devacour
TypeBox
JSON Schema Type Builder with Static Type Resolution for TypeScript
Install
Node
$ npm install @sinclair/typebox --saveDeno
import { Static, Type } from 'https://deno.land/x/typebox/src/typebox.ts'Usage
import { Static, Type } from '@sinclair/typebox'
const T = Type.String() // const T = { "type": "string" }
type T = Static<typeof T> // type T = stringOverview
TypeBox is a library that builds in-memory JSON Schema objects that can be statically resolved to TypeScript types. The schemas produced by this library are designed to match the static type checking rules of the TypeScript compiler. TypeBox allows one to create a unified type that can be statically checked by the TypeScript compiler and runtime asserted using standard JSON Schema validation.
TypeBox can be used as a simple tool to build up complex schemas or integrated into RPC or REST services to help validate JSON data received over the wire. TypeBox does not provide any JSON schema validation. Please use libraries such as AJV to validate schemas built with this library.
Requires TypeScript 4.3.5 and above.
License MIT
Contents
- Install
- Overview
- Example
- Types
- Modifiers
- Options
- Generic Types
- Reference Types
- Recursive Types
- Extended Types
- Strict
- Validation
- OpenAPI
Example
The following demonstrates TypeBox's general usage.
import { Static, Type } from '@sinclair/typebox'
//--------------------------------------------------------------------------------------------
//
// Let's say you have the following type ...
//
//--------------------------------------------------------------------------------------------
type T = {
id: string,
name: string,
timestamp: number
}
//--------------------------------------------------------------------------------------------
//
// ... you can express this type in the following way.
//
//--------------------------------------------------------------------------------------------
const T = Type.Object({ // const T = {
id: Type.String(), // type: 'object',
name: Type.String(), // properties: {
timestamp: Type.Integer() // id: {
}) // type: 'string'
// },
// name: {
// type: 'string'
// },
// timestamp: {
// type: 'integer'
// }
// },
// required: [
// "id",
// "name",
// "timestamp"
// ]
// }
//--------------------------------------------------------------------------------------------
//
// ... then infer back to the original static type this way.
//
//--------------------------------------------------------------------------------------------
type T = Static<typeof T> // type T = {
// id: string,
// name: string,
// timestamp: number
// }
//--------------------------------------------------------------------------------------------
//
// ... then use the type both as JSON schema and as a TypeScript type.
//
//--------------------------------------------------------------------------------------------
function receive(value: T) { // ... as a Type
if(JSON.validate(T, value)) { // ... as a Schema
// ok...
}
}Types
The following table outlines the TypeBox mappings between TypeScript and JSON schema.
┌────────────────────────────────┬─────────────────────────────┬────────────────────────────────┐
│ TypeBox │ TypeScript │ JSON Schema │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Any() │ type T = any │ const T = { } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Unknown() │ type T = unknown │ const T = { } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.String() │ type T = string │ const T = { │
│ │ │ type: 'string' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Number() │ type T = number │ const T = { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Integer() │ type T = number │ const T = { │
│ │ │ type: 'integer' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Boolean() │ type T = boolean │ const T = { │
│ │ │ type: 'boolean' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Null() │ type T = null │ const T = { │
│ │ │ type: 'null' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.RegEx(/foo/) │ type T = string │ const T = { │
│ │ │ type: 'string', │
