Why not .py to .rs #6710

📝 CodeRabbit Plan Mode

Generate an implementation plan and prompts that you can use with your favorite coding agent.

• Create Plan

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord or schedule a call!

The problem of trans-piling Python code into Rust is likely beyond the scope of this project.

Well, still posible. It is similar to a mypyc approach.
Rather than generating rust AST, we can pre-compile the python code to native using rustpython-jit at the moment.
Then it will be more approachable, and be able to fallback to python code when it is not compilable well.

I found this project: https://github.com/paiml/depyler. It mentions RustPython.
i appreciate all developers and this effort @youknowone

Couldn't you use cython to generate a C file and then use a combination of pyo3 and c2rust to convert it? At that point it's simpler to just compile the C file though, as you gain no safety or anything from using the rust version.

I checked Cython code and realized it is not true C, and it does not generate pure C code.
If Cython could generate pure C code, that would be good. Then I could convert .py to .c, compile it fully natively, achieve true C-level speed, and then connect it back to Python. @arihant2math

I really don't want to start this kind of discussion here, so it is not a real question.

What is true C code? If you have a code import logging; logging.info('hello'), you can turn it into C ast and compile it, but where the logging library come from? That's python library. And thats what Cython and mypyc do. You'd better check out what crystal lang achieved and limited.

And the goal of RustPython is neither transpiler nor crystal. It is obvious.

What about Nuitka?¹