Is there any guidance around using variables within

select

and

include

? Seems like it breaks type inference pretty hard. Cut an issue about it here: https://github.com/prisma/prisma/issues/13571

Hello, I'm trying to create a following/followers system by using a many-to-many relation, everything is fine with the schema side, but on the client I'm getting an error when connecting a new ID to the following value of a user. The error is the following:

Unknown arg `id` in data.following.connect.id for type FollowsWhereUniqueInput. Available args:

type FollowsWhereUniqueInput {
  followerId_followingId?: FollowsFollowerIdFollowingIdCompoundUniqueInput
}

EDIT: I did try to use the followerId_followingId argument in the connect object, but when using it, a different error appears saying it expected type FollowsFollowerIdFollowingIdCompoundUniqueInput but received String.

hi, everyone, I want to use vitest to test mock prisma api. Prisma’s official test case uses jest. How can I use vitest to implement this test? this is prisma unit test page: https://www.prisma.io/docs/guides/testing/unit-testing test uses vitest-mock-extended here my test code, mock not work and no error! import { PrismaClient } from ‘@prisma/client’; import { mockDeep, mockReset } from ‘vitest-mock-extended’; import { type DeepMockProxy } from ‘vitest-mock-extended/lib/Mock’; import prisma from ‘./client’; vi.mock(‘./client’, () => ({ __esModule: true, default: mockDeep(), })); // beforeEach(() => { // mockReset(prismaMock); // }); export const prismaMock = prisma as unknown as DeepMockProxy; it(‘test prisma’, async () => { const user = { id: 1, user_id: 1, created_time: new Date(), updated_time: new Date(), }; prismaMock.payingUser.create.mockResolvedValue(user); const res = prismaMock.payingUser.findMany(); console.log(res); });

Hi People, I need a help This is my schema

generator client {
  provider = "prisma-client-js"
}

datasource db {
  provider = "postgresql"
  url      = env("DATABASE_URL")
}
model Organisation {
  id             Int           @id @default(autoincrement())
  name           String        @unique
  slogan         String
  logoUrl        String?
  organisationId String        @unique
  juniorAdmins   JuniorAdmin[]
}

model JuniorAdmin {
  id           Int           @id @default(autoincrement())
  name         String
  number       BigInt
  photoUrl     String?
  password     String
  email        String
  Organisation Organisation? @relation(fields: [id], references: [id])
}

This is working fine when I add the first entry but when I add more of them I get this error

Why is there such a huge difference between duration of a query logged by prisma and results of explain analyze of them same query. Plan:

Planning Time: 2.072 ms
Execution Time: 186.830 ms

but the query in prisma took took 32441 ms

it can’t be network latency, because although the server and client / app are on different places they are not that far away

Hey Guys! Need help about how to connect GCP Cloud SQL with the unix socket. Here is my

DATABASE_URL

DATABASE_URL="<mysql://MY_USER:MY_PASSWORD@localhost/MY_DB?socket=/cloudsql/inbound-BLAH-BLAH-BLAH>"

But I got error as picture below. Any suggestion?

Hello here. How do you all solve over-fetching with prisma/graphql. I have a graphql

province

query that might include

districts

or not, depending on whether we requested for it. But my

Prisma

query for the resolver always includes it in the result, regardless of whether the user wants it or not. Isn't this overfetching?:

Query.ts

function province(_: any, args: any, context: any) {
  return context.prisma.province.findUnique({
    include: {districts: true},
    where: {
      id: args.id,
    },
  });
}

graphql.schema

type Query {
  province(id: ID!): Province!
}
type Province {
  id: ID!
  code: String!
  name: String!
  country_id: String
  country: Country
  districts: [District!]!
  created_at: String!
  created_by: String!
  last_modified_at: String!
  last_modified_by: String!
}

type District {
  id: ID!
  name: String!
  code: String!
  province_id: String!
  province: Province!
  created_at: String!
  created_by: String!
  last_modified_at: String!
  last_modified_by: String!
}

GraphQL Query that needs districts

query Query($provinceId: ID!){
  province(id: $provinceId) {
    code
    name
    districts {
      code
      name
    }
  }
}

GraphQL Query that does not need districts but still hits the same resolver

(Prisma loads districts that are not needed by the client) How can I solve this?

query Query($provinceId: ID!){
  province(id: $provinceId) {
    code
    name
  }
}

Hey all, does anyone have any good guides as to how we can test prisma endpoints?

