Home

Awesome

@garph/gqty

tRPC-style client for Garph, based on GQty

npm i @garph/gqty

Example:

schema.ts

import { g, buildSchema } from 'garph'

export const queryType = g.type('Query', {
  greet: g.string()
    .args({
      name: g.string().optional().default('Max'),
    })
    .description('Greets a person')
})

const schema = buildSchema({ g })

client.ts

import { InferClient, createClient } from '@garph/gqty'
import { createScalarsEnumsHash, createGeneratedSchema } from '@garph/gqty/dist/utils'
import { schema, queryType } from './schema'

type ClientTypes = InferClient<{ query: typeof queryType }>

export const { useQuery, ... } = createClient<ClientTypes>({
  generatedSchema: createGeneratedSchema(schema),
  scalarsEnumsHash: createScalarsEnumsHash(schema),
  url: 'http://localhost:4000/graphql'
})

// Needed for the babel plugin
export { schema as compiledSchema }

Using the client (React):

query.tsx

import { useQuery } from './client'

export default function Example() {
  const query = useQuery()
  return <p>{ query.greet({ name: 'Mish' }) }</p>
}

Subscriptions

With Server-Sent-Events

npm i graphql-sse
import { createClient as createSubscriptionsClient } from 'graphql-sse'

export const { useSubscription, ... } = createClient<ClientTypes>({
  generatedSchema: createGeneratedSchema(schema),
  scalarsEnumsHash: createScalarsEnumsHash(schema),
  url: 'http://localhost:4000/graphql',
  subscriptionClient: createSubscriptionsClient({
    url: process.env.NODE_ENV === 'production' ? `https://${process.env.NEXT_PUBLIC_VERCEL_URL}/api/graphql/stream` : 'http://localhost:3000/api/graphql/stream'
  })
})

Using the Babel plugin (alpha)

In production, you might want to use the babel plugin in order to replace the runtime dependencies (such as generatedSchema, scalarsEnumsHash) in your client config with statically-generated artefacts.

.babelrc

{
  "plugins": [["@garph/gqty/dist/plugin", {
    "clientConfig": "./utils/client.ts"
  }]]
}

Where clientConfig is the path to the file where you call createClient

Special thanks to Vicary of GQty project for early feedback and helping to make @garph/gqty possible