Update ruff

[micaeljtoliveira]

Closes #36

[micaeljtoliveira]

@manodeep @minghangli-uni One of you keen is reviewing this? Should be simple.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (216bddc) to head (2616acc).
⚠️ Report is 5 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main       #37   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            8         8           
  Lines          455       455           
=========================================
  Hits           455       455           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[minghangli-uni]

Thanks @micaeljtoliveira for this update! This change resolves the immediate version mismatch, but I dont think it's ideal on its own. Updating .pre-commit-config.yaml to v0.15.0 resolves it for now but the root cause remains - we are still maintaining two independent ruff paths. To avoid future drift, I'd prefer to consolidate it so ci runs the same pre-commit hooks as the local does?

Would you mind if I push a couple of commits to this PR?

[micaeljtoliveira]

@minghangli-uni Yes, this way there will be another mismatch in the future.

Would you mind if I push a couple of commits to this PR?

That depends on how you intent to solve the mismatch. I would like to avoid having the version fully pinned, as in that case there's no incentive to update it when new releases of ruff come out.

[minghangli-uni]

Thats fair - I am not suggesting we freeze ruff. The intent is to pin it in one place only, not multiple.

Using pre-commit keeps versions deterministic while still making upgrades easy through pre-commit autoupdate. That way ci and local runs stay aligned and formatting changes only happen when we intentially update but not implicitly via ci pulling a newer ruff.

[micaeljtoliveira]

As long as the decision of when to update ruff is out of the developers' hands, I'm fine with whatever the scheme is. Feel free to push your commits and I'll see if I like what I see ;)

[minghangli-uni]

@micaeljtoliveira I've pushed the changes

[micaeljtoliveira]

Thanks. Looks like the version is indeed fully pinned. I'm assuming your idea is to activate the pre-commit-ci app on this repo. I'm not a huge fan of having external apps creating PRs, but I can live with that. What's more annoying is not being able to have the auto-fix in the hook. Is there a way to avoid that?

[minghangli-uni]

Good question. What comes to my mind now is two options. One is to add something in the doc to tell people run these locally before pushing commits or when ci fails.

ruff format
ruff check --fix

Or to allow ci to push fixes automatically back to the branch?

[micaeljtoliveira]

Good question. What comes to my mind now is two options. One is to add something in the doc to tell people run these locally before pushing commits or when ci fails.

ruff format
ruff check --fix

Yes, we should document this better. There should be a section about the hooks and what they do.
But documentation or not, one still needs to type the command explicitly.

Or to allow ci to push fixes automatically back to the branch?

That, for me, is a definite no, as I find these kind of commits to obscure and pollute the git history.

Okay, no need to waste any more time with this. Current solution is not perfect but will do. I'll clean the history of the branch and will let you have a final look and approve.

[minghangli-uni]

Okay, no need to waste any more time with this. Current solution is not perfect but will do. I'll clean the history of the branch and will let you have a final look and approve.

Agree! Thanks @micaeljtoliveira

[minghangli-uni]

Hi @micaeljtoliveira Just checking whether this is ready for review.

[micaeljtoliveira]

@minghangli-uni Yes. I force-pushed the commits after a rebase, with some very minor changes to the history, last Friday. Should have pinged-up, sorry.

[minghangli-uni]

LGTM, thanks @micaeljtoliveira