Break down your big monolitic GraphQl schema.js file into multiple files following the structure that makes sense to you and your team. SchemaGlue.js will glue all those files together for you ♥ ʘ‿ʘ.
Make your code more readable and understandable by breaking down your monolithic GraphQL schema and resolver into smaller domain models. SchemaGlue.js will help glueing them back together.
Without SchemaGlue - Stuck With a Monolithic Schema
- src/
|__ schema.js
- index.js
- package.json
With SchemaGlue - Structure Your Schema At Will
- src/
|__ graphql/
|__ product/
| |__ schema.graphql
| |__ resolver.js
|
|__ variant/
|__ schema.graphql
|__ resolver.js
- index.js
- package.json
npm install schemaglue --save
NOTICE: All the following examples use webfunc and graphql-serverless. graphql-serverless makes developing GraphQl server locally dead easy while webfunc is a serverless web framework that allows to write Express-like server to deploy them on any serverless solution.
const glue = require('schemaglue')
// This will look for any js files under the 'src/graphql' folder that
// comply to a certain convention described below. Those files describe
// bits and pieces of the GraphQL schema and resolver that schemaglue will
// reassemble into a single 'schema' string and 'resolver' object.
const options = {
js: '**/*.js', // default
ignore: '**/somefileyoudonotwant.js'
}
const { schema, resolver } = glue('src/graphql', options)
PREREQUISITE - Install the following dependencies before running the following code:
npm install webfunc graphql-serverless schemaglue lodash --save
Project Structure Example
- src/
|__ schema.js
- index.js
- package.json
Where the schema.js file would contain the entire GraphQl schema and resolver definition:
// src/schema.js
const { graphqlError } = require('graphql-serverless')
const { makeExecutableSchema } = require('graphql-tools') // this dependency is automatically included in 'graphql-serverless'
const _ = require('lodash')
const schema = `
type Product {
id: ID!
name: String!
shortDescription: String
}
type Variant {
id: ID!
name: String!
shortDescription: String
}
type Query {
# ### GET products
#
# _Arguments_
# - **id**: Product id (optional)
products(id: Int): [Product]
# ### GET variants
#
# _Arguments_
# - **id**: Variant id (optional)
variants(id: Int): [Variant]
}
`
const productMocks = [{ id: 1, name: 'Product A', shortDescription: 'First product.' }, { id: 2, name: 'Product B', shortDescription: 'Second product.' }]
const productResolver = {
Query: {
products(root, { id }, context) {
const results = id ? productMocks.filter(p => p.id == id) : productMocks
if (results.length > 0)
return results
else
throw graphqlError(404, `Product with id ${id} does not exist.`)
}
}
}
const variantMocks = [{ id: 1, name: 'Variant A', shortDescription: 'First variant.' }, { id: 2, name: 'Variant B', shortDescription: 'Second variant.' }]
const variantResolver = {
Query: {
variants(root, { id }, context) {
const results = id ? variantMocks.filter(p => p.id == id) : variantMocks
if (results.length > 0)
return results
else
throw graphqlError(404, `Variant with id ${id} does not exist.`)
}
}
}
const executableSchema = makeExecutableSchema({
typeDefs: schema,
resolvers: _.merge(productResolver, variantResolver)
})
module.exports = {
executableSchema
}
// index.js
const { graphqlHandler } = require('graphql-serverless')
const express = require('express')
const app = express()
const { executableSchema } = require('./src/schema')
const graphqlOptions = {
schema: executableSchema,
graphiql: { // If you do not want to host any GraphiQl web interface, leave this property undefined.
endpoint: '/graphiql'
}
}
// Host a GraphQl API on 2 endpoints: '/' and '/graphiql'. '/graphiql' is used to host the GraphiQL web interface.
// If you're not interested in the GraphiQl web interface, leave the above 'graphqlOptions.graphiql' undefined and
// replace the path following ['/', '/graphiql'] with '/'.
app.all(['/', '/graphiql'], graphqlHandler(graphqlOptions))
// Starting the server
app.listen(4000, () => console.log('Server start. Go to http://localhost:4000/graphiql'))
Simply run node index.js
and then browse to http://localhost:4000/graphiql.
PREREQUISITE - Install the following dependencies before running the following code:
npm i express graphql-serverless schemaglue lodash --save
Project Structure Example
- src/
|__ graphql/
|__ product/
| |__ schema.graphql
| |__ resolver.js
|
|__ variant/
|__ schema.graphql
|__ resolver.js
- index.js
- package.json
This is just one example of how to structure the schema and resolver. schemaglue looks for all .js and .graphql files under a specific path.
