def handle_starttag(self, tag, attrs):
tag = tag.lower()
if tag == sequence[self.__s]:
self.__s += 1
if self.__s == len(sequence):
self.__s = 0
for v in filter(lambda x: x[0]=='href',attrs):
host = urlparse(v[1])[1]
try:
if self.__check and gethostbyname(host) != self.__q:
continue
except: continue
if self.__callback:
self.__callback(host)
else:
self.__s = 0
评论列表
文章目录