Barbarian Meets Coding
barbarianmeetscoding

WebDev, UX & a Pinch of Fantasy

oopjavascriptc-sharp

Object-oriented JavaScript for C# Developers

The Mastering the Arcane Art of JavaScript-mancy series are my humble attempt at bringing my love for JavaScript to all other C# developers that haven’t yet discovered how awesome this language and its whole ecosystem are. These articles are excerpts of the super duper awesome JavaScript-Mancy book a compendium of all things JavaScript for C# developers.

Through the last couple of years I have been working on improving my JavaScript-Fu. From looking down at JavaScript with contempt, I have come to love the language and its little idiosyncrasies. In this article, I will attempt to write the article I would have loved to read, the first time I set my mind to re-learn JavaScript the proper way.

I Love JavaScript

Ey You C# developer! Wanting to learn you some JavaScript? You’ve come to the right place! In this article you will learn how to map all those OOP C# concepts you have come to know and love to JavaScript, the ever growing language of the web! I will help you traverse the waters of Object-oriented programming by going through the pillars of OOP (encapsulation, inheritance and polymorphism) and trying to work out how they can be achieved in JavaScript.

Note: In this article I will use two types of code samples. Some that you can type in a text editor, and others that you can run directly in any JavaScript REPL (in the Chrome Dev Tools console for instance). In these second type of examples every line will be preceded by a > followed by the result of evaluating an statement, and that’s how you will be able to differentiate them. I recommend the Chrome Dev Tools REPL (Console) because it will let you copy/paste even the first type - biggish - code examples on the REPL.

Encapsulation

Encapsulation is the packing of data and functions into a single component.

The easiest way to achieve encapsulation in JavaScript is by using object literals:

var conan = {
  name: 'Conan, The Barbarian',
  'character class': 'barbarian',
  hp: 200,
  weapons: [
    {
      name: 'two-handed sword',
      type: 'sword',
      damage: '2d20+10',
      material: 'cimmerian steel',
      status: 'well maintained',
      equipped: true,
    },
  ],
  talk: function() {
    console.log('I am ' + this.name + ' !!!')
  },
}

This will look familiar to you since we also have object literals in C#:

var conan = new Barbarian
    {
    Name = "Conan, The Barbarian",
    CharacterClass = "barbarian",
    Hp = 200,
    Weapons = new List<Weapon>
    {
        new Weapon {
            Name = "Two-handed sword",
            Type = "sword",
            ...
        }
    }
};

// Somewhere else we have the class definition
public class Barbarian{
    public string Name {get; set;}
    public string CharacterClass {get; set;}
    ...
}

However if you look closely at the JavaScript example you will notice that there’s no class identifier. What!? Welcome to the Jungle! There are no classes in JavaScript! (Although that may not hold true soon enough…).

An object literal in JavaScript is an easy way to create a single instance of an object with whichever arbitrary properties that you desire. All object literals and (all objects in JavaScript for that matter) are instances of the base Object type.

> var conan = {
    name: "conan"
};
undefined
> conan instanceof Object
true

Ok, so what happens if you want to create multiple similar objects? Do you need to instantiate them via literal objects one at a time? Nope, JavaScript provides Constructor Functions for just this purpose.

Constructor Functions

Constructor functions are JavaScript functions that allow us to create multiple instances which share the same properties. They work as a recipe for object creation and they represent a custom type (yep, this starts feeling like a class). Let’s see an example!:

function Barbarian(name) {
  this.name = name
  this['character class'] = 'barbarian'
  this.hp = 200
  this.weapons = []
  this.talk = function() {
    console.log('I am ' + this.name + ' !!!')
  }
  this.equipWeapon = function(weapon) {
    weapon.equipped = true
    this.weapons.push(weapon)
  }
}

Notice how the constructor, not only initializes an object with a set of values like in C#, but it also determines which properties an object is going to have. Effectively it works both as a constructor and a class definition (if we were to draw a parallelism with C#).

To create a new instance of a Barbarian we use the new keyword just like in C#:

var conan = new Barbarian('Conan, the Barbarian')
conan.equipWeapon({
  name: 'two-handed sword',
  type: 'sword',
  damage: '2d20+10',
  material: 'cimmerian steel',
  status: 'well maintained',
})

This new instance is of type Barbarian and all its properties are publicly available:

> var conan = new Barbarian("Conan, the Barbarian");
undefined
> conan instanceof Barbarian
true
> conan instanceof Object
true
> conan.talk();
I am Conan, the Barbarian!!!
> conan.name;
"Conan, The Barbarian"
> conan.weapons;
[]

Data/Information Hiding

Information hiding is the principle of segregation of the design decisions in a computer program that are most likely to change, thus protecting other parts of the program from extensive modification if the design decision is changed.

In the previous function constructor example we saw how all properties that we defined in the constructor function where publicly available. We may not want to have all our object implementation details out in the open, right? How do we then achieve private properties in JavaScript? We take advantage of closures, functions that refer to free variables from an outer scope.

