👋 i'm trying to find a solution to a problem that i can't possibly be the first person to tackle but i can't really find any working examples of solutions and nothing great is coming to mind

we use prisma + apollo server and i'm trying to implement https://relay.dev/graphql/objectidentification.htm on a relatively mature schema

we don't really wrap any prisma objects in our own models and it's hard to see how i can possibly implement

__resolveType()

on a Prisma object

possible solutions i've thought about are wrapping prisma objects in something introspectable or changing our database IDs to contain the type so we can infer the type from id

neither really feels great tbh though

what i really want is just some properly on the objects prisma returns that can tell me what type of object it is (since reflection in typescript is so bad)

i'd also accept some built in support for annotating the types prisma generates using gross typescript reflection support but that doesn't seem to really be possible

Hello people

i have a question ! how i can hide password when i do this query ? please.

const prisma = new PrismaClient()
  const posts  = await prisma.post.findMany({
    include : {
      user : true
    },
    orderBy: {
      created_at: 'desc',
    },
    

  })

Hi friends is an upsertMany possible? right now i can onlything i would have to do multiple upserts using transaction?

is there any way to tell prisma where the schema.prisma file is? trying to do more deterministic bundling for lambda functions and it's not easy to put it in the same dir as the script (related: it would be really cool if prisma didn't use

__dirname

to find it because it doesn't exist in ESM)

I am building a serverless function using the serverless framework. However im having an issue with running it locally prisma/schema.prisma

generator client {
    provider      = "prisma-client-js"
    binaryTargets = ["native", "rhel-openssl-1.0.x"]
}

serverless.ts

package: {
        individually: true,
        patterns: [
            "!node_modules/.prisma/client/libquery_engine-*",
            "node_modules/.prisma/client/libquery_engine-rhel-*",
            "!node_modules/prisma/libquery_engine-*",
            "!node_modules/@prisma/engines/**",
        ],
    },

npx prisma generate && npm install

sls invoke local -f main

✖️ Error: ENOENT: no such file or directory, open ''/.esbuild/.build/node_modules/.prisma/client/schema.prisma'

What am i doing wrong here? attempting to follow this example: https://github.com/prisma/prisma-examples/tree/latest/deployment-platforms/aws-lambda • note: i am using the serverless

aws-nodejs-typescirpt

template which uses serverless-esbuild and not serverless-webpack

Question about middleware, I want to write another `Event`record into the db after a create query has succeeded, is the right place for this logic 🤔

Hello, should I have one export of PrismaClient() from index.ts or create a new one in each file where I want to use prisma?

I'm unable to nullify a many-to-one and update a one-to-one at the same time. Can someone please help? `Error: Unknown arg

team_id

in data.team_id for type PersonUpdateInput. Did you mean

team

?`

model Person {
  id         Int    @id
  first_name String
  last_name  String
  team_id    Int?

  personData PersonData?
  team       Team?       @relation(fields: [team_id], references: [id])
}

model PersonData {
  person_id      Int    @id
  favorite_color String

  person Person @relation(fields: [person_id], references: [id])
}

model Team {
  id   Int    @id
  name String

  people Person[]
}

await prisma.person.update({
  where: {
    id: 1,
  },
  data: {
    team_id: null,
    personData: {
      update: {
        favorite_color: 'Purple',
      },
    },
  },
});

hello, I upgraded prisma client from version 3.14 to version 4.0.0 (latest)

PrismaClientKnownRequestError:
[0] Invalid `prisma.product.findMany()` invocation in
[0] /app/server/data/index.js:148:37
[0]
[0]   145 }
[0]   146
[0]   147 const getTopProducts = async () => {
[0] → 148   const result = await prisma.product.findMany(
[0]   Inconsistent column data: Conversion failed: Value 2847246203 does not fit in an INT column, try migrating the 'revenue' column type to BIGINT
[0]     at RequestHandler.handleRequestError (/server/node_modules/@prisma/client/runtime/index.js:49670:13)
[0]     at RequestHandler.request (/server/node_modules/@prisma/client/runtime/index.js:49652:12)
[0]     at async PrismaClient._request (/server/node_modules/@prisma/client/runtime/index.js:50572:18)
[0]     at async getTopProducts (/server/data/index.js:148:18)
[0]     at async /server/routes/index.js:142:12 {
[0]   code: 'P2023',
[0]   clientVersion: '4.0.0',
[0]   meta: {
[0]     message: "Conversion failed: Value 2847246203 does not fit in an INT column, try migrating the 'revenue' column type to BIGINT"
[0]   }
[0] }

always used the Prisma SQLite connector, didn't have this problem before upgrading to version 4.0.0 here is the model

datasource db {
  provider = "sqlite"
  url      = env("DATABASE_URL")
}

generator client {
  provider = "prisma-client-js"
  previewFeatures = []
}

model Product {
...
revenue				Int
...
}

Hello, I am new to Prisma, and I have a Postgres RAW query which is nested Inner Queries, I wanted to know that how I can write that using Prisma? below is my RAW Query..

