A Library for setting up login routes in a Chalice app.
Below is an example of a basic application making use of a Cognito User Pool.
First set up a new Chalice app:
$ chalice new-project test-auth $ cd test-auth
Next we add chalice-cognito-auth
as a dependency:
$ echo "chalice-cognito-auth" >> requirements.txt
Now update the app.py
file to configure a default user pool handler.
from chalice import Chalice
import chalice_cognito_auth
app = Chalice(app_name='test-auth')
app.experimental_feature_flags.update([
'BLUEPRINTS',
])
user_pool_handler = chalice_cognito_auth.default_user_pool_handler()
app.register_blueprint(user_pool_handler.blueprint)
@app.route('/whoami', authorizer=user_pool_handler.auth)
def index():
return {
'username': user_pool_handler.current_user
}
This will create a UserPoolHandler
object using the environment variables
APP_CLIENT_ID
for the Cognito Userpool application client id. POOL_ID
for the ID of the Cognito Userpool itself. And AWS_REGION
for the
region. AWS_REGION
is set by the AWS Lambda runtime, but the other two we
need to set ourselves. Update the file .chalice/config.json
to look
something like the following:
{ "version": "2.0", "app_name": "test-auth", "environment_variables": { "APP_CLIENT_ID": "...client id here...", "POOL_ID": "...pool id here..." }, "stages": { "dev": { "api_gateway_stage": "api" } } }
Substitute the client id and pool id values for ones that match an existing cognito user pool you have and can use for testing.
Now deploy the application using:
$ chalice deploy Creating deployment package. Updating policy for IAM role: test-auth-dev Updating lambda function: test-auth-dev Updating lambda function: test-auth-dev-UserPoolAuth Updating rest API Resources deployed: - Lambda ARN: arn:aws:lambda:us-west-2:...:function:test-auth-dev - Lambda ARN: arn:aws:lambda:us-west-2:...:function:test-auth-dev-UserPoolAuth - Rest API URL: https://id.execute-api.us-west-2.amazonaws.com/api/
Now that it has been deployed we can access the API using the Rest API
URL. chalice-cognito-auth injects a login
route which accepts a POST
request with a JSON payload containing the two keys username
and
password
. Make sure your configured userpool has a user in it that can be
used for testing and send something like the following:
$ curl -X POST -H Content-Type:application/json https://id.execute-api.us-west-2.amazonaws.com/api/login -d '{"username":"StealthyCoin", "password": "secret"}' {"id_token":"...","refresh_token":"...","access_token":"...","token_type":"Bearer"}
The above JSON response contains all the tokens needed to send authorized
requests. To test our authorizer we will use the whoami
route which simply
takes a request and either rejects it if unauthorized, or sends back the
username associated with the request. To do this we will send a GET
request
with an Authorization
header with the value of our id_token
from the
result JSON above.
In my case:
$ curl -H Authorization:...id token here... https://id.execute-api.us-west-2.amazonaws.com/api/whoami {"username":"StealthyCoin"}
Which sends back JSON object with the username that goes with my id token.
To check that a requset with a bad authorization token is rejected, run the following curl command:
$ curl -H Authorization:foobar https://id.execute-api.us-west-2.amazonaws.com/api/whoami {"Message":"User is not authorized to access this resource"}