store.py 文件源码

def get_description_urls(html):
    from htmllib import HTMLParser
    from formatter import NullFormatter
    import urlparse, sgmllib
    try:
        parser = HTMLParser(NullFormatter())
        parser.feed(html)
        parser.close()
    except sgmllib.SGMLParseError:
        return []
    result = []
    for url in parser.anchorlist:
        if urlparse.urlparse(url)[0]:
            result.append(xmlescape(url))
    return result