FastAPI Demo - Postgres

A simple project to show a couple examples of defining an API using Tiangolo's FastAPI that connects to a PostgreSQL DB using Tiangolo's SQLModel.

About this demo data

I just got a CSV file that contains a list of all sports facilities in Andalusia, hosted in Junta de Andalcía's Open Data portal. Current file was updated on 1st of July 2021. You can check it out if there is any update here.

Required environment variables

Only the parameters needed to reach our PostgreSQL database:

Variable	Description	Example
PG_USER	PostgresSQL DB Username	postgres
PG_PASS	PostgresSQL DB Password	postgres
PG_HOST	PostgresSQL DB Host URL or DNS	localhost
PG_PORT	PostgresSQL DB Port	5432
PG_DB	PostgresSQL DB Name	fast-api-demo

Running with docker-compose (PostgreSQL + API)

You have a docker-compose file for testing this API.

docker-compose up

Note if it's the first time you run the DB, you will need to create and populate the sports_facilities table. You can copy both files under data folder to, for example to tmp folder, inside your postgresql container and, after that, load the data. Here you have some useful commands for doing that:

docker cp data/sports_facilities.sql fast-api-demo_postgres_1:/tmp
docker cp data/sports_facilities_census.csv fast-api-demo_postgres_1:/tmp
docker exec -it fast-api-demo_postgres_1  psql -U postgres -d fast-api-demo -f /tmp/sports_facilities.sql

Running in docker (only API)

# Building local image
docker build -t fast-api-demo .

# We are asuming we have a postgresql running in localhost, change this to reach your instance
docker run --rm \
-p 5000:5000 \
-e PG_USER=postgres \
-e PG_PASS=postgres \
-e PG_HOST=localhost \
-e PG_PORT=5432 \
-e PG_DB=fast-api-demo \
--entrypoint /usr/src/app/docker-entrypoint.sh \
fast-api-demo

API Doc

If you want to see this API documentation, you can take a look by accessing to your API endpoint and querying to /docs (http://localhost:5000/docs if you're running it locally, for example).

Running tests

We do use Pytest for running unit tests in this project. Note you need to install requirements_dev before. It is as easy as running this couple commands (from the root folder of this project):

# Running test
python -m pytest
# Check test coverage
python -m pytest --cov=.