We have closures in C# so I will not delve a lot on them but for this teeny tiny example:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

namespace ConsoleApplication1
{
    class Program
    {
        static void Main(string[] args)
        {
            var Add5 = AddX(5);
            var number = 0;
            Console.WriteLine(string.Format("Initial Number: {0}", number)); // Initial Number: 0
            number = Add5(number);
            Console.WriteLine(string.Format("Add 5 and it becomes: {0}", number)); // Add 5 and it becomes: 5
            number = Add5(number);
            Console.WriteLine(string.Format("Add 5 and it becomes: {0}", number)); // Add 5 and it becomes: 10

            var Add100 = AddX(100);
            number = Add100(number);
            Console.WriteLine(string.Format("Add 100 and it becomes: {0}", number)); // Add 100 and it becomes: 110
        }

        /// <summary>
        /// A function that creates another function that adds a given number X to another number
        ///
        /// The returned function is a closure over the free variable 'x'. As you can see in the
        /// example above, the created function retains the value of the variable it encloses.
        /// </summary>
        /// <param name="x"></param>
        /// <returns></returns>
        public static Func<int, int> AddX(int x)
        {
            return number => number + x;
        }

    }
}

In JavaScript we can create a closure like this:

function Barbarian(name) {
  // this variables are only available in the scope
  // of this function.
  // Remember, in JavaScript, only functions create new scopes!
  var weapons = [],
    characterClass = 'barbarian',
    hp = 200

  this.name = name
  this.talk = function() {
    console.log('I am ' + this.name + ' !!!')
  }
  this.equipWeapon = function(weapon) {
    weapon.equipped = true
    // Now we reference the weapons variable directly
    // creating a closure.
    // You cannot use this.weapons to access this variable any more
    // It is so private that the only way to access it is via closure
    weapons.push(weapon)
  }
}

Which makes the weapons, characterClass and hp variables effectively private: They are no longer accessible from the outside.

> var conan = new Barbarian("conan");
undefined
> conan.name;
"conan"
> conan.weapons;
undefined

Method Overloading

There is no such thing as method overloading in JavaScript, or at least method overloading as we know it in C#. If you define a method within an object twice, the second method will overwrite the first (ups!):

> var barbarian = {
    sayHi: function(){console.log("hi!";)},
    sayHi: function(){console.log("WaaaaZaaaa!");}
}
undefined
> barbarian.sayHi();
"WaaaaZaaaa!"

In fact, you can call any function with any number of arguments and JavaScript will just run with it:

> var barbarian = {
    sayHi: function(){console.log("hi!";)},
    sayHi: function(message, message2){console.log("hi! " + message + ", " + message2);}
}
undefined
> barbarian.sayHi();
"hi! undefined, undefined"
> barbarian.sayHi("I am Groot!");
"hi! I am Groot!, undefined"
> barbarian.sayHi("I am Groot!", "Waaaaa....")
"hi! I am Groot!, Waaaaa...."

How do we then provide support for method overloading in JavaScript? Well, we handle it in kind of a rudimentary way. Every function scope in JavaScript has a special variable called arguments, which behaves sort of like an array, but it is not an array, which contains all arguments passed to the function, and which can be used in the fashion below to mimic method overloading:

> var barbarian = {
    sayHi: function(){
        var args;
        if (arguments.length === 0) {
            console.log("hi!");
        }
        if (arguments.length === 1){
            console.log(arguments[0]);
        }
        if (arguments.length > 1){
            // this is like type casting the arguments array-like variable
            // to an actual array so that I can use the join method
            args = Array.prototype.slice.call(arguments);
            console.log("hi " + args.join(", "));
        }
    }
}
undefined
> barbarian.sayHi();
hi!
> barbarian.sayHi("I am Groot!");
I am Groot!
> barbarian.sayHi("I am Groot!", "Waaaaa....", "For Great Justice!")
hi! I am Groot!, Waaaaa...., For Great Justice!
  • Fun Fact: the length property of any function returns the number of arguments it expects (its arity). Isn’t that unexpected? XD

Object Augmentation

An interesting feature of JavaScript is that you can add or remove properties from objects at any point in time. This enables interesting object composition patterns like mixins.

You can add new properties to any object by using the . notation:

> var conan = { name : "conan"};
undefined
> conan.title = "the barbarian";
undefined
> conan.title
the barbarian
> conan.sayHi = function(){console.log("HI! I am " + this.name + ", " + this.title)};
undefined
> conan.sayHi()
Hi! I am conan, the barbarian

And remove properties by using the delete keyword:

> var conan = { name : "conan"};
undefined
> conan.name
conan
> delete conan.name
true // delete always returns true
> conan.name
undefined

Inheritance

In object-oriented programming (OOP), inheritance is when an object or class is based on another object or class, using the same implementation (inheriting from a class) or specifying implementation to maintain the same behavior (realizing an interface; inheriting behavior). It is a mechanism for code reuse and to allow independent extensions of the original software via public classes and interfaces.

[Inheritance](http://en.wikipedia.org/wiki/Inheritance_(object-oriented_programming) in JavaScript is slightly different than what we are accostumed to in C#. For one, there are no classes (ouch!), furthermore, inheritance doesn’t play as big a part in polymorphism since JavaScript is a dynamically typed language that relies in duck typing instead (as we will see in the next section). JavaScript is all about objects, and achieves inheritance not via class inheritance but via prototypical inheritance, that is, objects that inherit from other objects which act as prototypes (a sort of blueprint).

