Source Backend

This is the backend for the Source platform. You can find documentation on setup for development below.

Dependencies

• = openjdk version 11.

Development setup

• Pull the project from GitHub.
• Create a .env file containing the following required information:

SUPER_SECRET_KEY [string, minimum 32 characters]
GOOGLE_CLIENT_ID [string, from Google Console]
GOOGLE_CLIENT_SECRET [string, from Google Console]

• Run the provided shell script with ./nrepl.sh to start your nrepl server. For this, you will need the required nrepl alias in .clojure/deps.edn.
• In order to start your server, evaluate the server/start-server function in dev/dev.clj. When you have made changes, you can restart the server by evaluating server/restart-server.

Database Migration

Database migration is handled using the provided migrate.sh shell script. You should have a db directory and ensure your configuration is pointing to it. The default is .db/ in the root, this can be configured with the DATABASE_DIR variable in the environment.

The migration system makes use of the kepler/mallard library, take a look at their docs to find out more on how it works. The datasource you will be operating on is passed in as context to the run-up and run-down functions. e.g. ./migrate.sh [argument] The following command arguments are available:

• up: Run all up migrations.
• down: Undo all migrations. Executes in reverse order.
• redo: Rerun the last applied migration (runs down then up).
• undo: Undo the last applied migration.
• next: Run the next unapplied migration. Same as up, but runs only 1.

In order to make a new migration, add a migration.clj file to src/source/migrations/ with a numbered prefix (e.g. 002_new_table.clj) and implement the following functions:

• (defn run-up! [context]): Run database actions to take place when running up. e.g. Create tables and seed data.
• (defn run-down! [context]): Run database actions to take place when running down. e.g. Drop tables.

Database Seeding

• Create a json file with the following structure to add admin accounts with which to seed the database. The default name is admins.json, this can be configured with the ADMINS_PATH environment variable. You will need to hash your chosen password by evaluating the hash-password function provided in src/source/password.clj:

[
    {"email": "johndoe@admin.com", "password": "hashedadminpassword"}
]

Testing

We are making use of cognitect-labs/test-runner for running our tests. Run the provided shell script with ./test.sh to run all the tests. Unit tests are written and placed in clj files in tests/source-be/. These files should have the _test postfix (e.g. google_auth_test.clj) and tests should be written per namespace and cover all important functions.