lsg-radu / typebox

A type system in a box + json schema

typebox

A type system in a box + json schema

import { Type, Static } from "./typebox"

// type Foo = {
//   a: string,
//   b: number,
//   c: boolean
// }
const Foo = Type.Object({
  a: Type.String(),
  b: Type.Number(),
  c: Type.Boolean()
})

// statically resolve types
const foo: Static<typeof Foo> = { a: "hello", b: 2, c: true }

console.log(Foo)
// -> { type: 'object',
//     properties: 
//      { a: { type: 'string' },
//        b: { type: 'number' },
//        c: { type: 'boolean' } },
//     required: [ 'a', 'b', 'c' ] }

overview

typebox is a type builder library which allows developers to compose json schema objects and allow these json schema objects to be statically resolved as TypeScript types. The aim of this project is to find parity with respect to the TypeScript type checker and JSON schema, allowing for runtime type verification using standard json schema validation that lines up to TypeScripts structural type checker.

note: typebox doesn't is only concerned with generating valid JSON schema and static resolution and does not come with a JSON schema validator. For JSON schema validation, recommend the ajv library for JavaScript.

usage

typebox is written as a portable single file that can be copied into any typescript project. Copy the ./src/typebox.ts file into your project and you're all setup.

import { Type, Static } from "./typebox"

const TString = Type.String()

// resolve as type
type String = Static<typeof TString>

// as an argument
const f = (s: Static<typeof TString>) => { /* ... */ }

// and so on...

typebox > typescript

The following are the typebox to typescript mappings.

type	typbox	typescript
Any	const T = Type.Any()	type T = any
Null	const T = Type.Null()	type T = null
Number	const T = Type.Number()	type T = number
String	const T = Type.String()	type T = string
Boolean	const T = Type.Boolean()	type T = boolean
Object	const T = Type.Object({ name: Type.String() })	type T = {name: string}
Array	const T = Type.Array(Type.Object({ name: Type.String() }))	type T = {name: string}[]
Enum	const T = Type.Enum("yes", "no")	`type T = "yes"
Tuple	const T = Type.Tuple(Type.String(), Type.Number())	type T = [string, number]
Union	const T = Type.Union(Type.String(), Type.Number())	`type T = string
Intersect	const T = Type.Intersect(Type.String(), Type.Number())	type T = string & number
Literal	const T = Type.Literal("click")	type T = "click"
Pattern	const T = Type.Pattern(/foo/)	type T = string
Range	const T = Type.Range(0, 100)	type T = number

typebox > json schema

The following are the typebox to json schema mappings.

type	typbox	json schema
Any	const T = Type.Any()	{ }
Null	const T = Type.Null()	{ type: "null" }
Number	const T = Type.Number()	{ type: "number" }
String	const T = Type.String()	{ type: "string" }
Boolean	const T = Type.Boolean()	{ type: "boolean" }
Object	const T = Type.Object({ name: Type.String() })	{ type: "object": properties: { name: { type: "string" } }, required: ["name"] }
Array	const T = Type.Array(Type.Object({ name: Type.String() }))	{ type: "array": items: { type: "object": properties: { name: { type: "string" } }, required: ["name"] } }
Enum	const T = Type.Enum("yes", "no")	{ type: "string", enum: ["yes", "no"] }
Tuple	const T = Type.Tuple(Type.String(), Type.Number())	{ type: "array", items: [{type: "string"}, {type: "number"}], additionalItems: false, minItems: 2, maxItems: 2 }
Union	const T = Type.Union(Type.String(), Type.Number())	{ anyOf: [{ type: "string"}, {type: "number"}] }
Intersect	const T = Type.Intersect(Type.String(), Type.Number())	{ allOf: [{ type: "string"}, {type: "number"}] }
Literal	const T = Type.Literal("click")	{ type: "string", enum: ["click"] }
Pattern	const T = Type.Pattern(/foo/)	{ type: "string", pattern: "foo" }
Range	const T = Type.Range(0, 100)	{ type: "number", minimum: 0, maximum: 100 }