<html>
<head>
<title>Test</title>
</head>
<body class="body" style="color: red;">
<p id="1">Test</p>
<p id="2">Test</p>
</body>
</html><html><head><title></title></head><body><p></p><p></p></body></html>
import re
s = """
<html>
<head>
<title>Test</title>
</head>
<body class="body" style="color: red;">
<p id="1">Test</p>
<p id="2">Test</p>
</body>
</html>"""
print(''.join(re.findall(r'</?[a-z]\w*\b|>', s, flags=re.I | re.M)))
