I’m trying to replicate a REST API that I’ve built in the past and one part that got me thinking was if one of my tables had an array of objects. So for example, I have a table called Profile and it contains arrays Experience and Education that would strictly be under Profile but has its own fields as well but not its own table.
While I was adding fields in GraphQL, I bumped into this not really having a solid solution other than to create new types and then associating them with relationships and then have a resolver or a front-end make sure a Profile is created first before the Experience/Education portion is. I’m not sure if it’s the right way to do it or if there is a better way to do it. Down below is a snippet of what I ended up using… looking at the admin page, there are created tables for Profile, Experience and Education which is expected. But is there a way to only have just Profile and accomplish something similar? Or is this more of a way of life with GraphQL?
type Profile {
id: ID! @id
handle: String!
company: String
website: String
location: String
status: String!
githubUsername: String
experience: [Experience!] @relation(link: INLINE)
education: [Education!] @relation(link: INLINE)
}
type Experience {
id: ID! @id
title: String!
company: String!
}
type Education {
id: ID! @id
title: String!
company: String!
}
In Prisma, you can use embedded types. You would drop the @relation directive and add @embedded directives to the types you're embedding:
type Profile {
id: ID! @id
handle: String!
company: String
website: String
location: String
status: String!
githubUsername: String
experience: [Experience!]
education: [Education!]
}
type Experience @embedded {
title: String!
company: String!
}
type Education @embedded {
title: String!
company: String!
}
However, this only possible if you're using MongoDB for your database and there's some specific limitations listed in the docs when using embedded types.