So, I had to change the feedparser code to run it behind proxy. Basically the idea is that python by default doesn’t user environment variables http proxy. You have to specify it in the python code.

In feedparser code, url is accessed at the line # 1895. Now, opener has to have http settings so that it could access the url. For that, I added 2 lines before calling opener:

proxy_support = urllib2.ProxyHandler({"http" : "http://<username>:<passwd>@<proxyserver>:<proxyport>"});
opener = urllib2.build_opener(proxy_support)

Here is the final feedparser : FeedParser With Proxy Support