HI All, I am looking for a way to extend a MongoDB list with a few items. Currently the prisma-client does not allow this, does anyone know how to do this with the

$runCommand

escape hatch? In Mongo I can write a query that looks like the following:

db.students.update(
   { name: "joe" },
   { $push: { scores: { $each: [ 90, 92, 85 ] } } }
)

and this will give me the following object:

students: {
   name: "joe",
   scores: [100, 99,  90, 92, 85 ]
}

What I Have Tried First of all, iterating through the list items item-by-item is infeasible for our application. The example here is simple, but the list items in our application is larger. So I tried to use

$runCommandRaw

let rawQuery = {
          insert: "Student",
          bypassDocumentValidation: false,
          documents: [
            text
              ? {text, deltas}
              : {scores: {$push: {$each: deltas}}}
          ]
        }

      }
  newText = await context.prisma.$runCommandRaw(rawQuery)

How do I do this with Prisma-js? @Nurul @janpio

Hey -- I was building an integration for Prisma which allowed people to compile multiple Prisma files into a single Prisma file. Is there an integration for this which exists already?

Hey everyone, does anyone have an example of adding a property on a type which is an array of objects? Been stuck on this for a while, i essentially have a computed field which is a nested property.. My code looks like the following.

const posts = await db.post.findMany({
        ...paginateArgs,
        where: {
          galleryId: galleryId,
        },
        include: { images: true },
        orderBy,
      })

posts array has many posts (obviously) and the post essentially has a child array called

images

. I’d like to add a computed field called

url

to this

images array

, and retain type safety on this. I debated adding URL to the DB as a dummy field but that doesnt really sound correct to me. I’m sure im over complicating things, so im curious what the best route to do this is.

I decided to not use computed fields until they are supported by Prisma, for this exact reason...it's a type safety nightmare. I'd recommend adding the

url

field but computing the value beforehand using JS whenever you do an insert or update. You could write a helper function to do that so it's sure to be consistent.

I have a model with an optional field and want to call

findMany()

and return only results where that field has a value. What do I need to do for that?

Hi, In my model I have a field:

trial_ends_on   DateTime?

I want to set the default value of this to be 14 days from today. If I was setting it to now, I would add it as follows:

trial_ends_on   DateTime?   @default(now())

However, instead of now, I actually want to add a date which is 14 days from now. How can I do this with default value?

Hello here. I am trying to figure out if I can use

findUnique

with multiple args and only enable

case insensitive

on some of them. I have tried the below but I am getting the error as :

"Argument code: Got invalid value ",
            "{",
            "  equals: 'ZM',",
            "  mode: 'insensitive'",
            "}",
            "on prisma.findUniqueCountry. Provided Json, expected String.",

And my query looks like:

country: (_parent: unknown, args, context) => {
      return context.prisma.country.findUnique({
        where: {
          id: args.id || undefined,
          code:
            args.code ? {
              equals: args.code,
              mode: "insensitive",
            } : undefined,
          name:
            args.name ? {
              equals: args.name,
              mode: "insensitive",
            } : undefined,
        },
      });
    },

whats a better way to typesafe nested prisma models? e.g

interface ModelC {
  ...
}

Interface ModelB extends ModelC {
  ...
}

Interface ModelA extends ModelB {
  ...
}

Is there a way where we can run Prisma in an atomic transaction state … such that we can either decide to commit or rollback once we have run some code?

Hello everyone. I am struggling to make auto-completion work in the resolvers for the

context

. I am using

Apollo-Server

with

Prisma

and

code-generator

for resolver types. The args are type-safe and I can auto-complete in resolvers but I can't do so with

context

. In the resolver if I set the

context: GraphQLContext

I am getting the below errors but it works without the context type. Here is the error:

Type '(_root: {}, { id, code, name }: Partial<QueryCountryArgs>, context: GraphQLContext) => Prisma__CountryClient<Country | null>' is not assignable to type 'Resolver<Maybe<ResolverTypeWrapper<Country>>, {}, any, Partial<QueryCountryArgs>> | undefined'.

