This project consists in a code assessment challenge to wrap the public Open Weather API while caching it's returned data and exposing an endpoint to get a city's current weather by it's name, as well as an endpoint to retrieve the latest cached city weathers. The cached data lasts for a default of 5 minutes.
To run this project, you will need Docker installed. Then, from the directory of the cloned repository, run one of the following commands:
docker-compose up -d, for the development stage;docker-compose -f docker-compose.yml -f docker-compose.test.yml up -d, for running the automated tests.docker-compose -f docker-compose.yml -f docker-compose.prod.yml up -d, for the production stage.
For the development and production stage, the API will be exposed locally at the port 5000. In order to achieve testing that represents best the production stage, the API and the automated tests use the same container. Because of that, they currently can't be run simultaneously.
After running them once, the automated tests can be rerun by using docker start pytest.
DISCLAIMER: The automated tests currently are taking long to run due to the test_get_weather_cache_expiration test. Even though this test reduces the configured cache expiration time, it still needs to wait at least 60 seconds in order to reliably assert if a given saved cache expired, since that is the duration of the cycle in which MongoDB checks for documents with expired TTL.
/weather/<city_name>: Get the weather data for the specifiedcity_name. If available, cached data is used. Otherwise, new data is fetched from the Open Weather API and cached./weather?max=<max_number>: Get the weather data for all the cached cities, up to the latestmax_numberentries (if specified).
The system runs in two docker containers refering to the docker-compose services api-service, which contains the application, and db-service, which runs a MongoDB instance used for caching (automatically expiring with TTL).
At the application, the endpoints are defined at the api.py module using Flask, which in turn connects with the data.py module. The data.py module handles where to retrieve from and save data to by interacting with the resources modules open_weather.py and mongo_db.py.
The automated tests, which use pytest, also run in the container referring to the docker-compose service api-service, although not simultaneously with the full, exposed, API.