Valibot-Mikro

Defining MikroORM entities with Valibot!

const User = defineEntitySchema("User", {
	id: number([primaryKey()]),
	name: string(),
	address: oneToOne(() => Address),
	cars: manyToMany(() => Car),
})

Highlights

• 💪 Fully type safe with static type inference
• 🔮 No need for ts-morph magic or experimental decorators
• 🧩 Compositing entities rather than Inheriting
• 📦 Simple and easy-to-use API

Install

It is assumed that you have already installed MikroORM, if not, please read the MikroORM installation documentation first.

## use npm
npm install valibot valibot-mikro

## use yarn
yarn add valibot valibot-mikro

## use pnpm
pnpm add valibot valibot-mikro

Usage

Valibot-Mikro is a utils library for EntitySchema of MikroORM. It helps you to define your entities via EntitySchema in a type-safe way.

Basic Usage

Here we define a User entity with Valibot and Valibot-Mikro:

import { number, string, optional } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export const User = defineEntitySchema("User", {
	id: number([primaryKey()]),
	fullName: string(),
	email: string(),
	password: string(),
	bio: optional(string(), ""),
})

We need to use em.create() method to create a new instance of the entity:

const author = em.create(User, { fullName: 'name', email: 'email', password:'secret' });
await em.flush();

Composable entities

We can easily use the Spread syntax to combine entities:

import { string, nullable, date, number } from "valibot"
import { primaryKey, property, defineEntitySchema } from "valibot-mikro"

const BaseEntity = {
	id: number([primaryKey()]),
	createdAt: date(),
	updatedAt: date([property({ onUpdate: () => new Date() })]),
}

const CanSale = {
	prices: number([property({ columnType: "money" })]),
	inventory: number([property({ columnType: "int" })]),
}

export const Book = defineEntitySchema("Book", {
	...BaseEntity,
	...CanSale,
	title: string(),
	author: string(),
})

export const Flower = defineEntitySchema("Flower", {
	...BaseEntity,
	...CanSale,
	variety: string(),
	color: nullable(string()),
})

We can also easily use merge() of valibot to combine entities:

import { object, string, optional, nullable, date, merge, number } from "valibot"
import { primaryKey, property, defineEntitySchema } from "valibot-mikro"

const BaseEntity = object({
	id: number([primaryKey()]),
	createdAt: date(),
	updatedAt: date([property({ onUpdate: () => new Date() })]),
})

const CanSale = object({
	prices: number([property({ columnType: "money" })]),
	inventory: number([property({ columnType: "int" })]),
})

export const Book = defineEntitySchema(
	"Book",
	merge([
		BaseEntity,
		CanSale,
		object({
			title: string(),
			author: string(),
		}),
	]),
)

export const Flower = defineEntitySchema(
	"Flower",
	merge([
		BaseEntity,
		CanSale,
		object({
			variety: string(),
			color: nullable(string()),
		}),
	]),
)

Add more metadata

For some special properties, we need to add more meta information，this is where we need to use property(). For example, add the onUpdate hook for updatedAt:

import { date, number, object, optional } from "valibot"
import { defineEntitySchema, primaryKey, property } from "valibot-mikro"

export const BaseEntity = defineEntitySchema("BaseEntity", {
	id: number([primaryKey()]),
	createdAt: optional(date(), () => new Date()),
	updatedAt: optional(date([property({ onUpdate: () => new Date() })]), () => new Date()),
})

The defineEntitySchema method can take same options as EntitySchema constructor. For example, we can add tableName to the User entity:

import { object, string, number } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export const User = defineEntitySchema(
	{ name: "User", tableName: "user_table", indexes: [{ properties: ["email"] }] },
	{
		id: number([primaryKey()]),
		fullName: string(),
		email: string(),
		password: string(),
	},
)

Modeling Entity Relationships

Let's see how easy it is to define relationships:

import { string, number } from "valibot"
import { defineEntitySchema, primaryKey, manyToOne } from "valibot-mikro"

const Breeder = defineEntitySchema("Breeder", {
	id: number([primaryKey()]),
	name: string(),
})

const Giraffe = defineEntitySchema("Giraffe", {
	id: number([primaryKey()]),
	name: string(),
	breeder: manyToOne(() => Breeder),
})

Defining relationships using valibot-mikro is similar to traditional methods. Here we have 6 methods to define relationships, four of which are the same as the traditional methods:

• oneToOne
• manyToOne
• oneToMany
• manyToMany

and two methods that allow you to define relationships manually:

• collection: define a collection
• reference: define a reference

Circular Reference

TypeScript is generally able to derive types correctly for us, however, when it comes to circular references, TypeScript can't help.
It's time to give TypeScript a hand by telling it the correct type:

import { object, string, optional } from "valibot"
import { EntitySchema, Ref } from "@mikro-orm/core"
import { defineEntitySchema, primaryKey, manyToOne, withRelations, InferEntity, oneToMany } from "valibot-mikro"

const Breeder = defineEntitySchema("Breeder", {
	id: optional(string([primaryKey()]), () => nanoid()),
	name: string(),
	giraffes: oneToMany(() => Giraffe, { mappedBy: "breeder" }),
})

const GiraffeSchema = object({
	id: optional(string([primaryKey()]), () => nanoid()),
	name: string(),
})

interface IGiraffe extends InferEntity<typeof GiraffeSchema> {
	breeder: Ref<InferEntity<typeof Breeder>>
}
const Giraffe: EntitySchema<IGiraffe> = defineEntitySchema(
	"Giraffe",
	withRelations<IGiraffe>(GiraffeSchema, {
		breeder: manyToOne(() => Breeder),
	}),
)

In this example there are circular references to Breeder and Giraffe, and TypeScript can't infer their types correctly for us, so we need to manually declare the type of one of them and tell TypeScript that type.
In this example we declare Giraffe's type as IGiraffe and mark the Giraffe as EntitySchema<IGiraffe>. In addition, we use the withRelations function to minimize boilerplate code.

Optional or Nullable Properties

To define a nullable property, which means that the database is allowed to store null values, we can use nullable() or nullish():

import { nullable, nullish, number, string } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export const Flower = defineEntitySchema("Flower", {
	id: number([primaryKey()]),
	variety: nullish(string(), "iris"),
	color: nullable(string()),
})

Default values

In some scenarios where we don't need to store a null value in the database, but simply need a default value, we use optional(). The most common use case is createdAt:

import { optional, date, string } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export const User = defineEntitySchema("User", {
	id: number([primaryKey()]),
	createdAt: optional(date(), () => new Date()),
	fullName: string(),
	email: string(),
	password: string(),
})

valibot-mikro use onInit hook under the hood to set the default values.

Enums

Just follow the valibot to define the enum:

import { enum_, number } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export enum UserRole {
	ADMIN = "admin",
	USER = "user",
}

export const User = defineEntitySchema("User", {
	id: number([primaryKey()]),
	role: enum_(UserRole),
})

In most scenarios, a picklist would be more concise:

import { picklist, number } from "valibot"
import { defineEntitySchema, primaryKey } from "valibot-mikro"

export const User = defineEntitySchema("User", {
	id: number([primaryKey()]),
	role: picklist(["admin", "user"]),
})