│ │ │ pattern: 'foo' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Literal(42) │ type T = 42 │ const T = { │
│ │ │ const: 42 │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Array( │ type T = number[] │ const T = { │
│ Type.Number() │ │ type: 'array', │
│ ) │ │ items: { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Object({ │ type T = { │ const T = { │
│ x: Type.Number(), │ x: number, │ type: 'object', │
│ y: Type.Number() │ y: number │ properties: { │
│ }) │ } │ x: { │
│ │ │ type: 'number' │
│ │ │ }, │
│ │ │ y: { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ }, │
│ │ │ required: ['x', 'y'] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Tuple([ │ type T = [number, number] │ const T = { │
│ Type.Number(), │ │ type: 'array', │
│ Type.Number() │ │ items: [ │
│ ]) │ │ { │
│ │ │ type: 'number' │
│ │ │ }, { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ ], │
│ │ │ additionalItems: false, │
│ │ │ minItems: 2, │
│ │ │ maxItems: 2, │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ enum Foo { │ enum Foo { │ const T = { │
│ A, │ A, │ anyOf: [{ │
│ B │ B │ type: 'number', │
│ } │ } │ const: 0 │
│ │ │ }, { │
│ const T = Type.Enum(Foo) │ type T = Foo │ type: 'number', │
│ │ │ const: 1 │
│ │ │ }] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.KeyOf( │ type T = keyof { │ const T = { │
│ Type.Object({ │ x: number, │ enum: ['x', 'y'], │
│ x: Type.Number(), │ y: number │ type: 'string' │
│ y: Type.Number() │ } │ } │
│ }) │ │ │
│ ) │ │ │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Union([ │ type T = string | number │ const T = { │
│ Type.String(), │ │ anyOf: [{ │
│ Type.Number() │ │ type: 'string' │
│ ]) │ │ }, { │
│ │ │ type: 'number' │
│ │ │ }] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Intersect([ │ type T = { │ const T = { │
│ Type.Object({ │ x: number │ allOf: [{ │
│ x: Type.Number() │ } & { │ type: 'object', │
│ }), │ y: number │ properties: { │
│ Type.Object({ │ } │ a: { │
│ y: Type.Number() │ │ type: 'number' │
│ }) │ │ } │
│ }) │ │ }, │
│ │ │ required: ['a'] │
│ │ │ }, { │
│ │ │ type: 'object', │
│ │ │ properties: { │
│ │ │ b: { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ }, │
│ │ │ required: ['b'] │
│ │ │ }] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Record( │ type T = { │ const T = { │
│ Type.String(), │ [key: string]: number │ type: 'object', │
│ Type.Number() │ } │ patternProperties: { │
│ ) │ │ '^.*$': { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Partial( │ type T = Partial<{ │ const T = { │
│ Type.Object({ │ x: number, │ type: 'object', │
│ x: Type.Number(), │ y: number │ properties: { │
│ y: Type.Number() | }> │ x: { │
│ }) │ │ type: 'number' │
│ ) │ │ }, │
│ │ │ y: { │
│ │ │ type: 'number' │
│ │ │ } │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Required( │ type T = Required<{ │ const T = { │
│ Type.Object({ │ x?: number, │ type: 'object', │
│ x: Type.Optional( │ y?: number │ properties: { │
│ Type.Number() | }> │ x: { │
│ ), │ │ type: 'number' │
│ y: Type.Optional( │ │ }, │
│ Type.Number() │ │ y: { │
│ ) │ │ type: 'number' │
│ }) │ │ } │
│ ) │ │ }, │
│ │ │ required: ['x', 'y'] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Pick( │ type T = Pick<{ │ const T = { │
│ Type.Object({ │ x: number, │ type: 'object', │
│ x: Type.Number(), │ y: number │ properties: { │
│ y: Type.Number(), | }, 'x'> │ x: { │
│ }), ['x'] │ │ type: 'number' │
│ ) │ │ } │
│ │ │ }, │
│ │ │ required: ['x'] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Omit( │ type T = Omit<{ │ const T = { │
│ Type.Object({ │ x: number, │ type: 'object', │
│ x: Type.Number(), │ y: number │ properties: { │
│ y: Type.Number(), | }, 'x'> │ y: { │
│ }), ['x'] │ │ type: 'number' │
│ ) │ │ } │
│ │ │ }, │
│ │ │ required: ['y'] │
│ │ │ } │
│ │ │ │
└────────────────────────────────┴─────────────────────────────┴────────────────────────────────┘Modifiers
TypeBox provides modifiers that can be applied to an objects properties. This allows for optional and readonly to be applied to that property. The following table illustates how they map between TypeScript and JSON Schema.
