[main] Split shared Python library to libpython

[h-vetinari]

Partial backport of #842, specifically 6eeda71, as suggested in #885. I also picked up another useful commit from the dev branch while rebasing. Once this PR (and a rebased #885) are merged, I'm happy to backport the pair as combined PRs to the maintenance branches.

[conda-forge-admin]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe/meta.yaml) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe/meta.yaml:

• ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.
• ℹ️ The recipe is not parsable by parser conda-recipe-manager. The recipe can only be automatically migrated to the new v1 format if it is parseable by conda-recipe-manager.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/32334869920. Examine the logs at this URL for more detail.}

[h-vetinari]

Rebased after #888. Could you PTAL @conda-forge/python? :)

[h-vetinari]

@conda-forge-admin, please relint

[isuruf]

libpython should be a dependency of python for python<3.15 to not break anything downstream

[h-vetinari]

Hm. So I guess I need to switch the build order here. Conda can't figure out to build python before libpython and then run-depend on the latter for the former.

Also, quick question, I just stumbled upon https://github.com/conda-forge/libpython-feedstock/blob/main/recipe/meta.yaml (by looking through the package metadata for existing libpython builds, but I can't see those from the python_rc label) - do we need to guard against co-installation with that?

[isuruf]

We should move feedstocks to use distutils-activate-mingw and remove that output there.

[h-vetinari]

And repodata-patch existing users of libpython to point to distutils-activate-mingw and then mark existing libpython builds as broken?

[h-vetinari]

libpython should be a dependency of python

This turned into a pretty big operation (and likely still contains bugs). Not sure if this is along the lines of what you had in mind.

[h-vetinari]

Before spending more time on this, can you confirm whether the last commit here is of the shape & kind that you had in mind @isuruf?

[h-vetinari]

OK, this is freshly rebased and ready now. I also improved the git history around the build order switch. PTAL @isuruf

[h-vetinari]

I guess we should merge conda-forge/conda-forge-repodata-patches-feedstock#1236 at the same time or before (due to the bare libpython dependencies that some packages have). That PR is also ready.

[h-vetinari]

@isuruf, could you review here please? 🙏

[h-vetinari]

Ping @isuruf. It'd be nice to be able to avoid having to keep rebasing this PR for anything that touches the build scripts (e.g. risc-v support)

[isuruf]

Thanks