tags.py

#!/usr/bin/env python3

import os
import sys
import re

def extract_frontmatter(file_path):
    """Extract frontmatter from markdown files manually without using yaml.safe_load."""
    try:
        with open(file_path, 'r', encoding='utf-8') as file:
            content = file.read()
            
            # Look for frontmatter between --- delimiters
            frontmatter_match = re.search(r'^---\s*\n(.*?)\n---', content, re.DOTALL)
            
            if frontmatter_match:
                frontmatter_text = frontmatter_match.group(1)
                
                # Manually extract tags using regex instead of yaml parsing
                tags_match = re.search(r'tags:\s*\n((?:- .*\n)*)', frontmatter_text)
                if tags_match:
                    # Extract individual tags from list format
                    tags_list = re.findall(r'- (.*?)\s*$', tags_match.group(1), re.MULTILINE)
                    return {'tags': tags_list}
                
                # Check for single tag format (tags: value)
                single_tag_match = re.search(r'tags:\s*(.*?)\s*$', frontmatter_text, re.MULTILINE)
                if single_tag_match:
                    return {'tags': [single_tag_match.group(1)]}
                    
                return None
            return None
    except (IOError, UnicodeDecodeError):
        # If file can't be read or isn't text, return None
        return None

def search_files_with_tag(start_path, target_tag):
    """Recursively search for files containing the specified tag."""
    matching_files = []
    
    # Validate the start path exists
    if not os.path.exists(start_path):
        print(f"Error: Path '{start_path}' does not exist.")
        return matching_files
    
    # Walk through all directories and files starting from start_path
    for root, _, files in os.walk(start_path):
        for filename in files:
            file_path = os.path.join(root, filename)
            
            # Skip non-regular files (symlinks, etc.)
            if not os.path.isfile(file_path):
                continue
                
            # Extract frontmatter from the file
            metadata = extract_frontmatter(file_path)
            
            # Check if metadata contains tags
            if metadata and 'tags' in metadata:
                tags = metadata['tags']
                
                # Check if the target tag is in the list of tags
                if target_tag in tags:
                    matching_files.append(file_path)
    
    return matching_files

def main():
    # Check if the correct number of arguments is provided
    if len(sys.argv) != 3:
        print("Usage: python script.py <directory_path> <tag>")
        sys.exit(1)
    
    # Get the directory path and tag from command line arguments
    directory_path = sys.argv[1]
    tag = sys.argv[2]
    
    # Search for files with the specified tag
    matching_files = search_files_with_tag(directory_path, tag)
    
    # Print the results
    if matching_files:
        for file_path in matching_files:
            if not file_path.endswith('~'):
                print(f"{file_path}")
    else:
        print(f"No files found with the tag '{tag}'.")

if __name__ == "__main__":
    main()