On June 30 2020 the city of Gent migrated its Open Data portal to a new platform, so this hack is not needed anymore (and it won't work anyway). This repository is archived and read-only now.
Improved GeoJSON datasets for data.stad.gent
The Open Data portal of the City of Ghent has a collection of
datasets can can be downloaded in different formats, like XML, KML, JSON or GeoJSON. The problem with
the GeoJSON responses is that they normally only contain MultiPolygon
or MultiPoint
collections
of coordinates and very little extra information (it seems the City of Ghent is not using the latest version
of The DataTank, that fixes this).
This simple application reads the associated KML files to generate more complete GeoJSON responses.
How to use it
You can access the data at https://geogent.herokuapp.com/.
To make a request you just need to replace the City of Ghent DataTank domain datatank.stad.gent
with
geogent.herokuapp.com
. For example, the following URLs:
https://datatank.stad.gent/4/cultuursportvrijetijd/speelterreinen.geojson
https://datatank.stad.gent/4/mobiliteit/parkinglocaties.geojson
become:
https://geogent.herokuapp.com/4/cultuursportvrijetijd/speelterreinen.geojson
https://geogent.herokuapp.com/4/mobiliteit/parkinglocaties.geojson
For developers
geogent
uses Flask. Feel free to contribute to the project.
Application dependencies
The application uses Pipenv to manage Python packages:
$ pipenv install --dev
$ pipenv shell
Update dependencies (and manually update requirements.txt
):
$ pipenv update --dev && pipenv lock -r
Running the server
$ python runserver.py
Style guide
$ black geogent
Heroku
Use Heroku Containers to deploy the app:
$ heroku login
$ heroku stack:set container
$ git push heroku master