Hi all. Thanks for the invitation and having me here!

I have a question in case anybody can help me: Is there any way to filtering by Regular Expression pattern?

I noticed there is a request open from 2018… https://github.com/prisma/prisma1/issues/1766

Hey guys, shouldn't raw queries still work with the schema file? I'm having issues when writing a raw query using the model/fields names, if I use the table/mapped fields names it works. If I understand correctly this doc, model names and fields names should work correctly (https://www.prisma.io/docs/concepts/components/prisma-client/raw-database-access), or raw query is a 100% raw query and I misunderstood the doc?

Any updates on Geo Fields like Point

I have a MySql column with type

double(10,2)

. How Prisma can handle it? if i run

prisma migrate

it will override my database to

DOUBLE

are

upsert

and

connectOrCreate

somehow incompatible? 🤔 i've got something along the lines of:

const director = someBoolean ? { connectOrCreate: { create: { name: 'Steven Spielberg' }, where: { name: 'Steven Spielberg' }  } } : undefined
const movie = { contentId: 'some_id_here', title: 'Jaws', director }

const upsertResult = await prisma.movie.upsert({ create: movie, update: movie, where: { contentId: movie.contentId } })
// results in "The change you are trying to make would violate the required relation 'DirectorToMovie' between the 'Director' and 'Movie' models.

for reference, the Movie model lists a director as being optional, which is why I'm so confused at it complaining that it would break a required relation. Not sure if it's important, but I'm using sqlite as the underlying db

model Movie {
  id        Int @id @default(autoincrement())
  contentId String
  title     String
  director  Director?
}

model Director {
  id      Int @id @default(autoincrement())
  name    String @unique
  movie   Movie @relation(fields: [movieId], references: [id])
  movieId Int
}

Hello... Im new to prisma. Just getting started today. Now I just read the docs about include specially in one to many advantage. But im curious if I have the user and post relationship does user will return all the posts made by the user?? how will I do the first 5 posts only?

How accurate will be the db introspection ? should we trust it? or should we trust it in most cases but still needed a user correction when pulled?

Can some one help me on this problem?

Guys! I just bought new macbook m1. I have this issue and I couldn't solve this.

Hi All, any idea when

Fuzzy/Phrase full text search

will be available in prisma schema and client?

🤙

Hey, how can I prefix my prisma command with a dotenv command? It should be something like this:

npx dotenv -e .env.development prisma db seed --preview-feature

However I still get this error:

Environment variables loaded from .env
Error: This feature is currently in Preview. There may be bugs and it's not recommended to use it in production environments.
Please provide the --preview-feature flag to use this command.

Note that it also uses the wrong env. Found it: Just separate the two commands with --

Hi everyone previously in older version of Prisma (using with Docker) locally I could've run Prisma playground and directly query Data Base but now I can't, now when I run prisma migrate dev , I see the below log : Datasource "db": PostgreSQL database "MyDBName", schema "public" at "localhost:5432" I think it has to do sth with this -> localhost:5432 Could somebody help me with this ?

Hi, I got this error when executing

prisma.schema.findMany()

. I can't seem to find what caused this on the documentation. On the table, there is 3 data already.

"name": "NotFoundError",
"clientVersion": "2.25.0"

Is it possible to use a "include" from another table in an aggregate query?

I need to sum all columns in multiple tables.

For some reason when I came to update a few models I started getting this errror

Error: Cannot find module '@prisma/engines'

Tried

npm i @prisma/engine

but still gives the same error

Hi, how do i generate optional Prisma Client Types for this schema

// schema
model User {
  id         Int     @id @default(autoincrement())
  facebookId String? @unique
  googleId   String? @unique
  profileURL String?
}

//desired type generated on the client
export type User = {
  id: number
  facebookId?: string | null
  googleId?: string | null
  profileURL?: string | null
}

// what is really generated
export type User = {
  id: number
  facebookId: string | null
  googleId: string | null
  profileURL: string | null
}

anyone has a working example of getting Prisma up and running against AWS Aurora Serverless with CDK? Can not get it work, it does not connect within the VPC 🥲

`page not found`: https://github.com/prisma/prisma-examples/blob/latest/typescript/rest-express/prisma/dev.db

Hi everyone , is there any way to seed your Database multiple times without running prisma reset command I am using the command down below (but it works just one time after migration and you need to reset to run it again)

prisma db seed --preview-feature

And I'm following this page: https://www.prisma.io/docs/guides/database/seed-database

