Hi all -- I'm trying to write my own JWTs for Prisma auth. The docs say where they get the encoding secret from, but they don't say what the actual encoded body is. Does anybody know?

Just in case anyone else comes across this, I dug through the source and found that the body just describes the expiration timing: https://github.com/prisma/prisma/blob/master/server/libs/auth/src/main/scala/com/prisma/auth/Auth.scala#L46 It's a JSON w/ keys

exp

and

nbf

, which correspond to the expiration time in seconds and the current time ("not before") in seconds.

how to avoid inserting duplicates based on an existing field match

for example, I need to bulk insert a list of items where the item.title may already exist, I want to omit inserting an item if that title is found

Can I host a prisma backed app simpler than this? App has different subdomains for each customer, the same app server responds to them all but switches out the

prisma-client

(yoga server) that

apollo-client

connects to depending on the subdomain, each yoga server is a separate running instantiation of the same

yoga

codebase but points to a different

prisma-service

on the one

prisma-server

(is there a way to have one yoga server point to different endpoints like graphcool used to do via

/api/serviceId

?)… The

public

app engine service uses

nginx

and

s6

to manage running the front end app and many yoga servers on the same app engine… This diagram doesn’t even include the google cloud functions used for subscriptions or the google cloud storage used for the File API - what is the simpler way? I’m a front end developer, hence why I chose graph.cool originally… I’m totally open to suggestions for a proper, more scalable set up… most of the tutorials tell you how to get your prisma and yoga servers up but then leave it at that and don’t explain dealing with multiple endpoints, multiple apps, etc..

hi! Anyone familiar with apollo mailgun integration? Would like to handle auth (email/pass, email verify, social logins). Any resources I can leverage that I can be directed to?

hello! is there a way to configure the apollo graphql playground such that all of its requests are GET requests with the query/variables serialized into the URL query params?

(it appears to default to POST requests with a request body)

is there a way to connect two required fields in a nested mutation, for example,

Role

requires a list of

[User!]!

and a single `Customer`… When I signup a user I want to create the new

Customer

, it’s first

User

and an admin

Role

for that

Customer

and

User

- how would I do this?

const user = await ctx.db.createUser({
      firstname: firstname || undefined,
      lastname: lastname || undefined,
      email,
      password: hashedPassword,
      emailConfirmToken,
      emailConfirmed: false,
      inviteAccepted: true,
      customer: {
        create: {
          name: customerName
        }
      },
      role: {
        create: {
          name: "Admin",
          permissions: {
            connect: {
              allow: "SUDO"
            }
          }
        }
      }
    });

this creates the role on the user, but I also want the

Role

to be associated to the

Customer

created above it…? Do I have to create the

Role

in a separate request and then just connect it in the

customer

and

role

fields?

Hi everyone. If I want to perform raw SQL query via prisma client, should I use

prisma.request

(or

$request

) and

mutation {
  executeRaw(
    query: "..."
  )
}

?

Is there any Chance to get access to Prisma Admin before it's official launch 🤓🤓?

However I cant find the mentioned

request

or

$request

exposed in prisma client. Any clue?

I found out that the Graphql prisma access changes to

prisma.$graphql

, yet there are some problems in reading query string.

Updates: the following seems to be working

await prisma.$graphql(`
    mutation {
        executeRaw(query: "SELECT * FROM default$default.\\"User\\" LIMIT 10")
    }
`)

I must not add

/* GraphQL */

in front of the query string, otherwise prettier will eliminate the double backslashes, because the prisma server needs a single backslash to identify quotes inside

executeRaw

Thats kinda troublesome when you need to pass a string with escape characters through two layers of server 🤣

AND, after all the searchings on google, I found out that the effective way to perform case insensitive search is to create a lower-case slug when creating an object. Then I could easily use prisma-client’s

contains

to do what I want! Just for you guys’ reference if anyone is trying something like me.

is it possible to update the datamodel as a hot deployment?

Hi everyone, I’m a beginner on graphcool and i would like to know if there is a way to use a nodejs dependency which required nodejs >= 8 ?

