GraphQL Server Boilerplate

Based upon turboless

The repositories contains 2 lambda functions:

• Apollo: a simple hello world graphql server with no outside connection.
• Artemis: a graphql server that connects to a postgres database.

Both repositories shares a commom ts and eslint configuration from packages folder. Artemis uses a knex configuration also available on the packages folder.

Quick start (not so much)

Create and env file with the following variables:

• POSTGRES_HOST
• POSTGRES_USER
• POSTGRES_PASSWORD
• POSTGRES_DB

Check the docker-compose.yml file for the default values for local running.

Start the database with docker-compose up -d postgres. Create the tables and seeds with npm run migrations and npm run seeds.

Run npm run dev to run locally the lambdas.

Call

curl --location 'http://localhost:3001/dev/hello' \
--header 'x-apollo-operation-name: any' \
--header 'Content-Type: application/json' \
--data '{"query":"query any {\n    hello {\n        message\n    }\n}","variables":{}}'

and

curl --location 'http://localhost:3002/dev/hello' \
--header 'x-apollo-operation-name: any' \
--header 'Content-Type: application/json' \
--data '{"query":"query any {\n    hello {\n        id\n        name\n        field\n        created_at\n        updated_at\n    }\n}","variables":{}}'

Scripts

Each app has the basic scripts:

• build: build the js files
• lint
• lint:fix
• package: build the packages for serverless deployment
• dev: run the lambdas locally
• deploy: deploy the lambdas to aws

On the root folder you can run npm run <script> to run the script on all apps. You can run a single script on a single app with npm run <script> -w apps/<app>.

For example, you can run npm run lint and see an any warning on Artemis.

Contributing

Feel free to open a PR, file an issue. We'll happily look into it.