Batuhan Taskaya

> if you don't do the work :) :-) [inserts a november song] > My goal for parso is just to get it running so people can use it. My...

I started some stuff, but calculating the rules beforehand is a bit tricky. The thing that makes backtracking harder for us is error-recovery. If we'd assume all input is valid,...

I presume you found it on the `docs.python.org`, which is the stripped version of the official grammar. That rule actually has 2 alternatives (one is only used for error reporting...

> About the proposed patch: I'm currently trying a technique in Rust that might work for us as well. I will keep you posted. :eyes:

> The thing that makes backtracking harder for us is error-recovery I was reading [a paper about error-recovery for PEG parsers, especially the ones that used by IDEs for auto-completition](https://arxiv.org/pdf/1806.11150.pdf)...

status update: PEP 634 accepted with soft keywords

Yeah, that is what I kind of used (not entirely). Do you have any opinions regarding my initial approach? I'm thinking about also introducing a `~` operator to our grammar,...

Can you show like a concrete example?

> I'm a bit confused by the memory usage. Can you explain this? The difference seems quite big and the only thing that I can think is the memorization cost...

For what its worth, here is a (~decent looking) python parser written in Rust that I stumbled against a few days ago: https://github.com/ProgVal/rust-python-parser/