ajv-ts

Table of Contents

• ajv-ts
  • Table of Contents
  • Zod unsupported APIs/differences
  • Installation
  • Basic usage
  • Base schema
    • examples
    • custom
    • meta
  • JSON schema overriding
  • Defaults
  • Primitives
  • Constant values(literals)
  • String
    • Typescript features
  • Numbers
    • Types
      • Number
      • Int
    • Formats
      • int32
      • int64
      • float
      • double
    • Typescript features
  • BigInts
  • NaNs
  • Dates
  • Enums
    • Autocompletion
  • Native enums
  • Optionals
  • Nullables
  • Objects
    • .keyof
    • .extend
    • .merge
    • .pick/.omit
    • .partial
    • .required
    • .requiredFor
    • .partialFor
    • .passthrough
    • .strict
    • .dependentRequired
    • .rest
  • Arrays
    • .addItems
    • .element
    • .nonempty
    • .min/.max/.length/.minLength/.maxLength
    • Typescript features
    • .unique
    • .contains/.minContains
  • Tuples
  • unions/or
  • Intersections/and
  • Set
  • Map
  • any/unknown
  • never
  • not/exclude
  • Custom Ajv instance
  • custom shema definition
  • Transformations
    • Preprocess
    • Postprocess
  • Error handling
    • Error Map
    • refine

JSON schema builder like in ZOD-like API

TypeScript schema validation with static type inference!

Reasons to install ajv-ts instead of zod

1. Less code. zod has 4k+ lines of code
2. not JSON-schema compatibility out of box (but you can install some additional plugins)
3. we not use own parser, just ajv, which wild spreadable(90M week installations for ajv vs 5M for zod)
4. Same typescript types and API
5. You can inject own ajv instance!

We inspired API from zod. So you just can reimport you api and that's it!

Zod unsupported APIs/differences

1. s.date, s.symbol, s.void, s.void, s.bigint, s.function does not supported. Since JSON-schema doesn't define Date, Symbol, void, function, Set, Map as separate type. For strings you can use s.string().format('date-time') or other JSON-string format compatibility: https://json-schema.org/understanding-json-schema/reference/string.html
2. s.null === s.undefined - same types, but helps typescript with autocompletion
3. z.enum and z.nativeEnum it's a same as s.enum. We make enums fully compatible, it can be array of strings or structure defined with enum keyword in typescript
4. Exporting s isntead of z, since s - is a shorthand for schema
5. z.custom is not supported
6. z.literal === s.const.

Installation

npm install ajv-ts       # npm
yarn add ajv-ts          # yarn
bun add ajv-ts           # bun
pnpm add ajv-ts          # pnpm

Basic usage

Creating a simple string schema

import { s } from "ajv-ts";

// creating a schema for strings
const mySchema = s.string();

// parsing
mySchema.parse("tuna"); // => "tuna"
mySchema.parse(12); // => throws Ajv Error

// "safe" parsing (doesn't throw error if validation fails)
mySchema.safeParse("tuna"); // => { success: true; data: "tuna" }
mySchema.safeParse(12); // => { success: false; error: AjvError }

Creating an object schema

import { s } from "ajv-ts";

const User = s.object({
  username: s.string(),
});

User.parse({ username: "Ludwig" });

// extract the inferred type
type User = s.infer<typeof User>;
// { username: string }

Base schema

Every schema inherits these class with next methods/properties

examples

The examples keyword is a place to provide an array of examples that validate against the schema. This isn’t used for validation, but may help with explaining the effect and purpose of the schema to a reader. Each entry should validate against the schema in which it resides, but that isn’t strictly required. There is no need to duplicate the default value in the examples array, since default will be treated as another example.

Note: While it is recommended that the examples validate against the subschema they are defined in, this requirement is not strictly enforced.

Used to demonstrate how data should conform to the schema. examples does not affect data validation but serves as an informative annotation.

s.string().examples(["str1", 'string 2']) // OK
s.number().examples(["str1", 'string 2']) // Error
s.number().examples([1, 2, 3]) // OK
s.number().examples(1, 2, 3) // OK

custom

Add custom schema key-value definition.

set custom JSON-schema field. Useful if you need to declare something but no api founded for built-in solution.

