backend

Pollaris Backend

This directory is the Django app for the Pollaris backend.

If you're browsing the code, check out urls.py, search_views.py, and models.py to get started.

Data Model

To have results to serve you'll need to populate the database with data.

Different states assign polling locations in different ways, and Pollaris is able to handle day-of-election polling locations, early vote centers, and drop boxes. It can handle either one or multiple polling locations per precinct.

The data model has two main sections:

1. Mapping a person's address to the precinct they live in. This is represented in the tables StreetSegment and Zip9ToPrecinct.
2. Mapping a precinct to the polling location(s) for that precinct. This is represented in PollingLocation, PrecinctToPollingLocation, and the corresponding tables for early vote locations and drop boxes.

Local Development

To set up the basic environment variables needed, you can start with the example env variables. Run:

cp example.env .env

We use pipenv to manage dependencies. To install dependencies:

pipenv install -d

Start the DB:

docker-compose up -d

Run the migrations to initialize the database:

pipenv run python manage.py migrate

Run the server:

pipenv run python manage.py runserver

Run tests, not including tests marked as "slow":

pipenv run pytest -m "not slow"

Run all tests:

pipenv run pytest

Deployment

On the Warren campaign, we ran this app on AWS using Lambda and Aurora Postgres Serverless, and used Serverless to deploy.