The Life Cycle of an Object

A typical program creates many objects, which interact by sending messages. Through these object interactions, a program can implement a GUI, run an animation, or send and receive information over a network. Once an object has completed the work for which it was created, its resources are recycled for use by other objects.

Here's a small program, called CreateObjectDemo, [1] that creates three objects: one Point object and two Rectangle objects. [2]

[1] CreateObjectDemo.java, Rectangle.java, and Point.java are included on the CD and are available online. See Code Samples (page 174).

[2] This example contains code for the Point and Rectangle classes so that you can see code for simple classes. However, instead of our Point and Rectangle classes, you should use the ones provided by the Java platform in the java.awt package.

public class CreateObjectDemo { 

 public static void main(String[] args) { 
 // declare and create a point object and two rectangle objects 
 Point origin_one = new Point(23, 94); 
 Rectangle rect_one = new Rectangle(origin_one, 100, 200); 
 Rectangle rect_two = new Rectangle(50, 100); 

 // display rect_one's width, height, and area 
 System.out.println("Width of rect_one: " + rect_one.width); 
 System.out.println("Height of rect_one: " + rect_one.height); 
 System.out.println("Area of rect_one: " + rect_one.area()); 

 // set rect_two's position 
 rect_two.origin = origin_one; 

 // display rect_two's position 
 System.out.println("X Position of rect_two: " 
 + rect_two.origin.x); 
 System.out.println("Y Position of rect_two: " 
 + rect_two.origin.y); 

 // move rect_two and display its new position 
 rect_two.move(40, 72); 
 System.out.println("X Position of rect_two: " 
 + rect_two.origin.x); 
 System.out.println("Y Position of rect_two: " 
 + rect_two.origin.y); 
 } 
} 

After creating the objects, this program manipulates the objects and displays some information about them. Here's the output from the program:

Width of rect_one: 100 
Height of rect_one: 200 
Area of rect_one: 20000 
X Position of rect_two: 23 
Y Position of rect_two: 94 
X Position of rect_two: 40 
Y Position of rect_two: 72 

This section uses this example to describe the life cycle of an object within a program. From this, you can learn how to write code that creates and uses an object and how the system cleans it up.

Creating Objects

As you know, a class provides the blueprint for objects; you create an object from a class. Each of the following statements taken from the CreateObjectDemo program creates an object:

Point origin_one = new Point(23, 94); 
Rectangle rect_one = new Rectangle(origin_one, 100, 200); 
Rectangle rect_two = new Rectangle(50, 100); 

The first line creates an object from the Point class. The second and third lines each create an object from the Rectangle class.

Each statement has the following three parts:

1. Declaration: The code set in boldface in the previous listing are all variable declarations that associate a name with a type. When you create an object, you do not have to declare a variable to refer to it. However, a variable declaration often appears on the same line as the code to create an object.
2. Instantiation: The new keyword is an operator that creates the new object (allocates space for it).
3. Initialization: The new operator is followed by a call to a constructor. For example, Point(23, 94) is a call to Point's only constructor. The constructor initializes the new object.

Declaring a Variable to Refer to an Object

From the section Variables (page 67) in Chapter 3, you learned that to declare a variable, you write:

type name 

This notifies the Java platform that you will use name to refer to a particular item of data whose type is type .

In addition to the primitive types, such as int and boolean, provided directly by the Java programming language, classes and interfaces are also types. To declare a variable to refer to an object, you use the name of a class or an interface as the variable's type. The sample program uses both the Point and the Rectangle class names as types to declare variables:

Point origin_one = new Point(23, 94); 
Rectangle rect_one = new Rectangle(origin_one, 100, 200); 
Rectangle rect_two = new Rectangle(50, 100); 

Declarations do not create new objects. The code Point origin_one does not create a new Point object; it just declares a variable, named origin_one, that will be used to refer to a Point object. The reference is empty until assigned, as illustrated in Figure 45. An empty reference is known as a null reference.

Figure 45. A reference variable, such as origin_one, contains a null reference until a value is assigned to it.

To create an object, you instantiate a class with the new operator.

Instantiating an Object

The new operator instantiates a class by allocating memory for a new object. The new operator requires a single, postfix argument: a call to a constructor. The name of the constructor provides the name of the class to instantiate. The constructor initializes the new object.

The new operator returns a reference to the object it created. Often, this reference is assigned to a variable of the appropriate type. If the reference is not assigned to a variable, the object is unreachable after the statement in which the new operator appears finishes executing.

Initializing and Object

Here's the code for the Point [1] class:

[1] Point.java is included on the CD and is available online. See Code Samples (page 174).

public class Point { 
 public int x = 0; 
 public int y = 0; 
 //A constructor! 
 public Point(int x, int y) { 
 this.x = x; 
 this.y = y; 
 } 
} 

This class contains a single constructor. You can recognize a constructor because it has the same name as the class and has no return type. The constructor in the Point class takes two integer arguments, as declared by the code (int x, int y). The following statement provides 23 and 94 as values for those arguments:

Point origin_one = new Point(23, 94); 

The effect of the previous line of code can be illustrated in Figure 46:

Figure 46. The variable origin_one contains a reference to the Point object.