Example: If-Then-Else you cannot declare without custom method.

const myObj = s.object({
 foo: s.string(),
 bar: s.string()
}).custom('if', {
 "properties": {
   "foo": { "const": "bar" }
 },
 "required": ["foo"]
 }).custom('then', { "required": ["bar"] })

meta

Adds meta information fields in your schema, such as deprecated, description, $id, title and more!

Example:

const numSchema = s.number().meta({
  title: 'my number schema',
  description: 'Some description',
  deprecated: true
})

numSchema.schema // {type: 'number', title: 'my number schema', description: 'Some description', deprecated: true }

JSON schema overriding

In case of you have alredy defined JSON-schema, you create an any/object/number/string/boolean/null schema and set schema property from your schema.

Example:

import s from 'ajv-ts'

const SchemaFromSomewhere = {
 "title": "Example Schema",
  "type": "object",
  "properties": {
    "name": {
      "type": "string"
    },
    "age": {
      "description": "Age in years",
      "type": "integer",
      "minimum": 0
    },
  },
  "required": ["name", "age"]
}

type MySchema = {
  name: string;
  age: number
}

const AnySchema = s.any()
AnySchema.schema = SchemaFromSomewhere

AnySchema.parse({name: 'hello', age: 18}) // OK, since we override JSON-schema

// or using object
const Obj = s.object<MySchema>()
Obj.schema = SchemaFromSomewhere

Obj.parse({name: 'hello', age: 18}) // OK

Defaults

Option default keywords throws exception during schema compilation when used in:

• not in properties or items subschemas
• in schemas inside anyOf, oneOf and not (#42)
• in if schema
• in schemas generated by user-defined macro keywords.

This means only object() and array() buidlers are supported.

Example

import s from 'ajv-ts'
const Person = s.object({
  age: s.int().default(18)
})
Person.parse({}) // { age: 18 }

Primitives

import { s } from "ajv-ts";

// primitive values
s.string();
s.number();
s.boolean();

// empty types
s.undefined();
s.null();

// allows any value
s.any();
s.unknown();

Constant values(literals)

const tuna = s.const("tuna");
const twelve = s.const(12);
const tru = s.const(true);

// retrieve literal value
tuna.value; // "tuna"

String

includes a handful of string-specific validations.

// validations
s.string().maxLength(5);
s.string().minLength(5);
s.string().length(5);
s.string().format('email');
s.string().format('url');
s.string().regex(regex);
s.string().format('date-time');
s.string().format('ipv4');

// transformations
s.string().postprocess(v => v.trim());
s.string().postprocess(v => v.toLowerCase());
s.string().postprocess(v => v.toUpperCase());

Typescript features

from >=0.7.x

Unlike zod - we make typescript validation for minLength and maxLength. That means you cannot create schema when expected length are negative number or maxLength < minLength.

Here is few examples:

s.string().minLength(3).maxLength(1) // [never, "RangeError: MaxLength less than MinLength", "MinLength: 3", "MaxLength: 1"]

s.string().length(-1) // [never, "TypeError: expected positive integer. Received: '-1'"]

Numbers

includes a handful of number-specific validations.

s.number().gt(5);
s.number().gte(5); // alias .min(5)
s.number().lt(5);
s.number().lte(5); // alias .max(5)

s.number().int(); // value must be an integer

s.number().positive(); //     > 0
s.number().nonnegative(); //  >= 0
s.number().negative(); //     < 0
s.number().nonpositive(); //  <= 0

s.number().multipleOf(5); // Evenly divisible by 5. Alias .step(5)

Types

Number

Number - any number type

s.number()
// same as
s.number().number()

Int

Only integers values.

Note: we check in runtime non-integer format (float, double) and give an error.

s.number().int()
// or
s.number().integer()
// or
s.int()

Formats

Defines in ajv-formats package

int32

Signed 32 bits integer according to the openApi 3.0.0 specification

int64

Signed 64 bits according to the openApi 3.0.0 specification

float

float: float according to the openApi 3.0.0 specification

double

double: double according to the openApi 3.0.0 specification

Typescript features

from >= 0.8

We make validation for number type, format, minValue and maxValue fields. That means we handle it in our side so you get an error for invalid values.

