Small adjustments to the script should fix it

Write a small test that basically does the whole workflow but for a handful of stations.

Right now, the methodology of the scraper is rather rudimentary, and while we are able to pick up the names of the stations, they are sometimes not specific enough to resolve the accurate location of the monitoring station.

I am working on a way to reverse geocode each of the stations using once again a hacky Selenium solution and Mapdeveloper's geocode tool, which provides a set of coordinates, state/zip code, etc. given a direction. Many of the names are not precise enough though, and the obtained coordinates will be rather imprecise but at least work at the city/prefecture level.

We should find a solution to solve this by fetching the coordinate values directly from AQICN, since they must have them to display the stations on the map, but I am not sure they expose them to the user in the website.

The site changed ~3 months ago so the Selenium scraper just doesn't work right now.
Should re-write the instructions to adapt to the new website.

Github is removing support for Node 12, and several of the actions used in the GHA depend on it, so they're now broken.

• actions/setup-python@v2
• Gr1N/setup-poetry@v7

This is probably breaking the CI/CD of many of my repos, so better get it solved soon.