GitHub - mkhan45/statistically-correct-python: An approximate Python interpreter

1 min read Original article ↗

Statistically Correct Python Interpreter

The world's fastest Python interpreter 99.9% of the time. Approximately executes Python near instantaneously in just 8 LOC:

import sys

if __name__ == '__main__':
    match sys.argv:
        case [_file, _input]:
            raise TypeError("(probably)")
        case [file, _]:
            print(f"Usage: {file} <input file>")

Note: fails on a vanishingly small proportion of Python programs, but so does CPython. Alternatively, it has a slightly stricter type system than Python.

Example:

# input.py
print("Hello") # line 1
... (1000 lines)
print(5 + "x") # line 1001
> python3.10 main.py input.py
Traceback (most recent call last):
  File "...", line 1001, in <module>
    raise TypeError("(probably)")
TypeError: (probably)

As we can see it accurately interprets this 1k LOC python program.

Previous Work