PolyData

PolyData is a modular platform for ingesting, managing, and publishing data built on Django, Wagtail and PygeoAPI. Its architecture combines a Django backend with:

• Wagtail CMS for content and administration
• Catalog API, DRF for tabular data and Pygeoapi for geospatial services on top of Postgres with PostGIS and ParadeDB.
• OpenID Connect authentication (OIDC) using django-oidc-provider
• Internal apps (account, catalog, ingestion, main) containing business logic
• Docker compose ready for production deployment

Features

• OIDC authentication for users and relying-party clients
• REST API with OpenAPI/Swagger (drf-spectacular)
• OGC API service configurable via environment variables
• Data ingestion leveraging PostGIS and GeoAlchemy
• Elastic quality search with ParadeDB
• Flexible configuration through env vars
• Docker Compose orchestration of database and Django server

Requirements

• Docker & Docker Compose
• (Optional for local tasks) Python 3.14+
• PostgreSQL with PostGIS and ParadeDB extension (provided by the database container)

Quickstart

When running through Docker Compose, set the database host to the Compose service name database, not localhost.

1. Clone the repository:

   git clone https://github.com/tu-org/PolyData.git
   cd PolyData

2. Copy and edit the environment files:

   cp backend/.env.example backend/.env
   cp frontend/.env.example frontend/.env

3. Bring up the services:

   docker-compose up -d --build

   This will start the database (PostGIS and ParadeDB) and backend (Django/Gunicorn) containers.

   If you already have a local PostgreSQL volume created with the old /var/lib/postgresql/data mount layout, do not reuse it directly with PostgreSQL 18+ images. Remove the old development volume if it is disposable, or migrate it with pg_upgrade before reattaching data.

4. Create a Django superuser to access the admin:

   docker-compose exec backend uv run src/manage.py createsuperuser

5. Visit the key endpoints:

   • Django admin: http://127.0.0.1:8000/admin/
   • Wagtail admin: http://127.0.0.1:8000/cms/ (if enabled)
   • REST API: http://127.0.0.1:8000/api/
   • PygeoAPI: http://127.0.0.1:8000/geoapi/ (configurable via PYGEOAPI_BASE_PATH)
   • Swagger/OpenAPI: http://127.0.0.1:8000/api/schema/swagger-ui/

Development

Inside the backend container:

docker-compose exec backend bash
# then use manage.py or uv like any Django project
uv run src/manage.py migrate
uv run src/manage.py runserver 0.0.0.0:8000

Python dependencies are managed with UV. To add a package:

docker-compose exec backend uv add <package>

Project structure

• backend/ – main Django application
  • account/ – custom user model and auth
  • catalog/ – catalog logic and pygeoapi configuration
  • ingestion/ – data loading services
  • main/ – settings, urls and general views
• frontend/ – (optional) code for a decoupled client
• docker-compose.yaml – defines db and backend services

Important environment variables

• POSTGRES_HOST, POSTGRES_PORT, POSTGRES_DB, POSTGRES_USER, POSTGRES_PASSWORD
• SECRET_KEY, DEBUG, DJANGO_URL, FRONTEND_URL
• PYGEOAPI_TABLE, PYGEOAPI_BASE_PATH, PYGEOAPI_TITLE, etc.

See backend/src/main/settings/base.py for the full list with comments.

Deployment

You can deploy on any container platform. Be sure to:

1. Adjust environment variables for production (DEBUG=false, hosts, etc.).
2. Use an appropriate WSGI/ASGI backend (Gunicorn is already configured).
3. Mount persistent volumes for db and backend if needed.

Roadmap

• Celery task queue for asynchronous processing
• Fine grained permissions and roles
• [] Data versioning and history
• [] More ingestion connectors (CSV, APIs, etc.)
• [] Map creation and publishing tools
• [] DataStories and dashboards

Contributing

Pull requests are welcome! 🙌

1. Fork and branch with a feature/ or fix/ prefix.
2. Add tests and documentation when applicable.
3. Open a PR describing your change.
4. Respond to feedback and keep commits tidy.

License

This project is licensed under the GNU Affero General Public License. See the LICENSE file for details.

---

Thanks for using PolyData! 🌍