5 things that might surprise a JavaScript beginner/ OO Developer

Follow me on Twitter, happy to take your suggestions on topics or improvements /Chris

TLDR; this is not a critique against JavaScript, it's just acknowledging the fact that it differs a bit from OO languages and you can either curse JS Or you can use the patterns that are made possible through that, to your advantage.

I love the language but it works differently than other languages I'm used to.

Regardless if you are a beginner to JavaScript or beginner to programming there are things in JS that might surprise you. Just because it surprises you doesn't mean it's wrong, it's just different, quirky, or perfectly sane, depending on what your previous experience is. Each one of the upcoming topics deserves their own article or even book, almost, but here goes:

## -1- Really really equals

If you learned to code in some other language maybe Java, you've learned that one = means an assignment and == compares. In JavaScript, you have both === and == for comparing equality. Which one to use? What's the difference? == compare values only. Consider this example:

if('2' == 2) {} // true
1

It returns true when it's the same value but the type differs.

Look at this example now:

if('2' === 2) {} // false

if(2 === 2) {} // true
1
2
3

Above the === detects that '2' and 2 have different types and therefore evaluates to false. It's generally recommended to be using this way of doing comparison.

-2- There are many ways to create an object

In Java or C# you have a class. From that class, you can instantiate an object. It makes sense. JavaScript gives you more options. There you can create an object in the following way:

  • Using a class, There is the keyword class that you use to define fields, methods, getters/setters all within the context of a class. Here's an example:
class Person {
  constructor(n) {
    this.name = n;
  }

  getName() { return this.name; }
}
1
2
3
4
5
6
7
  • Object literal, You can define an object without defining a class. All you need is {}. It can look like so:

    const person = {
      name: 'chris',
      city: 'location',
      getAll() {
        return `${this.name} ${this.city}`;
      }
    }
    
    1
    2
    3
    4
    5
    6
    7
  • Object create, you can use the method Object.create() to create an object. It takes a prototype object to base it off of. Here's an example:

    const address = {
      city: '',
      country: ''
    } 
    
    const adr = Object.create(address);
    adr.city = 'London';
    adr.country = 'UK'
    console.log(adr.city); // London
    console.log(adr.country); // UK
    
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10

Block statements, look ma no scope

Block statements, if, for, while etc, don't create a local scope. That means whatever you create in there is accessible outside of the statement, like so:

for (var i =0; i< 10; i++) {
  console.log(i);
}

console.log(i);
1
2
3
4
5

The last console.log() will print 10. This might surprise you.

Yes, why is JS designed so that is i still alive?

Ask Brendan Eich, it's a feature 😃

To make JS behave like other languages you might know, you need to use a let or a const, like so:

for (let i = 0; i< 10; i++) {
  console.log(i);
}

console.log(i);
1
2
3
4
5

Running this code now states i is not defined. Why did this work? Well, let allows you to declare variables that are limited to the scope of a block statement. So it's the usage of the keyword let over var that does this rather than the block statement being given a scope. (Thanks to Will for this comment)

Phew

-3- Context, what's the value of this

You might have heard the jokes that no one knows what this is. Starting out with an empty file this is the global context. Consider the following code:

global.name = "cross";

function someFunction() {
  console.log(this.name);
}

someFunction();
1
2
3
4
5
6
7

Above we are assigning name to the variable global (that's what we call it in Node.js, on the frontend it would be window). The value of this comes from the global context.

Let's look at a different example below:

var object = {
  name: 'chris',
  getName() {
    console.log(`${this.name}`);
  }
}

object.getName();
1
2
3
4
5
6
7
8

Here, the value of this is the object itself, it knows what name is, i.e the value chris.

Changing context

We can change what this is. There are some helper methods in JavaScript that allows us to do that bind(), call() and apply(). Consider this example again but with object added:

global.name = "cross";

var object = {
  name: 'chris',
  getName() {
    console.log(`${this.name}`);
  }
}

function someFunction() {
  console.log(this.name);
}

someFunction();
1
2
3
4
5
6
7
8
9
10
11
12
13
14

We can alter this from the global context to that of object. Below we showcase how anyone of the mentioned methods can use this principle:

someFunction.bind(object)();
someFunction.call(object)
someFunction.apply(object)
1
2
3

It will now print chris, instead of cross.

These three methods are used in a little bit different ways normally but for this example, they are pretty equivalent.

The this confusion

Ok, so when are we actually confused what the value of this is? It happens in more than one place, but one common place is when we try to use a constructor function to create an object like so:

function Person(n) {
  this.name =  n || 'chris';
  function getName() {
    return this.name;
  }
  return {
   getName
  };
}

const person = new Person();
console.log(person.getName()) // undefined 
1
2
3
4
5
6
7
8
9
10
11
12

This is because the this changes for inner functions once you use new on it. There are different solutions to fixing this:

Solution 1 - this = that

A way to approach this is to make it remember the value of the outer this. Rewrite the above example to look like this:

function Person(n) {
  this.name =  n || 'chris';
  var that = this;
  function getName() {
    return that.name;
  }
  return {
   getName
  };
}

const person = new Person();
console.log(person.getName()) // 'chris'
1
2
3
4
5
6
7
8
9
10
11
12
13

It fixes the issue by introducing the that variable that remembers the value of this. But there are other solutions.