Prototypes

Every function in javascript has a prototype property that holds an object that will act as a prototype - will provide shared properties - for all objects created via that function and the new keyword. The prototype object has in its turn a constructor property that points back to the constructor function:

> function Barbarian(name) { this.name = name; }
undefined
> Barbarian.prototype
Barbarian {}
> Barbarian.prototype.constructor
function Barbarian(name) { this.name = name; }
> Barbarian.prototype = {}
undefined
> Barbarian.prototype

In the simplest inheritance scenario, you can create an inheritance hierarchy by setting the prototype property of a constructor function. By doing that, all objects instantiated using that constructor function will inherit properties and methods from the prototype object:

> function Barbarian(name){ this.name = name; }
undefined
> Barbarian.prototype.sayHi = function (){ console.log("Hi! I am " + this.name);}
undefined
> var krull = new Barbarian("krull");
undefined
> krull.sayHi()
Hi! I am krull

In a more common scenario you may have several custom prototypes in your inheritance hierarchy:

// Inheritance Hierarchy:
//   Barbarian -> DrawableGameObject -> GameObject -> Object

function GameObject() {
  // game object ctor
}
function DrawableGameObject() {
  // drawable object ctor
}

DrawableGameObject.prototype = Object.create(GameObject.prototype)
DrawableGameObject.constructor = DrawableGameObject

function Barbarian() {
  // barbarian ctor
}
Barbarian.prototype = Object.create(DrawableGameObject.prototype)
Barbarian.prototype.constructor = Barbarian
// copy paste code from the example of the Chrome Dev Tools console
> var conan = new Barbarian();
undefined
> conan instanceof Barbarian
true
> conan instanceof DrawableGameObject
true
> conan instanceof GameObject
true

It is important to note that each property defined in the constructor function will be part of every single object instance, whilst each property defined in the prototype will be shared amongst all object instances. Again:

  • constructor properties and methods are defined in each instance
  • prototype properties and methods are only defined in the prototype. This emulates static variables properties. Methods can still refer to a particular instance by using the this keyword.

A common idiom to avoid needing to write the Function.prototype.property each time you want to update the prototype is to assign the prototype to a new object:

> function Barbarian(name){ this.name = name; }
undefined
> Barbarian.prototype = {
  constructor: Barbarian
  sayHi: function(){console.log("Hi! I am " + this.name);}
}
undefined
> var conan = new Barbarian("conan");
undefined
> conan.sayHi()
Hi! I am conan
> conan instanceof Barbarian
true

In summary once more, because this was weird as hell the first time I saw it, to emulate a C# class you would need to create a constructor function that will contain the specific properties of each particular instance and a prototype that will provide a series of shared methods and static members that will be shared across of all instances of a given type.

A C# class is equivalent to a javascript constructor and prototype pair

The Prototype Chain

When we use prototype inheritance as a method of code reuse, to reduce duplication and optimize memory, a prototype chain is built between object and the different prototypes in the inheritance hierarchy. When we call a method on a particular object instance, the JavaScript runtime tries to find that method in the object itself and if it can’t find it, it follows the prototype chain until it does:

> function Barbarian(name){this.name = name;}
undefined
> Barbarian.prototype.ravage = function(){console.log(this.name + "attaaaaack!!");}
undefined
> var conan = new Barbarian("conan");
undefined
> conan isintanceof Barbarian
true
> Barbarian.prototype.isPrototypeOf(conan)
true
> conan.ravage === Barbarian.prototype.ravage
true
> conan isintanceof Object
true
> Object.prototype.isPrototypeOf(conan)
true
> conan.toString === Object.prototype.toString
true

In the example above, we can appreciate how within the conan object instance, the ravage method is part of the Barbarian.prototype and the toString method is part of the Object.prototype.

JavaScript prototype chain

Differentiating Between Object Properties and Prototype Properties

In order to know whether a property is part of the object itself or part of the prototype you can use the hasOwnProperty method:

> function Barbarian(name){this.name = name;}
undefined
> Barbarian.prototype.ravage = function(){console.log(this.name + "attaaaaack!!");}
undefined
> var conan = new Barbarian("conan");
undefined
> conan.hasOwnProperty("name")
true
> conan.hasOwnProperty("ravage")
false

Object Augmentation in Prototypes

In the same way we used object augmentation to add new properties to an object, we can use the same concept to add new properties to a prototype and, hence, make those properties available to any object that has that prototype:

> function Barbarian(name){this.name = name;}
undefined
> var conan = new Barbarian("conan");
undefined
> var logenNinefingers = new Barbarian("The Bloody Nine")
undefined
> var shivers = new Barbarian("Caul Shivers");
undefined
> shivers.ravage
undefined
> Barbarian.prototype.ravage = function(){console.log(this.name + " attaaaaack!!");}
> conan.ravage()
"conan attaaaaack!!"
> logenNinefingers.ravage()
"The Bloody Nine aattaaaaack!!"
> shivers.ravage()
"Caul Shivers attack!!"

And so, we can see how after adding the ravage method to the prototype, instances of Barbarian that had been created previously, suddenly and auto-ma-gi-ca-lly get a new method. :O

Method Overriding

In order to override a method in JavaScript you will need to provide a new implementation of the method in your constructor:

> function Barbarian(name){
    this.name = name;
}
undefined
> var conanTheBarbarian = new Barbarian("Conan The Barbarian");
undefined
> conanTheBarbarian.toString()
"[object Object]"
> function Barbarian(name){
    this.name = name;
    this.toString = function(){
        return "Barbarian: " + name;
    };
}
undefined
> var conanTheDestroyer = new Barbarian("Conan The Destroyer!!!");
undefined
> conan.toString()
"Barbarian: Conan The Destroyer!!!"

