swc_api contains classes and utilities for interacting with the Small World Community REST API.
swc_connection()
returns a request Session object configured to implement JWT refreshing, paging, and throttling.
The CommunityToken
class handles fetching JWT tokens for interacting with Community.
To install locally:
Checkout the project from Github:
> git clone https://github.com/citizensclimateeducation/community_api.git
Then (with the virtual environment active for the project you'll be using this with):
> pip install -e /path/to/community_api
To install via pypi:
> pip install swc-api
To create a session based on environmental variables, create a .env
file with the following variables set based on your Community instance:
export SWC_AUDIENCE=[community.yourdomain.org] export SWC_APP_ID=[Application ID for your OAuth Application] export SWC_USER_ID=[Authorized User for your OAuth Application] export SWC_SECRET=[OAuth secret for your Oauth Application]
Source the .env
file before running your application and you won't need to set these variables when instantiating the
api client:
> from swc_api import swc_connection > swc = swc_connection() > # make calls to the API endpoint > group_members = swc.get_all("groups/[groupId]/members", params={"embed": "user"})
However you can pass these variables in manually as well:
from swc_api import swc_connection, CommunityToken swc = swc_connection( community_domain="[your domain]", app_id="[OAuth App ID]", user_id=[Authorized User Id], app_secret="OAuth App Secret", ) group_members = swc.get_all("groups/[groupId]/members", params={"embed": "user"})
You can make regular requests
calls with the swc_connection
and will receive a standard Response
object back.
The session also has a get_all
method which will handle pagination for result sets longer than 500 objects. The
get_all
method will return a JSON response rather than a requests.models.Response
object.
An example of an update would be:
swc.post("groups/[id]/members", json={"userId": "[userId]"}) # response { 'id': '[membershipId]', 'groupId': '[groupId]', 'userId': '[userId]', 'status': '1', 'joinDate': '2020-07-20T12:43:07-07:00', 'invited': True }
Currently throttling and error handling is very basic. After each response is returned the session sleeps for the
duration indicated by SWC_API_RATE
. This means that the duration of the call is not taken into consideration, nor are
calls made in other threads. There are also no intelligent retry mechanisms built in.
Given those limitations, this should probably not be used for front-facing production applications, but for other purposes such as analyzing data, looking at fields provided in the API but not Community Reports or making bulk updates not available in the Community Admin Import Manager.