#!/usr/bin/env python
# encoding: utf-8

from HTMLParser import HTMLParser

class HTMLParser(HTMLParser):
    def handle_starttag(self, tag, attrs):
        if tag == 'a':
            for name, value in attrs:
                print name, value

html = '''<a href="中文" title="中文">中文</a>'''
parser = HTMLParser()
parser.feed(html)

html = '''<a href="&#20013;&#25991;" title="&#20013;&#25991;">&#20013;&#25991</a>'''
parser = HTMLParser()
parser.feed(html)

html = '''<a href="中&#25991;" title="中&#25991;">中&#25991;</a>'''
parser = HTMLParser()
parser.feed(html)

# vim: set sta sw=4 et: