How to log in to a website with urllib? - forms

Question

I am trying to log on this website: http://www.broadinstitute.org/cmap/index.jsp. I am using python 3.3 on Windows. I followed this answer https://stackoverflow.com/a/2910487/651779. My code:
import http.cookiejar
import urllib
url = 'http://www.broadinstitute.org/cmap/index.jsp'
values = {'j_username' : 'username',
'j_password' : 'password'}
data = urllib.parse.urlencode(values)
binary_data = data.encode('ascii')
cookies = http.cookiejar.CookieJar()
opener = urllib.request.build_opener(
urllib.request.HTTPRedirectHandler(),
urllib.request.HTTPHandler(debuglevel=0),
urllib.request.HTTPSHandler(debuglevel=0),
urllib.request.HTTPCookieProcessor(cookies))
response = opener.open(url, binary_data)
the_page = response.read()
http_headers = response.info()
It runs without erros, however the html in the_page is just the log in page. How can I log onto this page?

Answer 1

The site is using a JSESSIONID cookie to create session since HTTP requests are stateless. When you're making your request, you're not getting that session id first.
I sniffed a session to log into that site using Fiddler and found that the POST is made to a different URL, but it has that JSESSIONID cookie set. So you need to make a get to the URL first, capture that cookie using the cookiehandler, then POST to this URL:
post_url = 'http://www.broadinstitute.org/cmap/j_security_check'
You don't need to save the HTTP GET request at all, you can simply call opener.open(url), then in your code change the response line to this:
response = opener.open(post_url, binary_data)
Also the payload was missing the submit method. Here's the whole thing with the changes I suggest:
import http.cookiejar
import urllib
get_url = 'http://www.broadinstitute.org/cmap/index.jsp'
post_url = 'http://www.broadinstitute.org/cmap/j_security_check'
values = urllib.parse.urlencode({'j_username': <MYCOOLUSERNAME>,
'j_password': <MYCOOLPASSSWORD>,
'submit': 'sign in'})
payload = bytes(values, 'ascii')
cj = http.cookiejar.CookieJar()
opener = urllib.request.build_opener(
urllib.request.HTTPRedirectHandler(),
urllib.request.HTTPHandler(debuglevel=0),
urllib.request.HTTPSHandler(debuglevel=0),
urllib.request.HTTPCookieProcessor(cj))
opener.open(get_url) #First call to capture the JSESSIONID
resp = opener.open(post_url, payload)
resp_html = resp.read()
resp_headers = resp.info()
Any other requests using the opener you created will re-use that cookie and you should be able to freely navigate the site.