Inria Digital Accessibility MOOC API

This web application handles Inria MOOC Accessibility web player events. Its main purpose is to act as a gateway between the web client and an ElasticSearch server.

Requirements

• python =3.6
• Pipenv

Installation

pipenv install

For development purpose, you also should install additional packages:

pipenv install --dev

The application assumes ElasticSearch is used to save records. If no ElasticSearch server can be found, it will just output errors and return 500 errors.

Most straightforward installation requires Docker. Just cd into the application directory and run docker-compose up.
Running this command will start an ElasticSearch server on http://localhost:9200 and a Kibana instance on http://localhost:5601 to visualize events.

Linting

pipenv run lint # runs pycodestyle
pipenv run lint_fix # fixes malformed code

Run the application

Development

cd /path/to/mooc-accessibility-api
pipenv run serve_dev

Application will be served through built-in Flask server on port 3001.

Production

Gunicorn is provided as a dependency, and should be the prefered way to serve the application in production.

# numbers of workers depend on the machine hosting the application
pipenv run gunicorn --workers=3 --bind=127.0.0.1:8000 wsgi:application

Note About Logging

Logs are written to STDERR and STDOUT. This can be overridden by configuring the wsgi app container.

Environment Variables

The .env file hosts a few variables you may need to change before running the application.