Examples:

s.number().format('float').int() // error in type!
s.int().const(3.4) // error in type!
s.number().int().format('float') // error in format!
s.number().int().format('double') // error in format!

// ranges are also check for possibility

s.number().min(5).max(3) // error in range!
s.number().min(3).max(5).const(10) // error in constant!

BigInts

Not supported

NaNs

Not supported

Dates

Not supported, but you can pass parseDates in your AJV instance.

Enums

const FishEnum = s.enum(["Salmon", "Tuna", "Trout"]);
type FishEnum = s.infer<typeof FishEnum>;
// 'Salmon' | 'Tuna' | 'Trout'

const VALUES = ["Salmon", "Tuna", "Trout"] as const;
const FishEnum = s.enum(VALUES);

Autocompletion

To get autocompletion with a enum, use the .enum property of your schema:

FishEnum.enum.Salmon; // => autocompletes

FishEnum.enum;
/*
=> {
  Salmon: "Salmon",
  Tuna: "Tuna",
  Trout: "Trout",
}
*/

You can also retrieve the list of options as a tuple with the .options property:

FishEnum.options; // ["Salmon", "Tuna", "Trout"];

Native enums

Numeric enums:

enum Fruits {
  Apple,
  Banana,
}

const FruitEnum = s.enum(Fruits);
type FruitEnum = s.infer<typeof FruitEnum>; // Fruits

FruitEnum.parse(Fruits.Apple); // passes
FruitEnum.parse(Fruits.Banana); // passes
FruitEnum.parse(0); // passes
FruitEnum.parse(1); // passes
FruitEnum.parse(3); // fails

String enums:

enum Fruits {
  Apple = "apple",
  Banana = "banana",
  Cantaloupe, // you can mix numerical and string enums
}

const FruitEnum = s.enum(Fruits);
type FruitEnum = s.infer<typeof FruitEnum>; // Fruits

FruitEnum.parse(Fruits.Apple); // passes
FruitEnum.parse(Fruits.Cantaloupe); // passes
FruitEnum.parse("apple"); // passes
FruitEnum.parse("banana"); // passes
FruitEnum.parse(0); // passes
FruitEnum.parse("Cantaloupe"); // pass

Const enums:

The .enum() function works for as const objects as well. ⚠️ as const requires TypeScript 3.4+!

const Fruits = {
  Apple: "apple",
  Banana: "banana",
  Cantaloupe: 3,
} as const;

const FruitEnum = s.enum(Fruits);
type FruitEnum = s.infer<typeof FruitEnum>; // "apple" | "banana" | 3

FruitEnum.parse("apple"); // passes
FruitEnum.parse("banana"); // passes
FruitEnum.parse(3); // passes
FruitEnum.parse("Cantaloupe"); // fails

You can access the underlying object with the .enum property:

FruitEnum.enum.Apple; // "apple"

Optionals

You can make any schema optional with s.optional(). This wraps the schema in a Optional instance and returns the result.

const schema = s.string().optional();

schema.parse(undefined); // => returns undefined
type A = s.infer<typeof schema>; // string | undefined

Nullables

const nullableString = s.string().nullable();
nullableString.parse("asdf"); // => "asdf"
nullableString.parse(null); // => null
nullableString.parse(undefined); // throws error

Objects

// all properties are required by default
const Dog = s.object({
  name: s.string(),
  age: s.number(),
});

// extract the inferred type like this
type Dog = s.infer<typeof Dog>;

// equivalent to:
type Dog = {
  name: string;
  age: number;
};

.keyof

Use .keyof to create a Enum schema from the keys of an object schema.

const keySchema = Dog.keyof();
keySchema; // Enum<["name", "age"]>

.extend

You can add additional fields to an object schema with the .extend method.

const DogWithBreed = Dog.extend({
  breed: s.string(),
});

You can use .extend to overwrite fields! Be careful with this power!