┌────────────────────────────────┬─────────────────────────────┬────────────────────────────────┐
│ TypeBox │ TypeScript │ JSON Schema │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Object({ │ type T = { │ const T = { │
│ name: Type.Optional( │ name?: string, │ type: 'object', │
│ Type.String(), │ } │ properties: { │
│ ) │ │ name: { │
│ }) │ │ type: 'string' │
│ │ │ } │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Object({ │ type T = { │ const T = { │
│ name: Type.Readonly( │ readonly name: string, │ type: 'object', │
│ Type.String(), │ } │ properties: { │
│ ) │ │ name: { │
│ }) │ │ type: 'string' │
│ │ │ } │
│ │ │ }, │
│ │ │ required: ['name'] │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Object({ │ type T = { │ const T = { │
│ name: Type.ReadonlyOptional( │ readonly name?: string, │ type: 'object', │
│ Type.String(), │ } │ properties: { │
│ ) │ │ name: { │
│ }) │ │ type: 'string' │
│ │ │ } │
│ │ │ } │
│ │ │ } │
│ │ │ │
└────────────────────────────────┴─────────────────────────────┴────────────────────────────────┘Options
You can pass additional JSON schema options on the last argument of any given type. The following are some examples.
// string must be an email
const T = Type.String({ format: 'email' })
// number must be a multiple of 2
const T = Type.Number({ multipleOf: 2 })
// array must have at least 5 integer values
const T = Type.Array(Type.Integer(), { minItems: 5 })Generic Types
Generic types can be created using functions. The following creates a generic Nullable<T> type.
import { Type, Static, TSchema } from '@sinclair/typebox'
// type Nullable<T> = T | null
const Nullable = <T extends TSchema>(type: T) => Type.Union([type, Type.Null()])
const T = Nullable(Type.String()) // const T = {
// "anyOf": [{
// type: 'string'
// }, {
// type: 'null'
// }]
// }
type T = Static<typeof T> // type T = string | null
const U = Nullable(Type.Number()) // const U = {
// "anyOf": [{
// type: 'number'
// }, {
// type: 'null'
// }]
// }
type U = Static<typeof U> // type U = number | nullReference Types
Types can be referenced with Type.Ref(...). To reference a type, the target type must specify an $id.
const T = Type.String({ $id: 'T' }) // const T = {
// $id: 'T',
// type: 'string'
// }
const R = Type.Ref(T) // const R = {
// $ref: 'T'
// }It can be helpful to organize shared referenced types under a common namespace. The Type.Box(...) function can be used to create a shared definition container for related types. The following creates a Math3D container and a Vertex structure that references types in the container.
const Math3D = Type.Box({ // const Math3D = {
Vector4: Type.Object({ // $id: 'Math3D',
x: Type.Number(), // definitions: {
y: Type.Number(), // Vector4: {
z: Type.Number(), // type: 'object',
w: Type.Number() // properties: {
}), // x: { type: 'number' },
Vector3: Type.Object({ // y: { type: 'number' },
x: Type.Number(), // z: { type: 'number' },
y: Type.Number(), // w: { type: 'number' }
z: Type.Number() // },
}), // required: ['x', 'y', 'z', 'w']
Vector2: Type.Object({ // },
x: Type.Number(), // Vector3: {
y: Type.Number() // type: 'object',
}) // properties: {
}, { $id: 'Math3D' }) // x: { 'type': 'number' },
// y: { 'type': 'number' },
// z: { 'type': 'number' }
// },
// required: ['x', 'y', 'z']
// },
// Vector2: {
// type: 'object',
// properties: {
// x: { 'type': 'number' },
// y: { 'type': 'number' },
// },
// required: ['x', 'y']
// }
// }
// }
const Vertex = Type.Object({ // const Vertex = {
position: Type.Ref(Math3D, 'Vector4'), // type: 'object',
normal: Type.Ref(Math3D, 'Vector3'), // properties: {
uv: Type.Ref(Math3D, 'Vector2') // position: { $ref: 'Math3D#/definitions/Vector4' },
}) // normal: { $ref: 'Math3D#/definitions/Vector3' },
// uv: { $ref: 'Math3D#/definitions/Vector2' }
// },
// required: ['position', 'normal', 'uv']
// }Recursive Types
Recursive types can be created with the Type.Rec(...) function. The following creates a Node type that contains an array of inner Nodes. Note that due to current restrictions on TypeScript inference, it is not possible for TypeBox to statically infer for recursive types. TypeBox will infer the inner recursive type as any.
