INNER CODE UNIT · Python

url_encode_path

netkiller/netkiller.github.io · generate_sitemaps.py:28

def url_encode_path(path):
    """URL encode a file path."""
    parts = str(path).split('/')
    encoded_parts = [quote(part, safe='') for part in parts]
    return '/'.join(encoded_parts)

def extract_title(filepath):
    """Extract title from HTML file."""
    try:
        with open(filepath, 'r', encoding='utf-8', errors='ignore') as f:
            content = f.read(5000)  # Read first 5KB
            match = re.search(r'<title>(.*?)</title>', content, re.IGNORECASE | re.DOTALL)
            if match:
                title = match.group(1).strip()
                # Clean up title
                title = re.sub(r'\s+', ' ', title)
                return title[:200]  # Limit length
    except:

View source record →