Task Execution
==============

Tasks are executed on background using Celery -- function :py:func:`ppc_robot_worker.tasks.tasks.run_job`.
Task execution is described by the following diagram:

.. image:: images/ppc_robot_worker.tasks-sq.png

Before the task is executed, job definition is retrieved from the database -- this is not described in the diagram,
as the worker uses standard Django models.

When the definition is ready, ``run_job`` calls method
:py:meth:`ppc_robot_lib.tasks.task_factory.TaskFactory.create_task` (1). ``TaskFactory`` then selects appropriate
``taskType`` (given by the ``task_type_name`` and ``version`` parameters) and creates a new instance of
the report type (1.1). The factory then creates a new :py:class:`ppc_robot_lib.tasks.task.Task` instance with
reference to the given task and parameters (1.2) and returns it (1.3).

``run_job`` then passes the `Task` instance to :py:meth:`ppc_robot_lib.tasks.task_executor.TaskExecutor.execute` (2).
``TaskExecutor`` calls the :py:meth:`ppc_robot_lib.tasks.abstract_task_type.AbstractFunctionTaskType.execute` (2.1).

After the ``execute()`` method completes, ``run_job`` updates the job status and sends notification to the user.