fuelstats

fuelstats is an app that allows you to get differents dynamic stats around fuel prices in France. 🥖 feel free to open issues or pull requests if you want/need to.

Project pre-setup

1. install python>3.7
2. download your yearly archive from here
3. unzip your archive and add it to your folder of choice, and specify the path (PATH_TO_DATA) and the file name without the extension (DATA_ARCHIVE_NAME) separately in your .env file. (xml files only)

note : you will need an internet connection, a Sentry account (for bug tracking in production environments - optional, but you'll have to do some code adjusting if you decide not to use it) and a Google Maps API access (paid) for fuelstats to work correctly.

How-to

Clone this project

git clone https://github.com/ceIia/fuelstats

Get a Google Maps API key

1. Log in to the Google Cloud Platform Console
2. Setup a project with the following accesses : Places API, Distance Matrix API, Maps Static API and Geocoding API.
3. Make sure to secure the usage of your API key as it will be publicly available (limit the access to your domain-s).
4. Copy your API key and go on to the next step.

Setup API keys

1. Rename .env.sample to .env
2. Edit the file and configure the values

Example :

GOOGLE_MAPS_API_KEY=your_key_here <== replace the value after the equal sign

Install fuelstats dependencies (cd into root clone folder before)

pip3 install -r requirements.txt
(if it fails) python3 -m pip install -r requirements.txt

Start the production server

gunicorn3 --bind unix:fuelstats.sock -m 007 webizer:app -b 0.0.0.0:80 --timeout 300

I recommend you create a unix service to run the gunicorn instance for fuelstats so you can get better CI integration, error handling, clean logging and seamless runtimes.

Start the flask webserver (DEVELOPMENT ONLY)

FLASK_DEBUG=1 FLASK_APP=webizer.py python3 -m flask run --host=0.0.0.0

Only use the built-in Flask server for development purposes, use gunicorn for productionn purposes. If a critical error occurs, the whole flask server will crash, requiring manual restart. Using an actual WSGI server will provide stability and better logging.

---

this project was made possible thanks to the french government. #bigdata