Package Workflow not publishing to PyPI

[banillie]

probably something to do with updating secret keys.

[yulqen]

@banillie - what exactly is the problem? Not sure what "Package Workflow" means. Can help if you're able to describe what you're trying to do.

[banillie]

@yulqen I've created a .yml file in .github/workflows. This publishes Python 🐍 distributions 📦 to PyPI and TestPyPI.

It was working fine until Tuesday when I initiated the workflow to publish the latest version of the package. It ran as expected, but skipped the actual publishing to PYPI. The work log shows this https://github.com/banillie/analysis_engine/actions

Not sure what happened?! I was going to go back over the tutorial I used for setting up the workflow to refresh my memory on how it works.

Cheers.

[yulqen]

@banillie - I can't see the log containing the failure (the URL in your last post is wrong) but looking at your .yml file, there's a bit too much going on in there for me to work out and I've not used a similar workflow before. Depends what the error log says - if you weren't passing authentication at PyPI (i.e. if your secret was wrong), it would tell you. Strange that it fails silently.

[banillie]

@yulqen ok thanks. Will show you tomorrow! Another thing to add to the list!

[banillie]

@yulqen still ok for a coding chat later? What time is good for you?

[banillie]

@yulqen something to do with how I'm version tagging the code git tag v[no] and then git push --tags before or after the commit. The yml file will only push to PyPi if: startsWith(github.ref, 'refs/tags'). I've obviously forgotten the correct steps for tagging.

I've just removed the tagging requirement for now. Will read-up on git tagging when I have sometime.

[yulqen]

You might have to use semantic versioning format or something (v.1.0.0, v.0.1.0, etc). Looks like you're doing it right from a git point of view.

[banillie]

@yulqen do you have any time tomorrow for a coding chat? No problem if you don't dude. Only if it's suitable for you.

[yulqen]

Yep, should be good. Usual time.

[banillie]

Excellent. 👍 Cheers 👨 see you then.