I have a dto file that looks something like this:

export class CreateJobDto {
  @ValidateNested()
  @Type(() => CreateCustomerDto)
  readonly client: CreateCustomerDto;
}

My Job Schema looks like this:

import { Prop, Schema as NestSchema } from '@nestjs/mongoose';
@NestSchema({collection: 'Jobs'})
export class Job {
  @Prop({ type: Customer })
  client: Customer
}

My CreateCustomerDto looks like this:

export class CreateCustomerDto {
  @IsOptional()
  @IsArray()
  @ArrayNotEmpty()
  @IsMongoId({ each: true })
  readonly brand: string[];
}

My Customer Schema looks like this:

import { Prop, Schema as NestSchema } from "@nestjs/mongoose";
@NestSchema({collection: 'Customers'})
export class Customer extends Document {
  @Prop([{ type: Schema.Types.ObjectId, ref: 'Brand', default: undefined }])
  brand: Schema.Types.ObjectId[]
}

When a Customer is fetched using Mongoose I am using .populate(‘brand’) to populate the Customer object with the corresponding Brands.

The issue is that when I create a Job, it is expecting the client property to be a nested object but the schema validation fails because the Customer schema is only storing references and not populated objects.

Since I am using a mix of nested objects and references to objects what is considered best practice and how should this be solved?

I have created a different dto for the Customer which specifies what it should look like when the properties are populated. I could create a different Customer schema purely for validation in other documents but I am not sure whether that is considered good practice.