.merge

Equivalent to A.extend(B.schema).

const BaseTeacher = s.object({ students: s.array(s.string()) });
const HasID = s.object({ id: s.string() });

const Teacher = BaseTeacher.merge(HasID);
type Teacher = s.infer<typeof Teacher>; // => { students: string[], id: string }

.pick/.omit

Inspired by TypeScript's built-in Pick and Omit utility types, all object schemas have .pick and .omit methods that return a modified version. Consider this Recipe schema:

const Recipe = s.object({
  id: s.string(),
  name: s.string(),
  ingredients: s.array(s.string()),
});

To only keep certain keys, use .pick .

const JustTheName = Recipe.pick({ name: true });
type JustTheName = s.infer<typeof JustTheName>;
// => { name: string }

To remove certain keys, use .omit .

const NoIDRecipe = Recipe.omit({ id: true });

type NoIDRecipe = s.infer<typeof NoIDRecipe>;
// => { name: string, ingredients: string[] }

.partial

Inspired by the built-in TypeScript utility type Partial, the .partial method makes all properties optional.

Starting from this object:

const user = s.object({
  email: s.string(),
  username: s.string(),
});
// { email: string; username: string }
We can create a partial version:

const partialUser = user.partial();
// { email?: string | undefined; username?: string | undefined }
You can also specify which properties to make optional:

const optionalEmail = user.partial({
  email: true,
});
/*
{
  email?: string | undefined;
  username: string
}
*/

.required

Contrary to the .partial method, the .required method makes all properties required.

Starting from this object:

const user = z
  .object({
    email: s.string(),
    username: s.string(),
  })
  .partial();
// { email?: string | undefined; username?: string | undefined }

We can create a required version:

const requiredUser = user.required();
// { email: string; username: string }
You can also specify which properties to make required:

const requiredEmail = user.required({
  email: true,
});
/*
{
  email: string;
  username?: string | undefined;
}
*/

.requiredFor

Accepts keys which are required. Set requiredProperties for your JSON-schema

const O = s.object({
  first: s.number().optional(),
  second: s.string().optional()
}).requiredFor('first')

type O = s.infer<typeof O> // {first: number, second?: string}

.partialFor

Accepts keys which are partial. unset properties from required schema field in your JSON-schema

const O = s.object({
  first: s.number().optional(),
  second: s.string().optional()
}).required().partialFor('second')

type O = s.infer<typeof O> // {first: number, second?: string}

.passthrough

By default object schemas strip out unrecognized keys during parsing.

const person = s.object({
  name: s.string(),
});

person.parse({
  name: "bob dylan",
  extraKey: 61,
});
// => { name: "bob dylan" }
// extraKey has been stripped

Instead, if you want to pass through unknown keys, use .passthrough() .

person.passthrough().parse({
  name: "bob dylan",
  extraKey: 61,
});
// => { name: "bob dylan", extraKey: 61 }

.strict

By default JSON object schemas allow to pass unrecognized keys during parsing. You can disallow unknown keys with .strict() . If there are any unknown keys in the input - will throw an error.

const person = z
  .object({
    name: s.string(),
  })
  .strict();

person.parse({
  name: "bob dylan",
  extraKey: 61,
});
// => throws ZodError

.dependentRequired

The dependentRequired keyword conditionally requires that certain properties must be present if a given property is present in an object. For example, suppose we have a schema representing a customer. If you have their "credit card number", you also want to ensure you have a "billing address". If you don't have their credit card number, a "billing address" operty on another using the dependentRequired keyword. The value of the dependentRequired keyword is an object. Each entry in the object maps from the name of a property, p, to an array of strings listing properties that are required if p is present.

const Test1 = s.object({
  name: s.string(),
  credit_card: s.number(),
  billing_address: s.string(),
  }).requiredFor('name').dependentRequired({
    credit_card: ['billing_address'],
  })

/**
Test1.schema === {
    "type": "object",
    "properties": {
      "name": { "type": "string" },
      "credit_card": { "type": "number" },
      "billing_address": { "type": "string" }
    },
    "required": ["name"],
    "dependentRequired": {
      "credit_card": ["billing_address"]
    }
  }
*/

.rest

The additionalProperties keyword is used to control the handling of extra stuff, that is, properties whose names are not listed in the properties keyword or match any of the regular expressions in the patternProperties keyword. By default any additional properties are allowed.