And here is how I am setting up the server and declaring /instantiating the context:

import fs from "fs";
import path from "path";
import { ApolloServer } from "apollo-server-express";
import { ApolloServerPluginDrainHttpServer } from "apollo-server-core";
import { makeExecutableSchema } from "@graphql-tools/schema";
import { WebSocketServer } from "ws";
import { useServer } from "graphql-ws/lib/use/ws";
import { PrismaClient } from "@prisma/client";
import express, { Request } from "express";
import http from "http";
import {
  typeDefs as scalarTypeDefs,
  resolvers as scalarResolvers,
} from "graphql-scalars";
import { GraphQLSchema } from "graphql";
import { resolvers } from "./api/resolvers/resolvers";

const prisma = new PrismaClient();

export type GraphQLContext = {
  req: Request;
  prisma: PrismaClient;
};

export async function createContext(
  req: Request,
  prismaClient: PrismaClient
): Promise<GraphQLContext> {
  return {
    req,
    prisma: prismaClient,
  };
}

const typeDefs = fs.readFileSync(
  path.join(path.resolve(), "src/api/schema.graphql"),
  "utf-8"
);

const schema = makeExecutableSchema({
  typeDefs: [typeDefs, ...scalarTypeDefs],
  resolvers: { ...resolvers, ...scalarResolvers },
});

async function startApolloServer(
  gqlSchema: GraphQLSchema,
  prismaClient: PrismaClient
) {
  const app = express();

  // 1. Http Server
  const httpServer = http.createServer(app);

  // 2. Websocket Server
  const wsServer = new WebSocketServer({
    server: httpServer,
    path: "/",
  });

  const wsServerCleanup = useServer({ schema: gqlSchema }, wsServer);

  // 3. Apollo Server
  const server = new ApolloServer({
    schema,
    context: ({ req }) => {
      return createContext(req, prismaClient);
    },
    csrfPrevention: true,
    plugins: [
      // Proper shutdown for the HTTP server.
      ApolloServerPluginDrainHttpServer({ httpServer }),

      // Proper shutdown for the websocket server
      {
        async serverWillStart() {
          return {
            async drainServer() {
              await wsServerCleanup.dispose();
            },
          };
        },
      },
    ],
  });

  await server.start();
  server.applyMiddleware({
    app,
  });

  //   Modified server startup
  await new Promise<any>((resolve: any) =>
    httpServer.listen({ port: 4000 }, resolve)
  );
  console.log(`🚀 Server ready at <http://localhost:4000>${server.graphqlPath}`);
}

startApolloServer(schema, prisma);

And my resolvers fail when I do this:

country: (_root, { id, code, name }, context: GraphQLContext) => { // fails when this type is applied here
    return context.prisma.country.findUnique({
      where: {
        id: id || undefined,
        code: code || undefined,
        name: name || undefined,
      },
    });
  },

I am using graphql-iso-date for date custom scalar. why is the type of Date custom scalar specified as any type ? can’t i explicitly specify the type ?

hey, how you doing? I've been using TypeORM for a while, I cannot continue using it haha. What's the first resource besides official docs, you would recommend me to get a deep understanding of Prisma? Thx in advanced go gopher dance

Hi Everyone. Currently running a

prisma migrate deploy

on my production database. The command has been running for over 15 minutes, and all it is doing is dropping a unique index. Is this normal? What do i do? It's stuck in the "Applying Migration" state

Hi, Anyone help me out to select a Enum column using findUnique? When I give like below select: { status: true } I am getting ReferenceError: status is not defined Schema referrence status JobStatus @default(WAITING)

return this.prismaService.employeeOrderItem.update(
  The column `(not available)` does not exist in the current database.

why is prisma unable to tell me the column name

Hey, <!channel> - only two weeks left until Prisma Day! ⏳ The final schedule of talks and workshops is ready to view on our site and it is the bomb! 💥 Check out the schedule and get your ticket here. You can join Prisma Day online on the day, or choose to celebrate with us in James June Sommergarten in Berlin. We hope to see you there! 👋

👋 Hello, team!

I have been using Prisma with NodeJs for so long, I was so happy when I saw the Prisma client module for Golang and then the deprecation notice hit me so hard😭 Hope you guys will go back to maintaining it, you rule when it comes to ORMs in my very humble opinion

hi all