hawthorn-server

The hawthorn-server repository contains the source code for the backend, data-access layer of the Hawthorn project.

This backend process serves a GraphQL API on port :4000.

Software dependencies

The following software must be installed prior to running hawthorn-server:

• git - This is for version control
• Node.js - Node is used for running hawthorn-server and installing dependent JavaScript packages using the npm command.
• nodemon - This is for running your development server and hot-swapping for code changes
• Docker - PostgreSQL and Prisma (you only need to install Docker manually, not these other things)
• Prisma CLI - The prisma CLI tool is used for deploying the database schema to the database, and for generating an interface for our hawthorn-server to interact with the database through.

Set up a development environment

The following steps walk you through downloading hawthorn-server, installing some last minute packages, and running the server so that you can start coding. Please make a pull request if this process is incomplete or inaccurate, and open an issue if you cannot get things working.

Clone hawthorn-server

Download the source code and change directories into your freshly cloned hawthorn-server directory:

#If your `git` is configured with SSH:
git clone [email protected]:trex/hawthorn-server.git && cd hawthorn-server
#If your `git` is configured with HTTPS:
git clone https://github.com/trex/hawthorn-server.git && cd hawthorn-server

Create a .env file

The Docker containers use environment variables for configuring their services, these are read from a .env file.

Create .env in the project root directory:

touch .env

Paste the following contents into .env in your text editor:

# PostgreSQL
POSTGRES_USER=postgres
POSTGRES_PASSWORD=postgres

Download and run the Docker containers

The following command will download all of the required Docker images and get them running in the background. If it is your first time running this command, it can take a while to download the Docker images.

docker-compose up -d

You can make sure all of the containers are up and healthy with the following command, the output is included to show the expected result:

docker-compose ps

       Name                      Command               State                       Ports
-------------------------------------------------------------------------------------------------------------
server_db_1           docker-entrypoint.sh postgres    Up      0.0.0.0:5432->5432/tcp
server_prisma_1       /bin/sh -c /app/start.sh         Up      0.0.0.0:4466->4466/tcp

Deploy the database schema

The following command will deploy the database schema to the database, and generate JavaScript code that we will call to interact with our database:

prisma deploy

You can test that the schema was successfully deployed by navigating your browser to http://localhost:4466/. You should see a Prisma admin interface, click the "SCHEMA" and "DOCS" tabs on the righthand side to see the possible queries and mutations we can make against our data access layer, and the various data we can expect to be returned. This defines the interface hawthorn-server will use to query and mutate data in the database.

Install Node and JavaScript packages

The following command installs the required packages that are specified in package.json:

npm install

Run the API server

Finally, start the hawthorn-server process:

npm run dev

You can test that everything is working by navigating your browser to http://localhost:4000/. You should see a GraphQL admin interface, that looks similar to the Prisma admin interface.

You will notice that the "DOCS" and "SCHEMA" tabs on the right show much less information than they did in the Prisma interface. Here, the "SCHEMA" and "DOCS" tabs show the possible queries and mutations we can make against our GraphQL API layer. This defines the interface client applications (hawthorn-client) use to talk to hawthorn-server.

System layout

GraphQL - API layer

The graphql schema is defined in schema.graphql. The schema specifies various datatypes, and the how data may be queries or mutated through the GraphQL API.

The resolver functions are defined in index.js. Resolvers determine how to handle a requests to the GraphQL API.

Prisma - data access layer

Prisma deploys a database schema to our PostgreSQL database based on the schema definition found in datamodel.prisma.

Prisma offers an easy method of interacting with the database by generating JavaScript code that defines these interactions.

Updating the data model

Update datamodel.prisma to define changes to the data model. Then, deploy an updated database schema:

prisma deploy

Finally, generate an updated JavaScript client:

prisma generate

Docker

The following services run in docker containers:

• prisma client - ORM for primary application database
• postgresql - primary application data store.

Run these services in the background with the command:

# For development:
docker compose up -d
# For production:
docker-compose -f docker-compose.yml -f docker-compose.production.yml up -d