GAIA GeoDataLake — Example Client

Example scripts for uploading and downloading geodata from the GAIA GeoDataLake, using the CKAN API, OAuth2 authentication, and RabbitMQ notifications.

Setup

Requires Python >= 3.13 and uv.

uv sync

Environment configuration

Copy the template and fill in your credentials:

cp envs/env.template .env

The .env file must contain the following variables:

Variable	Default	Description
PROJECT	gaia	Project name (lowercase, no stopwords)
CKAN_URL	https://datalake.gaia-project.cloud	GeoDataLake API base URL
OAUTH_URL	https://auth.gaia-project.cloud	OAuth2 / FusionAuth base URL
OAUTH_USERNAME	—	Your username
OAUTH_PASSWORD	—	Your password
OAUTH_API_KEY	—	API key (provided separately)
OAUTH_APP_ID	—	Application ID (provided separately)
RMQ_HOSTNAME	bus.gaia-project.cloud	RabbitMQ hostname
RMQ_PORT	5671	RabbitMQ TLS port
RMQ_VHOST	example	RabbitMQ virtual host
RMQ_EXCHANGE	amq.topic	RabbitMQ topic exchange name
RMQ_QUEUE	qexample.geodatalake	RabbitMQ queue name
RMQ_USERNAME	—	RabbitMQ username
RMQ_PASSWORD	—	RabbitMQ password
RMQ_CA_FILE	(optional)	Path to the CA certificate PEM file

---

Upload resources to the geodata repository

The example uploads a mock dataset from data/datasets/NameDataset/.

• INSPIRE metadata: data/metadataINSPIRE_NameDataset.json — a JSON dict with all fields compliant to the INSPIRE metadata guidelines. All values are strings except spatial, which holds a GeoJSON geometry.
• Per-resource metadata: data/datasets/NameDataset_resource_metadata/ — one .json file per dataset file describing temporal validity, datatype ID, format, etc.

The upload runs in two steps:

1. Upload the INSPIRE metadata package.
2. Iteratively upload dataset files, each attached to the metadata package. If any upload fails the metadata package is deleted automatically.

uv run tools/upload.py

---

Download resources from the geodata repository

When a new resource is uploaded, the repository automatically sends a notification to the messaging bus. The download script:

1. Simulates sending a notification to the bus (for testing — this step is not needed in production).
2. Connects to the queue and listens for notifications.
3. Downloads each notified resource to outputs/.

uv run tools/download.py

Press Ctrl-C to stop listening.

---

Check the RabbitMQ connection

tools/rabbitmq.py is a lightweight connectivity smoke test. It connects to the broker, binds the configured queue to the exchange with a hello.# routing key, publishes a test message, and consumes it back. Useful to verify credentials and TLS configuration before running the main scripts.

uv run tools/rabbitmq.py

---

Useful links

• INSPIRE metadata guidelines
• Validate GeoJSON geometry
• CKAN API docs
• Example metadata table
• Postman collection