Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Microsoft CSharp Programming For The Absolute Beginner (2002) [eng]-1.pdf
Скачиваний:
46
Добавлен:
16.08.2013
Размер:
15.71 Mб
Скачать

Creating the Critter Class

From all the evidence, a clone would seem to be almost exactly like a critter. Here’s the code for the clone:

using System;

namespace CritClone

{

///<summary>

///The Clone is a very simple class

///Illustrates basic inheritance

///Andy Harris, 12/21/01

///</summary>

public class Clone: Critter

{

// there's nothing here! } // end class

}// end namespace

The most startling part of the Clone class is what isn’t there. The clone has no properties, methods, or constructors, yet it acts as if it has them. The Critter Viewer program uses the talk() method and changes the name property. The Clone class acts much like the Critter class, but I didn’t have to rewrite the critter’s characteristics. The key is in the way the class is derived. Look at this line of code:

public class Clone: Critter

This line defines Clone, but the colon followed by a class name indicates that the Clone is derived from Critter. In other words, I am starting from the Critter class, so the Clone class will start with all the characteristics of the Critter class. Without writing a single new line of code, I’ve made a new class related to an existing class.

Trap Having an exact duplicate of an existing class is pointless without modification. I wanted to show you the concept of inheritance with a clear example. In the rest of this chapter you will learn how to make modifications to your copy so that it has new behavior, methods, and properties.

Because Clone is derived from Critter, it inherits all of Critter’s characteristics. Clone has a talk() method because Critter does, and Clone is derived from Critter. Clone also has access to all of Critter’s other properties and methods.

Improving an Existing Class

Most of the time, you don’t need to build an object completely from the ground up. Usually, your object can be based on another object. The ability to make one class based on another class is the foundation of inheritance. For example, imagine that you are a car designer, and you have defined a car factory that produces a standard car. If you want to build taxis and police cars, you don’t have to start from scratch because your new car can inherit all the basic characteristics (wheels, doors, engines) of its parent class, the standard car. To make a taxi class, you start with a car class and add the characteristics that make it a taxi (a horn that blows constantly and the capability to splash pedestrians are requisite features).

113

Most projects involve using classes in as many as three ways. You use existing classes, such as the Console and Convert classes. You sometimes make entirely new objects, as I did with the Critter. Also, you frequently modify existing classes and add new capabilities to them.

Introducing the Glitter Critter

To illustrate how a new class can modify an existing class, look at the glitter critter. This is a variation of the Critter class that has a new method, shine(). You can see the glitter critter in all its glory in Figure 5.9.

Figure 5.9: The glitter critter is like normal critters, but it has a shine() method.

The GlitterCritter class also supports all the constructors of the Critter class. Because most of the behavior already belongs to the Critter class, the GlitterCritter code is dedicated mainly to the new elements of this new class:

using System;

namespace GlitCrit

{

///<summary>

///New version of Critter that adds a "shine" method

///</summary>

public class GlitterCritter : Critter

{

//overloaded constructors map to base (Critter) constructors public GlitterCritter(): base(){ }

public GlitterCritter(string name): base(name){ }

public GlitterCritter(string theName, int fullness, int happiness, int theAge)

: base(theName, fullness, happiness, theAge){ }

//new shine method public void shine(){

Console.WriteLine(name + " shines beautifully.");

} // end shine

} // end GlitterCritter

}// end namespace

The GlitterCritter sports three constructors and a method. It inherits everything else from its parent class, the Critter.

114

The joy of inheritance is that you don’t have to keep reinventing the wheel. You can often base your class on an existing class and have automatic use of all its methods. If the class you are deriving from is inherited from another class, you have access to all the properties and methods of the grandparent class as well. Returning to the car example, you can have a sedan class, which is a standard car. You can derive a heavy−duty sedan class with the same features as a sedan but modifications for the functions of a commercial vehicle. You can have a taxi class with features exclusive to taxis (the meter), features of the heavy−duty sedan class (a special suspension), and many features of the original sedan class (steering wheel, doors, and so on). After you build the heavy−duty sedan class, making new commercial vehicles is easier because all you have to worry about are those characteristics that distinguish a class from its parent.

In the Real World

Part of being an object−oriented programmer is knowing the humor. You’re now ready for the classic object−oriented programming joke:

How many object−oriented programmers does it take to change a light bulb?

None! You should inherit the change() method from the light bulb’s parent class!

Calling the Base Class’s Constructors

The code for the GlitterCritter class has several constructors that are very simple, but not much like the other methods you’ve come to know. For example, the default (no−parameters) constructor looks like this:

public GlitterCritter(): base(){ }

The colon after the constructor name lets you specify which constructor of the base (parent) class you want to call. In this case, if the user decides to instantiate GlitterCritter with no parameters, it calls the Critter class constructor, also with no parameters. The constructor has a pair of braces so that you can put code in it. However, inherited constructors often don’t need much code because most of the building is done in the parent class. You can often leave the constructor code blank in an inherited constructor. The only time this isn’t true is when you need to initialize a value or property that belongs to the child class but not to its parent. I added other constructors to the GlitterCritter class. They are very similar to the no−parameters version:

public GlitterCritter(string name): base(name){ }

public GlitterCritter(string theName, int fullness, int happiness, int theAge)

: base(theName, fullness, happiness, theAge){ }

Each constructor must have a different parameter signature. For the GlitterCritter, I decided to use exactly the same types of parameters as the original Critter, so I had each constructor map to the similar constructor of Critter. I didn’t need additional code in the GlitterCritter’s constructors, so I left them blank for now.

115

Adding Methods to a New Class

The remaining feature of the GlitterCritter class is the new method, shine(). Creating a new method in an inherited class is easy. Usually, you create the method just as you did in the original class. However, if the method existed in the base class, you must think more carefully about what you want your method to do.

Changing the Critter Viewer Again

Again, I modified the critter viewer to test my new critter’s behavior. Here’s the new version of the critter viewer:

using System;

namespace GlitCrit

{

///<summary>

///Another Critter Viewer

///Adding capabilities to an inherited class

///Andy Harris, 12/21/01

///</summary>

class CritViewer {

static void Main(string[] args) { CritViewer cv = new CritViewer();

} // end main

public CritViewer(){

GlitterCritter gc = new GlitterCritter("Sparky"); gc.shine();

Console.WriteLine(gc.talk());

Console.WriteLine("Please press Enter to continue");

Console.ReadLine();

} // end constructor

} // end CritViewer class

}// end namespace

The only new code in this program invokes the shine() method of the glitter critter.

Using Polymorphism to Alter a Class’s Behavior

Adding new methods to an inherited class is easy, but sometimes you don’t want to add a new behavior as much as modify an existing behavior. For example, the critter is a pleasant creature, but suppose that you want (for some bizarre reason) to make a grumpier critter. This bitter critter should have a talk() method, but that method should not be just like the standard cheerful Critter talk. The bitter critter’s talk() method should have the same purpose but should be written differently.

This situation is exactly what polymorphism is meant to solve.

When your new class has a method, and its parent has the same method with the same signature, you create an opportunity for confusion. If both Critter and BitterCritter have a talk() method, you must indicate which version should be invoked. C# allows you to specify that you wish to override a method of a parent class. The technique, called method overriding, also prevents you from overriding an existing method without knowing that you’re doing it. I’ll describe these mechanisms

116

Соседние файлы в предмете Программирование на C++