authentik/website/developer-docs/local-dev-environment.md

2.1 KiB

title slug
Getting started /

Backend

To create a local development setup for authentik, you need the following:

  • Python 3.9
  • pipenv, which is used to manage dependencies, and can be installed with pip install pipenv
  • PostgreSQL (any recent version will do)
  • Redis (any recent version will do)

For PostgreSQL and Redis, you can use the docker-compose file in scripts/. You can also use a native install, if you prefer.

To configure authentik to use the local databases, create a file in the authentik directory called local.env.yml, with the following contents

debug: true
postgresql:
  user: postgres

log_level: debug

Afterwards, you can start authentik by running ./manage.py runserver. Generally speaking, authentik is a Django application.

Most functions and classes have type-hints and docstrings, so it is recommended to install a Python Type-checking Extension in your IDE to navigate around the code.

Before committing code, run make lint to ensure your code is formatted well. This also requires pyright, which can be installed with npm.

Run make gen to run all unittests and generated an updated swagger document for any changes you made.

Frontend

By default, no transpiled bundle of the frontend is included. To build the UI, you need Node 12 or newer.

The Frontend also uses a generated API client to talk with the backend. To generate this client, you can use the openapitools/openapi-generator-cli CLI tool.

If you want to generate the client without installing anything, run this command:

docker run \
  --rm -v $(pwd):/local \
  openapitools/openapi-generator-cli generate \
  -i /local/schema.yml \
  -g typescript-fetch \
  -o /local/web/api \
  --additional-properties=typescriptThreePlus=true,supportsES6=true,npmName=authentik-api,npmVersion=1.0.0

To build the UI, run these commands:

cd web/
npm i
npm run build

If you want to make changes to the UI, run npm run watch instead.

To ensure the code is formatted well, run npx eslint . --fix and npm run lit-analyse.