utils/remove_tag.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
# ファイルから指定された単語を削除する
# Remove a specified word from a file

import pathlib
import re

def remove_word_from_file(file_path, word):
    with open(file_path, 'r', encoding='utf-8') as file:
        content = file.read()
    
    # Remove the word with the comma and space if there is one after it
    pattern = re.compile(r'\b' + re.escape(word) + r',?\s?')
    new_content = pattern.sub('', content)
    
    with open(file_path, 'w', encoding='utf-8') as file:
        file.write(new_content)

def remove_word_from_directory(directory, word):
    path = pathlib.Path(directory)
    for txt_file in path.rglob('*.txt'):
        remove_word_from_file(txt_file, word)

if __name__ == "__main__":
    import sys
    if len(sys.argv) != 3:
        print("Usage: python script.py <directory> <word>")
        sys.exit(1)
    
    target_directory = sys.argv[1]
    target_word = sys.argv[2]
    
    remove_word_from_directory(target_directory, target_word)