Can python mechanize handle HTTP auth?

Question

Mechanize (Python) is failing with 401 for me to open http digest URLs. I googled and tried debugging but no success.

My code looks like this.

import mechanize

project = "test"
baseurl = "http://trac.somewhere.net"
loginurl = "%s/%s/login" % (baseurl, project)
b = mechanize.Browser()
b.add_password(baseurl, "user", "secret", "some Realm")
b.open(loginurl)

Answer 1

Mechanize claims that the parameters should be uri, username and password as parameters, but you have four parameters. Four parameters are correct for urllib2.add_password, but then the first parameter should be the realm, not the uri.

http://wwwsearch.sourceforge.net/mechanize/

I'd try to change that first.

Does trac require digest? if not a next step could be to try using basic auth, as a test to see if that works, since you can add that with just addHeader:

browser = Browser()
browser.addHeader('Authorization', 'Basic %s:%s' % (user, pwd))

Answer 2

Depending on how complex your web automation project is, consider using iMacros. Unlike Mechanize it runs in the web browser, so it works with most websites out of the box.

I use both the free Firefox addon via a command line (on our Linux server) and the iMacros Scripting Edition via COM object (on a Windows VMware for sites with Flash and Java).

Command line (freeware & open source): http://wiki.imacros.net/iMacros_for_Firefox#Command_Line_Support

COM object (paid version required): wiki.imacros.net/Python

Tim

Answer 3

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

import mechanize

a=mechanize.Browser()
a.open("http://www.facebook.com/login.php")
a.select_form(nr=0)  #form number.
a["email"]="mailaddress"
a["pass"]="password"
a.submit()
print a