Thank you. Thank you. So interfaces, we've talked about them. And one thing we want to talk about now is a little bit of how to use them. So what are they used for, language-wise? And in language, why would you need an interface? And I already said, interfaces, they express some sort of a conceptual similarity between different types. So the idea is that if two types satisfy an interface, then they must be similar in some way that is important to the application. So one common, sort of, practical thing that you would use an interface for, is when you need a function, you want to write a function which takes multiple types as a parameter. So specifically, normally, a function, it takes, say, it takes an integer as its arguments, right? It can only take an integer. But what if you want to take an integer or a float? Something like this, maybe a integer, a float or a string, right? You want it to take multiple types. In different types, it'll do different things, but you want it to be able to take different types. You can use an interface for that. So giving an example, say I've got a function, foo ( ), and it's gotta take a parameter. And this parameter, it could be either type X or type Y. And I'm talking very generically now. I'll give you a concrete example in the next slide, but so I got this foo. It's going to take a parameter which is type X or type Y. I want it to take either type. So the way I can do that is I can define an interface called Z. And X as a parameter can be, the type it can be, in interface type Z. Then I define X and Y to satisfy Z. Right, so, X and Y are both satisfying the interface Z. Then foo(), since it can take anything that satisfies interface Z, it can take X and Y as its arguments. So this is a common way to use an interface. And basically, an interface in this way, it sort of generalizes, right? It hides the details of the differences between the types. It's like look, these two types are similar in the way that's important to me. And so your function could just take the interface. It needs to take anything that is similar in that way. So, to be a little bit more specific, I made up a problem about a pool in a yard, right? So, I have a backyard, and I want to put a pool in my yard. But the pool, before I can put the pool in the yard, it needs to fit in my yard. And it needs to be fenced, because I don't want my kids to fall into the pool, so I need fencing around the pool. And it needs to fit in the yard. So to fit in the yard, the total area this pool needs to be limited. Less than the area of the backyard. Also to fence it, I only have so much fence, because fence cost money. I only have a limited amount of fence. So the perimeter of this thing has to be limited to within some limit, depending on how much I can afford. So I need to determine if a particular pool shape satisfies this criteria, because I'm trying to go through a bunch of different pool shapes. And I want to pick one that satisfies these criterias. It's sort of small enough that it can fit in the yard. And also the perimeter's small enough so that I can afford to fence it. So I'm going to write this function called FitInYard (). And this is going to return a boolean. So it takes a shape as an argument. Some shape, maybe I got some triangular shape, I was pass that to FitInYard, and it returns true, if the shape satisfies the criteria. So if the area is small enough, and the pool is small enough, then it says true. If I pass it a shape like rectangle and it's a big rectangle and it doesn't fit in my yard then it will return false, okay? So that's what FitInYard is. Now the thing about fit in yard is it's got to take a shape as an argument. But I want it to take any shape as an argument. I don't care if it's a triangle, circle, square, rectangle, whatever, it can be any shape. I should take it as an argument. But I have to be able to compute the area and compute the perimeter, okay? So not any shape, its got to be a shape whose area and perimeter I can compute. So let's say the idea is should take rectangles, triangles, whatever. But a valid shape has to have an area method and a perimeter method, right? So if the shape, if I can't compute the area, then I won't be able to tell if it fits in my yard. Say it's a sphere or something, right? There's no area, it's a 3D shape, that's a 3D object, I can't compute area of a thing like that, right? So, that is not a valid shape that I want to try to fit in my yard. So, any shape that has area and perimeter, that's okay with me. So what I can do is I can define this interface for shapes that have the area and perimeter. So I make my Shape2D interface. We already talked about this. But I make my Shape2D interface. It specifies area and perimeter, which return float64. Then I can make my types, triangle type, rectangle type, whatever types I want. And as long as these types, I don't care how they define what data's inside them, as long as they have methods that use them as receiver types. That it has area method and perimeter method. So, a triangle, you got area that has triangle as a receiver method and also perimeter. Same thing for rectangles, it's got an area and a perimeter. As long as they have area and parameter, I should be able to take this as an argument. So they satisfy this interface Shape2D. So in my FitInYard implementation, you can see that the argument that it takes is called s, and its type is the interface type, Shape2D. So what that means is that this argument could be any type that satisfies that Shape2D interface. Like rectangle, triangle, whatever the types are. And it returns a boolean and all the function does is very simple. Just says if s.Area is less than 100, let's say 100 is the size of my backyard, right? And as our perimeter is less than 100, because that's all the fence I could afford, then return true, else return false. So a valid argument to this is any type that satisfies the shape to the interface. Now the empty interface is standard interfaces predefined and it just specifies no methods. So that means that any type can actually satisfy that interface. And what you use it for is when you want to have a function argument be any type. You don't want to restrict it at all in terms of the type that this function can accept. Then you just make it's type the empty interface. So as an example, we got this function PrintMe. And its val argument is the empty interface. That's how you specify the empty interface that I haven't read. So that means that val can just be any type. And all this does is just print it. So it will print any type you give it. You give it an int, float, string, whatever, it will just print that to the screen. Thank you.
