Nested array

While kysely is not an ORM and it doesn't have the concept of relations, we do provide helpers for fetching nested objects and arrays in a single query. In this example we use the jsonArrayFrom helper to fetch person's pets along with the person's id.

Please keep in mind that the helpers under the kysely/helpers folder, including jsonArrayFrom, are not guaranteed to work with third party dialects. In order for them to work, the dialect must automatically parse the json data type into JavaScript JSON values like objects and arrays. Some dialects might simply return the data as a JSON string. In these cases you can use the built in ParseJSONResultsPlugin to parse the results.

import type {
  ColumnType,
  Generated,
  GeneratedAlways,
  Insertable,
  Kysely,
  Selectable,
  SqlBool,
  Updateable,
} from 'kysely'

export interface Database {
  audit: AuditTable
  person: PersonTable
  person_backup: PersonTable
  pet: PetTable
  toy: ToyTable
  wine: WineTable
  wine_stock_change: WineStockChangeTable
}

interface AuditTable {
  id: Generated<number>
  action: string
}

interface PersonTable {
  id: Generated<number>
  address: { city: string } | null
  age: number | null
  birthdate: ColumnType<Date | null, string | null | undefined, string | null>
  created_at: GeneratedAlways<Date>
  deleted_at: ColumnType<Date | null, string | null | undefined, string | null>
  experience: { role: string }[] | null
  first_name: string
  gender: 'male' | 'female' | 'other' | null
  has_pets: Generated<'Y' | 'N'>
  last_name: string | null
  middle_name: string | null
  nicknames: string[] | null
  nullable_column: string | null
  profile: {
    addresses: { city: string }[]
    website: { url: string }
  } | null
  updated_at: ColumnType<Date | null, string | null | undefined, string | null>
  marital_status: 'single' | 'married' | 'divorced' | 'widowed' | null
}

interface PetTable {
  id: Generated<number>
  created_at: GeneratedAlways<Date>
  is_favorite: Generated<SqlBool>
  name: string
  owner_id: number
  species: Species
}

interface ToyTable {
  id: Generated<number>
  name: string
  pet_id: number
  price: number
}

interface WineTable {
  id: Generated<number>
  name: string
  stock: number
}

interface WineStockChangeTable {
  id: Generated<number>
  stock_delta: number
  wine_name: string
}

export type Person = Selectable<PersonTable>
export type NewPerson = Insertable<PersonTable>
export type PersonUpdate = Updateable<PersonTable>
export type Pet = Selectable<PetTable>
export type NewPet = Insertable<PetTable>
export type PetUpdate = Updateable<PetTable>
export type Species = 'dog' | 'cat' | 'hamster'

declare global {
  // @ts-ignore
  export class Buffer {
    static isBuffer(obj: unknown): obj is { length: number }
    static compare(a: Buffer, b: Buffer): number
  }
  export const db: Kysely<Database>
  export function functionThatExpectsPersonWithNonNullValue(
    person: Person & { nullable_column: string },
  ): void
}


import { jsonArrayFrom } from 'kysely/helpers/postgres'

const result = await db
  .selectFrom('person')
  .select((eb) => [
    'id',
    jsonArrayFrom(
      eb.selectFrom('pet')
        .select(['pet.id as pet_id', 'pet.name'])
        .whereRef('pet.owner_id', '=', 'person.id')
        .orderBy('pet.name')
    ).as('pets')
  ])
  .execute()

More examples

The API documentation is packed with examples. The API docs are hosted here, but you can access the same documentation by hovering over functions/methods/classes in your IDE. The examples are always just one hover away!

For example, check out these sections: