Static typing for Message, Services, and Actions

[InvincibleRMC]

Adds static typing to generated messages. Closes #157.

[InvincibleRMC]

Still need to fix a circular import issue.

[christophebedard]

Pulls: #206
Gist: https://gist.githubusercontent.com/christophebedard/9a392164fa167b98f745f9a505a40a73/raw/1cb2fe9a7ed734bbd4a459211613e40209a97991/ros2.repos
BUILD args:
TEST args:
ROS Distro: rolling
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/16562

• Linux
• Linux-aarch64
• Linux-rhel
• Windows

[christophebedard]

I ran full CI^ so that interfaces get generated, and so that hopefully the linters (including mypy) run on the generated files as part of tests in the various interface packages. The Rpr job doesn't really test much here.

[InvincibleRMC]

Thanks so much for the review!

I ran full CI^ so that interfaces get generated, and so that hopefully the linters (including mypy) run on the generated files as part of tests in the various interface packages. The Rpr job doesn't really test much here.

I was going to ask about this since I have a feeling this will expose some downstream bugs to be fixed.

[christophebedard]

Pulls: #206
Gist: https://gist.githubusercontent.com/christophebedard/23340b5f0c0b1e577396c407ded46165/raw/1cb2fe9a7ed734bbd4a459211613e40209a97991/ros2.repos
BUILD args:
TEST args:
ROS Distro: rolling
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/16581

• Linux
• Linux-aarch64
• Linux-rhel
• Windows

[InvincibleRMC]

Another old mypy bug.

import typing

class A:

    def __init__(self) -> typing.NoReturn:   # The return type of "__init__" must be None
        raise NotImplementedError()

Will add a # type: ignore for now. Cannot wait for rhel 10.

[christophebedard]

Pulls: #206
Gist: https://gist.githubusercontent.com/christophebedard/6ea01a76e36980e450b65455f81bcd72/raw/1cb2fe9a7ed734bbd4a459211613e40209a97991/ros2.repos
BUILD args:
TEST args:
ROS Distro: rolling
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/16607

• Linux
• Linux-aarch64
• Linux-rhel
• Windows

[InvincibleRMC]

rhel strikes again 😢

[christophebedard]

@ros-pull-request-builder retest this please

[christophebedard]

Alright, let's try a full CI run again.

[christophebedard]

Pulls: #206
Gist: https://gist.githubusercontent.com/christophebedard/1f6a02ffe82ec8f80bdc504b48fab103/raw/1cb2fe9a7ed734bbd4a459211613e40209a97991/ros2.repos
BUILD args:
TEST args:
ROS Distro: rolling
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/16612

• Linux
• Linux-aarch64
• Linux-rhel
  • Linux-rhel
• Windows

[InvincibleRMC]

@christophebedard the rhel CI crashed can you restart it when you get a chance.

[christophebedard]

It usually gets automatically retried if it fails due to an infra issue, but it's not always easy to find the corresponding job.

ci_linux-rhel 3922 was retried with 3925, which also failed due to an infra issue and was retried with 3929: https://ci.ros2.org/job/ci_linux-rhel/3929/. There's a ros2topic test failure, which is slightly suspicious. It's hard to tell what went wrong, but it could be due to the potential crosstalk we've been seeing. Let's retry it: Linux-rhel

[christophebedard]

The RHEL job had some different test failures; I don't think it's related to these changes. The first one seems to be Zenoh router-related and the second one seems to be a time-related flake.

[christophebedard]

@sloretz do you want to give this another look? Otherwise, I would go ahead and approve + merge it.

[InvincibleRMC]

Thanks so much for all the reviews and finally getting this merged!

[christophebedard]

I created a new release (0.26.0): ros/rosdistro#47213.

The timing is great, because a Rolling sync was just completed, so we'll have time to let the buildfarm rebuild downstream packages and fix anything that breaks before the next sync.

[InvincibleRMC]

Feel free to ping me about any downstream problems you find.