SELECT * FROM tableName WHERE id NOT IN (SELECT columnName FROM tableName WHERE (startDateInputFromUser between startDate and endDate) or (endDateInputFromUser between startDate and endDate))

here in WHERE clauses the

startDateInputFromUser

&

endDateInputFromUser

it's from the User Input & other

startDate

&

endDate

is the Table Column Name... can anyone help me how I can write this query in the Prisma ORM?

Hi everyone. I recently created a simple REST API (using Express, Prisma and PostgreSQL) which works fine. Am now currently trying to write some unit tests but am wondering how I can create a test environment with it's own test database. Is it possible to programmatically change the DATABASE_URL in the

schema.prisma

file? For example: Given DEVELOPMENT, TEST, PRODUCTION environments, I would maybe create this kind of logic;

const { DEVELOPMENT, PRODUCTION, LOCAL } = require("./envTypes");
let DB_URL;
switch (process.env.NODE_ENV) 
{  case DEVELOPMENT:    
     DB_URL = process.env.DB_URL_DEV;    
      break;  
   case PRODUCTION:    
     DB_URL = process.env.DB_URL_PROD;    
     break;   
   case LOCAL:    
     DB_URL = process.env.DB_URL_LOC;    
     break;  
   default:    
      DB_URL = process.env.DB_URL;
}

Are there any docs on how to properly setup OpenTelemetry instrumentation in

@prisma/client

? I enabled the

previewFeatures

flag and see

request

spans, but they dont include any attributes

I have a mock created like the following:

import { PrismaClient } from "@prisma/client";
import { mockDeep, mockReset, DeepMockProxy } from "jest-mock-extended";
import { PrismaService } from "@services/prisma.service";

jest.mock("@services/prisma.service", () => ({
    __esModule: true,
    PrismaService: mockDeep<PrismaClient>(),
}));

beforeEach(() => {
    mockReset(PrismaMock);
});

export const PrismaMock = PrismaService as unknown as DeepMockProxy<PrismaClient>;

and then in my test i am using it... However its still connecting to the live (local) db.

PrismaMock.nft_collection.create.mockResolvedValueOnce({});

const [response, error] = await CreateCollection(payload);

Not sure whats going on, why Prisma is not being mocked?

Has anyone had a chance to try the new Prisma Metrics API in 3.15.0 yet? • You can watch a demo video here: • Metrics documentation is here: https://www.prisma.io/docs/concepts/components/prisma-client/metrics Please let me know if you give it a go and what you think of it!

Hello, I’m having trouble modeling a pair in my schema. e.g. I’m trying to model

Pair

being a unique pair of

Tokens

— so far my schema is:

model Token {
  id      Int    @id @default(autoincrement())
  chainId Int
  address String
  name    String

  @@unique([chainId, name])
}

model Pair {
  id                  Int                   @id @default(autoincrement())
  chainId             Int
  name                String
  token0Id            Int
  token1Id            Int

  @@unique([chainId, name])
}

Any hints on how to do this?

It feels like it should be 1:many? Or is it many:many?

What is the best way to handle nested one to many self relations? I tried:

model Item {
  id             String   @id @default(uuid())
  relatedItemsId String[]
  relatedItems   Item[]   @relation("RelatedItems", fields: [relatedItemsId], references: [id])
}

But it throwing and error in my schema file.

