Semiotic Analysis of Test Items

Development

With a working version of Python 3.8 and Pipenv:

1. Install dependencies (use sync instead of install to use Pipfile.lock instead, and ensure a deterministic environment).

   $ pipenv sync --dev
   

2. Install pre-commit hooks.

   $ pipenv run pre-commit install
   

3. Create a database (if not created yet) and migrate. Only PostgreSQL is supported. Database settings must be specified in a .env file (see .env_tempate for details).

   $ pipenv run python manage.py migrate
   

4. Create a superuser.

   $ pipenv run python manage.py createsuperuser
   

5. Collect static files from installed apps.

   $ pipenv run python manage.py collectstatic
   

6. Start development server and go to http://localhost:8000/admin/.

   $ pipenv run python manage.py runserver localhost:8000
   

Testing

• Tests

  $ pipenv run python manage.py test
  

• Linting and formatting

  $ pipenv run pre-commit run --all-files
  

Docker

A production-ready deployment can be brought up with just docker-compose up --build [-d]. This will start a container for the django app and another for the PostgreSQL database server. Note that the app is served by gunicorn on port 8000 which is exposed to the docker host, and the /static and /media roots are mounted from the host filesystem, as this is intended to be run behind a webserver on the docker host which serves up /static and /media and reverse-proxies the app.

Running the Django development server is possible with a command like the example below, but note that the docker image is compiled without installing the Django dev. dependencies.

docker-compose run --rm --name sati-app -p 8000:8000 django python manage.py runserver 0.0.0.0:8000

To run with DEBUG=True (and have the dev. server serve static assets) it is necessary to install the --dev dependencies. This can be done (if needed for some reason) by getting a shell:

docker-compose run --rm --name sati-app -p 8000:8000 django ash

and using the following commands:

apk add gcc python3-dev musl-dev;
pipenv install --deploy --system --ignore-pipfile --dev
export DEBUG=True
./manage.py runserver 0.0.0.0:8000

See the notes in .env_tempate for further details.