Solution 2 - Arrow function

function Person() {
  this.name = 'chris';
  
  const getName = () => {
    return this.name;
  }

  return {
    getName
  }
}

const person = new Person();
console.log(person.getName()) // 'chris'
1
2
3
4
5
6
7
8
9
10
11
12
13
14

The above replaces the function keyword for an arrow function =>.

Solution 3 - Use a closure

The third solution is to use a so-called closure. This involves not using the new keyword but relies on the fact that JavaScript barely needs to use this. Consider the below code:

function Person() {
  var name = 'chris';

  const getName = () => {
    return name;
  }

  return {
    getName
  }
}

const person = Person();
console.log(person.getName()) // 'chris'
1
2
3
4
5
6
7
8
9
10
11
12
13
14

Above this has been completely removed. We are also NOT using new. IMO this is the most JavaScript-like pattern to use.

Solution 4 - put method on the prototype

In this approach, we use a class:

function Person() {
  this.name = 'chris';
}

Person.prototype.getName = function() {
  return this.name;
}

const person = new Person();
console.log(person.getName()) // 'chris'
1
2
3
4
5
6
7
8
9
10

This is a good solution for more than one reason. It solves the this problem but it also makes sure that the method is only created once, instead of once per instance.

Solution 5 - use a class

This is quite close to the fourth solution:

class Person {
  constructor() {
    this.name = 'chris'
  }

  getName() {
    return this.name;
  }
}

const person = new Person();
console.log(person.getName()) // 'chris'
1
2
3
4
5
6
7
8
9
10
11
12

For this article to be crazy long I can't name all the possible cases where this isn't what you think it is. Hopefully, these solutions offer you an insight into when it goes wrong and approaches to fix it.

-4- const works, but not the way you might think

There's the const keyword, we've seen how it creates a local scope. But wait, there's more 😃 The word const makes you think it will always have this value, it's a constant, unchanging etc. Weeell.. Looking at the following example:

const PI = 3.14 // exactly :)
PI = 3;
1
2

The above gives me the error Assignment to a constant variable.

So I can't change it, goood

Let's look at another example:

const person = {
  name: 'chris'
}

person.name = 'cross'; 
1
2
3
4
5

This works without a problem 😃

Wait whaaat. You said the value couldn't change?

Did I say that? I didn't say that. I said the word const sounds like it. What const means is that there is a read-only reference, i.e the reference can't be reassigned. I never said it can't be changed. Look at this for clarification:

const person = {
  name: "chris",
};

person = {
  name: 'chris'
}
1
2
3
4
5
6
7

The above gives an error. Cannot assign to a constant variable.

Hmm ok, can I make it immutable?

Well you can use Object.freeze() like so:

Object.freeze(person)

person.name = "cross"; 

console.log(person.name) // 'chris'
1
2
3
4
5

Great so there is a way to make everything immutable

Weeell.

Well what?

It only freezes on the first level. Consider this code:

const person = {
  name: "chris",
  address: {
    town: 'London'
  }
};

Object.freeze(person)

person.name = "cross"; 
person.address.town = 'Stockholm';

console.log(person.address.town) // Stockholm
1
2
3
4
5
6
7
8
9
10
11
12
13

But but.. is there no way to freeze it?

You would need a deep freeze algorithm for that. Ask yourself this though, do you need that? I mean in most cases your constants are usually primitives.

...

To be fair this is a little how const works in other languages as well. I mean in C# it's static readonly if you want something immutable and locked reference, in Java you need final.

Yea yea whatever

## -5- There's life after function invocation

Let's look at the following piece of code:

function aFunction() {
  let name = 'chris';
  console.log(name) // prints chris
}

console.log(name)
1
2
3
4
5
6

Nothing special with it, it doesn't know what name is in the last console.log() cause it's outside the function. Let's modify it slightly:

function aFunction() {
  let name = "chris";
  return {
    getName() {
      return name;
    },
    setName(value) {
      name = value;
    }
  }
}

const anObject = aFunction();
console.log(anObject.getName());
anObject.setName("cross");
console.log(anObject.getName());
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

At this point it prints chris calling getName(), ok you might think it was bound to a value. Then you call setName() and lastly you call getName() again and this time it prints cross. So why is this surprising? Well think about how a function normally works, you call it and the variables in it seize to exist. Now look at the above code again and notice that the name variable seems to still exist long after the function has stopped executing. This is not really surprising if you compare it to a language like Objective-c for example. You are then used to referencing counting, if some part of the code is no longer referencing something it is garbage collected. You are clearly still referencing it via the anObject variable.

But still, if you come from an OO background you might be used to objects holding a state and that the state lives on the object itself. In this case, name lives in the lexical environment outside of the object, that's trippy right? 😉

The easiest way to think about this one is object creation with private variables. It's also how I create objects more and more these days.. Nothing wrong with classes though, whatever floats your boat 😃

Summary

I'd love your comments on other things that may surprise you/trip you up or makes your life better. Cause that is true for me about a lot of things JavaScript - I type a lot less.