import re

import os

def cleanup_text_file(filepath):

"""

Cleans up a text file by removing various artifacts.

"""

try:

with open(filepath, 'r', encoding='utf-8') as f:

text = f.read()

# Remove comments (e.g., # ... , // ...)

text = re.sub(r'#.*', '', text)

text = re.sub(r'//.*', '', text)

# Remove empty lines

text = text.split('\n')

text = [line for line in text if line.strip()]

text = '\n'.join(text)

# Remove extra whitespace (multiple spaces, tabs, newlines)

text = re.sub(r'\s+', ' ', text)

# Remove trailing whitespace

text = text.rstrip()

# Remove specific unwanted characters/patterns (customize as needed)

text = re.sub(r'\[.*?\]', '', text) # Remove bracketed text

text = re.sub(r'\(.*?\)', '', text) # Remove parenthesized text

with open(filepath, 'w', encoding='utf-8') as f:

f.write(text)

print(f"Cleaned: {filepath}")

except FileNotFoundError:

print(f"Error: File not found: {filepath}")

except Exception as e:

print(f"Error cleaning {filepath}: {e}")

def cleanup_directory(directory):

"""

Cleans up all text files in a directory.

"""

for filename in os.listdir(directory):

if filename.endswith(".txt"):

filepath = os.path.join(directory, filename)

cleanup_text_file(filepath)

if __name__ == '__main__':

# Example usage:

# cleanup_directory("/path/to/your/text/files") # Clean all .txt files in the specified directory

# cleanup_text_file("my_text_file.txt") # Clean a single file

pass