Does anyone have an idea how to debug this?

prisma generate
Generating schema 27ms
 ▸    Syntax Error: Expected Name, found  }

How can we handle separate files for resolvers and schema? I would like to have one file per concern. Eg User.graphql and UserResolver.js Then Comment.graphql and CommentResolver.js

At prisma team: is there sources for https://www.npmjs.com/package/prisma-generate-schema ?

Hi, can you do transactions on a MySQL db when using Prisma?

not sure about the etiquette here, but is it ok to ask for a problem i am running into. Its looking like a bug, but not entirely sure, if its some error on my part.

I have a mutation

async signup(parent, args, ctx, info) {
    // lowercase their email
    args.email = args.email.toLowerCase();
    // hash their password
    const password = await bcrypt.hash(args.password, 10);
    const user = await ctx.db.mutation.createUser(
      {
        data: {
          ...args,
          password
        }
      },
      info
    );
    return {
      user, // is always undefined
      token: utils.getToken(user)
    };
  }

this doesn't work if my schema is like this.

type SignupResponse {
  token: String!
  user: User!
}

type Mutation {
  signup(name: String!, email: String!, password: String!): SignupResponse!
}

However it works if i stop returning custom return type and start returning

User

What am i missing

Hey all, I'm running in to the strangest issue. I have a very simple Yoga back end that I've set up to play around with some graphQL concepts and it's working great locally. However when I deploy to now.sh I'm seeing an issue where upon creating an entry it's actually creating two entries in the database each with a unique id. I've tested this on the playground of my deployed server to rule out the front end being an issue and I'm seeing the same thing. Running:

mutation  {
  createListItem(title:"this is a test of a new item") {
    id
    title
  }
}

Returns one item. But then, when running:

query {
  listItems(first:4 skip:0){
    id
    title
  }
}

I see two items returned with the same title and different ID's. My mutation is about as simple as it gets

createListItem(parent, args, context, info) {
      return context.prisma.createListItem({ title: filter.clean(args.title) })
    }

so I'm just not sure what could be going wrong here. Again, it's working fine locally and the only real difference between the two is the NODE_ENV. Both the local and deployed versions are connecting to the same Prisma demo database. If I go directly to the playground of the prisma database I don't get the same issue. Any help figuring out where to start would be appreciated beyond belief

Hi guys I have 1 question, does I need graphql-yoga to bind with prisma to write custom resolvers for my Queries and Mutations ?

Hi guys. I consider running graphql-yoga and prisma on the same ec2 instance using docker-compose. The db on aws rdb. Is that a good idea? Currently I do not need much scaling and I find docker-compose simpler than e.g. yoga on lambda and prisma on fargate cluster. What are your thoughts?

Hi guys! really need help with caching. I trying use graphql-yoga and just setup cacheControl but... that look like not work... or maybe i'm not understand some

it's possible to use memcache like store in store option? that will work?

my config

server.start({
      debug: true,
      cacheControl: {
        defaultMaxAge: 5
      },
      port: process.env.PORT,
      playground: process.env.PLAYGROUND
    },
    () => log(`🚀  Server is running on localhost:${process.env.PORT}`)

but when i send request to server i see what no caching in this moment

in my schema i have some think like

type Address @cacheControl(maxAge: 86400) {...fields...}

but anyway request will directed to db

Simple question. In javascript I have a resolver like

const resolvers = {
    Query: {      
        laws(parent, args, context, info) {            
            return context.prisma.query.laws(
                { where: args.where, skip: args.skip, first: args.first, orderBy: args.orderBy },
                info,
            )
        },        
    },
}

, but i struggle to get it working in typescript :

export const Query: QueryResolvers.Type = {
  ...QueryResolvers.defaultResolvers,
 
  laws: (parent, args, ctx) => ctx.db.laws({where: args.where})

}

throws the error

Property 'where' does not exist on type '{}'.

