Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

Python's Requests OAuth2 (Open Authentication) plugin

branch: master
README.md

requests-oauth2

This plugins adds OAuth v2.0 support to @kennethreitz well-known requests library.

requests-oauth2 wants to provide the simplest and easiest way to do OAuth2 in Python. OAuth2 is several orders of magnitude easier to do than old OAuth1.0, so this is basically a simple connection initialization library. If you are looking for a way of doing OAuth 1.0 see requests-oauth

Author: Miguel Araujo Licence: BSD

Usage with Facebook API

Initialize the connection handler. It accepts this parameters. authorization_url and token_url are optional and have defaults.

from requests_oauth2 import OAuth2
OAuth2(client_id, client_secret, site, redirect_uri, [authorization_url='oauth/authorize'], [token_url='oauth/token'])

An example for facebook would be:

oauth2_handler = OAuth2(client_id, client_secret, "https://www.facebook.com/", "http://yoursite.com/webhook", "dialog/oauth", "oauth/access_token")

Get the url to redirect the user to for consenting OAuth2 application usage using authorize_url. This method can be passed a scope, which defines the permissions your application will have with that user. If not passed, an empty string will be used, which in some providers means default privileges:

authorization_url = oauth2_handler.authorize_url('email')

You can pass named parameters to authorize_url. Some OAuth2 providers allow extra parameters for configuring authorization. For example in google api:

authorization_url = oauth2_handler.authorize_url('https://www.googleapis.com/auth/books', response_type='code')

Once the user clicks in this authorization_url. He will be requested to log in, if he wasn't, and consent access to the application. After granting access, user will be redirected to http://yoursite.com/webhook?params. params are a list of GET params. If everything went right they should at least contain a param named code, you will need to parse it and pass it to the connection handler.

The code will be used to request an access token, necessary for all following requests to the API you do. Sometimes the site for authorization is different to the site for user consent (token_url). You can change the site in between doing:

oauth2_handler.site = "https://graph.facebook.com/"

Finally we have to get an access token passing the code we got from the OAuth provider, for that we use get_token. This method also accepts extra named parameters that you may need:

response = oauth2_handler.get_token(code)

Response can be a dictionary or None, if everything went right it should contain at least an access_token key. It will usually contain other interesting parameters such as expiring time. We can now do API calls, all of them should contain the access_token as a parameter. Thus we can generate a requests session, to avoid passing the parameter every time.

oauth2_client = requests.session(params={'access_token': response['access_token']})
oauth2_client.get('https://graph.facebook.com/me')

Next

From here you can code your own binding for your favorite API the way you like. This will usually imply persisting the access token mapped to some user's information, so you can replicate the session on every request. Also you will have to handle error situations and token expiration, for sure requests will help you tackle this task.

There are also many API bindings available that will start requesting you an access token, but won't do the OAuth2 initialization handling for you, then requests-oauth2 will prove useful.

Interesting readings

Something went wrong with that request. Please try again.