data-registry-api

Running the server

Running Locally

You should have python 3.9 or above installed and docker if you wish to use local docker db server. You can start a local mysql via ./docker_db/docker_db.sh start and the run alembic upgrade head to create the db tables. You'll likely want/need to specify the following environment variables either in your .env file (don't commit to vcs) or in your cli invocation:

• DATA_REGISTRY_DB_CONNECTION - the connection string to the database, e.g. export DATA_REGISTRY_DB_CONNECTION=mysql+pymysql://dataregistry:dataregistry@localhost:3307/dataregistrywill work if you're using the docker_db script
• DATA_REGISTRY_S3_BUCKET - the s3 bucket to use for storing the files (use dig-data-registry-qa unless you need prod data)
• GOOGLE_OAUTH_CLIENT_ID - this is necessary if you will be testing authentication look this up here.
• GOOGLE_OAUTH_CLIENT_SECRET - same as above
• GOOGLE_OAUTH_REDIRECT_URI - http://local.kpndataregistry.org:3000/login/callback

To set up dependencies via virtual environment: python -m venv venv source venv/bin/activate pip install -r requirements.txt

Override the db via an environment variable if you are using a local db: export DATA_REGISTRY_DB_CONNECTION=mysql+pymysql://dataregistry:dataregistry@localhost:3307/dataregistry

To run the server: python3 -m dataregistry.main serve

FastAPI will automatically generate a swagger/open page at http://localhost:5000/docs which you can use to test/explore all the API endpoints.

For unit tests:

• Start a local mysql via ./docker_db/docker_db.sh start
• if you want to override the test db's url, set the env var DATA_REGISTRY_TEST_DB_CONNECTION otherwise it will default to the docker db defaults.
• From project root run pytest or you can run tests from Pycharm so long as you set the working directory to the project root.

Running Docker

Python dependencies can be a pain. If you just need a running a API server so you can do front end work, using Docker is a decent option. Depending on your needs you will need to specify some or all of the environments variables listed above. You can do that by creating an .env in the dataregistry directory (keep this out of source control) or specifying the variables with the -e option in the docker run invocation below.

docker build -t <name> .

docker run --rm --name dr_container -p 5000:5000 <name>

With some environment variables specified: docker run -e DATA_REGISTRY_DB_CONNECTION=<> --rm --name dr_container -p 5000:5000 <name> Docker will use environment variables in your .env file too so you can override everything there if you like.

Deploying the server

• Pushes/PR merges to main will deploy to QA via Github actions.
• Publishing a tag pointing to a commit in the main branch will deploy that version to production