Statsector API

Dev environment setup

Running the app in dev mode

We use poetry to manage dependencies. If you don't have poetry installed, you can run pip install poetry, or install via e.g. Homebrew.

gcloud components update
gcloud auth login
gcloud auth application-default login
gcloud config set run/region europe-west1
gcloud config set project cloud-composer-243010
poetry install
poetry shell
uvicorn app.main:app --reload

Running the tests

Unit Tests

poetry run python -m unittest discover -s tests

Performance Tests

The following tests can be run to test the performance of the API with e.g. caching enabled.

Testing the geo to statsector conversion

poetry run python -m tests.perftest_geo

Testing the address to statsector conversion

Warning: Do not test this with too high loads and with caching disabled, as it will result in a lot of requests to the Google Maps API, and increase costs.

poetry run python -m tests.perftest_address

Deploying to Cloud Run

Make sure you ran the above steps first, as this also sets your config for Cloud Run.

gcloud auth configure-docker
gcloud components install docker-credential-gcr
poetry export -f requirements.txt --output requirements.txt
gcloud run deploy statsector --port 8080 --source .

Issues

➜  statsector git:(main) ✗ gcloud run deploy statsector --port 8080 --source .       
ERROR: Error in retrieving repository from Artifact Registry.
ERROR: (gcloud.run.deploy) INVALID_ARGUMENT: Request contains an invalid argument.

This resulted from cloud build not having permissions: was made Artifact registry admin, but in the end it was a wrong region setup.

Providing API access

1. Create a service account in Google Cloud IAM
2. Create a json key for the service account and download it
3. Grant invoker access to the service account on the Cloud Run service

Run your own statsector container

To run locally without any Python knowledge / requirements, build the docker image like this:

docker build . -t statsector

Then, you can start your container like this:

docker run -p 8000:8080 statsector

(or use any other docker tools you prefer)

Once your container is up-and-running, you will see this:

INFO:app.main:Starting with geo_coding disabled
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:8080 (Press CTRL+C to quit)

And you can query statsectors like this:

curl 'http://localhost:8000/get-statsector/?lat=50.940&lon=4.204'
{
  "sector_id": "23052A0PA",
  "sector_name": "MERCHTEM-ZUID"
}

(or using the HTTP client of your language and choice)

Things to note:

• The statsector API endpoints are unauthenticated. If you want to run this in production, you will have to deal with this yourself.
• By default the app is Starting with geo_coding disabled. If you need geocoding (?address=A real street with number, zipcode city), you need to configure Google Cloud credentials.