check_pycodestyle.py

from __future__ import annotations

import os
import subprocess

SELECT = [
    "E111",
    "E112",
    "E113",
    "E114",
    "E115",
    "E116",
    "E117",
    "E201",
    "E202",
    "E203",
    "E211",
    "E221",
    "E222",
    "E223",
    "E224",
    "E225",
    "E226",
    "E227",
    "E228",
    "E231",
    "E251",
    "E252",
    "E261",
    "E262",
    "E265",
    "E266",
    "E271",
    "E272",
    "E273",
    "E274",
    "E275",
]


def take_after(s: str) -> str:
    """Given filename:204:22, return everything after the first colon."""
    s = s.split(":", 1)[1]
    s = s.replace(
        "`# `",
        "'# '",
    )
    s = s.replace(
        "`#`",
        "'#'",
    )
    s = s.replace(
        "for block comment",
        "before block comment",
    )
    s = s.replace(
        "Insert at least two spaces before an inline comment",
        "at least two spaces before inline comment",
    )
    return s


def take_comparable(s: str) -> str:
    s = s.lower()
    # Ignore location, I guess? The issue here is really for "Over-indented (comment")
    # where we use the start of the indentation, but pycodestyle uses the start of the
    # comment.
    if "over-indented" in s:
        s = s.split(":", 2)[-1]
    return s


def case_insensitive_compare(a: list[str], b: list[str]) -> bool:
    return [take_comparable(line) for line in a] == [
        take_comparable(line) for line in b
    ]


def pycodestyle(fname: str) -> list[str]:
    output = subprocess.getoutput(f"pycodestyle {fname} --select {','.join(SELECT)}")
    if output:
        return [take_after(line) for line in output.split("\n") if ":" in line]
    else:
        return []


def ruff(fname: str) -> list[str]:
    output = subprocess.getoutput(
        f"./target/debug/ruff {fname} --select {','.join(SELECT)} --no-cache"
    )
    if output:
        return [take_after(line) for line in output.split("\n") if ":" in line]
    else:
        return []


if __name__ == "__main__":
    dirname = "../scipy/scipy"

    # Walk over every file in the directory.
    seen = False
    for root, dirs, files in os.walk(dirname):
        for filename in files:
            # If the file is a python file, check if it contains a docstring.
            if filename.endswith(".py"):
                filepath = os.path.join(root, filename)

                # Local optimization: skip until you see a specific failing file...
                if False:
                    if filepath == "../scipy/scipy/stats/tests/__init__.py":
                        seen = True
                    if not seen:
                        continue

                # Run pycodestyle on the file.
                pycodestyle_errors = pycodestyle(filepath)
                ruff_errors = ruff(filepath)

                if case_insensitive_compare(pycodestyle_errors, ruff_errors):
                    print(f"OK: {filepath}")
                else:
                    print(f"File: {filepath}")

                    # Print the first error that differs.
                    for i in range(min(len(pycodestyle_errors), len(ruff_errors))):
                        if pycodestyle_errors[i].lower() != ruff_errors[i].lower():
                            print(f"pycodestyle: {pycodestyle_errors[i]}")
                            print(f"ruff:         {ruff_errors[i]}")
                            break

                    # Print the remaining errors.
                    if len(pycodestyle_errors) > len(ruff_errors):
                        for error in pycodestyle_errors[len(ruff_errors) :]:
                            print(f"pycodestyle: {error}")
                    else:
                        for error in ruff_errors[len(pycodestyle_errors) :]:
                            print(f"ruff:         {error}")

                    exit(1)