// src/graphql/product/schema.graphql
type Product {
id: ID!
name: String!
shortDescription: String
}
type Query {
# ### GET products
#
# _Arguments_
# - **id**: Product id (optional)
products(id: Int): [Product]
}
// src/graphql/product/resolver.js
const { graphqlError } = require('graphql-serverless')
const productMocks = [{ id: 1, name: 'Product A', shortDescription: 'First product.' }, { id: 2, name: 'Product B', shortDescription: 'Second product.' }]
exports.resolver = {
Query: {
products(root, { id }, context) {
const results = id ? productMocks.filter(p => p.id == id) : productMocks
if (results.length > 0)
return results
else
throw graphqlError(404, `Product with id ${id} does not exist.`)
}
}
}
// src/graphql/variant/schema.graphql
type Variant {
id: ID!
name: String!
shortDescription: String
}
type Query {
# ### GET variants
#
# _Arguments_
# - **id**: Variant id (optional)
variants(id: Int): [Variant]
}
// src/graphql/variant/resolver.js
const { graphqlError } = require('graphql-serverless')
const variantMocks = [{ id: 1, name: 'Variant A', shortDescription: 'First variant.' }, { id: 2, name: 'Variant B', shortDescription: 'Second variant.' }]
exports.resolver = {
Query: {
variants(root, { id }, context) {
const results = id ? variantMocks.filter(p => p.id == id) : variantMocks
if (results.length > 0)
return results
else
throw graphqlError(404, `Variant with id ${id} does not exist.`)
}
}
}
// index.js
const { graphqlHandler } = require('graphql-serverless')
const express = require('express')
const app = express()
const { makeExecutableSchema } = require('graphql-tools')
const glue = require('schemaglue')
const { schema, resolver } = glue('src/graphql')
const executableSchema = makeExecutableSchema({
typeDefs: schema,
resolvers: resolver
})
const graphqlOptions = {
schema: executableSchema,
graphiql: { // If you do not want to host any GraphiQl web interface, leave this property undefined.
endpoint: '/graphiql'
}
}
// Host a GraphQl API on 2 endpoints: '/' and '/graphiql'. '/graphiql' is used to host the GraphiQL web interface.
// If you're not interested in the GraphiQl web interface, leave the above 'graphqlOptions.graphiql' undefined and
// replace the path following ['/', '/graphiql'] with '/'.
app.all(['/', '/graphiql'], graphqlHandler(graphqlOptions))
// Starting the server
app.listen(4000, () => console.log('Server start. Go to http://localhost:4000/graphiql'))
Simply run node index.js
and then browse to http://localhost:4000/graphiql.
By default, schemaglue uses .js files to define resolvers and .graphql files to define schemas. If your project uses Typescript, it is possible to change the default behavior to target resolvers that use .ts files:
const { schema, resolver } = glue('src/graphql', { mode: 'ts' })
You can go even further by defining your own custom globbing rule if you want to only include certain resolver files:
const { schema, resolver } = glue('src/graphql', { mode: 'product/*.js' })
The above will only get resolvers defined under the src/graphql/product/
folder.
In some cases, you might want to ignore some specific files under the schema folder (by default, SchemaGlue will take into account all .js files). SchemaGlue adds support to ignore files or folders using globbing:
const { schema, resolver } = glue('./src/graphql', { ignore: '**/somefile.js' })
This will take into account all .js files under the folder ./src/graphql, excluding all somefile.js files. The ignore property supports both a single string or an array of strings.
// Ignore more than 1 specific .js file
const { schema:schema1 } = glue('./src/graphql', { ignore: ['**/somefile.js', '**/someotherfile.js'] })
// Ignore all files under the ./src/graphql/variant folder
const { schema:schema1 } = glue('./src/graphql', { ignore: 'variant/**' })
You can specify a custom glob to select your resolver files with different extensions.
const { schema, resolver } = glue('./src/graphql', { js: '**/*.ts' })
If you're not familiar with this concept, check out this great article GraphQL Tour: Interfaces and Unions.
In our case, we're interested is knowing how to structure our code with unions or interfaces. If your schema is small, I would recommend to manage your unions and intefaces definitions inside a single schema.js file per model:
- src/
|__ graphql/
|__ product/
| |__ schema.graphql
| |__ resolver.js
|
|__ variant/
|__ schema.graphql
|__ resolver.js
- index.js
- package.json
Let's take the schema.graphql and resolver.js under src/graphql/product/ as an example:
schema.graphql
union Product = Bicycle | Racket
type Bicycle {
id: ID!
brand: String!
wheels: Int!
}
type Racket {
id: ID!
brand: String!
sportType: SportType!
