change from urllib to requests

[anandghegde]

I say we remove urllib and urllib2 and use requests instead.

It has many advantages

check it out here

http://docs.python-requests.org/en/latest/index.html

[boyska]

Well, it sure is an impressive library. But speaking of our real needs, I have to say that we mostly do GET requests, building search query. This usually boils down to these 3 lines:

search = "%(artist)s %(title)s" % song_metadata
query = urllib.urlencode({'q': search})
url = 'http://www.lyricsite.org/search.php?%s' % query
page = lxml.html.parse(urllib2.urlopen(url))

How would this become using requests?

search = "%(artist)s %(title)s" % song_metadata
r = requests.get('http://www.lyricsite.org/search.php', {'q': search})
page = lxml.html.parse(r.raw)

Well, one line saved. Just a bit more elegant, but the difference is not that much

My opinion is that, on "basic" GET we just use urllib and urllib2.
If, however, you feel that sometimes urllib is just too boring to use, don't be afraid using requests: it will become a dependency for just that retriever, so it's not a big problem.

Am I missing some point about requests advantages?
Just my 2 cents, feel free to say me I'm just wrong