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

class GrandfatheredRemovalProcessor(FileProcessor):
    def __init__(self, options: ProcessorOptions, pattern: str):
        super().__init__(options)
        self.pattern = pattern
    
    def process_content(self, content: str) -> str:
        # Remove occurrences of the grandfathered content pattern
        content = re.sub(self.pattern, '', content)
        # Normalize whitespace and commas
        content = re.sub(r'\s+,', ',', content)
        content = re.sub(r',\s+', ',', content)
        return re.sub(r'\s+', ' ', content).strip()

def main():
    target_dir = sys.argv[1] if len(sys.argv) > 1 else '.'
    
    options = ProcessorOptions(
        recursive=True,
        dry_run=False,
        file_extensions={'.txt', '.tags'}
    )
    
    # Define the pattern for grandfathered content (modify as needed)
    grandfathered_pattern = r'grandfathered_content_pattern'
    
    processor = GrandfatheredRemovalProcessor(options, grandfathered_pattern)
    processor.process_directory(Path(target_dir))

if __name__ == "__main__":
    main()