399 robot is stuck in infinite loop using translate by with wait=true

[FabienDanieau]

The bug was due to the if wait that was not supposed to be triggered with unreachable comment (i.e. there is nothing to wait for). It is fixed by a elif

    response = self._goto_stub.GoToCartesian(request)
    if response.id == -1:
        self._logger.error(f"Target pose:\\n {target} \\nwas not reachable. No command sent.")
    if wait:
        self._logger.info(f"Waiting for movement with {response}.")
        while not self._is_goto_finished(response):
            time.sleep(0.1)
        self._logger.info(f"Movement with {response} finished.")
    return response

I've added this mechanism to all goto functions, even if the goto_joints don't raise unreachable errors for now.
Also there is a timeout in the wait function to be sure that we don't wait forever if something goes bad on the server side (timeout is set to duration + 1 sec.)
+ little bit of refactoring

[github-actions]

☂️ Python Coverage

current status: ✅

Overall Coverage

Lines	Covered	Coverage	Threshold	Status
965	880	91%	0%	🟢

New Files

No new covered files...

Modified Files

File	Coverage	Status
tests/units/online/test_advanced_goto_functions.py	100%	🟢
TOTAL	100%	🟢

updated for commit: db01002 by action🐍

[glannuzel]

I think we should change how timeouts are triggered

[glannuzel]

I think timeout for the gotos should be checked once the movement has begun only.

If we do:

pose2 = get_pose_matrix([0.4, -0.2, -0.1], [0, -90, 0])
reachy.goto_posture("default")
reachy.r_arm.goto_from_matrix(pose2, wait=True)

Then the timeout is triggered, whereas the goto was just queued

[FabienDanieau]

The timeout is now the sum of the duration of all pending movements on the part

[glannuzel]

Awesome!