Max Bachmann

Yes I think this would cover all of my needs and would be a lot more concise than my current implementation with `scikit-build` :+1: I will start a branch porting...

I migrated this project to scikit-build-core. The implementation for rapidfuzz still requires a little more work. However everything required should be in scikit-build-core now.

Quite possible that there is an error in the type hints. Can you send me a complete sample that allows me to reproduce this warning? What's the tool you use...

I am asking because this is no complete sample. This sample would raise the following warnings: - unidecode doesn't exist - col doesn't exist - patterns doesn't exist

I can reproduce the warning in pycharm. This works as expected in mypy and pyright. I can't see anything wrong with the type hints either. So I assume some of...

Yes the readme isn't really correct for `thefuzz` anymore. 1) `thefuzz` is MIT licensed 2) it doesn't have the same performance difference `fuzzywuzzy` had The reason for both of those...

> Using --noconsole when exe is being made towards the ends exe gets deleted by AV. Is that already the issue? Or is there a separate issue with the packaging...

Given that there was no update and this works fine in CI I will close this issue for now

A couple small things I noticed: 1) both the rapidfuzz and scikit-build-core file list an incorrect license 2) the rapidfuzz version lists a version 3.9.7. I suppose this should be...

Did you validate that that your patched version properly builds the package and doesn't just fall back to the pure Python version?