Here's the code for the Rectangle [1] class, which contains four constructors:

[1] Rectangle.java is included on the CD and is available online. See Code Samples (page 174).

public class Rectangle { 
 public int width = 0; 
 public int height = 0; 
 public Point origin; 

 //Four constructors 
 public Rectangle() { 
 origin = new Point(0, 0); 
 } 

 public Rectangle(Point p) { 
 origin = p; 
 } 

 public Rectangle(int w, int h) { 
 this(new Point(0, 0), w, h); 
 } 

 public Rectangle(Point p, int w, int h) { 
 origin = p; 
 width = w; 
 height = h; 
 } 
 //A method for moving the rectangle 
 public void move(int x, int y) { 
 origin.x = x; 
 origin.y = y; 
 } 

 //A method for computing the area of the rectangle 
 public int area() { 
 return width * height; 
 } 
} 

Each constructor lets you provide initial values for different aspects of the rectangle: the origin, width, and height. You can set the values for all three or none. If a class has multiple constructors, they all have the same name but a different number of arguments or different typed arguments. The Java platform differentiates the constructors, based on the number and the type of the arguments. When the Java platform encounters the following code, it knows to call the constructor in the Rectangle class that requires a Point argument followed by two integer arguments:

Rectangle rect_one = new Rectangle(origin_one, 100, 200); 

This call initializes the rectangle's origin variable to the Point object referred to by origin_one. The code also sets width to 100 and height to 200. Now there are two references to the same Point object; an object can have multiple references to it, as shown in Figure 47:

Figure 47. Multiple references can refer to the same object.

The following line of code calls the constructor that requires two integer arguments, which provide the initial values for width and height. If you inspect the code within the constructor, you will see that it creates a new Point object whose x and y values are initialized to 0:

Rectangle rect_two = new Rectangle(50, 100); 

The Rectangle constructor used in the following statement doesn't take any arguments, so it's called a no-argument constructor:

Rectangle rect = new Rectangle(); 

If a class does not explicitly declare any constructors, the Java platform automatically provides a no-argument constructor, called the default constructor, that does nothing. Thus, all classes have at least one constructor.

This section talked about how to use a constructor. The section Providing Constructors for Your Classes (page 186) explains how to write constructors for your classes.

Using Objects

Once you've created an object, you probably want to use it for something. You may need information from it, want to change its state, or have it perform some action. You can use an object in one of two ways:

• Manipulate or inspect its variables
• Call its methods

Referencing an Object's Variables

The following is the general form of a qualified name, which is also known as a long name:

objectReference.variableName 

You may use a simple name for an instance variable when the instance variable is in scopethat is, within code for the object's class. Code that is outside the object's class must use a qualified name. For example, the code in the CreateObjectDemo class is outside the code for the Rectangle class. So to refer to the origin, width, and height variables within the Rectangle object named rect_one, the CreateObjectDemo class must use the names rect_one.origin, rect_one.width, and rect_one.height, respectively. The program uses two of these names to display the width and the height of rect_one:

System.out.println("Width of rect_one: " + rect_one.width); 
System.out.println("Height of rect_one: " + rect_one.height); 

Attempting to use the simple names width and height from the code in the CreateObjectDemo class doesn't make sensethose variables exist only within an objectand results in a compiler error.

Later, the program uses similar code to display information about rect_two. Objects of the same type have their own copy of the same instance variables. Thus, each Rectangle object has variables named origin, width, and height. When you access an instance variable through an object reference, you reference that particular object's variable. The two objects rect_one and rect_two in the CreateObjectDemo program have different origin, width, and height variables.

The first part of the variable's qualified name, objectReference , must be a reference to an object. You can use the name of a reference variable here, as in the previous examples, or you can use any expression that returns an object reference. Recall that the new operator returns a reference to an object. So you could use the value returned from new to access a new object's variables:

int height = new Rectangle().height; 

This statement creates a new Rectangle object and immediately gets its height. In essence, the statement calculates the default height of a Rectangle. Note that after this statement has been executed, the program no longer has a reference to the created Rectangle, because the program never stored the reference in a variable. The object is unreferenced, and its resources can be recycled by the Java platform.

A word about Variable Access

The direct manipulation of an object's variables by other objects and classes is discouraged because it's possible to set the variables to values that don't make sense. For example, consider the Rectangle class from the previous section. Using that class, you can create a rectangle whose width and height are negative, which, for some applications, doesn't make sense.

Ideally, instead of allowing direct manipulation of variables, a class would provide methods through which other objects can inspect or change variables. These methods ensure that the values of the variables make sense for objects of that type. Thus, the Rectangle class would provide methods called setWidth, setHeight, getWidth, and getHeight for setting and getting the width and the height. The methods for setting the variables would report an error if the caller tried to set the width or the height to a negative number. The other advantage of using methods instead of direct variable access is that the class can change the type and the names of the variables it uses for storing the width and the height without affecting its clients.

