Tau - the debate planner response cannon

Tau is the debatecore debate tournament planner project's response cannon - also known as a backend.

Deployment and local development

Environment setup

Set environment variables via .env or your shell. The following example .env file is geared for both production and development:

DATABASE_URL=postgres://tau:tau@localhost:5432/tau
SECRET=CENTRUMRWLYSONOSTARPOZNANCDNSBCD4L52SPM
DOCKER_DB_ROOT_PASSWORD=superdoopersecretpasswordthatcannotbeleaked
DOCKER_DB_PASSWORD=wedoingsecurityinhere
FRONTEND_ORIGIN=https://example.com
PORT=2023

Required variables

• DOCKER_DB_ROOT_PASSWORD will be used as the password for the database root user.
• DATABASE_URL is used for db connection. During development, this is postgres://tau:tau@localhost:5432/tau.
• FRONTEND_ORIGIN will be used as an allowed origin for the purpose of CORS. Must be a valid URL.

Optional configuration

• SECRET will be used as additional high entropy data used for generating tokens. By default, tau uses system entropy and the current UNIX timestamp.
• PORT will be used as the port the server listens on. The default is 2023.

Local development

In this scenario it is assumed, that you run the project on your local machine and use a database container for compile-time queries validation.

Note: sqlx validates queries at compile time, so being connected to a database is required to compile the crate.

architecture-beta
    group your_machine(server)[Your machine]
	
		service backend_container(server)[Backend] in your_machine
	group docker[Docker] in your_machine
	    service db_container(database)[Database] in docker
	backend_container:R -- L:db_container
	
	service env(disk)[env file or shell variables] in your_machine
	env:T -- B:backend_container
	env:T -- B:db_container

Loading

Prerequisites: docker, cargo (Rust), uv (Python).

Once you configure your environment, you can run (in your project directory):

docker compose --profile dev up -d  		# To start the database container
cargo install sqlx-cli && sqlx migrate run  # To perform sqlx migrations (recommended way)
cargo run                           		# To run the crate

Commit hook

It is advisable to run git config --local core.hooksPath .githooks/ to configure commit hooks included in this repository.

Note: the hooks require you to be connected to the database when committing.

Deployment

In this scenario, you run fully functional backend split into two containers: first with a database and second with a server.

architecture-beta
    group your_machine(server)[Your machine]
	
	group docker[Docker] in your_machine
	    service backend_container(server)[Backend] in docker
	    service db_container(database)[Database] in docker
	    
	backend_container:R -- L:db_container

	service env(disk)[env file or shell variables] in your_machine
	env:T -- B:backend_container
	env:T -- B:db_container

Loading

Prerequisites: docker.

Once you configure your environment, you can simply run:

docker compose --profile prod

Documentation

• API documentation: Once the project is built, you can access the API documentation at /swagger-ui.
• ER diagram: a database ER diagram can be found in docs/er_diagram.pdf. If you have your commit hooks configured, it will be automatically updated.