mock-relay-server

A tool for using simple data input to mock a GraphQL schema in a way that allow for full CRUD access

Example

import { getMockedResolvers } from 'mock-relay-server'
import { makeExecutableSchema } from 'graphql-tools'

// Your full schema in GraphQL type language
const typeDefs = `
  type User {
      id : ID!
      name: String
  }
  ...
`

// The data to initialize your mock database
const users = [
    {
        id: '1',
        name: 'User One',
    },
    {
        id: '1',
        name: 'User Two',
    },
]

const mocks = {
    User: {
        // connections are created based on any types that are provided a
        // a data key, in this case only user
        // this is your interface with the mock database
        data: users,
    },
    Query: {
        // this resolver functions similarly to resolvers in graphql-tools
        // with the only additional wrinkle that you're provided connections
        // as a way to interface with the database
        resolver: connections => ({
            getUser: async (query, { userId }) => connections.User.getNode(userId),
            getUsers: connections.User.paginate,
        }),
    },
    Mutation: {
        resolver: connections => ({
            updateUser: connections.Widget.update,
            addUser: connections.Widget.create,
            deleteUser: connections.Widget.delete,
        }),
    },
}

const resolvers = getMockedResolvers(mocks)

// at this point, you have a fully valid graphQL schema that can be
// be used to locally run queries
const schema = makeExecutableSchema({ typeDefs: [typeDefs], resolvers })

Creating mocks

The mocks in mock-relay-server are very similar to the resolver map from graphql-tools with a couple of key differences. First, you can provide an optional data key which gives you full CRUD on the array of items provided. Second, instead of providing resolver directly, it's provided on a resolver key that returns a function that provides you the initialized connections and returns the final resolver map.

Connections API

Connections are used to make it easy to implement resolvers. Most of the time, you should be able to use one of the helpers on a connection with out needing any configuration.

create / update / delete

These all should all work out of the box with very little customization. The only requirement is that the mutation that they're mocking only accepts an input object as a variable

getNode

Accepts a global id(the combination of the id and the nodes type) and returns the associated node

paginate

Accepts standard relay compliant input variables and returns the expected relay compliant connection format