Hello, everybody! Today I’m going to show you how to use classes in TypeScript. There’s a lot to cover about classes, so for this video we’re going to stick to the basics, and then in future videos we’ll dig into the really advanced stuff.
At their core, classes are a way of describing what data you want certain objects to hold, along with methods that act on that data in various ways. There are a lot of other features built on top of this, but that is the core.
If you come from a language like C# or Java, classes in TypeScript are very similar to what you worked with there. Though there are some important differences in the low-level details, which we’ll be covering in later videos.
Unlike Java or C#, however, there’s nothing in TypeScript that is forcing you to use classes. So if you prefer to write your code in some other way, you are perfectly free to do so. Classes are just another tool you have in your toolbelt as a TypeScript developer.
So let’s start off with a simple example class.
Let’s say we’re making game and we’re planning on having multiple players and a decent amount of logic common to the players to handle things like movement, displaying the players, and so on. This sounds like a good candidate for a class, so let’s create a class called Player.
Creating a class is very simple, we’ll write the class keyword followed by our class name, and then the contents of the class are contained within curly brackets. If you saw the previous video on interfaces, you’ll notice that class definitions and interface definitions share a lot of parallels, which you’ll see more of going forward.
Let’s pause for a moment and note that this is actually a complete class. It’s empty, so it’s not very useful, but before we start putting things in it, I want to show you how to create an instance of this class.
We’ll create a new variable called “p”, and set it equal to “new Player()”.
The “new” keyword is how you indicate that you’d like a new instance of an object that’s defined by a certain class. Player is obviously the name of your class, and your parentheses indicate that this is a function call.
Now, we haven’t actually defined any function here, so what function is it calling? Well, all classes have what are called “constructor” functions, and if you don’t create one explicitly, TypeScript will create one for you. We’ll go into more detail on this in a bit.
Let’s log the variable and see what it looks like.
You’ll see that Node.js is actually very helpful here, showing us the class name, and then the empty curly brackets will contain the data in our object once we actually start adding some fields to our class.
So, how do we add things to our Player class?
The first and easiest way is to simply write the name of the variable, and then assign to it whatever value you want.
Here we’ll say each player needs and x and a y position, so we’ll set both of those equal to zero.
If we run our program again, you’ll see that I have a player with an x and y of 0. Excellent!
These fields are just like the fields of any other object. For example, I can set “p.x” equal to 1, and that will do what you expect.
A note to those from other programming languages: Fields in TypeScript are all publicly accessible by default.
Just to prove to you that we have an isolated instance of our object, let’s create a second variable called p2, and log that one out as well.
And you’ll see it’s only our original object that has changed, just as we would expect.
Going back to our class definition, note that if you wanted to be explicit about the type that you wanted a field to be, you can indicate that in the usual way by simply writing a colon and then the type that you’d like.
What we’ve done so far works well for fields that will always start off at some default value, but what about a case where there isn’t necessarily a correct default value?
For example, let’s say we wanted to give each Player a name. We’ll add the name field and say that it should be a string.
But, what’s a good default for that? Well, an empty string might not make much sense there. Neither would a random name. Plus then we’d have to remember to change it properly after instantiating it.
Probably what makes most sense would be for us to force the name to be specified at the creation of the object. We mentioned previously that there was a constructor function being created for us, let’s define it ourselves now so that we can use it to require a name.
You can do that by going to the class definition, typing “constructor”, parentheses and then curly brackets for the constructor function body.
We’ll add “name” as a parameter to this function, and we’ll assign “this.name” equal to the “name” parameter.
You’ll see that now TypeScript is giving us errors about the places where we’re creating instances of this class because we’re not providing a name.
Let’s name the first one “Tyler” and the second one “Fred”. And now TypeScript is happy again.
A couple points of clarification on constructor functions. First, the special “this” keyword refers to the particular instance of the class that you’re currently working with. Here in the constructor, that’s the newly created object that we’re building up.
Second, the way these default values for x and y are set by TypeScript is by moving the assignment to the very top of your constructor function.
So assigning a value to them on the same line as their definition is directly equivalent to writing “this.x = 0” at the top of the constructor function.
In general, since assigning the default value where the field is declared is more compact than the alternative, I tend to favor doing that in my own code. If you prefer to leave the declaration as only the types, and initialize all the defaults in the constructor that’s a perfectly valid stylistic choice.
Note that if you’re initializing something in the constructor, the field declaration must have a type on it, because otherwise TypeScript will not be able to infer what type it is that you want.
Okay, so we’ve got some data that we’re holding in our objects, let’s create a function to act on that data.
Let’s say we don’t think the output of console.log is quite to our liking, so we’d like some other way of displaying our Players.
Well, we can create a new method by specifying the name, the arguments in parentheses, and then we can specify the return type with a colon after the parentheses.
And we’ll type out a call to console.log() that formats things a little more nicely.
And we’ll run that, and hey! Things are looking pretty good.
One thing you’ll notice down below is that the list of functions that are part of an object aren’t included in the default console.log() output in Node.js. Generally this is fine, since the data will be more important to see, but this is something you should be aware of.
Let’s tie things into the previous video by showing how we can make interfaces and classes work together.
Since we’re working on a game here, I imagine we’ll have a number of different types of objects that we’ll want to treat generically in some kind of way. Perhaps we’ll use some common fields to do collision detection or to draw them or whatever else.
To define what we need, we can create an interface listing what we’ll need from each game object.
Let’s say this is a 3D game, so we’ll need an x y and z position.
In order to indicate that we want our player to be compatible with this interface, we can use the “implements” keyword followed by the name of the interface. Note that if you had multiple interfaces, you can separate them by commas.
Now we’re getting an error about our player not implementing the interface correctly because it doesn’t have the “z” field, so let’s add that.
And now everything is good.
Down lower in the code, we can create an array of GameObjects, and put our two player objects in that array. Since our Player class properly implements the GameObject interface, TypeScript is perfectly happy with this.
Note that the “implements” keyword on our class definition isn’t actually required as you might be used to from other languages. If we remove it, everything is still fine. The difference here is that if we were to remove our “z” field, we now get an error down in our array rather than on the class itself.
So why would you want to use the implements keyword if you can get away without it?
The biggest reason is that it will tell you while you’re focused on just the class definition itself whether or not the class properly implements the interface.
It also serves as great documentation for your future self and other programmers that the specific class is intentionally conforming to some interface.
A third reason, is that some frameworks like Angular will check objects for specific methods. If they exist, Angular will call them, if they don’t, Angular won’t. Including the “implements” keyword and the relevant interface for that specific Angular feature will cause TypeScript to validate that you are actually including that method properly. If not, you’ll get an error, which is helpful.
Okay, we’ll pause here and in future videos we’ll be covering things like inheritance, private, public, and protected fields, static properties, accessors, the nitty gritty details on the implementation details behind classes, and much more.
If you’ve enjoyed this video and want to go deeper, I’m working on a course on TypeScript. You can find it at https://typescriptbyexample.com I’m still working on it but if you scroll down to the bottom of the page leave your email address there I’ll shoot you an email when that course is released.
That’s it for this video! Thank you so much for watching, I will see you in the next one.