graphql-community / graphql-directive-uid

Generate uid for GraphQL schema

graphql-directive-uid

Unique id directive for GraphQL schema.

The problem

Generating unique id for each graphql schema type may be a daunting task.

This solution

The @uid is a GraphQL directive allowing users to generate unique IDs for their queries. All benefits of GraphQL cache with minimum effort.

Table of Contents

• Installation
• Usage
• More on data-testids
• Examples
• LICENSE

Installation

This module is distributed via npm which is bundled with node and should be installed as one of your project's devDependencies:

use npm:

npm install --save graphql-directive-uid

or yarn :

yarn add graphql-directive-uid

This library has a peerDependencies listing for:

• graphql
• graphql-tools

Usage

import { makeExecutableSchema } from 'graphql-tools';
import graphqlDirectiveUid from 'graphql-directive-uid';
import typeDefs from './schema.graphql';
import resolvers from './resolvers';

export default makeExecutableSchema({
  typeDefs,
  resolvers,
  schemaDirectives: {
    uid: graphqlDirectiveUid, // the name of directive can be as you like
  },
});

GraphQL schema:

type Person @uid(from: ["personId", "name"]) {
  personId: Int
  name: String
}

type Query {
  persons: [Person]
}

GraphQL query:

query getPersons {
  persons {
    uid
    personId
    name
  }
}

Examples

You'll find examples of how to use it in the test directory.

Feel free to contribute more!

LICENSE

MIT

graphql-directive-uid's Issues

Error with graphiql schema

Error: Cannot return null for non-nullable field __Field.isDeprecated.

Add getDirectiveDeclaration

Names for all graphql directive

Add all necessary files related with contributing notes

Question: Ability to retrieve a UID in a mutation resolver

I'm trying to use the uid as a key to mutate data in my resolver, since the uid only lives in the directive and is output via a graphql query i'm having a hard time designing the best way to link the uid back to the data outside of the graph. ie.

const schema = gql`
directive @uid(from: [String]) on OBJECT

type Element @uid(from: ["email", "name"]) {
  name: String!
  email: String!
}

type Query {
  elements: [Element]
}

mutation DeleteElement($uid:String) {
  deleteElement(uid:$uid) : Boolean
}`;

const elements = [{name:"bill", email: "[email protected]"}]
const resolvers = {
  Query: {
    elements: () => elements
  },
  Mutations: {
    deleteElement: (parent, {uid}) => {
      // how am I supposed to find the index in elements without redoing the hash work you do in the resolve function of UniqueIdDirective
   
    }
  }
}

I understand I could redo the work that is done here UniqueId.visitObject()fields[].resolve
but this seems like a hack.

Should this uid be used for anything but caching? Is there another way to resolve the uid outside of the graph? Has anyone done something else to resolve this problem?

Readme.md is out of date

TOC links point to wrong or missing section headers

Setup README

Integrate circleci

Basically, runeslint and test.

What with auto release from master branch?

Publish to npm

@czystyl can you take care of that? 🙂

Add tests

We need to figure out (or find, haven't looked yet!) some nice abstractions on top of which we can write tests for other directives too.

package.json "main" property points to missing file

"main" should point to index and node.js engines should be set in package.json to avoid babel requirement.