Caddy is a small application to harvest land parcel legal description fields and index address data via minimal a Django application, and expose it as a searchable API for geocoding.
The recommended way to set up this project for development is using
uv
to install and manage a Python virtual environment.
With uv installed, install the required Python version (see pyproject.toml). Example:
uv python install 3.12
Change into the project directory and run:
uv python pin 3.12
uv sync
Activate the virtualenv like so:
source .venv/bin/activate
To run Python commands in the activated virtualenv, thereafter run them like so:
python manage.py
Manage new or updated project dependencies with uv also, like so:
uv add newpackage==1.0
This project uses environment variables (in a .env file) to define application settings.
Required settings are as follows:
DATABASE_URL=postgres://USER:PASSWORD@HOST:PORT/NAME
SECRET_KEY=ThisIsASecretKey
AZURE_ACCOUNT_NAME=azureaccountname
AZURE_ACCOUNT_KEY=azureaccountsecret
AZURE_CONTAINER=containername
Run the frontend application with python geocoder.py (the default port
is 8080, which can be overridden by defining a PORT environment variable.
Run Django console commands manually:
python manage.py shell_plus
The shack application contains a single model, Address. This model is used to store the relevant address fields of the cadastre dataset, plus each land parcel's centroid and spatial bounds. A utility script in shack/utils.py is used to query the cadastre WFS layer to mirror the data. Address fields are rendering into a single document, stored in the address_text field.
The full text search in this project leverages PostgreSQL features for parsing and normalising text documents. A custom Django migration has been written to create a tsvector column in the Address table to store precalculated tsvector values for each address document, plus a GIN index on that field and a database trigger to update the index on insert or update.
The API endpoint to geocode addresses is a custom view that uses raw SQL to query the tsvector field and return any results. Returned data is deliberately limited to a "human readable" address, object centroid and bounding box.
Further reference: http://www.postgresql.org/docs/current/static/textsearch.html
This project includes the following pre-commit hooks:
Pre-commit hooks may have additional system dependencies to run. Optionally install pre-commit hooks locally like so:
pre-commit install
Reference: https://pre-commit.com/