winds.mobi - real-time weather observations

winds.mobi: Paraglider pilot, kitesurfer, check real-time weather conditions of your favorite spots on your smartphone, your tablet or your computer.

winds-mobi-providers

Python scripts that get the weather data from different providers and save it in a common format into mongodb. This project use Google Cloud APIs to compute any missing station details (altitude, name, timezone, ...). Google Cloud API results are cached with redis.

Dependencies

• python 3.10 and poetry
• mongodb 4.4
• redis
• Google Cloud API key

See settings.py

Run the project with docker compose (simple way)

Create a .env file from .env.template which will be read by docker compose:

• fill GOOGLE_API_KEY with you own Google Cloud API key
• optionally fill the missing secrets for each provider

Then start the external services and the providers scheduler:

• docker compose --profile=scheduler up --build

Some providers need winds-mobi-admin running to get stations metadata.

Run the project locally on macOS

Install dependencies

• brew install openssl

• export LDFLAGS=-L/usr/local/opt/openssl/lib

• brew install libpq

• export PATH=/usr/local/opt/libpq/bin:$PATH

• brew install mysql-client

• export PATH=/usr/local/opt/mysql-client/bin:$PATH

Python environment

• poetry install
• poetry shell

Create a .env.localhost file from .env.localhost.template which will be read by dotenv for our local commands:

• fill GOOGLE_API_KEY with you own Google Cloud API key
• optionally fill the missing secrets for each provider

External services with docker compose

Create a .env file from .env.template which will be read by docker compose.

Then start the external services:

• docker compose up

Run the scheduler

• dotenv -f .env.localhost run python run_providers.py

Run only a provider

• dotenv -f .env.localhost run python providers/ffvl.py

Contributing

Checking the code style

Format your code: poetry run black .

Run the linter: poetry run flake8 .

Add a new provider to winds.mobi

You know a good weather station that would be useful for many paraglider pilots or kitesurfers?

Awesome! Fork this repository and create a pull request with your new provider code. It's easy, look at the following example:

providers/my_provider.py

import arrow
import requests

from winds_mobi_provider import Provider, StationStatus, ureg, Q_, Pressure


class MyProvider(Provider):
    provider_code = "my-provider"
    provider_name = "my-provider.com"

    def process_data(self):
        self.log.info("Processing MyProvider data...")
        # data = requests.get(
        #     "https://api.my-provider.com/stations.json", timeout=(self.connect_timeout, self.read_timeout)
        # ).json()
        data = [{
            "id": "station-1",
            "name": "Station 1",
            "latitude": 46.713,
            "longitude": 6.503,
            "status": "ok",
            "measures": [{
                "time": arrow.now().format("YYYY-MM-DD HH:mm:ssZZ"),
                "windDirection": 180,
                "windAverage": 10.5,
                "windMaximum": 20.1,
                "temperature": 25.7,
                "pressure": 1013,
            }]
        }]
        for station in data:
            winds_station = self.save_station(
                provider_id=station["id"],
                short_name=station["name"],
                name=None,  # Lets winds.mobi provide the full name with the help of Google Geocoding API
                latitude=station["latitude"],
                longitude=station["longitude"],
                status=StationStatus.GREEN if station["status"] == "ok" else StationStatus.RED,
                url=f"https://my-provider.com/stations/{station['id']}",
            )

            measure_key = arrow.get(station["measures"][0]["time"], "YYYY-MM-DD HH:mm:ssZZ").int_timestamp
            measures_collection = self.measures_collection(winds_station["_id"])

            if not self.has_measure(measures_collection, measure_key):
                new_measure = self.create_measure(
                    for_station=winds_station,
                    _id=measure_key,
                    wind_direction=station["measures"][0]["windDirection"],
                    wind_average=Q_(station["measures"][0]["windAverage"], ureg.meter / ureg.second),
                    wind_maximum=Q_(station["measures"][0]["windMaximum"], ureg.meter / ureg.second),
                    temperature=Q_(station["measures"][0]["temperature"], ureg.degC),
                    pressure=Pressure(station["measures"][0]["pressure"], qnh=None, qff=None),
                )
                self.insert_new_measures(measures_collection, winds_station, [new_measure])
        self.log.info("...Done !")


def my_provider():
    MyProvider().process_data()


if __name__ == "__main__":
    my_provider()

And test it

Start the external services:

• docker compose up

Build a Docker image containing your new provider providers/my_provider.py:

• docker build --tag=winds.mobi/my_provider .

Then run your provider inside a container with:

• docker run -it --rm --env-file=.env --network=winds-mobi-providers --entrypoint=python winds.mobi/my_provider -m providers.my_provider

To avoid building a new image on every change, you can mount your local source to the container directory /opt/project with a Docker volume:

• docker run -it --rm --env-file=.env --network=winds-mobi-providers --volume=$(pwd):/opt/project --entrypoint=python winds.mobi/my_provider -m providers.my_provider

Licensing

Please see the file called LICENSE.txt