Or in one of the prototypes in the prototype chain:

> function Barbarian(name){
    this.name = name;
}
undefined
> var conan = new Barbarian("Conan The Barbarian");
undefined
> conan.toString()
"[object Object]"
> Barbarian.prototype.toString = function (){
        return "Barbarian: " + name;
};
undefined
> conan.toString()
"Barbarian: Conan The Destroyer!!!"

This is called shadowing just like in C#. Additionally, you can call your base prototype method by making use of JavaScript apply and call functions as illustrated in the example below:

// Create an inheritance hierarchy:
//   BarbarianKing -> Barbarian -> Object
function Barbarian() {
  this.title = 'the Scavenger from the Highlands'
}
Barbarian.prototype.warCry = function(message) {
  console.log(message)
  console.log('Waaaaaaaaaaahhhhh!')
}

function BarbarianKing() {
  this.title = 'the King of Thieves'
}
BarbarianKing.prototype = Object.create(Barbarian.prototype)
BarbarianKing.prototype.constructor = BarbarianKing

// Override method warCry and call method from super prototype
BarbarianKing.prototype.warCry = function(message) {
  Barbarian.prototype.warCry.call(this, message)
  console.log('I am the ' + this.title + '!!!!')
}
// execute previous file by copy-pasting on chrome dev tools console
> var conan = new Barbarian()
undefined
> conan.warCry("Good morning, how are you doing good sir? Now Die!")
Good morning, how are you doing good sir? Now Die!
Waaaaaaaaaaaaaaaaah!
> var kingConan = new BarbarianKing()
undefined
> kingConan.warCry("!!!!hhhhhhaaaaaaW")
!!!!hhhhhhaaaaaaW
Waaaaaaaaaaaaaaaaah!
I am the King of the Thieves!

If you are not familiar with the apply and call methods, you just need to know that they let you call a function and explicitly specify the object that this will refer to within the scope of the function:

> var conan = {
    name: "conan",
    sayHi: function(message){console.log(message + " " + this.name);}
}
undefined
> var logen = {
    name: "logen"
}
undefined
> conan.sayHi("hello I am")
hello I am conan
> conan.sayHi.call(logen, /* message */ "hello I am")
hello I am logen
> conan.sayHi.apply(logen, /*arguments*/["hello I am"])
hello I am logen
> conan.sayHi("hello I am")
hello I am conan

In the previous example you can see how, even though we have defined the sayHi method in the conan object, by using apply and call we can borrow the method and use it on a different object logen. For additional information, take a look at the awesome MDN docs on apply and call.

Emulating Classical Inheritance in JavaScript

Building on what we have learned so far, we can emulate classical inheritance as we known it in C# by:

  1. Making use of prototypical inheritance to ensure that any instance of an object inherits methods from its prototypes
  2. Making sure that each constructor function calls its base type constructor function using call or apply so that any instance of an object contains all properties defined in each and every constructor.

I’ll illustrate this through an example:

// Inheritance Hierarchy:
//   Barbarian -> DrawableGameObject -> MobileGameObject

function Position(x, y) {
  this.x = x
  this.y = y
}
Position.prototype.toString = function() {
  return '[' + this.x + ',' + this.y + ']'
}

function MobileGameObject(position) {
  this.position = position
}
MobileGameObject.prototype.move = function(newPosition) {
  console.log('moving ' + this + ' to ' + newPosition)
  this.position = newPosition
}

function DrawableGameObject(position, sprite) {
  // call base type constructor function
  MobileGameObject.apply(this, position)
  this.sprite = sprite
}
// establish prototypical inheritance
// between DrawableGameObject and MobileGameObject
DrawableGameObject.prototype = Object.create(MobileGameObject.prototype)
DrawableGameObject.prototype.constructor = DrawableGameObject
DrawableGameObject.prototype.draw = function() {
  console.log('drawing sprite: ' + this.sprite)
  // draw sprite
}

function Barbarian(name, position, sprite) {
  // call base type constructor function
  DrawableGameObject.call(this, position, sprite)
  this.name = name
}
// establish prototypical inheritance
// between Barbarian and DrawableGameObject
Barbarian.prototype = Object.create(DrawableGameObject.prototype)
Barbarian.prototype.constructor = Barbarian
Barbarian.prototype.toString = function() {
  return this.name + '!!!'
}
// copy paste code from the example of the Chrome Dev Tools console
> var conan = new Barbarian("Conan", new Position(0,0), "conan.jpg");
undefined
> conan.move(new Position(5,5))
moving Conan!!! to [5,5]
> conan.draw()
drawing sprite: conan.jpg

Polymorphism

_Polymorphism_ allows a function to be written to take an object of a certain type T, but also work correctly if passed an object that belongs to a type S that is a subtype of T

Polymorphism in JavaScript will be the least of your problems. As a dynamically typed language, JavaScript exhibits what is known as duck typing, whereby an object’s semantics are based on the object’s own methods and properties and not on the inheritance chain or interface implementations (like in C#).

