@hugorodriguesqw/mini-schema

schema.js

A proof of concept schema validator for javascript. It's quite usual, so much so that I'm using it as a validator for my neural network setup.

#Instalation

• using npm:

npm install @hugorodriguesqw/mini-schema

• using yarn:

yarn install @hugorodriguesqw/mini-schema

• using unpkg:

<script src="https://unpkg.com/@hugorodriguesqw/mini-schema@latest/schema.js"></script>

#Usage

import {Schema, SchemaItem} from '@hugorodriguesqw/mini-schema'

const schema = new Schema({ ... SchemaItem ... })

schema.validate({ data }).then().catch()

#Schema Syntax

Each item within the schema must be wrapped with a SchemaItem and configured using the parameters found here individually.

const schema = new Schema({
  age: new SchemaItem({ type: Number }),
  weight: new SchemaItem({ type: Number }),
  name: new SchemaItem({ type: String }),
  address: new SchemaItem({ type: String }),
})

schema.validate({
  age: 21,
  weight: 64.3,
  name: 'Hugo Rodrigues',
  address: 'somewhere',
})

#SchemaItem Configuration

Each schematic item has its own configuration. It is passed in the SchemaItem constructor to generate a valid schematic item. Follow the valid settings below:

parameter	type	defaults	required	description
type	Any	-	false	Type of field (Number, Object, Classes, Etc)
required	Boolean	true	false	required field
instance	Boolean	false	false	validate as instance
custom	Function	-	false	custom additional validate function
defaults	Any	-	false	default value (same type)
children	Object	-	false	Children of field (object or array)

#Item. Type

It defines the typing of the item in question. Generally, primitive constructors are used to validate them like Number, String, Symbol, Object, etc. Custom classes can also be used as per this example.

const schema = new Schema({
  age: new SchemaItem({ type: Number }),
  person: new SchemaItem({ type: Person }),
})

schema.validate({
  age: 21,
  person: Person,
}) // valid! return {age: 21, person: Person}

#Item. Instance

This changes the behavior of the Schema to validate instances instead of the original constructor.

const schema = new Schema({
  me: new SchemaItem({type: Person, instance: true})
})

schema.validate({
  me: new Person({....})
}) // valid! return {me: #person }

#Item. Required

By default this parameter is true. It defines whether the field is required or not.

const schema = new Schema({
  age: new SchemaItem({ type: Number }),
  anything: new SchemaItem({ required: false }),
})

schema.validate({ age: 21 }) // valid! return {age: 21}

#Item. Defautls

If there is no value in the field, it will be replaced by defaults. It is important to note that it must match the type field. This parameter only applies to fields required=false

const schema = new Schema({
  age: new SchemaItem({ type: Number }),
  name: new SchemaItem({ type: String, required: false, defaults: 'John Doe' }),
})

schema.validate({ age: 21 }) // valid! return {age: 21, name: 'John Doe'}

#Item. Custom

An additional function to check, in addition to the informed parameters, if the field is valid.

function customValidation (age) {
  return age < 20 // boolean
}

const schema = new Schema({
  age: new SchemaItem({type: Number, custom: customValidation ),
})

schema.validate({ age: 21 }) // invalid!

// catch output:

{
   "validator": {
       "name": "age",
       "required": true,
       "instance": false,
       "type": Number,
       "custom": customValidation
   },
   "parameter": 21,
   "reason": "custom"
}

#Item. Children

If it is an object, the child elements must be declared inside the children key of SchemaItem. It must be declared as an object and a new schema must be generated to process it. This allows you to validate child properties, rather than just validating the parent as an Object.

const schema = new Schema({
  address: new SchemaItem({
      type: Object,
      children: {
        number: new SchemaItem({type: Number}),
        street: new SchemaItem({type: String})
      }
  }),
})

schema.validate({
  address: {
    number: 138,
    street: 'Adelaide Street'
  }
}) // valid!

// then output:

{
   "address": {
       "number": 138,
       "street": 'Adelaide Street',
    }
}