Using mechanize to do website authentication

[trihaitran]

I am trying to write a web scraper and am having trouble accessing pages that require authentication. I am attempting to utilise the mechanize library, but am having difficulties. The site I am trying to login is http://www.princetonre view.com/Login3.aspx?uid badge=

user: bugmenot2008@ya hoo.com
pass: letmeinalready

Previously I did something similar to another site: schoolfinder.co m. Here is my code for that:

import cookielib
import urllib
import urllib2
 
cj = cookielib.CookieJar()
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
resp = opener.open('http://schoolfinder.com') # save a cookie
 
theurl = 'http://schoolfinder.com/login/login.asp' # an example url that sets a cookie, try different urls here and see the cookie collection you can make !
body={'usr':'greenman','pwd':'greenman'}
txdata = urllib.urlencode(body) # if we were making a POST type request, we could encode a dictionary of values here - using urllib.urlencode
txheaders =  {'User-agent' : 'Mozilla/4.0 (compatible; MSIE 5.5; Windows NT)'} # fake a user agent, some websites (like google) don't like automated exploration
 
 
try:
    req = urllib2.Request(theurl, txdata, txheaders) # create a request object
    handle = opener.open(req) # and open it to return a handle on the url
    HTMLSource = handle.read()
    f = file('test.html', 'w')
    f.write(HTMLSource)
    f.close()
 
except IOError, e:
    print 'We failed to open "%s".' % theurl
    if hasattr(e, 'code'):
        print 'We failed with error code - %s.' % e.code
    elif hasattr(e, 'reason'):
        print "The error object has the following 'reason' attribute :", e.reason
        print "This usually means the server doesn't exist, is down, or we don't have an internet connection."
        sys.exit()
 
else:
    print 'Here are the headers of the page :'
    print handle.info() # handle.read() returns the page, handle.geturl() returns the true url of the page fetched (in case urlopen has followed any redirects, which it sometimes does)
 

This method does not work on the Princeton Review site however. Interestingly I cannot even get mechanize to access the schoolfinder.co m site. Here is the code I am using:

#!/usr/bin/env python
# -*- coding: UTF-8 -*-
import mechanize
 
theurl = 'http://www.princetonreview.com/Login3.aspx?uidbadge='
mech = mechanize.Browser()
mech.open(theurl)
 
mech.select_form(nr=0)
mech["ctl00$MasterMainBodyContent$txtUsername"] = "[e-mail]"
mech["ctl00$MasterMainBodyContent$txtPassword"] = "letmeinalready"
results = mech.submit().read()
 
f = file('test.html', 'w')
f.write(results) # write to a test file
f.close()
 

This code is so short and I just cannot figure out what I am doing wrong. What is incorrect about this? Thank you in advance.