If you need to set additionalProperties=false use strict method

const Test = s.object({
  street_name: s.string(),
  street_type: s.enum(["Street", "Avenue", "Boulevard"])
}).rest(s.string())

Test.schema === {
  "type": "object",
  "properties": {
    "street_name": { "type": "string" },
    "street_type": { "enum": ["Street", "Avenue", "Boulevard"] }
  },
  "additionalProperties": { "type": "string" }
}

Arrays

const stringArray = s.array(s.string());
type StringArray = s.infer<typeof stringArray> // string[]

Or it's invariant

const stringArray = s.string().array();
type StringArray = s.infer<typeof stringArray> // string[]

Or you can pass empty schema

const empty = s.array()

type Empty = s.infer<typeof empty> // unknown[]

.addItems

push(append) schema to array(parent) schema.

Example:

import s from 'ajv-ts'

const empty = s.array()
const stringArr = empty.addItems(s.string())

stringArr.schema // {type: 'array', items: [{ type: 'string' }]}

.element

Use .element to access the schema for an element of the array.

stringArray.element; // => string schema, not array schema

.nonempty

If you want to ensure that an array contains at least one element, use .nonempty().

const nonEmptyStrings = s.array(s.string()).nonempty();
// the inferred type is now
// [string, ...string[]]

nonEmptyStrings.parse([]); // throws: "Array cannot be empty"
nonEmptyStrings.parse(["Ariana Grande"]); // passes

.min/.max/.length/.minLength/.maxLength

s.string().array().min(5); // must contain 5 or more items
s.string().array().max(5); // must contain 5 or fewer items
s.string().array().length(5); // must contain 5 items exactly

Unlike .nonempty() these methods do not change the inferred type.

Typescript features

from >=0.7.x

Unlike zod - we make typescript validation for minLength and maxLength. That means you cannot create schema when expected length are not positive number or maxLength < minLength.

Here is few examples:

s.string().array().minLength(3).maxLength(1) // [never, "RangeError: MaxLength less than MinLength", "MinLength: 2", "MaxLength: 1"]

s.string().array().length(-1) // [never, "TypeError: expected positive integer. Received: '-2'"]

.unique

Set the uniqueItems keyword to true.

const UniqueNumbers = s.array(s.number()).unique()

UniqueNumbers.parse([1,2,3,4]) // Ok
UniqueNumbers.parse([1,2,3,3]) // Error

.contains/.minContains

Tuples

Unlike arrays, tuples have a fixed number of elements and each element can have a different type.

const athleteSchema = s.tuple([
  s.string(), // name
  s.number(), // jersey number
  s.object({
    pointsScored: s.number(),
  }), // statistics
]);

type Athlete = s.infer<typeof athleteSchema>;
// type Athlete = [string, number, { pointsScored: number }]

A variadic ("rest") argument can be added with the .rest method.

const variadicTuple = s.tuple([s.string()]).rest(s.number());
const result = variadicTuple.parse(["hello", 1, 2, 3]);
// => [string, ...number[]];

unions/or

includes a built-in s.union method for composing "OR" types.

This function accepts array of schemas by spread argument.

const stringOrNumber = s.union(s.string(), s.number());

stringOrNumber.parse("foo"); // passes
stringOrNumber.parse(14); // passes

Or it's invariant - or function:

s.number().or(s.string()) // number | string

Intersections/and

Intersections are "logical AND" types. This is useful for intersecting two object types.

const Person = s.object({
  name: s.string(),
});

const Employee = s.object({
  role: s.string(),
});

const EmployedPerson = s.intersection(Person, Employee);

// equivalent to:
const EmployedPerson = Person.and(Employee);

// equivalent to:
const EmployedPerson = and(Person, Employee);

Though in many cases, it is recommended to use A.merge(B) to merge two objects. The .merge method returns a new Object instance, whereas A.and(B) returns a less useful Intersection instance that lacks common object methods like pick and omit.

const a = s.union(s.number(), s.string());
const b = s.union(s.number(), s.boolean());
const c = s.intersection(a, b);

type c = s.infer<typeof c>; // => number

Set

Not supported

Map

Not supported

any/unknown

Any and unknown defines {} (empty object) as JSON-schema. very useful if you need to create something specific

never

Never defines using {not: {}} (empty not). Any given json schema will be fails.

not/exclude

Here is a 2 differences between not and exclude.