}
enum SportType {
TENNIS
SQUASH
}
type Query {
# ### GET products
#
# _Arguments_
# - **id**: Product id (optional)
products(id: Int): [Product]
}
resolver.js
const { graphqlError } = require('graphql-serverless')
const productMocks = [{
id: 1,
brand: 'Giant',
wheels: 2
},{
id: 2,
brand: 'Prince',
sportType: 'TENNIS'
}]
exports.resolver = {
Query: {
products(root, { id }, context) {
const results = id ? productMocks.filter(p => p.id == id) : productMocks
if (results.length > 0)
return results
else
throw graphqlError(404, `Product with id ${id} does not exist.`)
}
},
Product: {
__resolveType(obj, context, info) {
return obj.wheels ? 'Bicycle' :
obj.sportType ? 'Racket' : null
}
}
}
Notice you need to define a resolveType method for the Product type under the exports.resolver
const glue = require('schemaglue')
const { schema } = glue('src/graphql', {
schemas:[`
type Product {
id: ID
name: String
}
type Query {
products(id:ID, name:String): [Product]
}`,
`
enum VehicleEnum {
car
bike
bus
}
type Mutation {
createCar: Car
}
`]
})
There are some edge cases where some software engineers may want to express their GraphQL schema using .js
or .ts
files rather than .graphql
files (e.g., Typescript support).
In that case, you will have to transform your files as follow:
product.graphql
example:
type Product {
id: ID!
name: String!
shortDescription: String
}
type ProductNameChangedMsg {
id: ID!
name: String
}
type Query {
products(id: Int): [Product]
}
type Mutation {
productUpdateName(id: Int, name: String): ProductNameChangedMsg
}
type Subscription {
productNameChanged(id: Int): ProductNameChangedMsg
}
product.js
example:
exports.schema = `
type Product {
id: ID!
name: String!
shortDescription: String
}
type ProductNameChangedMsg {
id: ID!
name: String
}
`
exports.query = `
products(id: Int): [Product]
`
exports.mutation = `
productUpdateName(id: Int, name: String): ProductNameChangedMsg
`
exports.subscription = `
productNameChanged(id: Int): ProductNameChangedMsg
`
As explained in the Typescript Support & Custom Globbing section, Schemaglue supports resolvers written in Typescript. However, Typescript is less than friendly with the .graphql
file extension. The main challenge occurs when the Typescript code is compiled (for example into a dist
folder). You'll notice that the .graphql
files are omitted, therefore breaking your code in production. This article explains how to use Webpack to include those .graphql
files: How to resolve import for the .graphql file with typescript and webpack.
However, for those of you who do not wish to mess up with more Webpack configuration, there is a solution with a trade-off. Schemaglue supports GraphQL schema file written in plain JS or TS as explained in the Replacing .graphql
files with plain .js
or .ts
files section. You will loose in readability as your GraphQL schema will have to be expressed as raw Javascript strings. To make this work, update the your code as follow:
Before:
const { schema, resolver } = glue('src/graphql', { mode:'ts' })
After:
const { schema, resolver } = glue('src/graphql', { mode:'**/*.[jt]s' })
This example above helps reading .ts
files and .js
files at the same time. This is required as your .ts
files are compiled to .js
files in your dist
folder.
If you choose that convention, the GraphQL schema will be properly exported to your dist
folder.
We're always super excited to accept pull-requests. We're not too strict about how you implement your fixes or desired features, but we would still highly appreciate if you could:
- Lint your code using the command
npm run eslint
before pushing your pull-request. - Make sure you've added a unit test under the
test
folder (more about that in the following section). - Make sure all the unit tests pass by running
npm test
All unit tests are located under the test
folder. There, you will notice 3 main components:
graphql
folder: This folder contains a series of mock files that you should use to test schemaglue. Feel free to add your own, but please, stick to the already defined naming convention.index.js
: This file is the main unit test file. You will most probably add your unit test there. Please, stick with the naming convention in that file.utils.js
: This file contains the unit tests for the src/utils.js code. Please, stick with the naming convention in that file.
To run the test, simply execute this command: npm test
.
We are Neap, an Australian Technology consultancy powering the startup ecosystem in Sydney. We simply love building Tech and also meeting new people, so don't hesitate to connect with us at https://neap.co.
Our other open-sourced projects:
- webfunc: Write code for serverless similar to Express once, deploy everywhere.
- now-flow: Automate your Zeit Now Deployments.
- graphql-serverless: GraphQL (incl. a GraphiQL interface) middleware for webfunc.
- schemaglue: Naturally breaks down your monolithic graphql schema into bits and pieces and then glue them back together.
- graphql-s2s: Add GraphQL Schema support for type inheritance, generic typing, metadata decoration. Transpile the enriched GraphQL string schema into the standard string schema understood by graphql.js and the Apollo server client.
- graphql-authorize: Authorization middleware for graphql-serverless. Add inline authorization straight into your GraphQl schema to restrict access to certain fields based on your user's rights.
- react-native-game-engine: A lightweight game engine for react native.
- react-native-game-engine-handbook: A React Native app showcasing some examples using react-native-game-engine.
- aws-cloudwatch-logger: Promise based logger for AWS CloudWatch LogStream.
Copyright (c) 2018, Neap Pty Ltd. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
- Neither the name of Neap Pty Ltd nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NEAP PTY LTD BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.