Interesting question for everyone, do people use

ID

or

Id

? For example when adding a relation, do you do

postId

or

postID

? I personally find myself doing

ID

but I noticed that the VSCode extension defaults to

Id

Hello everybody, have a few questions: I’ve got 2 schemas: prisma schema and app schema. Prisma generated types for me, also I’ve generated types for the app schema using

graphql-code-generator

. 1. How to properly ‘type’ the resolvers ? I typed the resolvers itself with types generated from the

graphql-code-generator

for example:

const updateUser: MutationResolvers['updateUser'] = <resolver code>

But the problem is - return type from the resolver (

prisma.user.update

call ) is User (prisma) and it does not include the relations, and here is the conflict ( ’Type ‘User’ is missing the following properties from type ‘User’: adminOf, employeeOf, accounts’ or similar ones ). Basically for every model/type I’ve got 2 type definitions and in many cases they differ. Here is the User type in the app schema:

type User {
  email: String!
  firstName: String!
  lastName: String!
  phone: String
  billingAddress: Address
  shippingAddress: Address
  adminOf: [Company!]!
  employeeOf: [Company!]!
  accounts: [Account!]!
  adminOfVendors: [Vendor]
}

Here is typings for the User model (prisma):

export type User = {
  id: string
  emailVerificationToken: string | null
  passwordResetToken: string | null
  superUser: boolean
  email: string
  firstName: string
  lastName: string
  password: string
  phone: string | null
  billingAddress: Prisma.JsonValue | null
  shippingAddress: Prisma.JsonValue | null
}

Here is typings for the User type (graphql-code-generator):

export type User = {
  __typename?: 'User';
  email: Scalars['String'];
  firstName: Scalars['String'];
  lastName: Scalars['String'];
  phone?: Maybe<Scalars['String']>;
  billingAddress?: Maybe<Address>;
  shippingAddress?: Maybe<Address>;
  adminOf: Array<Company>;
  employeeOf: Array<Company>;
  accounts: Array<Account>;
  adminOfVendors?: Maybe<Array<Maybe<Vendor>>>;
};

Maybe there is some universal practice of typing the API with 2 main type files (for the prisma datamodel and app schema) ? I’m new to the typescript. 2. If my app schema defines some fields which are relations (User example):

adminOf: [Company!]!
  employeeOf: [Company!]!
  accounts: [Account!]!
  adminOfVendors: [Vendor]

I do need to use

include

in the Prisma Client in every resolver that should return User type? Because they are not returned by default. 3. T*ype resolvers:* For example I’ve got

Company

type and it has field

admin

which is of type User. When I write type resolvers for the

Company

type, specifically for the

admin

field, do I also need to use

include

in it (to get relations of the User) ? Example:

const resolvers = {
  Query: {
    // ... 
  },
  Mutation: {
    // ... 
  },
  Company: {
    admin: (parent, args, context) => {
       return context.prisma.company.findUnique({
         where: { id: parent.id }
       }).user()

       // Or something like that
       return context.prisma.user.findUnique({
        where: { id: parent.userId },
        include: {
          some_relation_field: true
        }
      })
     }
   },
}

Sorry for a lot questions, thank you.

Is it possible to use

connect

and update the connecting relation at the same time? Kinda like how you can use

create

to create a new relation.

const createPost = await prisma.post.create({
  data: {
    title: "How to make croissants",
    author: {
      create: {
        email: "<mailto:viola@prisma.io|viola@prisma.io>",
        name: "Viola"
      }
    }
  }
})

I kinda want the same, but like...

const createPost = await prisma.post.create({
  data: {
    title: "How to make croissants",
    author: {
      connect: {
        id: 10,
        data: {
          email: "<mailto:viola@prisma.io|viola@prisma.io>",
          name: "Viola"
        }
      }
    }
  }
})

I know that exact API is not available, but is there something simlar?

I can connect to my Aurora Serverless with this raw SQL lib: https://www.npmjs.com/package/serverless-postgres but the same setup is not working with Prisma…. do not know what is happening here…. 🥲

Hi All just noticed this while preparing for a prisma related presentation: https://www.prisma.io/docs/getting-started/quickstart The package.json file in the quickstart project is missing "@types/node": "^15.12.4" so it throws an error when trying to run.