-
-
Notifications
You must be signed in to change notification settings - Fork 481
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
sagemath-standard: fix manifest #37286
Conversation
Documentation preview for this PR (built with commit ac629c5; changes) is ready! 🎉 |
@mkoeppe I think this is also easy and uncontroversial. The workaround for setuptools_scm >= 8 is I think much better (and maybe we should also remove the previous workaround, but I'm not sure and it's certainly harmless to apply both). |
I'll take a look |
In several other setup.py files, we have another workaround that somehow didn't make it into pkgs/sagemath-standard. Take a look at https://github.com/sagemath/sage/blob/develop/pkgs/sage-docbuild/setup.py#L13 |
The workaround here just replaces with a noop the function The other seems a workaround for a different issue, I don't understand how it works and its implications. The warning "LookupError" seemed harmless. |
This seems to work well, thanks! The next time I touch the setup.py files in the other disitributions, I'll check if some of the other workarounds for setuptools_scm can be replaced by this. |
Hi guys, this is the only ticket included in 10.4.beta4 that I think may be related to what I just found. When I create a sdist for sagemath-standard-10.4.beta3 there are no python files (.py) in the sage folder. Only cython files. The only python files present in the sdist are at the top level (setup.py) and inside bin (sage-num-threads.py and sage-startuptime.py). Do you have any ideas? |
Indeed the 10.4.beta3 sdist (also on PyPI) is broken in this way. I'll investigate. |
How do you create the sdist? Is the repo bootstrapped? What version of setuptools? Can you try reverting each one of the two changes by hand to see which one, if any, is the culprit? |
I have a simple fix in #37804 |
Looks like something in #36964 is at fault. |
This PR fixes two issues with the manifest for sagemath-standard:
The use of
global-include <PATTERN>
is inconvenient since it willinclude any file matching the pattern anywhere; replace it with
recursive-include sage <PATTERN>
which is what we really mean.The workaround already there is not good for setuptools_scm >= 8.
Implement a more robust workaround (should work for any version of
setuptools_scm but I have not tested, so I did not remove the old
workaround -- it's harmless to apply both).
📝 Checklist