But this is probably best illustrated through an example. Imagine a roguelike RPG, and a game scene with our mighty hero bravely venturing herself inside an eery dungeon, risking her life for adventure and profit, turning a corner into a host of ravenous ghouls…

function Character(name){
    this.name = "",
    this.draw = function(){ // draw a character }
}
function Zombie(){
    this.eatBrains = function(){ // mmmmm braaains }
    this.draw = function() { // draw a zombie }
}
function Skeleton(){
    this.creepyLaugh = function(){ // muhahahaha }
    this.draw = function() { // draw a skeleton}
}
// etc

var redSonja = new Character("red sonja");
var zombie = new Zombie();
var skeleton = new Skeleton();
// etc

var scene = {
    entities :[
        redSonja, ghouls, zombie, skeleton, orc, dragon, priest, warlock, walls, pottery, tapestries
    ],
    draw() : {
        entities.forEach(function(entity){
            entity.draw(); // duck typing style polymorphism right here
        });
    }
}

So, as long as an object has the draw method, it will behave as something that can be drawn regardless of inheritance. And thus the popular saying that you may have heard in the past…

If it walks like a duck, swims like a duck and quacks like a duck, I call it a duck.

Appendix A. Classes in JavaScript, Welcome ECMAScript 6

Somewhere above in this article I said something like ”… there are not such thing as classes in JavaScript…” and I was telling the truth. You may be interesting in knowing that classes are a part of the specification of the next version of JavaScript - ECMAScript6 Harmony.

class MobileGameObject {
  constructor(position) {
    this.position = position
  }
  move(newPosition) {
    // draw sprite
  }
}

class DrawableGameObject extends MobileGameObject {
  constructor(position, sprite) {
    super(position)
    this.sprite = sprite
  }
  draw() {
    // draw sprite
  }
}

class Barbarian extends DrawableGameObject {
  constructor(name, sprite) {
    super(new Position(0, 0), sprite)
    this.name = name
  }
  toString(message) {
    //...
  }
}

It looks awesome, doesn’t it? (Even though it is just syntactic sugar over all the concepts you have seen in this article).

For more information regarding ECMAScript 6 Harmony take a look at @lukehoban great summary at GitHub

Appendix B. A Note About this in JavaScript

To us C# developers used to the reliable value of this, JavaScript can be a little bit crazy and confusing. The only thing you need to know to be able to use this without fear in JavaScript is to forget how it works in C# and understand the following:

The this parameter refers to an object that’s implicitly associated with the function invocation as is termed the function context. (…)

(…)as it turns out, what this parameter points to isn’t, as in Java, defined by how the function is declared, but by how it is invoked.

So:

  • If a function is invoked as a method in the context of an object (e.g. conan.sayHi()) this will point to the object itself. (Like in C#)
  • If a function is invoked as a function (e.g. helloWorld()) this will be set to window. If you are using strict mode ("use strict";), this will be undefined saving you from unexpected errors.
  • If a function is invoked as a constructor with the new keyword then this will be a new object.
  • If a function is invoked with call or apply we can set the this to point at whatever our heart desires
> var conan = {whosThis : function(){console.log(this);}}
undefined
// 1) invoke as a method
> conan.whosThis()
Object {whosThis: function}
// 2) invoke as a function
> var whosThis = conan.whosThis
undefined
> whosThis()
Window {top: Window, window: Window, location: Location, external: Object, chrome: Object…}
// 3) invoke as a constructor
> var o = new whosThis();
conan.whosThis {}
// 4.a) invoke with call
> whosThis.call({a: "1", b: "2"})
Object {a: "a", b: "b"}
// 4.b) invoke with apply
> whosThis.apply({a: "1", b: "2"})
Object {a: "a", b: "b"}

For more detailed information about the uses of this in JavaScript visit the MDN documentation. In addition to the above, you can use the bind method to create functions that are for ever bound to a specific object.

In Closing

So finally! Now that you are able to master the unwieldy JavaScript OOP arcane and take advantage of all the freedom and craziness this beautiful language has to offer, go off and build something awesome!

Fare thee well friend! ^ _ ^

Interested in Learning More JavaScript? Buy the Book!

Are you a C# Developer interested in learning JavaScript? Then take a look at the JavaScript-mancy book, a complete compendium of JavaScript for C# developers.

a JavaScriptmancy sample cover

Have a great week ahead!! :)

More Articles In This Series

Other Awesome Resources

P.S.1. I had envisioned writing something more about higher level constructs to organize your code, and common design patterns, drawing parallels to the .NET world but I fear this article would have become undesirably long. I hope you will share my opinion and won’t be to hard on me if I leave it for a future article.

P.S.2. Still Not Liking JavaScript? Take a look at TypeScript then! :)

testingunit-testingtdd

Write 30% Less Test Code With AutoFixture

I have been following Mark Seemann for a while ‘cause this guy is as smart as it gets. He wrote a book on dependency injection, released two awesome courses on unit testing in Pluralsight, and has written the AutoFixture unit testing library (with many other contributors).

AutoFixture is a .NET library that helps you write unit tests with less code by providing:

  • Test data generators for primitive and complex types (in a constrained non-deterministic way - the data generation algorithm can be customized)
  • Test data builders for any type
  • AutoMocking of dependencies
  • I repeat, A U T O M O C K I N G of dependencies

and in words of Mark Seemann:

AutoFixture is an open source framework for .NET designed to minimize the ‘Arrange’ phase of your unit tests. Its primary goal is to allow developers to focus on what is being tested rather than how to setup the test scenario, by making it easier to create object graphs containing test data.

