NoiseWorks

Repository to hold code for noise case management software.

This repository houses a Django project, running against a PostgreSQL database.

Local Development

Natively

In summary, with more details below:

1. Install poetry, postgres and postgis
2. script/setup to install dependencies
3. Create a noiseworks db e.g. createdb noiseworks
4. Update the DATABASE_URL in .env to point to this e.g. DATABASE_URL='postgis://USERNAME@localhost/noiseworks'
5. poetry run script/server to run a server

Code dependencies

This repository is using poetry for python package management. There is an installer for it (though they're switching to a new one, which sounds better), but there are also hundreds of open issues for it. It may be fine.

I installed poetry by installing pipx – https://pypa.github.io/pipx/ – and then installed poetry using that – pipx install poetry. For stable poetry, this may apparently have an issue with respect to multiple python versions, but it is unclear. In the master poetry docs, the warning has been removed. We are not really bothered about multiple python versions for this project.

Anyway, once you have poetry installed, script/setup will install the dependencies.

Running

Private settings are read from environment variables, which can be listed in a local .env file.

• poetry run script/server to run migrate and the dev server

The first time you run script/server, you will be asked if you want to run via Docker or natively. (Your choice is stored in .env as the DEVENV variable.)

Docker

1. Run script/server

The first time you do this, you will be asked if you want to run via Docker or natively. (Your choice is stored in .env as the DEVENV variable.) Running with Docker, script/server will remove stopped containers, remove the node_modules volume (so it stays up to date with any changes) and runs docker-compose up which creates database and web containers, with the web container running migrate and the Django development server.

If there has been a change to a Python or Node package since your local Docker image was last built, you’ll need to rebuild the image:

docker-compose build
script/server  # Outside docker

Adding fake data

1. Get hold of a text file of UPRNs, one UPRN per line.

2. Ensure you have a MAPIT_API_KEY in your .env file.

3. Ensure you have ADDRESS_API_URL and an ADDRESS_API_KEY in your .env file.

4. Run ./manage.py loaddata action_types_hackney

5. Run ./manage.py add_random_cases --uprns UPRN_FILE --number 100 --commit

If you are running inside Docker you will want to enter a shell in the web container first: docker-compose exec web bash

Tests

• script/test or script/test --coverage to run the tests (100% coverage at present).

Tests are written with pytest; the current examples should show uses of test client, fixtures, request mocking, settings override, and so on.

Using the shell

By default, ./manage.py shell is configured to use ptpython. This provides autocomplete, vi-mode and a bunch of other useful features.

Run ./manage.py shell -i python to just use the normal python shell.

Contributing

• black is run for code tidying: poetry run black .

• Static files are handled by django-compressor with django-libsass. SCSS files are automatically compiled and dealt with.