Skeletons

• Skeletons is a pure javascript library that helps you validate data structure in programming/tesing.

• Define your rules, validate data sources and parameters/arguments to make sure they meet your expectations.

Usage

You might bother with data validation.

(notice: there were some mistakes in Readme examples and were fixed at version 0.0.6)

if(typeof data!=='object'||data===null) throw 'options must be an object'
if(!data.name) throw '...'
if(!data.id) throw '...'
if(typeof data.age === 'number' && !isNaN(data.age)) { /* ... */ }
//....

Skeletons provide an intuitive way to define a rule that makes sure all data meet your expectations, lets you focus on other things in programming.

const Skeletons = require('skeletons')
//schema
const user_schema = {
  name: String,
  id: Skeletons.String({
    validator: (val)=>val.length===7
  }),
  friends: Skeletons.Array({
    item: {
      name: String,
      id: Skeletons.String({
        validator: (val)=>val.length===7
      })
    }
  }),
  age: Skeletons.Number({
    allowNaN: false
  }),
  grownup: Skeletons.Boolean({
    validator: (val, source) => val === (source.age>=18)
  })
}
//data
const data = {
  name: 'Tim',
  id: 'djfo1k3',
  friends: [
    {
      name: 'Alex',
      id: 'zkfap1d',
    },
    {
      name: 'Leo',
    }
  ],
  age: 17,
  grownup: true
}
 
// define rule
const rule = new Skeletons(user_schema)
 
//validate data
rule.validate(data)
/*
 Skeletons Warn: [Unexpected Type] at data['friends'][1]['id']: expect [string] but got [undefined]
 Skeletons Warn: [Value invalid] at data['grownup']: validation failed
*/
rule.valid //false

Version

• 0.1.0
  • Add es module support

Installation

nodejs

as a dependency or devDependency

npm i --save skeletons
npm i -D skeletons

//CommonJs
const Skeletons = require('skeletons')
 
// ES Module
import Skeletons from 'skeletons'

browser

the script for browsers at /dist/skeletons.min.js

unpkg cdn : https://unpkg.com/skeletons@:version/dist/skeletons.min.js

Document

• document
• static function
• Some examples

Must Know

null

In javascript, null is a primitive type, however, typeof null is

typeof null // 'object'

Skeletons use Skeletons.typeof instead of typeof to check types.

When defining schema to be an object but got null, validation will fail.

To define a null value schema, use Skeletons.Null().

array

In javascript, an array is also an object

typeof [] // 'object'

However, Skeletons will distinguish the array from other objects.

To define an array schema, use Skeletons.Array() or array literal schema.

function

typeof function(){} // 'function'

Although typeof function return 'function', it is worth mentioning that function is also a Function Object, however, Skeletons will distinguish the function from the object.

So you cannot define an 'object' schema but got 'function' at data.

To define a function schema, use Skeletons.Function().

undefined

For every undefined value, validation will fail.

To allow undefined value, see options.required.

Define Schema

Passing a schema to create a rule:

let rule_ex1 = new Skeletons(schema)
 
//or assign schema later
let rule_ex2 = new Skeletons()
rule_ex2.schema = schema

There are four types of schema

• premitive types function
• object literal
• array literal (version 0.0.8)
• call Skeletons static function

premitive types function

• String: define a string
• Number: define a number
• Boolean: define a boolean
• Symbol: define a symbol

new Skeletons(Number).validate(5)

Only premitive types function, do not use other functions like Object, Array.

object literal

use object literal to define keys/values and deeper layers.

{
  key1: <schema>,
  key2: {
    key21: <schema>
    key22: {
      key221: <schema>
    }
  }
}

This type of schema defines an object that has exactly keys.

Validation for data that missing keys or has extra keys will fail.

const rule = new Skeletons({
  x: Number,
  y: Number,
})
// missing keys/properties
rule.validate({
  x: 1,
})
//Skeletons Warn: [Unexpected Type] at data['y']: expect [number] but got [undefined]
 
// has extra keys/properties
rule.validate({
  x: 1,
  y: 2,
  z: 3,
})
//Skeletons Warn: [Unknown Property] at data : property 'z' not defined in schema

To allow a undefined property, see options.required.

To allow dynamic keys in object, use MapObject.

To check keys that are defined and also ignore keys that are not defined, see Skeletons.Object : set options.extraKey to true.

array literal

To define an array that has exact numbers of elements.

[<schema>,<schema>,<schema>...]

Here's example

To defined an array that has repeated elements, use Skeletons.Array().

call Skeletons static function

These functions allow you to pass options to define a flexible rule.

• Skeletons.Number()
• Skeletons.String()
• Skeletons.Boolean()
• Skeletons.Null()
• Skeletons.Any()
• Skeletons.Array()
• Skeletons.Object()
• Skeletons.Function()
• Skeletons.MapObject()
• Skeletons.Symbol()

See Skeletons static function for more options you can use.

Validation

create rule & validate

After new a rule member, call method validate(data) to validate data

let schema = Boolean
let rule = new Skeletons(schema)
 
rule.validate(1) //this also return rule itself

set options

set default options to rule.default

let rule = new Skeletons(Boolean, {
  console: true, //use console to show validate warnings
  throw: false, //throw validate warning
  dataName: 'datasource', //data name show in warning message
  schemName: 'mySchema' //schema name show in warning message
})
.validate(1)
//Skeletons Warn: [Unexpected Type] at datasource : expect [boolean] but got [undefined]

set options only for this validation, if not set, use rule.default as default options

rule.validate(1 ,{
  console: false,
  throw: false,
  dataName: 'datasource'
  schemName: 'mySchema'
})

If data isn't valid, skeletons will show the warning message in the console or throw error depends on your setting.

However, if there's a schema problem (defined a wrong schema), skeletons will always throw an error.

notice : Skeletons will not discover schema erorrs until rule.validate() is called.

rule.valid

rule.valid will set to true or false after every validation.

rule.validate(1)
rule.valid //false
rule.validate(false)
rule.valid //true

rule.warnings

After rule.validate(), rule.warnings will be an array contains informations about invalid data.

for example in Usage, warnings should look like below :

[
  {
    code: 0, //code to identify warning type
    log: 'expect [string] but got [undefined]', // message
    type: '[Unexpected Type]',
    depth: [ 'friends', 1, 'id' ] //object keys show where validation fail:
    // at data['friends'][1]['id']
  },
  {
    code: 2,
    log: 'validation failed',
    type: '[Value invalid]',
    depth: [ 'grownup' ]
  }
]

more about warnings and Document.