What does this means in terms of that application you are developing right now if you were to adopt AutoFixture just this very moment?

  • Write less unit test code
    • Bye writing mother objects
    • Bye writing your own test data builders
    • Bye to manually providing dependencies for your system under test
  • But wait Joe! There’s more!
    • By using a SUT (system under test) factory we effectively isolate all tests from changes in the SUT constructor. You will probably have handled this via factory methods, builders or fixture objects. But with AutoFixture and automocking, adding a new dependency to a constructor will not break any test!
    • Being able to automock dependencies means that you don’t need to explicitly mock/stub dependencies that are not used in a given test, which in turns means that tests become more readable, intentional and deliberate. The code that it is there takes active part in the test. All crud is removed, gone, caput.

Installing AutoFixture

You can easily start using AutoFixture in your project via Nuaet:

>Install-Package AutoFixture
>Install-Package AutoFixture.NUnit2
>Install-Package AutoFixture.AutoRhinoMock

Note: I am using Nunit and RhinoMocks in this article but there is support for many other testing frameworks and mocking libraries.

Where each of these packages consist in the following:

  • AutoFixture: this is AutoFixture itself
  • AutoFixture.NUnit2: this is an improved integration between AutoFixture and NUnit that allows AutoFixture to inject test data and SUT directly in NUnit Test methods
  • AutoFixture.AutoRhinoMock: this allows auto mocking of dependencies via Rhino.Mocks

Using AutoFixture

I am going to go fast like lightning through some of the interesting features provided by AutoFixture, brace yourself:

Auto Mocking SUT Dependencies

public class Calculator{
	private readonly ILogger logger;
    private readonly IEmailSender emailSender;
    private readonly IArgumentParser argumentParser;
    private readonly IBellsAndWistles bellsAndWistles;

    public Calculator(ILogger logger, IArgumentParser argumentParser, IEmailSender emailSender, IBellsAndWistles bellsAndWistles){...}
    public int Add(string arguments){
	   Tuple<int,int> operands = argumentParser.Parse(arguments)
       var sum = operands.Item1 + operands.Item2;

       logger.Info(string.Format("I just added a couple of numbers {0} + {1} = {2}", operands.Item1, operands.Item2, sum));
       emailSender.Send(new AdditionWasDoneNotificationEmail(operands, sum));
       bellsAndWistles.Alarm(operands, sum);

       return sum;
	}
}

[TestFixture]
public class CalculatorTests{

    private IFixture fixture;

    [SetUp]
    public void SetUp(){
       fixture = new Fixture().Customize(new AutoRhinoMockCustomization());
    }

    [TestCase("1 1",1,1,2)]
    [TestCase("2 2",2,2,4)]
    [TestCase("3 2",3,2,5)]
    public void AddTwoIntegers_ShouldPerformAddition(string arguments, int op1, int op2, int expectedSum)
    {
       // Arrange
       var argumentParser = fixture.Freeze<IArgumentParser>();
       argumentParser.Stub(p => p.Parse(arguments)).Return(new Tuple<int,int>(op1, op2));

       var calculator = fixture.Create<Calculator>();
       // Act
       var sum = calculator.Add(arguments);
       // Assert
       Assert.That(sum, Is.EqualTo(expectedSum));
    }

    // more tests

}

###Test Builders For Any Type

var camilla = fixture.Build<User>()
					  .With(u => u.FirstName, "Camilla")
                      .With(u => u.SecondName, "For Great Justice")
  				      .Without(u => u.Roles)
					  .Without(u => u.ParentCompany)
                      Create(); // other fields filled with data generated

###Generating Test Data

// for complex types
var camilla = fixture.Create<User>(); // all fields are filled with data generated
// for primitive types
var number = fixture.Create<int>();

Note For The Intrepid: AutoFixture doesn’t work well if there are cycles in the object graph. If a complex type has cycles, you will need to tell AutoFixture to leave that property alone (OmitAutoProperties) or set it explicitely.

Integration with NUnit

If we do a full integration with NUnit we can remove the Arrange stage and inject the test data directly as arguments:

[Test, AutoData]
public void IntroductoryTest(int expectedNumber, MyClass sut)
{
    // Act
    int result = sut.Echo(expectedNumber);
    // Assert
    Assert.Equal(expectedNumber, result);
}

And even auto mock SUT dependencies:

[Test, AutoData]
[AutoRhinoMockData]
public void IntroductoryTest(int expectedNumber, MyClass sut)
{
    // Act
    int result = sut.Echo(expectedNumber);
    // Assert
    Assert.Equal(expectedNumber, result);
}

// in this case you need to define a new attribute
internal class AutoRhinoMockDataAttribute : AutoDataAttribute
{
    internal AutoRhinoMockDataAttribute() : base(new Fixture().Customize(new AutoRhinoMockCustomization())){}
}

Mind Officially Blown Away. Booom… Boooom… Mindbloooown with AutoFixture Booom… Boooom… Booooom…

It even has testimonials. Open-source library with testimonials?!? Whatz?!?:

“I’ve introduced AutoFixture to my developers (at www.gab.de ) some time ago. We’ve been using it successfully with xunit in multiple projects all across the .NET technology stack. We also use it for feeding dummy data to the UI when developing prototypes. That saved us quite some time. Florian Hötzinger, GAB Enterprise IT Solutions GmbH

