thanks!

Not entirely sure if this is the right channel, but is there a way to have a multi-line string or multi-line comment in a Prisma schema? I'm building a generator and I want to be able to pass in a string with configuration options like this:

generator myThing {
    provider = "node ./dist/index.js"
    config = `
my config goes here
it is multiple lines
    `
}

Or like this if necessary:

generator myThing {
    /**
my config goes here
it is multiple lines
    */
    provider = "node ./dist/index.js"
}

But I can't figure out a way to do it. Right now the best option seems to be to have a separate docstring comment for each line, but that gets annoying to type if there's a lot of lines:

generator myThing {
    /// my config goes here
    /// it is multiple lines
    provider = "node ./dist/index.js"
}

I don't suppose there is any ETA on supporting

ON CONFLICT ...

for upserts?

Hi Prisma-Users! I have very basic question on how to create joins the PRISMA-way. I did not find a clear guidance in the docs: Imagine the ER in the attachment. We have

areas

and

epics

. An

area

has many

epics

and they are related through an

areaId

on epics. I have a controller method

area/<area-uid>/epics

in my API. What’s the best way to get all epics for a given area ? I tried two ways: 1. Two function calls

const area = await this.findOne(uid)
return this.prisma.epic.findMany({ where: { area } })

--> this creates two independent SQL quereies 2. One function call

return (await this.prisma.area.findUnique({ where: { uid }, select: { epics: true } }))['epics']

Works, but feels quirky. --- Is there a better solution? Like a simple right join? Thanks for feedback. I am sure this is a rookie thing. Cheers Anton

hello, I would like to kindly ask if there is a way to apply a helper method when you try to get nested data with a relational db. the helper method applies some data transformation

return this.prisma.test.findUnique({
  where: {
    id: params.id,
  },
  include: {
    test1: {
      include: {
        test2: transformationHelper()
      }
    }
  }

it seems something like that ^, possibly there is a better approach

Is there an efficient way to do a bulk upsert? ultimately I would like to do a

createMany

-- but there is a chance that some rows will duplicate on unique constraint. In those cases I'd like it to just be graceful eg. skip that row or update etc. The best idea i have is to run many concurrent upsert operations, but this could be very inefficient with thousands of rows.

Anyone have a better idea?

I thought about simply removing those rows first (not ideal), but then there is no way to efficiently remove them either 😅 (as far as I know) without running many concurrent delete operations

hi guys, im very curious, did prisma did not ever have hooks like oncreate or beforecreate or beforeupdate or afterupdate or before/afterDelete?

setting

log: ["query"]

is nice and helpful to print out the SQL queries, is there a way to print out the values along with the placeholders when it executes the query?

How to add a new field to a Prisma Model? I asked the full question on stack: https://stackoverflow.com/questions/70210307/how-to-add-a-field-to-a-model-in-prisma-graphql This has to be the simplest thing you can do in Prisma but I still manage to get errors, why is Prisma so complicated?

in JS when I do a query that fails because of a missing column, I don’t see any exception or error message instead the lambda just hangs until timeout. is this a known issue?

When I use the

upsert()

method as a

findorCreate

implementation, I get

{ count: 0 }

. This is only happening when a user is already in the database. Can someone tell me why this is happening? Below is a snippet of the method call.

const user = await prisma.user.upsert({
	where: { email: profile.emails[0].value },
	update: {},
	create: {
        email: profile.emails[0].value,
		name: profile.displayName,
	},
});

Does anyone have an efficient way to implement hashids with Prisma? We have a schema that uses autoincrementing IDs throughout but would like to use hashids to encode/decode IDs when they become "public facing". Right now, we have a rather hacky internal library that tracks all of our known ID property names and encodes/decodes them using hashids.

Hi everyone, Entry level dev with an easy one 🙂 I'm missing something about how

?

this works with a relation, or in general. I have a model

Issue

with a field

assignedTo

that takes a relation

User?

as an optional field.

model Issue {
  id          Int      @id @default(autoincrement())
  createdAt   DateTime @default(now())
  title       String
  description String
  createdBy   User     @relation(name: "createdByUser", fields: [createdById], references: [id])
  createdById Int

  assignedTo   User? @relation(name: "assignedTo", fields: [assignedToId], references: [id])
  assignedToId Int?
}

However, I can't seem to find a way to write a query to

create

or

update

that doesn't demand an

Int

at

assignedTo

without the query failing. For example, the query below won't work unless there is an

Int

at

input.isssue.assignedTo.id

.

const newIssue = await db.issue.create({
      data: {
        title: input.issue.title,
        description: input.issue.description,
        createdBy: {
          connect: {
            id: ctx.session.userId,
          },
        },
        assignedTo: {
          connect: {
            id: Number(input.issue.assignedTo.id),
          },
        },
      },
    })

I tried passing in

null

and

undefined

directly but both are rejected by the schema. Or should there be separate create and update queries for optional fields?

if(input.issue.assignedTo.id) { createIssueWithAssignedUser()}
  else {createIssue()}

Hello, I am new to prisma and I am trying to get a basic new model to work but any operation on it errors out with

Field does not exist on enclosing type.

I am using prisma version 3.6 with this schema definition.

model KeepAlives {
  id       Int      @id @default(autoincrement())
  tfield   String
  seededAt DateTime @default(now())

  @@map("keep_alives")
}

I get the error when trying to execute this line:

*const* keepAliveArr *=*

*await* prisma.keepAlives.findMany()

.

