How to use urllib2 to get a webpage using SSLv3 encryption

Question

I\'m using python 2.7 and I\'d like to get the contents of a webpage that requires sslv3. Currently when I try to access the page I get the error SSL23_GET_SERVER_HELLO, and

Answer 1

Since I was unable to do this using urllib2, I eventually gave in and moved to using the libCurl bindings like @Bruno had suggested in the comments to pastylegs answer.

Answer 2

I realize this response is a few years too late, but I also ran into the same problem, and didn't want to depend on libcurl being installed on a machine where I ran this. Hopefully, this will be useful to those who find this post in the future.

The problem is that httplib.HTTPSConnection.connect doesn't have a way to specify SSL context or version. You can overwrite this function before you hit the meat of your script for a quick solution.

An important consideration is that this workaround, as discussed above, will not verify the validity of the server's certificate.

import httplib
import socket
import ssl
import urllib2

def connect(self):
    "Connect to a host on a given (SSL) port."

    sock = socket.create_connection((self.host, self.port),
                                    self.timeout, self.source_address)
    if self._tunnel_host:
        self.sock = sock
        self._tunnel()

    self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_TLSv1)

httplib.HTTPSConnection.connect = connect

opener = urllib2.build_opener()
f = opener.open('https://www.google.com/')

*Note: this alternate connect() function was copy/pasted from httplib.py, and simply modified to specify the ssl_version in the wrap_socket() call

Answer 3

SSL should be handled automatically as long as you have the SSL libraries installed on your server (i.e. you shouldn't have to specificially add it as a handler)

http://docs.python.org/library/urllib2.html#urllib2.build_opener

If the Python installation has SSL support (i.e., if the ssl module can be imported), HTTPSHandler will also be added.

Also, note that urllib and urllib2 have been merged in python 3 so their approach is a little different