“I have used AutoFixture for 3 years, it’s a vital tool in my TDD toolbox, a real time-saver. Setting up maintainable and robust unit tests with AutoFixture is easy and straightforward - highly recommendable” Mads Tjørnelund Toustrup, Senior .Net Developer

“Autofixture is more than just another test data generator. It helps me to write tests faster, which are robust against changes in my production code. Moreover, with Autofixture I can focus the tests on the behavior I want to check which why they are easier to read and understand.” Hendrik Lösch, Saxonia Systems AG

It is time to Experiment Thyselve! Go NuGet Install-Package AutoFixture Now!

booksbarbaric-book-reviewux

3 Things I Learned From Seductive Interaction Design, Creating Playful, Fun and Effective User Experiences - A Barbaric Book Review

With Barbaric Book Reviews I bring you interesting reviews and useful insights from awesome books that I have read. I also bring myself the magic of everlasting memory so that I don’t forget these tidbits of knowledge as I grow old and wither.

I should read more UX books. We! [Developers] should read more UX books. They are filled with wondrous jewels of information about the arcane arts of psychology and human behavior, of inestimable value when crafting effective and engaging user experiences. They are also beautiful, I haven’t seen a UX/Design book that wasn’t in itself a wonderful and inviting artpiece of a book. (ehrr… Jaime, always sooo enthusiastic)

We recently started a book club and a library at work. In the book club we read a book a month and afterwards we discuss it (cool), in the library we have shared our books with each other for the betterment of the Universe and for great justice. That’s how I got to borrow this great book titled Seductive Interaction Design: Creating Playful, Fun and Effective User Experiences by @stephenanderson from my good friend (and UX whiz) Erik Claesson and Voilà! The article!

Read on →
dev-talk-mondayprogrammingdev

Dev Talk Monday: Architecting Large JavaScript Single Page Apps With Knockout.js

Dev Talk Monday is the series that brings you awesome dev talks every Monday

This weekend I had the chance to see a superb talk by Steven Sanderson on building large scale SPAs (Single Page Application) with Knockout.js. This talk is invaluable if you are spending any amount of time working with Knockout.js and are looking for tips on how to bring your front-end to 2014. You will learn:

Pretty awesome.

Also pretty awesome the fact that you can see all ndc 2014 talks on vimeo (and all ndc talks from all time back to the prehistoric age)

tddtestingunit-testing

Thoughts on Unit Testing And TDD: Test Behavior, Not Implementation

Ladies and gentlemen: The blog post you are about to read is based on a true story/e-mail. Only the names of classes and methods have been changed to protect the privacy and innocence of the source code. The facts remain unchanged and the contents unadultered…

Disclaimer: I love automated testing and I have been a practicioner of TDD since 2010.

When I refer to unit tests in this article I use the most strict definition of unit tests, that is, test a unit of code - a public method of a class - in complete isolation - by stubbing or mocking all its dependencies. When I refer to integration tests I mean testing classes that collaborate together. When I refer to acceptance tests I mean end-to-end tests.

Last Friday I sent a code review (what a wonderful invention the code review) to a friend and colleague, let’s call him D… hmm… Dumbledore! Yes! That will do. I sent a code review to Dumbledore and continued coding away. After a brief period of time Dumbledore dilligently went through my code and sent me the review back. I impatiently opened it and read:

It looks like you are testing an implementation detail.

Which started a great conversation about programming in general (what a wonderful invention the code review) and about testing behavior vs testing implementation in particular. I kept thinking about the testing behavior not implementation conundrum after work which lead me to write an email the day after and this blog post today. I think that some times I need time to think before I can articulate a good answer to stuff. Particularly with some techniques and practices that I have been doing for a while and I have internalized so much that I almost don’t question any more.

Here is what I concluded after some thought. The email read like this (foreground slowly fades into the past, the screen now in fuzzy shades of gray, Gmail):

I think that indeed the most ideal objective when testing is to test behavior, that’s what you really want to focus on when writing automated tests because it is the only real/final type of test that is going to tell you whether or not your application behaves in the way you want. But that means that, to really test behavior in complex systems you are going to need to write acceptance/integration tests (black-box tests), that is, expensive tests (expensive to setup and expensive to run). For instance, in the particular case that we were talking about in the code review, in order to write a real test of behavior I would have needed to write a black box test that would have tested the service, the inbox manager, any other inbox manager collaborators and probably a repository, NHibernate and the database. This is a huge investment and that’s why I usually like to write unit tests (test unit of code in isolation) instead when I am developing, particularly if I feel that time is running out XD (although in the best case scenario I would write an acceptance test and then TDD my way into a working implementation).

It is quite difficult to test behavior when writing a unit test for a complex class that has many collaborators, that in turn have more collaborators, that have in turn yet more collaborators. For instance,

  1. If you have a calculator with an add method that takes a couple of integers, then no problem, you can easily test behavior.
  2. If you have a console app that takes string arguments and adds them (string calculator kata?), you may have a class that does the arithmetic operations and an argument parser. You can test both individually for behavior and the whole console app with black box testing. Still no problem.
  3. If you have a huge application with tons of classes that depend on each other then testing gets harder. Black box testing becomes hard to setup and slow to run, so you lose one of the most useful/important characteristics of automated testing: short feedback loops. Something that quickly tells you that your code is [not] working as expected, and does it fast.