Hello, is it possible to write some sort of "smart update" query?

ra-data-prisma

plugin for React Admin (to connect with Nexus backend) does a thing where you supply new and previous data, it compares them field-wise and only sends a GQL request with setting only new values. So if I have an object

{ id: 1, name: 'John', lastName: 'Doe' }

and I pass a new object

{ id: 1, name: 'John', lastName: 'Smith' }

it automatically sends a query like

{ lastName: { set: 'Smith' } }

and doesn't touch any unchanged fields. Do we have anything like this in Prisma? Or is the "unnecessary update" overhead negligible?

hi guys, anyone using prisma with planetscale? im getting some very cryptic error, wondering if anyone have similar issues with planetscale

await this.prisma.invoiceItems.create(
  Error occurred during query execution:
ConnectorError(ConnectorError { user_facing_error: None, kind: QueryError(Server(ServerError { code: 3140, message: "target: unigate.-.primary: vttablet: (errno 3140) (sqlstate 22032) (CallerID: unsecure_grpc_client): Sql: \"insert into InvoiceItems(id, invoice_id, item, paid, accepted_coins, createdAt, currency, updatedAt, amount, paid_amount) values (:v1, :v2, :v3, :v4, :v5, :v6, :v7, :v8, :v9, :v10)\", BindVars: {}", state: "22032" })) })

Hi guys, I am currently having the following problem and hope you can point me to the right direction. I have a Lambda function that among other things writes to my AWS RDB database. This is working fine except for the case when one of the non-nullable input variables is null. My understanding is that in this case Prisma would throw an error. But it just hangs and the Lambda function times out after 60 seconds as follows (taken from AWS Cloudwatch logs) without any other error or notice:

2021-12-09T02:11:42.536Z 38e30598-f091-5333-936d-5fb3975e21d6 Task timed out after 62.06 seconds

Is that normal behavior or do I need to change my settings so that an error is thrown by Prisma? This is my table definition:

model tbl_Log {
  id Int @id @default(autoincrement())
  type LogType
  message String?
  createdAt DateTime
  user tbl_User @relation(fields: [userId], references: [id])
  userId String
  orderSnapshot Json
  amount Float
}

And I am creating the DB entry via:

await prisma.tbl_Log.create({
      data: {
        type: log.type!,
        message: log.message!,
        createdAt: log.createdAt!,
        userId,
        orderSnapshot: log.orderSnapshot!,
        amount: log.amount ?? 0,
      },
    });

If

amount

is

null

, the Lambda function hangs at this step, if

amount

has a

number

it works (with 0, too).

Is there a way to mix a raw statement into a Prisma Client operation?

Hi there, is there a type i can use for specifying a method return type where I use a findUnique with dinamyc include options passed in. For Example let's imagine i want a function that always add comments when i try to find a post:

function getPost(postId: string, includeDef: Prisma.PostInclude) Promise<Post> {
  const include = {
    ...includeDef,
    comments: true
  }
  return await this.prisma.post.findUnique({
    where: {id: postId},
    include
  })
}

The problem is if i call it this way

const post = getPost(postId, {owner: true})
// trying to do something with post.owner will cause typescript to complain that property owner doesn't exist

what type can i use instead of Promise<Post> to make it work properly ?

Hello everyone. How to change toExpNeg in a Decimal type field by default. For me, when the value is 0.00000002, 2e-8 is returned

Is there any way to reduce the dependency size of prisma in my lambda functions? 7.5MB of minified JS is causing some brutal cold starts

hi, I'm looking at migrating our plain SQL database driver to prisma for the type safety, one question so far - can Prisma Client work properly with a partitioned table in postgresql? i've seen this github issue about this issue, where one of the prisma folk says it's "probably" transparent for the prisma client, but I'm not sure what this actually means

Hi all — quick question, If I wanted to write an object typed with a prisma model into a

JsonObject

, E.G: see below. How would I do this if the model includes dates? I am currently getting an error due to a type mismatch since a JSON object doesn’t accept date values? but a date should just be parsed to a string I believe, do I need to do this manually so TS won’t complain?

model User { // user schema here}

import { User } from '@prisma/client';

const myObject: User = prisma.user.findUnique(// query here);

prisma.modelWithJsonField.create({
  data: { jsonField: myObject }
});

Hey all! Has anyone had success running Prisma on AWS Graviton EC2 instances? What binrary target do I need to use for the engine? We currently have these options set:

binaryTargets = ["native", "linux-musl", "linux-arm-openssl-1.1.x"]

but then once we deploy to AWS we get

Error: Unknown PRISMA_QUERY_ENGINE_LIBRARY undefined. Possible binaryTargets: darwin, darwin-arm64, debian-openssl-1.0.x, debian-openssl-1.1.x, rhel-openssl-1.0.x, rhel-openssl-1.1.x, linux-arm-openssl-1.1.x, linux-arm-openssl-1.0.x, linux-musl, linux-nixos, windows, freebsd11, freebsd12, openbsd, netbsd, arm, native or a path to the query engine library.

We are using Prisma 3.0.1, I know this is older we cannot upgrade (without some more investigtion) because of this issue https://github.com/prisma/prisma/issues/10512 We have multiple prisma schemas in our Nx repo and upgrading to the newest prisma seems to break them. cause it is looking in a hard coded spot for the schema. Ideas?

Is there a way to set up a relation with a table using a non-unique key? Trying to find a way to avoid writing this as two separate queries and I don't have say in changing the existing database.

const entityStates = await this.prisma.entityState.findMany({
    select: { id: true, entityTypeID: true },
    where: {
        name: state
    }
})


const entityTypes = await this.prisma.entityType.findMany({
    select: { id: true },
    where: {
        name: 'Enrollment',
        id: { in: entityStates.map( state => state.entityTypeID ) }
    }
})

The logic I'm looking for would look something like this

const entityTypes = await this.prisma.entityType.findMany({
    select: { id: true },
    where: {
        name: 'Enrollment',
        EntityStates: {
            where: {
                name: state
            }
        }
    }
})