I have then tried to do something like

interface LawsArgs{
  where?: lawsArgs.LawWhereInput;
  orderBy?: lawsArgs.LawOrderByInput;
  skip?: <http://lawsArgs.Int|lawsArgs.Int>;
  after?: String;
  before?: String;
  first?: <http://lawsArgs.Int|lawsArgs.Int>;
  last?: <http://lawsArgs.Int|lawsArgs.Int>;
}

export const Query: QueryResolvers.Type = {
  ...QueryResolvers.defaultResolvers,
  laws: (parent, args:LawsArgs, ctx) => ctx.db.laws({where: args.where})

}

but still no luck.. (I am a typescript newbie :-()

Hi all, Im having this weird issue where I'm missing a role (whose type is an enum) field from my user object in the datamodel.

type User {
  id: ID! @unique
  email: String @unique
  firstName: String!
  middleName: String
  lastName: String!
  password: String!
  role: Role!
  orders: [Order!]!
}

enum Role { 
  Admin
  Owner
  Physician
  Patient
  Provider
  Disabled
}

when I

createUser()

I've manually added the Role, and I have tried to also put it as a

@default

directive, however, when I call back or fetch on user the Role value is not there thus interfering with my permissions etc. Would any one able to shed some light? I am using prisma 1.17

Hi Prisma team, I managed to make a simple POC https://github.com/Sach97/prisma-go-subscription-experiments with a subscription query against the prisma ws endpoint with the same datamodel from example repo. My subscribe only takes a query string in argument for the moment and return a go channel. It is a toy program but I hope it can help the team. I'll try to play with the ast of the query now 🙂. Design question, once I have something nice, should I fork prisma/prisma-client-lib-go with an interface let's say like this

type Client struct {
	Endpoint string
	Secret   string
    Session *Session
	GQLClient *graphql.Client
}

type Session struct {
	ws      *websocket.Conn
	errChan chan error
}

Or fork both prisma-client-lib for the channel part and machinebox/graphql for the websocket connection part ?

Will graphqlgen have a more formal announcement and will there be more complete docs?

Question: I know it isn't official, but is there any way to use prisma for an exisiting MongoDB setup?

Hey guys, I’m a UI designer and developer which recently came across the need to display data on an interface using graphQL, while I know little about data handling I came across graphql-request and was able to console.log data

The question is now that I don’t know how to grab the data to display across the interface, if anyone could point me some resources on a way to do it, that would appreciated

the log

I’m having an issue with using prisma deploy with the graphql-cli. Here’s what I’m doing: -Set up new graphql app via graphql-cli, run prisma init to add a heroku hosted prisma server. When I run prisma deploy, I get a “mutiple graphql version error”

Hi guys, is it possible that I setup Prima on top the rest API only without changing exiting code base?

hi all, i am trying to figure out the simplest structure for data entry for my application. I am building past setlists for concerts, so my models look like this:

type Song @model {
  id: ID! @isUnique
  name: String!
}

type ConcertSong @model {
  id: ID! @isUnique
  song: Song @relation(name: "Songs")
  concerts: [ConcertSong!]! @relation(name: "ConcertSongs")
}

type Concert @model {
  id: ID! @isUnique
  venue: Venue! @relation(name: "ConcertVenue")
  start: DateTime!
  setlist: [ConcertSong!]! @relation(name: "ConcertSongs")
  opener: Artist 
}

data entry will be adding a bunch of songs, then a bunch of concerts, then a bunch of ConcertSong's... is there a way to add ConcertSongs without having to find the id (something more human readable). also, is there a better way to do a bunch of data entry like this than mutations? there will likely be hundreds of concerts and songs. Thanks!

Does anybody have issues with nested mutations when updating many entities at once? Like using

connect

inside

updateManyPosts

. It returns the correct “count” of modified entities but nothing has changed in the DB 🤔 Edit: seems that it’s already reported https://github.com/prisma/prisma/issues/2010

Will prisma-binding be updated to graphql 14? I'm getting warnings because now graphql-yoga installs graphql 14 and other dependencies have graphql "^0.13.0"

Hi there.

While querying with prisma client is great, I want to implement search function with it.