Improve line parsing

[gforcada]

The raw lines are being processed by a huge regular expression that needs to capture every small detail of it. This is the main factor of the slowness, as checked with a rather small file (50k lines only) and performance tools.

Are there other, faster, ways to process it? :thinking:

[gforcada]

https://ohadravid.github.io/posts/2023-03-rusty-python/ sounds like a good inspiration 🦀 🎉