Skip to main content

farrow-schema

farrow-schema is a powerful and extensible schema builder library.

Installation#

# via npm
npm install --save farrow-schema
# via yarn
yarn add farrow-schema

API#

import {
Schema, // abstract class inherited by all schema
List, // List type constructor
ObjectType, // abstract class of object type schema
Int, // Int type
Float, // Float type
ID, // ID type
Struct, // Struct type constructor
Nullable, // Nullabel type constructor
Union, // Union type constructor
Intersect, // Intersection type constructor
Literal, // Literal type constructor
Record, // Record type constructor
Json, // Json type
Any, // Any type
Strict, // Strict type constructor
NonStrict, // NonStrict type constructor
Tuple, // Tuple type constrcutor
pick, // pick object/struct keys to create a new object/struct type
omit, // omit object/struct keys to create a new object/struct type
keyof, // get the keys of object/struct
} from 'farrow-schema'
// create transformer for transforming schema to another type
import { createTransformer } from 'farrow-schema/transformer'
// schema validator
import { Validator } from 'farrow-schema/validator'

Usage#

import * as Schema from 'farrow-schema'
import { Validator } from 'farrow-schema/validator'
const { ObjectType, List, ID, Float, Nullable, Struct, Union, Intersect, Literal, Json, Any, Record } = Schema
// define User Object, it supports recursive definition
class User extends ObjectType {
id = ID
name = String
orders = List(Order) // order list type
}
// define Order Object
class Order extends ObjectType {
id = ID
product = Product
user = User
}
// define Product Object
class Product extends ObjectType {
id = ID
title = String
description = String
price = Float
}
// define AppState Object
class AppState extends ObjectType {
descriptors = {
a: Boolean,
// a light way to construct struct type
b: Struct({
c: {
d: List(Nullable(String)),
},
}),
}
struct = Struct({
a: Number,
b: String,
c: {
deep: {
d: List(Boolean),
},
},
})
nullable = Nullable(List(Number))
union = Union(List(Number), List(String), List(Boolean))
intersect = Intersect(Struct({ a: String }), Struct({ b: Boolean }))
record = Record(Product)
literal = Literal(12)
json = Json
any = Any
getUser = User
getOrder = Order
// supports { [Schema.Type]: SchemaCtor }
getProduct = {
[Schema.Type]: Product,
description: 'get product',
}
}
type T0 = Schema.TypeOf<AppState>
type T1 = Schema.TypeOf<User>
type T2 = Schema.TypeOf<Product>
const result0 = Validator.validate(Product, {
id : 'product id'
title : 'product title'
description : 'product description'
price : 1000.1
})
if (result0.isOk) {
console.log(result0.value)
}

ValidatorType#

it's useful to build your own validator-type with custom validate function.

import { ValidatorType } from 'farrow-schema/validator'
class DateType extends ValidatorType<Date> {
validate(input: unknown) {
if (input instanceof Date) {
return this.Ok(input)
}
if (typeof input === 'number' || typeof input === 'string') {
return this.Ok(new Date(input))
}
return this.Err(`${input} is not a valid date`)
}
}
class EmailType extends ValidatorType<string> {
validate(input: unknown) {
if (typeof input !== 'string') {
return this.Err(`${input} should be a string`)
}
if (/^example@farrow\.com$/.test(input)) {
return this.Ok(input)
}
return this.Err(`${input} is not a valid email`)
}
}

RegExp#

Given a regexp for creating a validator-type

import { RegExp, createSchemaValidator } from 'farrow-schema/validator'
let Reg0 = RegExp(/123/)
let Reg1 = RegExp(/abc/i)
let validateReg0 = createSchemaValidator(Reg0)
let validateReg1 = createSchemaValidator(Reg1)
expect(assertOk(validateReg0('123'))).toBe('123')
expect(() => assertOk(validateReg0('12'))).toThrow()
expect(assertOk(validateReg1('abc'))).toBe('abc')
expect(assertOk(validateReg1('ABC'))).toBe('ABC')
expect(() => assertOk(validateReg1('cba'))).toThrow()