So, I'm having issues with Cloudflare Pages: [pages:err] ReferenceError: DATABASE_URL is not defined. Attempted to access binding using global in modules. I read a bit about 4.2.0 making it possible to override the db url through the PrismaClient constructor (see https://github.com/prisma/prisma/issues/13771). Is this possible yet?

am i doing something wrong here? i randomly get this error in my tests

[test:*backend] Invalid `prisma.$queryRaw()` invocation:
[test:*backend] 
[test:*backend] 
[test:*backend]   Raw query failed. Code: `22P03`. Message: `db error: ERROR: incorrect binary data format in bind parameter 1`
[test:*backend]  ❯ RequestHandler.handleRequestError node_modules/@prisma/client/runtime/index.js:49670:13
[test:*backend]  ❯ RequestHandler.request node_modules/@prisma/client/runtime/index.js:49652:12
[test:*backend]  ❯ Proxy._request node_modules/@prisma/client/runtime/index.js:50572:18
[test:*backend]  ❯ EmailCandidateSentRepository.getUnsentEmails packages/service/src/repo/repository/emailCandidateSent.ts:89:17
[test:*backend]      87|       ${userId ? sql` AND "userId"=${userId}::uuid` : Prisma.empty}
[test:*backend]      88|     `
[test:*backend]      89|     const rows = await prisma.c.$queryRaw<{ id: string }[]>`
[test:*backend]        |                 ^
[test:*backend]      90|       SELECT id FROM "EmailCandidateSent"
[test:*backend]      91|         WHERE ${whereFilters}

Is there a way to change the connection URL between connections? I’m trying to set up database test isolation.

Hi all, A recent change to heroku postgresql (https://devcenter.heroku.com/changelog-items/2446) is causing us some problems because extensions are now by default install in the

heroku_ext

namespace, and can only be installed in the

heroku_ext

namespace. Our application uses a custom schema

mobile

in our datasource definition

schema=mobile

in data source url as per Postgresql connection documentation in prisma. The problem is that because we are using a custom schema, it looks like the Quaint db driver fully clobbers the

search_path

to just

set search_path = mobile

which makes our migrations fail when looking for installed extensions. A short form of the fully failing migration is: And then this is our output:

Prisma schema loaded from db/schema.prisma
Datasource "db": PostgreSQL database "d3o13esjuu8i9a", schema "mobile" at "<http://ec2-3-222-74-92.compute-1.amazonaws.com:5432|ec2-3-222-74-92.compute-1.amazonaws.com:5432>"
18 migrations found in prisma/migrations
Applying migration `20220323211357_create_mobile_schema` // this is CREATE SCHEMA IF NOT EXISTS mobile;
Applying migration `20220323223149_enable_extensions`   // this contains CREATE EXTENSION IF NOT EXISTS "citext" - and succeeds
Applying migration `20220324172504_create_users`
Error: P3018
A migration failed to apply. New migrations cannot be applied before the error is recovered from. Read more about how to resolve migration issues in a production database: <https://pris.ly/d/migrate-resolve>
Migration name: 20220324172504_create_users
Database error code: 42704
Database error:
ERROR: type "citext" does not exist // I'm guess this is because the citext extension is now in the `heroku_ext` namespace which is not in the search_path
Position:
  0
  1 -- CreateTable
  2 CREATE TABLE "users" (
  3     "id" UUID NOT NULL DEFAULT gen_random_uuid(),
  4     "email" CITEXT NOT NULL,

DbError { severity: "ERROR", parsed_severity: Some(Error), code: SqlState(E42704), message: "type \"citext\" does not exist", detail: None, hint: None, position: Some(Original(101)), where_: None, schema: None, table: None, column: None, datatype: None, constraint: None, file: Some("parse_type.c"), line: Some(270), routine: Some("typenameType") }

I'm pretty sure that is because at the time that the create table is happening, the search path is only

mobile

and not

mobile, heroku_ext

Is this the right forum to post this or should I be posting this to the GitHub Issues for prisma, or to the prisma-engines repos?

I have a repo that is dedicated for my database and making migration schema changes. So far in a separate microservice i have an api that simply introspects the database to get the latest table schemas and works with the orm. Rather than introspecting the database, does it make sense to publish the generated prisma client to a private npm package? so instead my api microservice, rather than introspecting the database to look for latest changes, it could just update to the latest npm version of the package? wondering if anyone else has gone down a similar usecase or rather than republishing the entire PrismaClient, is it better to publish just the schema.prisma and hae the other microservice run prisma generate?

Hi all 👋, I am trying to deploy my service to heroku but i keep getting the error:

Error: Not Found <https://binaries.prisma.sh/all_commits/369b3694b7edb869fad14827a33ad3f3f49bbc20/debian-openssl-3.0.x/query-engine.gz>

Whenever it tries to run

npx prisma generate

as part of my build script. I have tried installing the cli globally on heroku still i get same error. Anyone encountered this before? please how do i resolve it? Thanks