action_tutorials_py

Python action tutorial code

README

Action Server

In the constructor for FibonacciActionServer, an action server is created with a callback that is called when a goal is accepted.

self._action_server = ActionServer(
            self,
            Fibonacci,
            'fibonacci',
            self.execute_callback)

There are three types of callbacks:

A goal_callback can optionally be added to conditionally accept or reject the goal, however, by default the goal is accepted.
A cancel_callback can also optionally be added to conditionally accept or reject the cancel goal request, however, by default the cancel goal request is accepted.
The execute_callback calculates the Fibonacci sequence up to order and publishes partial sequences as feedback as each item is added to the sequence.

The thread sleeps for 1 second between the calculation of each item in order to represent a long-running task. When execution is complete, the full sequence is returned to the action client.

Action Client

In the constructor for FibonacciActionClient, an action client for the fibonacci action is created:

self._action_client = ActionClient(self, Fibonacci, 'fibonacci')

A goal of type Fibonacci is created with order 10. The goal is sent asynchronously with callbacks registered for the goal response and the feedback:

self._send_goal_future = self._action_client.send_goal_async(
            goal_msg,
            feedback_callback=self.feedback_callback)

self._send_goal_future.add_done_callback(self.goal_response_callback)

Within the goal_response_callback, if the goal is accepted, the goal result is requested asynchronously. A callback is registered for receiving the result:

self._get_result_future = goal_handle.get_result_async()

self._get_result_future.add_done_callback(self.get_result_callback)

There are two types of callbacks:

The feedback_callback logs the partial sequences as they are received.
The get_result_callback logs the full Fibonacci sequence.