[setup.py] avoid relative import of _info.py #25
Closed
gph82
wants to merge 1 commits from gph82:master
into master
Loading…
Reference in new issue
There is no content yet.
Delete Branch 'gph82:master'
Deleting a branch is permanent. It CANNOT be undone. Continue?
Hi! I think you now prefer issues/PRs in codeberg, right?
Also in a conda environment and with conda's pip, I get the same
ModuleNotFoundError: No module named '_info'
error, whichi is probably having to do with the absolute vs relative import.
I tested this fix locally and it worked. I am not an expert in these packaging issues and the interplay packaga managers with the different setuptools version they use, but I think this is the way to go.
Hi @gph82 yes, Codeberg is preferred, thanks for the report and the PR!
I'm a bit unsure where the issue comes from and why the change fixes it: it seems to remain an absolute import (there's no relative import like
._info
).I'll test it on my side an come back to you as soon as I figure it out for myself.
I just checked by uploading your fix on the test Pypi server and I'm afraid that, as I expected, it does not solve the issue: it leads to
I don't know why it works locally but I guess conda does something spacific during the pip install...
I think I found a fix, will make a new PR and ask you to check it.
Before I do it, can you check that
pip install --index-url https://test.pypi.org/simple/ mpl_chord_diagram==0.3.2
works in your conda env?hm...yeah, I'm also not sure.
Just to be sure:
Thanks!
I used a new conda environment with Python 3.6 for which I'm pretty sure I have nothing installed anywhere... I'm not very familiar with conda but I think that should be close enough to a fresh install.
My fix seems to work, it is packaged as 0.3.2 on Pypi test, does it also work for you? (
pip install --index-url https://test.pypi.org/simple/ mpl_chord_diagram==0.3.2
)Hi, I'm testing this on different envs (with and without conda) and preparing a better report that makes sense SMH, because it's too much info. But hang in there, I'm interested in getting this sorted out ;)
OK, I'm closing this PR for now, can you summarize what you'll find in a new issue?
Sure!