utils/remove_extra_whitespace.py

from utils.file_processor import FileProcessor, ProcessorOptions
import re
from pathlib import Path

class WhitespaceProcessor(FileProcessor):
    def process_content(self, content: str) -> str:
        # Replace multiple spaces with a single space
        content = ' '.join(content.split())
        # Replace multiple newlines with a single newline
        return '\n'.join(line.strip() for line in content.split('\n'))

def main():
    options = ProcessorOptions(
        recursive=True,
        dry_run=False,
        file_extensions={'.txt', '.caption'}
    )
    
    processor = WhitespaceProcessor(options)
    processor.process_directory(Path('.'))

if __name__ == "__main__":
    main()