Max Bachmann

Yes I forgot, that when `v1.9.1` was released, only pypy3.7/pypy2.7 existed. So for any newer versions of pypy there is no version working on windows yet. I will add support...

`v2.1.0` includes a relatively fast pure python fallback of the algorithms when the C extension fails to compile. It should work pretty well especially under pypy. Check the [release notes](https://github.com/maxbachmann/RapidFuzz/releases/tag/v2.1.0)...

I just realized I forgot to add the fallback implementation to the setup.py (#234) :facepalm:

The fallback should work on pypy in `v2.1.2`. It is quite a bit slower than the C Extension even when using PyPy, but it is the fastest pure Python implementation...

The pure Python version works now and Python 3.11 is supported as well. The C++ version on PyPy will be supported once it is fixed in scikit-build. However this is...

The api will follow the following scheme: ```python class Compare: __init__(scorer, *): def set_seq1(a): -> single def set_seq1_list(a): -> many def set_seq2(b): -> single def set_seq2_list(b): -> many def all()...

I am still interested in adding this feature. However I do not expect that I will have the time to implement this until later this year.

@mikegerber as a first step I updated the algorithm to use the alignment algorithm by Heikki Hirrö ([A Note on Bit-Parallel Alignment Computation](https://www.semanticscholar.org/paper/A-Note-on-Bit-Parallel-Alignment-Computation-Hyyr%C3%B6/bbd0a7afaf7c395d9ea3c04d99ca7c0ab30fbf60). This improves memory usage by 33% from...

This is now supported for `process.cdist` using the `workers` argument.

So far multiprocessing is only supported by `process.cdist`.