INNER CODE UNIT · Python

extract_title

netkiller/netkiller.github.io · generate_sitemaps.py:34

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:
        pass
    return None

def generate_sitemap_xml(html_files):
    """Generate sitemap.xml."""
    lines = ['<?xml version="1.0" encoding="UTF-8"?>']

View source record →