Setup a GraphQL server in seconds using graphql-serve

GraphQL Serve

TL;DR — you can run a GraphQL backend without code using the graphql-serve package

Learning GraphQL is difficult, more so with all the boilerplate that is needed to create applications using it. The Graphback project is driven by the need to simplify GraphQL adoption.

The motivation of the Graphback project has always been to automatically generate boilerplate code for server. The project is a collection of packages that:

• Decorate schema with out of the box data queries and mutations basing on data models.
• Generate GraphQL resolvers either at runtime or as static code in your project.
• Generate database migrations from data models for relational databases.

All of these features can be accessed through the graphback-cli package.

Introducing GraphQL serve

The newest addition to the Graphback family of packages starting with 0.11.0 is graphql-serve.

While graphback-cli operates typically on code graphql-serve provides a full GraphQL server embedded in the command line tool.

It generates a fully-featured GraphQL server using Node.js, Apollo server and MongoDB database. GraphQL serve supports GraphQL config for further customisation. For example you can add an additional Offix plugin that will provide conflict resolution and data sync queries.

Usage

Install the graphql-serve package with npm or yarn:

npm i -g graphql-serve

or

yarn add graphql-serve

All you have to do then, is to define your data models. Data models are one or more files with the .graphql extension. Here are the contents of an example model file, let’s say Note.graphql:

""" @model """
type Note {
  id: ID!
  title: String!
  description: String
  likes: Int
}

Graphback uses these files to generate a GraphQL schema and resolvers. The definition of the model is similar to a GraphQL schema except it has some annotations that are used by Graphback to determine which types need a table in the database as well as relationships between those tables. All of these are covered in the Graphback docs here.

Put this in the current working directory and run the following command(for yarn users, please prepend yarn to all the commands below):

$ gqls serve . --port=8080

Snap!

This command takes the data model files in the current working directory and starts a GraphQL server that listens on port 8080. It also has the GraphQL Playground from Apollo GraphQL so you can test queries by going to the endpoint on a web browser.

If you just want to see the resulting generated GraphQL schema, use the print-schema command:

$ gqls print-schema .

This will only print the GraphQL schema generated from the data model files in the current working directory.

Why we created GraphQL serve

With just one command, you can have a GraphQL server running on your machine that you can use to test a frontend application or just test GraphQL. GraphQL serve supports data loading and has comprehensive support for all types of relationships

Unlike mocking alternatives, the MongoDB database used by default in graphql-serve persists through queries. We also provide a way to configure your own database using graphql-config, this is documented here. Thank you for reading!

Try graphql-serve today and let us know what you think!