pandocfilters
pandocfilters copied to clipboard
Request to publish wheels for this package
Hi! Thank you for making this package ❤️
I am using a build tool called bazel which only works with python wheels: https://github.com/tubular/rules_pygen
It seems that wheels are currently not published for the pandocfilters package. Please could you publish wheels for this package?
The instructions at https://pythonwheels.com/ may be helpful, it says that once you have installed the wheel package, where you would normally run python setup.py sdist, instead you can run python setup.py sdist bdist_wheel.
Duplicated. See #55.
For pure Python packages, to create a wheel along with source distribution:
(venv) $ pip install --upgrade pip setuptools wheel
(venv) $ python setup.py sdist bdist_wheel
# See dist/*.whl
To upload wheels:
(venv) $ pip install twine
(venv) $ twine upload dist/*
Duplicated. See #55.
As mentioned in the other thread, I barely maintain this now. I actually never use Python and don't know the first thing about its packaging infrastructure. It would really be better if someone else could take over maintenance of the pandocfilters package; I just have too much on my plate as it is...
I can help maintaining it if you want.
@ickc that would be great, I'll give you privileges on this repo.
Thanks!
Do you want me to manage the package at PyPI too? In that case you'll need to invite me as maintainer at PyPI as well. The handle is the same (ickc.)
A few things I probably will be working on:
- go through the issues/PRs
- add CI testing, perhaps improve the "Compatibility" section in the README
- may be add type annotations, which means that at some point we're going to drop Python2
- do a final release for Python2 first before dropping it
- improve the dist as mention in this issue
- may work with conda-forge recipe maintainer(s) to add run constrains matching pandoc versions.
Does that sound good?
I think it will be more about maintaining than adding any new things. As I'm helping maintaining panflute too, that would be the recommended tool to use. But there exists an ecosystem (especially in the backend like RMarkdown, jupyter) that depends on this so this should be continued to be maintained in the foreseeable future.
Great, I've invited you as maintainer.
That all sounds good to me.