Start of Tutorial > Start of Trail > Start of Lesson

Search

What Is an Interface?

In English, an interface is a device or system that unrelated entities use to interact. According to this definition, a remote control is an interface between you and a television set, the English language is an interface between two people, and the protocol of behavior enforced in the military is the interface between people of different ranks. Similarly, a Java interface is a device that unrelated objects use to interact with one another. Java interfaces are probably most analogous to protocols (an agreed-upon behavior). In fact, other object-oriented languages have the functionality of Java's interfaces, but they call their interfaces protocols.

A Java interface defines a set of methods but does not implement them. A class that implements the interface agrees to implement all of the methods defined in the interface, thereby agreeing to certain behavior.

---

Definition:  An interface is a named collection of method definitions (without implementations). An interface can also include constant declarations.

---

Interfaces are best understood through examples, so let's look at a concrete example of an interface and two classes that use it to interact. Then we'll talk more about interfaces in the abstract and clear up some common confusion.

Example: AlarmClock and Sleeper

This example is fairly simple, but it shows you how to create and use an interface. It also gives you some insight as to why you need them and how to decide when to use an interface versus when to use a class or an abstract class.

The AlarmClock class is a service provider-- it notifies objects after a certain amount of time has elapsed.

To get on AlarmClock's list of "sleepers," an object must do two things:

1. Ask the alarm clock to wake it up.
2. Implement the wakeUp method.

To satisfy the first requirement, an object calls AlarmClock's letMeSleepFor method, which is implemented like this:

public synchronized boolean letMeSleepFor(Sleeper theSleeper,
                                          long time) {
    int index = findNextSlot();
    if (index == NOROOM) {
        return false;
    } else {
        sleepers[index] = theSleeper;
        sleepFor[index] = time;
        new AlarmThread(index).start();
        return true;
    }
}

If AlarmClock has space, then it registers the sleeper, starts a new AlarmThread for it, and returns true. After the specified amount of time has elapsed the AlarmClock will call theSleeper's wakeUp method.

This leads to the second requirement. An object that wants to use AlarmClock must implement the wakeUp method (so that AlarmClock can call it to notify the object after the time has elapsed). But how is this enforced? It's enforced through the data type of the object being registered.

The first argument to the letMeSleepFor method is the object that wants to get woken up. The data type of this argument is Sleeper, which is the name of this interface:

public interface Sleeper {
    public void wakeUp();

    public long ONE_SECOND = 1000;	// in milliseconds
}

The Sleeper interface defines the wakeUp method but does not implement it. It also defines two useful constants. Classes that implement this interface "inherit" the constants and must implement wakeUp.

Any object that is a Sleeper (and can therefore be passed into letMeSleepFor) implements this interface. This means it implements all of the methods defined by the interface. Thus a Sleeper object implements the wakeUp method, thereby satisfying AlarmClock's second requirement.

For example, check out the following small class that implements the Sleeper interface. The GUIClock class is an applet that displays the current time and uses an AlarmClock object to wake it up every minute so that it can update its display:

class GUIClock extends Applet implements Sleeper {
    . . .
    public void wakeUp() {
        repaint();
        clock.letMeSleepFor(this, ONE_SECOND);
    }
}

Here's the GUIClock applet running:

---

Note: Because some old browsers don't support 1.1, the above applet is a 1.0 version (here is the 1.0 code; here's the 1.1 code). To run the 1.1 version of the applet, go to example-1dot1/GUIClock.html. For more information about running applets, refer to About Our Examples.

---

Now that you've seen an interface in action, we'll answer some of the inevitable questions.

Why Can't I Just Use an Abstract Class?

At this point, many programmers wonder how an interface differs from an abstract class. An interface is simply a list of unimplemented, and therefore abstract, methods. Wouldn't the following Sleeper class do the same thing as the Sleeper interface?

abstract class Sleeper {
    public abstract void wakeUp();
}

No. The two are not equivalent. If Sleeper is an abstract class, then all objects that wish to use AlarmClock must be instances of a class inherited from Sleeper. However, many objects that wish to use AlarmClock already have a superclass. For example, the GUIClock is an Applet; it must be an applet to run inside a browser. But Java doesn't support multiple inheritance. So GUIClock can't be both a Sleeper and an Applet. Hence, you use an interface instead.

This is the practical explanation of the problem. The conceptual explanation is this: AlarmClock should not force a class relationship on its users. It doesn't matter what their class is. It simply matters that they implement a specific method.

Oh! So Interfaces Provide for Multiple Inheritance?

Often interfaces are touted as an alternative to multiple class inheritance. While interfaces may solve similar problems, interface and multiple class inheritance are quite different animals, in particular:

• A class inherits only constants from an interface.
• A class cannot inherit method implementations from an interface.
• The interface hierarchy is independent of the class hierarchy. Classes that implement the same interface may or may not be related through the class hierarchy. This is not true for multiple inheritance.

Yet, Java does allow multiple interface inheritance. That is, an interface can have multiple superinterfaces.

So Tell Me, What Can I Use Interfaces For?

You use an interface to define a protocol of behavior that can be implemented by any class anywhere in the class hierarchy. Interfaces are useful for the following:

• Capturing similarities between unrelated classes without artificially forcing a class relationship
• Declaring methods that one or more classes are expected to implement
• Revealing an object's programming interface without revealing its class. (Objects such as these are called anonymous objects and can be useful when shipping a package of classes to other developers.)

Start of Tutorial > Start of Trail > Start of Lesson

Search