dict[Any, Any])

   Bases: :py:obj:`dask.distributed.WorkerPlugin`


   A plugin to configure logging on each worker.


   .. py:method:: setup(worker: dask.distributed.Worker) -> None

      Configure logging on the worker.

      :param worker: the dask worker



.. py:exception:: RunnerError

   Bases: :py:obj:`Exception`


   Exception class for the runner.


.. py:class:: BaseRunner(params: dict, sync_wk: collections.abc.Callable, n_workers: int = 1)

   An ABC for the task runners.


   .. py:method:: make_promise(task: list[Any]) -> None
      :abstractmethod:


      Log a new task to the list of task to tackle.



   .. py:method:: execute_promises() -> Any
      :abstractmethod:


      Execute the list of promises.



   .. py:method:: n_workers() -> int
      :abstractmethod:


      Return the number of workers in the runner.



.. py:class:: AsIORunner(params: dict, sync_wk: collections.abc.Callable, n_workers: int = 1)

   Bases: :py:obj:`BaseRunner`


   A task runner class based on asyncIO.

   An runner that relies on asyncio to schedule
   a tasks concurently in worker processes.
   Tasks are added to an internal queue from
   which worker can take them and put the results
   back into result queue.


   .. py:method:: add_task(task: list[Any]) -> None
      :async:


      Append a task to the queue.



   .. py:method:: make_promise(task: list[Any]) -> None

      A synchronous wrapper to add_task.



   .. py:method:: run_tasks() -> list[Any]
      :async:


      Create worker tasks and run.



   .. py:method:: execute_promises() -> Any

      A synchronous wrapper to run_tasks.



   .. py:method:: n_workers() -> int

      Return the number of workers in the runner.



.. py:class:: DaskRunner(params: dict, sync_wk: collections.abc.Callable, n_workers: int = 1)

   Bases: :py:obj:`BaseRunner`


   A task runner class based on Dask.

   An runner that relies on dask to schedule
   a tasks concurently in workers.


   .. py:attribute:: dask_backend


   .. py:method:: make_promise(task: list[Any]) -> None

      Append a task to the internal task list.



   .. py:method:: just_delay(obj: Any) -> Any

      Delay an object.



   .. py:method:: execute_promises() -> Any

      Execute a list of promises.

      :param list_of_p: a list of dask promises

      :returns: A list with the return argument of each promised task.

      :raises Exception if compute fails (raise internal error):



   .. py:method:: n_workers() -> int

      Return the number of workers in the runner.



.. py:function:: get_runner_type(params: dict) -> type[BaseRunner]

   Create an engine from parameters.