Building GraphQL servers in golang

This tutorial will take you through the process of building a GraphQL server with gqlgen that can:

Return a list of todos
Create new todos
Mark off todos as they are completed

You can find the finished code for this tutorial here

Set up Project

Create a directory for your project, and initialise it as a Go Module:

mkdir gqlgen-todos
cd gqlgen-todos
go mod init github.com/[username]/gqlgen-todos

Next, create a tools.go file and add gqlgen as a tool dependency for your module.

//go:build tools

package tools

import (
	_ "github.com/99designs/gqlgen"
)

To automatically add the dependency to your go.mod run

go mod tidy

By default you’ll be using the latest version of gqlgen, but if you want to specify a particular version you can use go get (replacing VERSION with the particular version desired)

go get -d github.com/99designs/gqlgen@VERSION

Building the server

Create the project skeleton

go run github.com/99designs/gqlgen init

This will create our suggested package layout. You can modify these paths in gqlgen.yml if you need to.

├── go.mod
├── go.sum
├── gqlgen.yml               - The gqlgen config file, knobs for controlling the generated code.
├── graph
│   ├── generated            - A package that only contains the generated runtime
│   │   └── generated.go
│   ├── model                - A package for all your graph models, generated or otherwise
│   │   └── models_gen.go
│   ├── resolver.go          - The root graph resolver type. This file wont get regenerated
│   ├── schema.graphqls      - Some schema. You can split the schema into as many graphql files as you like
│   └── schema.resolvers.go  - the resolver implementation for schema.graphql
└── server.go                - The entry point to your app. Customize it however you see fit

Define your schema

gqlgen is a schema-first library — before writing code, you describe your API using the GraphQL Schema Definition Language. By default this goes into a file called schema.graphqls but you can break it up into as many different files as you want.

The schema that was generated for us was:

type Todo {
  id: ID!
  text: String!
  done: Boolean!
  user: User!
}

type User {
  id: ID!
  name: String!
}

type Query {
  todos: [Todo!]!
}

input NewTodo {
  text: String!
  userId: String!
}

type Mutation {
  createTodo(input: NewTodo!): Todo!
}

Implement the resolvers

When executed, gqlgen’s generate command compares the schema file (graph/schema.graphqls) with the models graph/model/*, and, wherever it can, it will bind directly to the model. That was done already when init was run. We’ll edit the schema later in the tutorial, but for now, let’s look at what was generated already.

If we take a look in graph/schema.resolvers.go we will see all the times that gqlgen couldn’t match them up. For us it was twice:

func (r *mutationResolver) CreateTodo(ctx context.Context, input model.NewTodo) (*model.Todo, error) {
	panic(fmt.Errorf("not implemented"))
}

func (r *queryResolver) Todos(ctx context.Context) ([]*model.Todo, error) {
	panic(fmt.Errorf("not implemented"))
}

We just need to implement these two methods to get our server working:

First we need somewhere to track our state, lets put it in graph/resolver.go. The graph/resolver.go file is where we declare our app’s dependencies, like our database. It gets initialized once in server.go when we create the graph.

type Resolver struct{
	todos []*model.Todo
}

Returning to graph/schema.resolvers.go, let’s implement the bodies of those automatically generated resolver functions. For CreateTodo, we’ll use the crypto.rand package to simply return a todo with a randomly generated ID and store that in the in-memory todos list — in a real app, you’re likely to use a database or some other backend service.

func (r *mutationResolver) CreateTodo(ctx context.Context, input model.NewTodo) (*model.Todo, error) {
	randNumber, _ := rand.Int(rand.Reader, big.NewInt(100))
	todo := &model.Todo{
		Text: input.Text,
		ID:   fmt.Sprintf("T%d", randNumber),
		User: &model.User{ID: input.UserID, Name: "user " + input.UserID},
	}
	r.todos = append(r.todos, todo)
	return todo, nil
}

func (r *queryResolver) Todos(ctx context.Context) ([]*model.Todo, error) {
	return r.todos, nil
}

Run the server

We now have a working server, to start it:

go run server.go

Open http://localhost:8080 in a browser. Here are some queries to try, starting with creating a todo:

mutation createTodo {
  createTodo(input: { text: "todo", userId: "1" }) {
    user {
      id
    }
    text
    done
  }
}

And then querying for it:

query findTodos {
  todos {
    text
    done
    user {
      name
    }
  }
}

Don’t eagerly fetch the user

This example is great, but in the real world fetching most objects is expensive. We don’t want to load the User on the todo unless the user actually asked for it. So lets replace the generated Todo model with something slightly more realistic.

First let’s enable autobind, allowing gqlgen to use your custom models if it can find them rather than generating them. We do this by uncommenting the autobind config line in gqlgen.yml:

# gqlgen will search for any type names in the schema in these go packages
# if they match it will use them, otherwise it will generate them.
autobind:
 - "github.com/[username]/gqlgen-todos/graph/model"

And add Todo fields resolver config in gqlgen.yml to generate resolver for user field

# This section declares type mapping between the GraphQL and go type systems
#
# The first line in each type will be used as defaults for resolver arguments and
# modelgen, the others will be allowed when binding to fields. Configure them to
# your liking
models:
  ID:
    model:
      - github.com/99designs/gqlgen/graphql.ID
      - github.com/99designs/gqlgen/graphql.Int
      - github.com/99designs/gqlgen/graphql.Int64
      - github.com/99designs/gqlgen/graphql.Int32
  Int:
    model:
      - github.com/99designs/gqlgen/graphql.Int
      - github.com/99designs/gqlgen/graphql.Int64
      - github.com/99designs/gqlgen/graphql.Int32
  Todo:
    fields:
      user:
        resolver: true

Next, create a new file called graph/model/todo.go

package model

type Todo struct {
	ID     string `json:"id"`
	Text   string `json:"text"`
	Done   bool   `json:"done"`
	UserID string `json:"userId"`
	User   *User  `json:"user"`
}

And run go run github.com/99designs/gqlgen generate.

If you run into this error package github.com/99designs/gqlgen: no Go files while executing the generate command above, follow the instructions in this comment for a possible solution.

Now if we look in graph/schema.resolvers.go we can see a new resolver, lets implement it and fix CreateTodo.

func (r *mutationResolver) CreateTodo(ctx context.Context, input model.NewTodo) (*model.Todo, error) {
	randNumber, _ := rand.Int(rand.Reader, big.NewInt(100))
	todo := &model.Todo{
		Text:   input.Text,
		ID:     fmt.Sprintf("T%d", randNumber),
		UserID: input.UserID,
	}
	r.todos = append(r.todos, todo)
	return todo, nil
}

func (r *todoResolver) User(ctx context.Context, obj *model.Todo) (*model.User, error) {
	return &model.User{ID: obj.UserID, Name: "user " + obj.UserID}, nil
}

Finishing touches

At the top of our resolver.go, between package and import, add the following line:

//go:generate go run github.com/99designs/gqlgen generate

This magic comment tells go generate what command to run when we want to regenerate our code. To run go generate recursively over your entire project, use this command:

go generate ./...