Hello everyone. I saw that some of you have faced such issue

Variable '$data' cannot be non input type

but did someone find out what's going on and how to fix it? I see the error when running the following mutation:

return ctx.prisma.updateManyProductVariants({
      where: { product: { id: productId } },
      data: { deletedAt: new Date().toISOString() }
    });

It's working fine in playground e.g.

mutation {
  updateManyProductVariants(where:{
    product:{
      id:"cjolpj9ez01zq0a926dvb0v2t"
    }
  }, data: {
    deletedAt: "2018-11-17T17:20:13.998Z",
  }) {
    count
  }
}

Hello, new to the channel and Prisma - wondering how to do filtering with Enums. (I don't see a lot of documentation on using Enums for roles and permissions, so if there's a better way, I'd love to hear about it). Basically I'm trying to write a query that pulls all users that have two different enum statuses. My schema:

type Query {
  candidates(role: Role!, accountStatus: AccountStatus!): [User]!
}

type User {
  id: ID!
  firstName: String!
  lastName: String!
  roles: [Role!]!
  accountStatus: AccountStatus!
}

My resolver:

candidates(parent, args, ctx, info) {
    return ctx.db.query.users({
      where: {
        role: args.role,
        accountStatus: args.accountStatus,
      },
    }, info);
  },

My query:

query CANDIDATES_QUERY {
  candidates(role: FULL_MEMBER, accountStatus: ACTIVE) {
    id
    firstName
    lastName
  }
}

Hello. When using

graphqlgen

with prisma do we still run

prisma generate

?

Seems like Prisma generated is incompatible with graphqlgen generated...

Type 'import(".../src/generated/graphqlgen").MutationResolvers.UserUpdateInput' is not assignable to 
Type 'import(".../src/generated/prisma-client/index").UserUpdateInput'.
Types of property 'name' are incompatible.
Type 'string | null' is not assignable to type 'string | undefined'.
Type 'null' is not assignable to type 'string | undefined'. [2322]

Any ideas?

This is causing the above error. I think I'm doing something wrong but cant pin-point it.

any help is greatly appreciated

Hola a todos! Hoy les traigo una herramienta la cual permite crear una noticia falsa que se hace pasar como una aplicación de Facebook y capturar los datos ingresados por el usuario. Esta tiene como intención generar consciencia de lo fácil que es engañar mediante el uso de un tipo de ingeniería social, caracterizado por intentar adquirir información confidencial de forma fraudulenta (como puede ser una contraseña, información detallada sobre tarjetas de crédito u otra información bancaria). El cibercriminal, conocido como phisher, se hace pasar por una persona o empresa de confianza en una aparente comunicación oficial electrónica, por lo común un correo electrónico, o algún sistema de mensajería instantánea o incluso utilizando también llamadas telefónicas. Proyecto: https://github.com/MontoyaAndres/FacebookNews Vídeo:

Say I have a User model mapped to an actual DB table. But I want "extend" that user by pulling some extra meta data from an external API. Are there any examples of this online to look at? I tried adding my custom/extra fields to my graphql.schema file and got some resolver functions auto generated to process those fields, but not sure how to write the code in an optimized way so that I am not hitting the API for each and every field. Is there a better pattern to do this?

how to connect existing database from mysql

Hi, does anyone have idea whether

executeRaw

mutations support transactions?

Hello Guys recently I encounter some bug which is effect my project. I have initialize prisma@1.20.7. Which is latest version. After i update my datamodel and deploy with prisma deploy It does not update prisma.graphql with my latest datamodel. Which is very disappointing. But When ever I use prisma@1.14.2 It working perfect. Please go through my concern and fix this.

is anyone really using this prisma setup(https://www.prisma.io/docs/1.20/get-started/01-setting-up-prisma-existing-database-TYPESCRIPT-t003/) … its full of bugs. and is incomplete documentation.

TSError: ⨯ Unable to compile TypeScript:
generated/prisma-client/index.ts(348,72): error TS2345: Argument of type '{ typeDefs: string; endpoint: string; }' is not assignable to parameter of type '{ typeDefs: string; endpoint: string; secret?: string; models: Model[]; }'.
  Property 'models' is missing in type '{ typeDefs: string; endpoint: string; }'.