Backfill distutils.log.Log for compatibility

[abravalheri]

in pypa/setuptools#3693 we learned that some projects rely on distutils.log.Log and users are asking for distutils.log.Log to be added back to ensure compatibility is not broken.

I am not sure what is the best way forward, but I am providing this PR if pypa/distutils' maintainers are interested in the "backfill" approach.

• I found just a few usages in https://grep.app/search?q=from%20distutils.log%20import%20Log,
  • However it seams that the main problem is users importing numpy.disutils without pinning setuptools.
• I am not sure about adding tests. I noticed that all the tests about deprecated distutils.log behaviour were removed.

[abravalheri]

The error in ci_setuptools is probably solved by pypa/setuptools#3701

[rgommers]

Thanks @abravalheri, this PR is much appreciated.

However it seams that the main problem is users importing numpy.disutils without pinning setuptools.

This is also for usage at runtime, and we've got issues in e.g. Debian which simply ships only one setuptools version, so there is nothing to pin. Released numpy versions are out there using Log, can't yank those. So this seems like a good way to keep things from falling apart. Can be done without tests imho, it's just a band-aid.

[abravalheri]

Thank you very much @rgommers.

The way I manually tested this was:

git clone https://github.com/abravalheri/distutils --branch setuptools-issue-3693
git clone https://github.com/pypa/setuptools
rm -rf setuptools/setuptools/_distutils
cp -r distutils/distutils setuptools/setuptools/_distutils
virtualenv .venv
.venv/bin/python -m pip install -U pip ./setuptools
.venv/bin/python -m pip install -m numpy
.venv/bin/python -c '
import distutils

assert "setuptools/_distutils" in next(iter(distutils.__path__))

from numpy.distutils import log

log.warn("hello world")

import logging
assert isinstance(log.Log(), logging.Logger), f"{log.Log.__mro__=}"
'

After that I can see a coloured `hello world` message, so I am assuming the patch works well.

[jaraco]

Brilliant. Thank you.