-
-
Notifications
You must be signed in to change notification settings - Fork 75
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Github action to run pytest and coverage #191
Github action to run pytest and coverage #191
Conversation
On my fork the tests run and pass for all versions jagerber48#1 |
Could this be merged? I don't have privaleges to do so. |
I was hoping to get the GitHub actions to actually run before merging but that requires more permissions than I had before. I’ll check if I have more permissions now to make them run. |
From the way I interpret https://github.com/orgs/community/discussions/25746, we can't get python-package.yml to run for a PR for testing until some python-package.yml is already in the main branch. An alternative for testing would be pushing to a separate branch in the repo. It seems like GitHub Actions are enabled in the repository settings now. Perhaps we should just merge this and follow up if something doesn't work? Perhaps we should confirm what matrix to run tests on though. Are we happy just testing on Linux and testing 3.8 through 3.12? Do we want to run tests on Windows and Mac? The full matrix might be overkill but testing at least one Python version on Windows and Mac might be good. Should we remove the appveyor config and integration which tests on Windows and the .travis.yml config which was testing Linux (but seems to be disabled already) here? |
@wshanks ok I see. I was suspicious there was something funny with this being the first time the workflow is showing up. I'm tempted to create another branch to push to and test it there, but merging and following also sounds reasonable (it worked in my fork). In the other threads I think we've been converging on testing 3.8-3.12 I don't see any reason not to test on Windows and Mac, I can modify the matrix and see how it goes.
I don't have familiarity with either of those. I see the appveyor check failing. Where does the I guess I need to learn (or we need to define) what standard practice is in this repo for merging into master and eventually making releases. e.g. updating the changelog etc. |
Somehow this problem didn't show up on my fork. See jagerber48#1. There the PR adds the new |
I think this is why it works on your fork. It can't run across forks until it is on the main branch. |
This reverts commit d0f28ff.
If I include windows and mac in the matrix in parallel then there are 12 cases. Feels like kind of a lot, but what are there downsides to having too many test configurations? For some reason the failing appveyer check went away. |
Ok, I made a local branch and merged this PR into that one, now I made a new PR from the local branch into local master and the tests run: Now we should decide exactly how we want the tests to run. |
Adds a github action that runs pytest and coverage against python versions 3.8 - 3.12.
I think this will require some configuration on the repo side to allow the actions to run.