const Node = Type.Rec(Self => Type.Object({ // const Node = {
id: Type.String(), // $id: 'Node',
nodes: Type.Array(Self), // $ref: 'Node#/definitions/self',
}), { $id: 'Node' }) // definitions: {
// self: {
// type: 'object',
// properties: {
// id: {
// type: 'string'
// },
// nodes: {
// type: 'array',
// items: {
// $ref: 'Node#/definitions/self'
// }
// }
// }
// }
// }
type Node = Static<typeof Node> // type Node = {
// id: string
// nodes: any[]
//
function visit(node: Node) {
for(const inner of node.nodes) {
visit(inner as Node) // Assert inner as Node
}
}Extended Types
In addition to JSON schema types, TypeBox provides several extended types that allow for function and constructor types to be composed. These additional types are not valid JSON Schema and will not validate using typical JSON Schema validation. However, these types can be used to frame JSON schema and describe callable interfaces that may receive JSON validated data. These types are as follows.
┌────────────────────────────────┬─────────────────────────────┬────────────────────────────────┐
│ TypeBox │ TypeScript │ Extended Schema │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Constructor([ │ type T = new ( │ const T = { │
│ Type.String(), │ arg0: string, │ type: 'constructor' │
│ Type.Number(), │ arg1: number │ arguments: [{ │
│ ], Type.Boolean()) │ ) => boolean │ type: 'string' │
│ │ │ }, { │
│ │ │ type: 'number' │
│ │ │ }], │
│ │ │ returns: { │
│ │ │ type: 'boolean' │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Function([ │ type T = ( │ const T = { │
| Type.String(), │ arg0: string, │ type : 'function', │
│ Type.Number(), │ arg1: number │ arguments: [{ │
│ ], Type.Boolean()) │ ) => boolean │ type: 'string' │
│ │ │ }, { │
│ │ │ type: 'number' │
│ │ │ }], │
│ │ │ returns: { │
│ │ │ type: 'boolean' │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Promise( │ type T = Promise<string> │ const T = { │
│ Type.String() │ │ type: 'promise', │
│ ) │ │ item: { │
│ │ │ type: 'string' │
│ │ │ } │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Undefined() │ type T = undefined │ const T = { │
│ │ │ type: 'undefined' │
│ │ │ } │
│ │ │ │
├────────────────────────────────┼─────────────────────────────┼────────────────────────────────┤
│ const T = Type.Void() │ type T = void │ const T = { │
│ │ │ type: 'void' │
│ │ │ } │
│ │ │ │
└────────────────────────────────┴─────────────────────────────┴────────────────────────────────┘Strict
TypeBox includes the properties kind and modifier on each underlying schema. These properties are used to help TypeBox statically resolve the schemas to the appropriate TypeScript type as well as apply the appropriate modifiers to an objects properties (such as optional). These properties are not strictly valid JSON schema so in some cases it may be desirable to omit them. TypeBox provides a Type.Strict() function that will omit these properties if nessasary.
const T = Type.Object({ // const T = {
name: Type.Optional(Type.String()) // kind: Symbol(ObjectKind),
}) // type: 'object',
// properties: {
// name: {
// kind: Symbol(StringKind),
// type: 'string',
// modifier: Symbol(OptionalModifier)
// }
// }
// }
const U = Type.Strict(T) // const U = {
// type: 'object',
// properties: {
// name: {
// type: 'string'
// }
// }
// }Validation
TypeBox does not provide JSON schema validation functionality, so users will need to select an appropriate JSON Schema validator for their language or framework. TypeBox targets JSON Schema draft 2019-09 so any validator capable of draft 2019-09 should be fine. A good library to use for validation in JavaScript environments is AJV. The following example shows setting up AJV 7 to work with TypeBox.
