Traditional Object Oriented Programming in JavaScript (Part 2)

Page content

This is part 2 of our 2 part tutorial on “Traditional” Object Oriented programming in JavaScript. The tutorial was written by the course staff in 2020, but updated in 2021 and 2022.

In the first part of this tutorial (Traditional Object Oriented Programming in JavaScript (Part 1)), we showed you how to use JavaScript’s class syntax; this time, we’ll take a look at how we can use it to do the things you’re used to in object oriented programming: inheritance, polymorphism, and more.

For the later workbooks, you’ll need to understand inheritance and polymorphism. The code Framework uses these features a lot. The other parts of the workbook (static and private members) are useful, and can help you understand how THREE.js works.

Inheritance

Recall our first Rectangle class from last time:

1
2
3
4
5
6
7
8
class Rectangle {
    constructor(x, y, height, width) {
        this.x = x;
        this.y = y;
        this.height = height;
        this.width = width;
    }
}

When we wanted to add a draw method to it, we had to modify the original class. But what if we want two types of Rectangles: one which fills itself, and one which only draws the border? The simplest way we can do this is with inheritance, which we can use with the extends keyword as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
class SolidRectangle extends Rectangle {
 constructor(x, y, height, width) {
  super(x, y, height, width);
 }
 
 draw(context) {
        context.fillRect(this.x, this.y, this.height, this.width);
    }
}

class BorderRectangle extends Rectangle {
 constructor(x, y, height, width) {
  super(x, y, height, width);
 }
 
 draw(context) {
        context.strokeRect(this.x, this.y, this.height, this.width);
    }
}

The super keyword is a special JavaScript keyword that is used in methods. When it is used in a method (including a constructor), it refers to the same method in the parent class, allowing you to call the parent class’ method from within the child class’ method. So, for example, in line 3 of the above example, the call super(x, y, height, width) calls the parent class (Rectangle) constructor from inside the child class (SolidRectangle) constructor.

In a child class constructor, super must always be called before this is used. Basically, this is not defined until the call to super. Suppose that we will exteded SolidRectangle to allow it to store the color of the rectangle as well. You can use super in methods other than the constructor, without a restriction on having it before you use this. For example:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
class SolidColoredRectangle extends SolidRectangle {
 constructor(x, y, height, width, color) {
  this.color = color; // ReferenceError!
  super(x, y, height, width);
 }
 
 draw(context) {
  context.save();
  context.fillStyle = this.color;

        // Fills the rectangle with the parent class method
  super.draw(context); 
  context.restore();
 }
}

Because the parent constructor was called after this was used above, we get a reference error when we try to create the class! This is easy to fix by changing the order of the lines in the constructor.

1
2
3
4
5
6
7
8
class SolidColoredRectangle extends SolidRectangle {
 constructor(x, y, height, width, color) {
  super(x, y, height, width);
  this.color = color; // No error this time
 }
 
    // draw does not need to be changed
}

The draw method works fine (we can use this before super because it is not a constructor).

Polymorphism

Polymorphism in JavaScript works just like it does in other languages. In the code below, we draw all the different types of rectangles in the display list with a single forEach loop (see the documentation); each object draws itself as specified by its class’s draw method.

1
2
3
4
5
6
7
8
displayList = [
    new SolidRectangle(0, 0, 10, 10),
    new BorderRectangle(10, 10, 10, 10),
    new SolidColoredRectangle(5, 5, 10, 10, "blue"),
];

// Assume that some context exists
displayList.forEach(item => item.draw(context));

Static Properties

The static keyword can be used to define static properties and methods of a JavaScript class; these belong to the class itself rather than to an individual instance of it. See the following example (it uses a template string, a very handy JavaScript trick). Note the line static creationCount = 0; it is equivalent to StaticTest.creationCount = 0, and thus let is not necessary. This notation can be used to define class properties (and, by omitting static, instance properties as well) outside the constructor.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
class StaticTest {
    static creationCount = 0;
    
    constructor() {
        StaticTest.creationCount++; 
    }
    
    static saySomething() {
        console.log("Hello, world!");
    }
    
    toString() {
        return `Creation count: ${StaticTest.creationCount}`;
    }
}

console.log(new StaticTest().toString()) // Prints "Creation count: 1
console.log(new StaticTest().toString()) // Prints "Creation count: 2
console.log(new StaticTest().toString()) // Prints "Creation count: 3
StaticTest.saySomething() // Prints "Hello, world!"

// Don't do the following
console.log(new StaticTest().creationCount) // Prints "undefined"
// Generates a TypeError; the object does not have 
// a saySomething() method
new StaticTest().saySomething() 

The last two lines don’t work properly because they try to access a static (class) variable and method as properties of an instance of the class. Thus; as they are properties of the class and not the instance, the creationCount property of an instance of the class does not exist; similarly, the saySomething() method of the instance does not exist, explaining the TypeError.

Private Members, Getters, and Setters

ES6 does not support true private properties (there is currently a stage 3 proposal that is likely to be included in future JavaScript versions), but a common convention is to begin properties that should be kept private with an underscore. ES6 has special syntax for getters and setters, as shown below.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
class DirectoryTracker {
    _directoryStack = [];
    
    // Pushes the directory onto the stack
    set directory(directory)  {
        this._directoryStack.push(directory);
    }
    
    // Returns the current directory
    get directory() {
        return this._directoryStack[this._directoryStack.length - 1];
    }
    
    // Go back to the last directory
    pop() {
        this._directoryStack.pop();
        
        // Referring to this.directory inside the class also
        // translates into a call to the getter
        return this.directory; 
    }
}

let tracker = new DirectoryTracker();

// We should not do this, but it is permitted by JavaScript
console.log(tracker._directoryStack); // Prints "[]"

tracker.directory = "foo";
tracker.directory = "bar";
console.log(tracker.directory); // Prints "bar"

// Don't do this, but it is interesting to take a look
console.log(tracker._directoryStack); // Prints "['foo', 'bar']"

console.log(tracker.pop()); // Prints "foo"

The get and set keywords allow us to track the history of previous directories in a way which is transparent to the user. When we access tracker.directory, JavaScript makes a call to the getter we defined instead of directly accessing the directory property of tracker, and, similarly, when we assign tracker.directory above, a call is made to the setter.

Further Reading

See the Mozilla reference page for more information.

PREV: Part 1: Traditional OOP: Inherritence, Polymorphism, …