• Update the VOTER_FILE and DATABASE_OUT parameters in import_voters.py (DATABASE_OUT should be a temporary file that you inspect before moving it to database.json).
• Run the python3 -m car.script.import_voters script.
  • Expects an Alabama Secretary of State-format voter file export.
  • Writes a single turf containing all voters (other future scripts will create other turfs).
• Restart the app service to pick up the new database.

Some todos:

• Title case voter and street names for display (not all caps)
• Display age, not birthdate (we only get age from SOS)
• Allow merging (not just overwriting) databases; this may be a separate script.

Once you have imported voters you will probably want to geocode their doors so you can cut turf.

You can use the car.script.geocode_voters script for this. It references the geocode repository. Getting data sources for geocoding and defining a correct CompositeGeocoder is unfortunately left as an exercise for the reader.

Use the car.script.export_geocoded_voters script to create geocoded_doors.geojson. Open this file as a vector layer in QGIS.

Create a SpatiaLite layer called "turfs" with two fields: name (string) and car_id (int), with Polygon geometry. Draw turfs as you like.

When you are done cutting turf, run TURF_DATA_PATH=/path/to/your/layer_db.sqlite python3 -m car.script.update_voter_turfs to match doors to turfs, and move the doors/voters into those turfs. You will need to restart the web app to see the results. Note that you don't have to cover all voters with a turf; voters not in a turf will remain in the default "All Voters" turf.

The update_voter_turfs script also reorders doors in turfs. If you are in a grid city and are cutting griddy turfs, it will work basically perfectly. If you are not in a grid city, the lazy-TSP algorithm will try its best but probably fail quite miserably. Good luck! :3

python -m venv .venv
source .venv/bin/activate
pip install .
python3 app.py

uv run app.py