In case number 3, the case we find ourselves at. You have classes with tons of nested collaborators. You can, and should write acceptance tests, but the feedback loop for these tests is soooooooo slow. You write them, build, start backend, “ups, I made a stupid error”, fix, build, start backend, dammit, another error, and so on and so forth… that’s why, testing with acceptace tests is painful, and that’s where the value of unit tests lie. I can write a test, red, write code, green, refactor, green, write test, red, write code, green, refactor, green and so on super fast so that when I am ready to run the acceptance test I am in a position where I am pretty sure that it will pass.

And this brings us to the problem, How can I write unit tests for complex systems so I can test behavior instead of an implementation?. And that is a tricky question that requires consideration. In the case we are discussing about, I very distinctly tested an implementation:

ShouldGetTheFoldersViaTheInboxManager_WhenCalled (or something like that) where I asserted that a collaborator’s method had been called.

I could make it more behavior-y (and that’s an improvement indeed) by changing its name and body:

ShouldGetTheFoldersIntheUsersInbox_WhenCalled where I would setup the inbox manager stub and test that the service returns the aforementioned folders.

This is an improvement indeed but it still has a problem. I am testing an implementation disguised as a behavior. Why? Because I have explicitly set up the stub. (This test is the equivalent to the mock assertion, I just changed the side of the coin, particularly in this case where the only thing the method does is delegate)

In order to really test behavior in this particular case I think we would need to do a couple of things:

  1. Answer the question: What part of my code is just an implementation? How many collaborators and how deep are just part of this implementation detail? That would probably end up in us dividing the application in additional submodules with a more strict separation (which would be a good thing). We usually achieve this through projects/assemblies/layers, but it would be probably interesting to make use of more private classes and start thinking more deeply about when we use DI (dependency inversion) and IoC (inversion of control) to make an explicit separation between implementation details and the contract/service a given module fulfills/provides.
  2. Since we are using IoC heavily, we would need to improve our testing development environment to be able to inject all the dependencies that are part of the implementation, so that you won’t need to do that manually, and thus save time and pain.

Those are the things I can think about when reflecting about this problem, but there is a lot of people that [knows more about this than I do](http://blog.ploeh.dk/tags.html#Unit Testing-ref) that have much more to say.

Another different question is, is there value in unit tests that test an implementation? or is there value in testing units of code in isolation (because I think that most cases they are the same thing)?

The pros that I see in writing unit tests are the ones that I have always seen:

  • test ensures the code behaves as I want it to behave
  • fast to write, fast to run, short feedback loop, good value for price
  • find bugs with super-high granularity (if a unit test shows a bug, you pretty much know which class it is in, you don’t need to go throw all classes nor debug it, that’s priceless)
  • it forces you to write very simple classes and increase modularity and separation of concerns (because writing unit tests for fat, complex classes with tons of collaborators is horrible. We have a natural tendency to avoid pain)
  • increases developer confidence to refactor
  • documentation for my future self and other developers
  • if something changes that affects the unit under test, you will know it (regardless if it is a bug or a new implementation xD)

The cons are the ones that I have learnt to live with and that have made me reflect on improving the way I test:

  • they can be brittle in the sense that changes in the implementation may break your tests
  • they are not as trustworthy as higher level tests. You can make assumptions when stubbing collaborators that do not match reality and end up testing… something different

For instance, in the case of the calculator console app, you could have a calculator class that would call an argument parser to parse the arguments and then call the arithmetic unit to perform the arithmetics on those parameters. You could write a test to test that the calculator behaves in that way. GivenTwoNumbers_ShouldReturnItsSum and you would setup two stubs to do the parsing and the arithmetic calculation. The calculator class itself doesn’t do much, it justs delegates argument parsing to an argument parser class and arithmetic calculations to the other class. If you write a test to verify that the calculator behaves in that way, you are pretty much testing that algorithm: 1) parse 2)arithmetics, but more explicitely, there’s some IArgumentParser that is going to parse the arguments and some IArithmeticsUnit that is going to add them. What do you gain by writing this test? this particular test is of marginal value if you compare it to the unit tests for the arithmetic and the parser classes but you still gain something: 1) you know that the class works in the way you want, without the test, how would you know it works at all? 2) If you run the app and there’s a bug, you are going to be pretty sure that the bug is not in the calculator but in other dependencies.

And that’s pretty much it… I have to think about other specific examples within our application MagicFlow …

hmm… this could be a blog post xDDD

(foreground distorts into reality, welcome back to the present) And that was the email. Hope you found it interesting and feel free to comment and share your thoughs about unit testing and TDD!

I have been thinking for a while about writing more articles regarding my experience and journey with unit testing and TDD, so expect more of them coming. And particularly at this point where I am starting to question the definition of unit tests in isolation I learnt from the art of unit testing. All hope is not lost, I was just a liiiiitle bit slower than Roy :)

I used to feel that a ‘unit’ was the smallest possible part of a code base (a method, really). But in the past couple of years I’ve changed my mind. Here’s how I define a unit test, as of October 2011:

A unit test is an automated piece of code that invokes a unit of work in the system and then checks a single assumption about the behavior of that unit of work.

A unit of work is a single logical functional use case in the system that can be invoked by some public interface (in most cases). A unit of work can span a single method, a whole class or multiple classes working together to achieve one single logical purpose that can be verified.

Some Interesting references