使用mechanize的自定义身份验证处理程序

4 投票
1 回答
2555 浏览
提问于 2025-04-15 18:35

我想用 python-ntlm 来配合 mechanize.Browser() 使用。我已经成功让 HTTPNtlmAuthHandler 在 urllib2 和 mechanize.urlopen() 中工作了,但在用 Browser() 的时候却不行。

这是我用来进行 urlopen 的代码:

passman = mechanize.HTTPPasswordMgrWithDefaultRealm()
passman.add_password(None, url, user, password)
auth_NTLM = HTTPNtlmAuthHandler.HTTPNtlmAuthHandler(passman)

opener = mechanize.build_opener(auth_NTLM)
mechanize.install_opener(opener)
mechanize.urlopen(baseurl)

根据要求的错误追踪信息

harrisony@lithium:~$ python sitefoo.py 
now running mechanize.urlopen
<addinfourl at 169181868 whose fp = <httplib.HTTPResponse instance at 0xa15858c>>


now running mechanize.Browser then br.open
Traceback (most recent call last):
  File "sitescreaper.py", line 21, in <module>
    br.open(baseurl)
  File "/usr/lib/python2.6/dist-packages/mechanize/_mechanize.py", line 209, in open
    return self._mech_open(url, data, timeout=timeout)
  File "/usr/lib/python2.6/dist-packages/mechanize/_mechanize.py", line 261, in _mech_open
    raise response
mechanize._response.httperror_seek_wrapper: HTTP Error 401: Unauthorized

1 个回答

1

可能还有更好的方法,但我能让它工作的唯一办法就是去掉HTTPRobotRulesProcessor这个处理器,因为它不知怎么的阻止了HTTPNtlmAuthHandler的调用。

注意:下面的代码也顺便去掉了ProxyHandler,以便绕过代理服务器——如果不需要的话可以去掉。

passman = mechanize.HTTPPasswordMgrWithDefaultRealm()
passman.add_password(None, baseurl, user, password)
auth_NTLM = HTTPNtlmAuthHandler.HTTPNtlmAuthHandler(passman)

browser = mechanize.Browser()
browser.add_handler(auth_NTLM)

handlersToKeep = []
for handler in browser.handlers:
    if not isinstance(handler, (mechanize._auth.ProxyHandler, 
                                mechanize._urllib2_support.HTTPRobotRulesProcessor)):
        handlersToKeep.append(handler)
browser.handlers = handlersToKeep

browser.open(url)

撰写回答