graphql-query-gen
v1.3.2
Published
Node.js module to generate queries with random input data from graphQL endpoint or schema
Downloads
6,122
Maintainers
Readme
graphql-query-gen
Node.js module to generate queries from graphQL endpoint or from schema text.
It reads graphql SDL using introspection query and then generates queries/mutations with random input data. Additionally lists down inputs, types as well. Also, there you get schema string and some statistics data as well.
Installation
npm install graphql-query-gen
Live Demo
Graphql Query Generator is a chrome extension built using this. You can install and try it to see what this node module can do for you.
Usage
const qGen = require('graphql-query-gen');
const options = {};
// Work with endpoint
qGen.processEndpoint("endpoint-url", options).then(result => console.log(result));
// work with schema
const s = `
type Character {
id: ID
name: String
}
type Jedi {
id: ID
side: String
}
type Droid {
id: ID
model: Model
}
type Model {
key: String
}
input TestInput {
key1: String
key2: Int
}
union People = Character | Jedi | Droid
type Query {
allPeople(input: TestInput, input2: String): [People]
}
`;
const result = graphqlQueryGen.processSchema(s, options);
// work with SDL
const fs = require('fs');
let rawdata = fs.readFileSync('sdl.json');
const sdl = JSON.parse(rawdata);
const result = graphqlQueryGen.processSDL(sdl.data, options);
// or you can do like below to handle errors as well
const qGen = require('graphql-query-gen');
try {
qGen.generateQueries(
"http://graphql-endpoint-url/",
{
depth: 5,
indentBy: 2
}
).then(
result => console.log(result)
// or do what you want to do with it
);
} catch (err) {
console.log(err.message);
}
Options
{
debug: false, // Boolean [Default is false] -> Would print additional log message to help in debugging if true
filter: null, // String [Default is null ] -> You can give a query or mutation name or part of it
responseDepth: 7, // Number [Default is 5] -> For query/mutation result the nesting level of fields
inputDepth: 8, // Number [Default is 7] -> For query/muation input the nesting level of fields
spacer: ' ', // String [Default is ''] -> To indent query/mutation the space character (e.g. to print on HTML page you can use )
indentBy: 2, // Number [Default is 4] -> The number of spacer to use for indentation.
inputVariables: true, // Boolean [Default is false] -> In generated query input would be in form or variable if true, else inline input.
duplicatePercentage: 75 // Number [Default is 75] -> for types and input check and list for duplicates based on this threshold value.
}
Output
Sample output looks like following
{
operations: [
{ name: 'Query', options: [Array] },
{ name: 'Mutation', options: [] },
{ name: 'Subscription', options: [] }
],
types: [
{ name: 'Character', definition: '{\n id: ID\n name: String\n}' },
{ name: 'Jedi', definition: '{\n id: ID\n side: String\n}' },
{ name: 'Droid', definition: '{\n id: ID\n model: Model\n}' },
{ name: 'Model', definition: '{\n key: String\n}' }
],
inputs: [
{
name: 'TestInput',
definition: '{\n key1: String\n key2: Int\n}'
}
],
statistics: {
counts: { queries: 1, mutations: 0, subscriptions: 0, types: 4, inputs: 1 },
suggestions: { duplicates: [Array] }
},
schema: 'SCHEMA_AS_STRING'
}
TODO
- Subscriptions are not processed and will resturn as empty as of now, planned for future release
- Fragments support is limited for now, plan to enhance in future release
- Better logging, current version has console.log and console.debug only