remove_special_chars.py

#!/usr/bin/env python3
# -*- coding: ISO8859-15 -*-

# This might need some adjustments:
# Save file with the correct encoding and
# make sure that the chars to be replaced
# are encoded correctly in this file.

import os
import sys
import subprocess
import re
from shutil import move


# inspired by
# https://stackoverflow.com/a/241506/1272072
# make sure the chars that need to be replaced
# are encoded correctly in this file!
def comment_remover(text):
    def replacer(match):
        s = match.group(0)
        r = s.replace("Ä", "Ae")
        r = r.replace("Ö", "Oe")
        r = r.replace("Ü", "Ue")
        r = r.replace("ä", "ae")
        r = r.replace("ö", "oe")
        r = r.replace("ü", "ue")
        return r

    pattern = re.compile(r"\/\*(\*(?!\/)|[^*])*\*\/", re.DOTALL | re.MULTILINE)
    text = re.sub(pattern, replacer, text)

    pattern = re.compile(r'//.*?$', re.DOTALL | re.MULTILINE)
    return re.sub(pattern, replacer, text)


def process_files(f):
        bakfile = f + ".umlaut"
        move(f, bakfile)
        with open(bakfile, "r", encoding="ISO8859-15") as infile:
            with open(f, "w", encoding="ISO8859-15") as outfile:
                outfile.write(comment_remover(infile.read()))


RELEVANT_FILE_ENDINGS = [".c", ".cpp", ".h", ".hpp"]


def is_relevant_file(file):
    for ending in RELEVANT_FILE_ENDINGS:
        if not file.startswith(".") and file.endswith(ending):
            return True
    return False


def proces_directory(dir):
    for root, dirs, files in os.walk(dir):
        for file in files:
            if is_relevant_file(file):
                file_path = os.path.join(root, file)
                ret = subprocess.run(["file", file_path], stdout=subprocess.PIPE)
                if b"ISO-8859" in ret.stdout:
                    print(f"Processing {file_path}")
                    process_files(file_path)


if __name__ == "__main__":
    input_path = sys.argv[1]

    if os.path.isfile(input_path):
        process_files(input_path)
    else:
        proces_directory(input_path)