Junior Django Developer application challenge

A Django application for generating dummy data by user-defined schemas with arbitrary fields.

Requirements

Check challenge requirements in the requirements folder. Don't expect perfection (especially regarding the frontend side and git), yet there is some decency.

Also, check Tasks in the Projects tab for the planned tasks.

Try online

Go to https://datagen-challenge.herokuapp.com/ (give a minute if stalled - Heroku bootup) and use the next credentials:
username: preview
password: githubtest

Local run setup

1. Run docker compose up

2. export DJANGO_SETTINGS_MODULE=config.settings.local_docker

3. run worker celery --workdir datagen -A config worker -l INFO or export WORKER_LOCAL=True to call celery task as run() (inprocess) instead of delay()

Git hooks for development setup

Check the content of the Makefile and pre-commit. Run make add_git_hooks and then import somewhere (in user wide environment variables (as hooks often are run outside of IDE environment scope or in /.env) Git Guardian key as described here.

Heroku deployment

There is no requirements.txt, so to upload to Heroku: add poetry build pack Then add services and their credentials to environment variables (check settings.heroku module):

• Heroku CloudAMQP addon or whatever for RabbitMQ
• PostgreSQL URI (Heroku provides a free addon)
• Amazon S3 (walkthrough) for storing static files and generated datasets (privately)

Why there are so much code

1. Requirement that different fields can have different parameters.
2. I wanted to use Django form validators, so I could have error handling for free.
3. Just a few days before I experimented with Faker, so had an idea that there could be a hundred different fields with different parameters (and even different languages). YAGNI...

Having the first two aspects, I needed to have a separate form/model (whatever entity) for handling each field. So my first solution was to use multiple forms for validation and JSON field for storage https://github.com/furioness/django-challenge-planeks/tree/old_JSON_fields - a relatively simple solution.

Later, I got feedback that is better to store those fields in separate models, and there is too much code for such a task.

Got it! Now we have an even more complex solution :)

But what they probably meant is just to have a single model for storing all fields with all required parameters (I eventually searched GitHub for other solutions). That indeed is much more simple, but I never thought about it because of the third aspect.