Closures in JavaScript

What are closures?

I consider closures in JavaScript as an advanced topic. It is one of the topics that gets asked in interviews a lot.
Understanding closures would be more comfortable if you have read my previous blog posts or know about scope in JavaScript.

Function scope in JavaScript means that a variable declared inside a function is only accessible in that function. It is, however, accessible to the function and any of its child functions. Closures take this one step further. Closure makes sure that the parent scope is available to the child function even after the parent function has finished execution.

Example

function outer() {
    const outerVariable = "outer";
    function inner() {
        const innerVariable = "inner";
        console.log(`${outerVariable} ${innerVariable}`); // outer inner
    }
    inner();
}

outer();

I created and executed the outer function above. This action creates and invokes inner function. inner function successfully logs the variable it has declared and the variable in the parent function. It is expected since the child function has access to the parent function's scope.

Now let us not call the inner function and instead return it.

function outer() {
    const outerVariable = "outer";
    function inner() {
        const innerVariable = "inner";
        return (`${outerVariable} ${innerVariable}`);
    }
    return inner;
}

const returnFunction = outer();
console.log(returnFunction); // Function returnFunction

Variable returnFunction is a function because that is what was returned by outer. No surprises.

🚨At this point, execution of outer function has finished, and the return value has been assigned to a new variable.

This is key. JavaScript garbage collection should have removed all traces of outerVariable because that is what happens when a function is popped off the stack and finished executing. Let us run returnFunction.

function outer() {
    const outerVariable = "outer";
    function inner() {
        const innerVariable = "inner";
        return (`${outerVariable} ${innerVariable}`); // outer inner
    }
    return inner;
}

const returnFunction = outer();
console.log(returnFunction); // Function returnFunction
console.log(returnFunction()); // outer inner

Surprise surprise! It can still log the value of a variable in its parent function (which has finished executing).

JavaScript garbage collection does not clear out variables of a function if that function has child functions returned. These child functions can run later and are fully eligible to access the parent's scope by the principal of lexical scoping.

This behavior of garbage collection is not only limited to child function. A variable is not garbage collected as long as anything maintains a reference to it.

Real World Example

Let's say that I am programming about a car. This car can accelerate like a real-world car, and whenever it does accelerate, the speed of the car increases.

function carMonitor() {
    var speed = 0;

    return {
        accelerate: function () {
            return speed++;
        }
    }
}

var car = new carMonitor();
console.log(car.accelerate()); // 0
console.log(car.accelerate()); // 1
console.log(car.accelerate()); // 2
console.log(car.accelerate()); // 3
console.log(car.accelerate()); // 4

You can see how the speed of the car is provided by carMonitor, and it is accessible by accelerate function. Every time I call accelerate, not only can it access the said variable but increment it from last value and return it.

Creating Private Variables using Closures

Let's take the example for carMonitor.

function carMonitor() {
    var speed = 0;

    return {
        accelerate: function () {
            return speed++;
        }
    }
}

var car = new carMonitor();
console.log(car.accelerate()); // 0
console.log(car.accelerate()); // 1
console.log(car.accelerate()); // 2
console.log(car.accelerate()); // 3
console.log(car.accelerate()); // 4
console.log(speed); // speed is not defined

You can see that the variable is a private variable for function carMonitor, and it is only accessible by the child function accelerate. No one outside can access it. One could argue that it is expected due to the function scope. It is private to carMonitor, and it is private to each new instance of carMonitor.

Every instance maintains its copy and increments it!

This should help you realize the power of closures!

function carMonitor() {
    var speed = 0;

    return {
        accelerate: function () {
            return speed++;
        }
    }
}

var car = new carMonitor();
var redCar = new carMonitor()
console.log(car.accelerate()); // 0
console.log(car.accelerate()); // 1
console.log(redCar.accelerate()); // 0
console.log(redCar.accelerate()); // 1
console.log(car.accelerate()); // 2
console.log(redCar.accelerate()); // 2
console.log(speed); // speed is not defined

car and redCar maintain their own private speed variables, and speed is not accessible outside.

We are enforcing the consumer to use the methods defined on the function or class rather than accessing the properties directly (which they should not). This is how you would encapsulate your code.

I hope the article cleared up any doubts about closures in JavaScript!

Common Interview Question

Here is an interview question about closures that is asked fairly frequently.

What do you think the output of the following piece of code will be:

for (var i = 0; i <= 5; i++) {
    setTimeout(function () {
        console.log(i);
    }, 1000);
}

If you guessed numbers from 0 to 5 with a one-second gap, you are in for a surprise. By the time 1 second has passed for the setTimeout, i has a value of 6 at the time of invocation! We would like to use the value of i from time of creation and IIFE + closures will help you do that.

for (var i = 0; i <= 5; i++) {
    (function (i) {
        setTimeout(function () {
            console.log(i);
        }, 1000);
    })(i);
}

There is another way to solve this issue. Using the let keyword.

var in our loop used to declare i creates a function scope. This results in one binding shared for all loop iterations. When the six timers complete, they all use the same variable with a finally settled value of 6.

let has block scope and when used with a for loop creates a new binding for each iteration of the loop. Each timer in the loop gets a different variable with a different value from 0 to 5.

for (let i = 0; i <= 5; i++) {
    setTimeout(function () {
        console.log(i);
    }, 1000);
}

And now the output will be numbers from 0 to 5. If you are targeting ES5, use the IIFE plus closures method. If you have access to ES6, use the let keyword method.

This is what Babel does under the hood when it transpiles ES6 code to ES5. It transsforms the above let based code to a combination of closure + IIFE!