Skip to content

Conversation

@tony-p
Copy link

@tony-p tony-p commented Nov 20, 2025

Fixes bug that ament clang tools don't work on windows

Credit to @traversaro for pointing to the proposed fix

… and move to common implementation in ament_lint

Signed-off-by: Tony Paulussen <tony.paulussen@wefabricate.com>
…ang* execution

Signed-off-by: Tony Paulussen <tony.paulussen@wefabricate.com>
@tony-p tony-p force-pushed the fix/ament-clang-tools-windows-support branch from f7ccc78 to 5568e8f Compare November 20, 2025 13:26
Signed-off-by: Tony Paulussen <tony.paulussen@wefabricate.com>
Signed-off-by: Tony Paulussen <tony.paulussen@wefabricate.com>
Signed-off-by: Tony Paulussen <tony.paulussen@wefabricate.com>
@tony-p tony-p requested a review from ahcorde November 20, 2025 21:29
@tony-p
Copy link
Author

tony-p commented Nov 20, 2025

A general question, what is the best way of dealing with linting (flake/pep257) errors. If I run the tests locally, it throws up 100s of errors (more than what jenkins shows) is that because they are the only regressions? What is the best way to go about fixing them? If I run black or isort for simple fixes it changes way more than what I have touched which also doesn't seem ideal.

TLDR; is there a simple way of processing these style errors?

@ahcorde
Copy link
Contributor

ahcorde commented Nov 24, 2025

Pulls: #563
Gist: https://gist.githubusercontent.com/ahcorde/d4bf250d2d0f2565746a0eaa3a037793/raw/46291b259a37cb6d5fff9db63331ef97cb867186/ros2.repos
BUILD args: --packages-above-and-dependencies ament_clang_format
TEST args: --packages-above ament_clang_format
ROS Distro: rolling
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/17588

  • Linux Build Status
  • Linux-aarch64 Build Status
  • Linux-rhel Build Status
  • Windows Build Status

@tony-p
Copy link
Author

tony-p commented Nov 24, 2025

@ahcorde I see the windows test is failing due to not being able to find clang format (https://ci.ros2.org/job/ci_windows/26030/consoleFull), are the clang tools installed on the windows runners? I see they are not in the pixi environment for windows development https://raw.githubusercontent.com/ros2/ros2/refs/heads/rolling/pixi.toml

@ahcorde
Copy link
Contributor

ahcorde commented Dec 5, 2025

@tony-p Dependency was added here ros2/ros2#1776

@ahcorde
Copy link
Contributor

ahcorde commented Dec 5, 2025

  • Windows Build Status

@ahcorde
Copy link
Contributor

ahcorde commented Dec 8, 2025

@tony-p can you check in the last build that everything is as expected ?

@tony-p
Copy link
Author

tony-p commented Dec 9, 2025

@tony-p can you check in the last build that everything is as expected ?

@ahcorde The logs show that it still didn't find clang format. I double checked on my windows and that works

00:32:00 test\test_execution.py:23: in test_clang_format_execution
00:32:00 
00:32:00     assert rc == 0, 'Clang format found issues'
00:32:00 
00:32:00 E   AssertionError: Clang format found issues
00:32:00 
00:32:00 E   assert 1 == 0
00:32:00 
00:32:00 ---------------------------- Captured stderr call -----------------------------
00:32:00 
00:32:00 Could not find 'clang-format' / 'clang-format-3.8' / 'clang-format-3.7' / 'clang-format-3.6' / 'clang-format-3.5' / 'clang-format-3.4' / 'clang-format-3.3' executable

In the logs it shows a pixi list command seems to be run, but doesn't show any output, so I can't really see what packages are installed (or should I look somewhere else?)

21:33:46 Step 18/36 : RUN pixi --color never --no-progress -q list
21:33:46  ---> Using cache
21:33:46  ---> edb348f4655b

I also see the fetching of the pixi.toml seems to use a cache, could it be that it also got something out of date?

21:33:46 Step 16/36 : RUN powershell -noexit irm https://raw.githubusercontent.com/ros2/ros2/refs/heads/%ROS_DISTRO%/pixi.toml -OutFile pixi.toml
21:33:46  ---> Using cache
21:33:46  ---> c4515bc73215

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants