npm

Ready to take your JavaScript development to the next level? Meet npm Enterprise - the ultimate in enterprise JavaScript. Learn more »

schemaglue

4.0.4 • Public • Published

SchemaGlue · NPM Tests License Neap

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.

What It Does

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

Install

npm install schemaglue --save

How To use It

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.

In Short

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)

Without SchemaGlue

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 { app } = require('webfunc')
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 
eval(app.listen('app', 4000))

Simply run node index.js and then browse to http://localhost:4000/graphiql.

With SchemaGlue

PREREQUISITE - Install the following dependencies before running the following code:

npm install webfunc 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 { app } = require('webfunc')
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: { 
    endpoint: '/graphiql' 
  }
}
 
app.all(['/', '/graphiql'], graphqlHandler(graphqlOptions))
 
 
eval(app.listen('app', 4000))

Simply run node index.js and then browse to http://localhost:4000/graphiql.

Typescript Support & Custom Globbing

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.

Ignoring Certain Files

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/**' })

Custom Extensions

You can specify a custom glob to select your resolver files with different extensions.

const { schema, resolver } = glue('./src/graphql', { js: '**/*.ts' })

Interesting Examples

Unions & Interfaces

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

Contribute

General Guidance

We're always super excited to accept pull-requests. We're not too nazi on how you implement your fixes or desired features, but we would still highly appreciate if you could:

  1. Lint your code using the command npm run eslint before pushing your pull-request.
  2. Make sure you've added a unit test under the test folder (more about that in the following section).
  3. Make sure all the unit tests pass by running npm test

THANKS A LOT YOU AWESOME GOD COMMITTER!

Unit Testing

All unit tests are located under the test folder. There, you will notice 3 main components:

  1. 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.
  2. 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.
  3. 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.

This Is What We re Up To

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:

Web Framework & Deployment Tools

  • webfunc: Write code for serverless similar to Express once, deploy everywhere.
  • now-flow: Automate your Zeit Now Deployments.

GraphQL

  • 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 & React Native

Tools

License

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.

Neap Pty Ltd logo

install

npm i schemaglue

Downloadsweekly downloads

1,604

version

4.0.4

license

BSD-3-Clause

homepage

github.com

repository

Gitgithub

last publish

collaborators

  • avatar
Report a vulnerability