tabs to spaces

tabstospaces.py

#!/usr/bin/env python
"""
convert tabs to spaces and also prune 0xa0 characters, if any.
https://gist.github.com/urbanij/3b295f72e191958c533683e798437a53
Sat Nov 13 18:13:11 CET 2021

----
Fri Mar  4 09:53:51 CET 2022
also remove trailing whitespaces:
"""

import unicodedata
import sys
import pathlib
import re

def main(backup=False):
    try:
        filename = sys.argv[1]
    except Exception as e:
        print("Missing file. `./tabstospaces.py <file>`")
        exit(1)

    # reads
    with open(filename, 'r') as f:
        content = f.read()

    # writes a backup
    if backup:
        with open(filename+'.bak', 'w') as f:
            f.write(content)

    # do the work
    content = content.replace('\t', ' '*4)

    content = unicodedata.normalize("NFKD", content) # strips 0xa0 from the string

    TRAILING_WHITESPACES = r"( +)\n"
    content = re.sub(TRAILING_WHITESPACES, "\n", content)

    # writes new file
    with open(filename, 'w') as f:
        f.write(content)

if __name__ == '__main__':
    main()