We might just want to revert the ruff part of the change with a fixup commit and merge this to get the usable updates...

We might just want to revert the ruff part of the change with a fixup commit and merge this to get the usable updates...

Yeah, something in the docs part is grumpy though… seemingly setuptools related:

  File "/home/runner/work/setuptools-pyproject-migration/setuptools-pyproject-migration/.tox/docs/lib/python3.10/site-packages/setuptools/__init__.py", line 35, in <module>
    __version__ = setuptools.version.__version__
AttributeError: partially initialized module 'setuptools' has no attribute 'version' (most likely due to a circular import)

pickling environment... done
checking consistency... done
preparing documents... done
copying assets... copying static files... done
copying extra files... done
done
writing output... [ 50%] history
writing output... [100%] index
generating indices... genindex done
writing additional pages... search done
dumping search index in English (code: en)... done
dumping object inventory... done
build finished with problems, 1 warning.
docs: exit 1 (5.02 seconds) /home/runner/work/setuptools-pyproject-migration/setuptools-pyproject-migration/docs> python -m sphinx -W --keep-going . /home/runner/work/setuptools-pyproject-migration/setuptools-pyproject-migration/build/html pid=1882
.pkg: _exit> python /home/runner/.local/lib/python3.10/site-packages/pyproject_api/_backend.py True setuptools.build_meta
  docs: FAIL code 1 (18.77=setup[13.75]+cmd[5.02] seconds)
  evaluation failed :( (18.87 seconds)

Error: Process completed with exit code 1.

Seemingly, some sleuthing (read: pasting the error into Google Search) yields this: pytorch/pytorch#69904

More specific: pypa/setuptools#3744 (comment)

Okay, so fix made more complicated because setuptools older than 60.10.0 does NOT like Python 3.12 (due to a long deprecated feature being removed), and 60.10.0 itself not liking Python 3.6.

Oh interesting - I was running into that error on my branch for #58 and thought it had something to do with my changes 🤦

What do you think about splitting out the fix for that import error in a separate PR? After all, it's not related to the pre-commit changes, just something that happened to occur around the same time. Also, cutting off eight major versions of setuptools seems pretty drastic; I wonder if there's a way to fix it with a lighter touch... although I suppose getting it working is better than waiting to find the "right" way.

Oh interesting - I was running into that error on my branch for #58 and thought it had something to do with my changes 🤦

What do you think about splitting out the fix for that import error in a separate PR? After all, it's not related to the pre-commit changes, just something that happened to occur around the same time.

Yep, I have no issues doing that… I'll get right onto it.

Also, cutting off eight major versions of setuptools seems pretty drastic; I wonder if there's a way to fix it with a lighter touch... although I suppose getting it working is better than waiting to find the "right" way.

Yes well, I think it all depends on how quickly upstream identify and fix the issue. I was of the thought of "get it working" knowing this was meant to be "temporary". (Yes, fully aware how permanent a "temporary" fix can be.)

Could you rebase this on main? (or otherwise drop the commits that were taken care of in #60)