Could someone please tell me where I can find the code from prisma, which auto generates the graphql schema from a database?

Do you mean the GraphQL API, or the datamodel (introspection)?

With

graphql get-schema

I can download the schema from the endpoint, I would like to study the code which generates this schema based on the database.

Got it! The code resides here: https://github.com/prismagraphql/prisma/tree/master/server It's probably easiest to step through the code when executing a test. Here are instructions for setting up your dev environment to do so: https://github.com/SHASHANK-PRO-05/prisma/blob/11372ff7eb18cca9c4e2c186c75c12898ae85c74/server/CONTRIBUTING.md

Thank you very much! Amazing support 🙂

You're welcome! 🙂 Love to help out in the community

So if I understood this correctly, the code in

/server/servers/api/src/main/scala/com/prisma/api/schema

is responsible for creating a database specific schema, right?

Yup. Specifically, the SchemaBuilder: https://github.com/prismagraphql/prisma/blob/master/server/servers/api/src/main/scala/com/prisma/api/schema/SchemaBuilder.scala I am curious what you want to achieve, can you share more about this?

All right, I thought so, but I don't know Scala, so I wasn't sure. I need to write (Java) software that creates the GraphlQL schema for an existing database. In principle almost exactly like the schema you get with "graphlql get-schema" from the Prisma application, but it should only contain requests and should generally not be as extensive as with Prisma. So I wanted to look at the code of Prisma to see how exactly the database (i.e. relationships, attributes etc.) are mapped in the schema. I would also like to use jOOQ for modelling the database in Java and create with the help of these jOOQ classes the graphQL schema.

Interesting. What speaks against actually using Prisma? Are you missing something specifically?

Prisma works great, but I only need a really small part of it, so it would be a good idea to do it myself. Another question, I've looked at the schema generation now and in principle the schema is generated from the file

datamodel.graphql

, right? So the input for the schema generator is the data model and the complete schema is created from this file?