• not method wrap given schema with not
• exclude(schema) - add not keyword for incoming schema argument

Example:

import s from 'ajv-ts'

// not
const notAString = s.string().not() // or s.not(s.string())

notAString.valid('random string') // false, this is a string
notAString.valid(123) // true

// exclude
const notJohn = s.string().exclude(s.const('John'))

notJohn.valid('random string') // true
notJohn.valid('John') // false, this is John

// advanced usage

const str = s.string<'John' | 'Mary'>().exclude(s.const('John'))
s.infer<typeof str> // 'Mary'

Custom Ajv instance

If you need to create a custom AJV Instance, you can use create or new function.

import addKeywords from 'ajv-keywords';
import schemaBuilder from 'ajv-ts'

const myAjvInstance = new Ajv({parseDate: true})

export const s = schemaBuilder.create(myAjvInstance)

// later:
s.string().dateTime().parse(new Date()) // 2023-10-05T19:31:57.610Z

custom shema definition

If you need to append something specific to you schema, you can use custom method.

const condition = s.any() // schema: {}

const withIf = condition.custom('if', {properties: {foo: {type: 'string'}}})

withIf.schema // {if: {properties: {foo: {type: 'string'}}}}

Transformations

Preprocess

function thant will be applied before calling parse method, It can helps you to modify incomining data

Be careful with this information

const ToString = s.string().preprocess(x => {
  if(x instanceof Date){
    return x.toISOString()
  }
  return x
}, s.string())

ToString.parse(12) // error: expects a string

ToString.parse(new Date()) // 2023-09-26T13:44:46.497Z

Postprocess

function thant will be applied after calling parse method.

const ToString = s.number().postprocess(x => String(x), s.string())

ToString.parse(12) // after parse we get "12" 12 => "12". 

ToString.parse({}) // error: expects number. Postprocess has not been called

Error handling

Error handling and error maps based from official package ajv-errors. You can check in out from there.

Defines custom error message for not valid schema.

const S1 = s.string().error('Im fails unexpected')
S1.parse({}) // throws: Im fails unexpected

Error Map

You can define custom error map. In most cases you can pass just a string for invalidation. Also, you can pass error map.

Example:

import s from 'ajv-ts'

const s1 = s.string().error({ _: "any error here" })

s.parse(123) // throws "any error here"

s.string().error({ _: "any error here", type: "not a string. Custom" })

s.parse(123) // throws "not a string. Custom"

const Obj = s
  .object({ foo: s.string(),})
  .strict()
  .error({ additionalProperties: "Not expected to pass additional props" });

Obj.parse({foo: 'ok', bar: true}) // throws "Not expected to pass additional props"

const Schema = s
    .object({
      foo: s.integer().minimum(2),
      bar: s.string().minLength(2),
    })
    .strict()
    .error({
      properties: {
        foo: "data.foo should be integer >= 2",
        bar: "data.bar should be string with length >= 2",
      },
    });
Schema.parse({ foo: 1, bar: "a" }) // throws: "data.foo should be integer >= 2"

refine

Inspired from zod. Set custom validation. Any result exept undefined will throws(or exposed for safeParse method).

import s from 'ajv-ts'
// example: object with only 1 "active element"
const Schema = s.object({
active: s.boolean(),
name: s.string()
}).array().refine((arr) => {
  const subArr = arr.filter(el => el.active === true)
  if (subArr.length > 1) throw new Error('Array should contains only 1 "active" element')
})

Schema.parse([{ active: true, name: 'some 1' }, { active: true, name: 'some 2' }]) // throws Error