$ npm install ajv ajv-formats --save//--------------------------------------------------------------------------------------------
//
// Import the 2019 compliant validator from AJV
//
//--------------------------------------------------------------------------------------------
import { Type } from '@sinclair/typebox'
import addFormats from 'ajv-formats'
import Ajv from 'ajv/dist/2019'
//--------------------------------------------------------------------------------------------
//
// Setup AJV validator with the following options and formats
//
//--------------------------------------------------------------------------------------------
const ajv = addFormats(new Ajv({}), [
'date-time',
'time',
'date',
'email',
'hostname',
'ipv4',
'ipv6',
'uri',
'uri-reference',
'uuid',
'uri-template',
'json-pointer',
'relative-json-pointer',
'regex'
]).addKeyword('kind')
.addKeyword('modifier')
.addKeyword('discriminator')
//--------------------------------------------------------------------------------------------
//
// Create a TypeBox type
//
//--------------------------------------------------------------------------------------------
const User = Type.Object({
userId: Type.String({ format: 'uuid' }),
email: Type.String({ format: 'email' }),
online: Type.Boolean(),
}, { additionalProperties: false })
//--------------------------------------------------------------------------------------------
//
// Validate Data
//
//--------------------------------------------------------------------------------------------
const ok = ajv.validate(User, {
userId: '68b4b1d8-0db6-468d-b551-02069a692044',
email: 'dave@domain.com',
online: true
}) // -> okReference Types
Referenced types can be added to AJV with the ajv.addSchema(...) function. The following moves the userId and email property types into a Type.Box(...) and registers the box with AJV.
//--------------------------------------------------------------------------------------------
//
// Shared Types
//
//--------------------------------------------------------------------------------------------
const Shared = Type.Box({
UserId: Type.String({ format: 'uuid' }),
Email: Type.String({ format: 'email' })
}, { $id: 'Shared' })
//--------------------------------------------------------------------------------------------
//
// Setup Validator and Register Shared Types
//
//--------------------------------------------------------------------------------------------
const ajv = addFormats(new Ajv({}), [...])
.addKeyword('kind')
.addKeyword('modifier')
.addKeyword('discriminator')
.addSchema(Shared) // <-- Register Shared Types
//--------------------------------------------------------------------------------------------
//
// Create a TypeBox type
//
//--------------------------------------------------------------------------------------------
const User = Type.Object({
userId: Type.Ref(Shared, 'UserId'),
email: Type.Ref(Shared, 'Email'),
online: Type.Boolean()
}, { additionalProperties: false })
//--------------------------------------------------------------------------------------------
//
// Validate Data
//
//--------------------------------------------------------------------------------------------
const ok = ajv.validate(User, {
userId: '68b4b1d8-0db6-468d-b551-02069a692044',
email: 'dave@domain.com',
online: true
}) // -> ok
Please refer to the official AJV documentation for additional information.
OpenAPI
TypeBox can be used to create schemas for OpenAPI, however users should be aware of the various differences between the JSON Schema and OpenAPI specifications. Two common instances where OpenAPI diverges from the JSON Schema specification is OpenAPI's handling of string enum and nullable. The following shows how you can use TypeBox to construct these types.
import { Type, Static, TNull, TLiteral, TUnion, TSchema } from '@sinclair/typebox'
//--------------------------------------------------------------------------------------------
//
// Nullable<T>
//
//--------------------------------------------------------------------------------------------
function Nullable<T extends TSchema>(schema: T): TUnion<[T, TNull]> {
return { ...schema, nullable: true } as any
}
const T = Nullable(Type.String()) // const T = {
// type: 'string',
// nullable: true
// }
type T = Static<typeof T> // type T = string | null
//--------------------------------------------------------------------------------------------
//
// StringUnion<[...]>
//
//--------------------------------------------------------------------------------------------
type IntoStringUnion<T> = {[K in keyof T]: T[K] extends string ? TLiteral<T[K]>: never }
function StringUnion<T extends string[]>(values: [...T]): TUnion<IntoStringUnion<T>> {
return { enum: values } as any
}
const T = StringUnion(['A', 'B', 'C']) // const T = {
// enum: ['A', 'B', 'C']
// }
type T = Static<typeof T> // type T = 'A' | 'B' | 'C'