However, in practical situations, it sometimes makes sense to allow direct access to an object's variables. For example, both the Point class and the Rectangle class allow free access to their member variables by declaring them public. This keeps these classes small and simple. Also, it keeps them generally useful. Some applications might allow rectangles with negative widths and heights.

The Java programming language provides an access control mechanism whereby classes can determine what other classes can have direct access to its variables. A class should protect variables against direct manipulation by other objects if those manipulations could result in values that don't make sense for objects of that type. Changes to these variables should be controlled by method calls. If a class grants access to its variables, you can assume that you can inspect and change those variables without adverse effects. To learn more about the access control mechanism, refer to the section Controlling Access to Members of a Class (page 193) in Chapter 5. Also, by making the variables accessible, they become part of the class's API, which means that the writer of the class should not change their names or their types.

Calling an Object's Methods

You also use qualified names to call an object's method. To form the qualified name of a method, you append the method name to an object reference, with an intervening period (.). Also, you provide, within enclosing parentheses, any arguments to the method. If the method does not require any arguments, use empty parentheses.

objectReference.methodName(argumentList); 

or

objectReference.methodName(); 

The Rectangle class has two methods: area to compute the rectangle's area and move to change the origin. Here's the CreateObjectDemo code that calls these two methods:

System.out.println("Area of rect_one: " + rect_one.area()); 
... 
rect_two.move(40, 72); 

The first statement calls rect_one's area method and displays the results. The second line moves rect_two because the move method assigns new values to the object's origin.x and origin.y.

As with instance variables, objectReference must be a reference to an object. You can use a variable name, but you also can use any expression that returns an object reference. The new operator returns an object reference, so you can use the value returned from new to call a new object's methods:

new Rectangle(100, 50).area() 

The expression new Rectangle(100, 50) returns an object reference that refers to a Rectangle object. As shown, you can use the dot notation to call the new Rectangle's area method to compute the area of the new rectangle.

Some methods, such as area, return a value. For methods that return a value, you can use the method call in expressions. You can assign the return value to a variable, use it to make decisions, or control a loop. This code assigns the value returned by area to a variable:

int areaOfRectangle = new Rectangle(100, 50).area(); 

Remember, invoking a method on a particular object is the same as sending a message to that object. In this case, the object that area is invoked on is the rectangle returned by the constructor.

A Word about Method Access

The methods in our Point and Rectangle classes are all declared public, so they are accessible to any other class. Sometimes, a class needs to restrict access to its methods. For example, a class might have a method that only subclasses are allowed to call. A class can use the same mechanism to control access to its methods as it uses to control access to its variables. To learn more about the access control mechanism, refer to the section Controlling Access to Members of a Class (page 193).

Cleaning Up Unused Objects

Some object-oriented languages require that you keep track of all the objects you create and that you explicitly destroy them when they are no longer needed. Managing memory explicitly is tedious and error prone. The Java platform allows you to create as many objects as you want (limited, of course, by what your system can handle), and you don't have to worry about destroying them. The Java runtime environment deletes objects when it determines that they are no longer being used. This process is called garbage collection.

An object is eligible for garbage collection when there are no more references to that object. References that are held in a variable are usually dropped when the variable goes out of scope. Or, you can explicitly drop an object reference by setting the variable to the special value null. Remember that a program can have multiple references to the same object; all references to an object must be dropped before the object is eligible for garbage collection.

The Garbage Collector

The Java runtime environment has a garbage collector that periodically frees the memory used by objects that are no longer referenced. The garbage collector does its job automatically, although, in some situations, you may want to run the garbage collection explicitly by calling the gc method in the System class. For instance, you might want to run the garbage collector after a section of code that creates a large amount of garbage or before a section of code that needs a lot of memory.

Finalization

Before an object gets garbage-collected, the garbage collector gives the object an opportunity to clean up after itself through a call to the object's finalize method. This process is known as finalization.

Most programmers don't have to worry about implementing the finalize method. In rare cases, however, a programmer might have to implement a finalize method to release resources, such as native peers, that aren't under the control of the garbage collector.

The finalize method is a member of the Object class, which is the top of the Java platform's class hierarchy and a superclass of all classes. A class can override the finalize method to perform any finalization necessary for objects of that type. If you override finalize, your implementation of the method should call super.finalize as the last thing it does. The section Overriding and Hiding Methods (page 204) talks more about how to override methods.

Summary of Objects

You create an object from a class by using the new operator and a constructor. The new operator returns a reference to the object that was created. You can assign the reference to a variable or use it directly.

A class controls access to its instance variables and methods by using the Java platform's access mechanism. Instance variables and methods that are accessible to code outside of the class that they are declared in can be referred to by using a qualified name. The qualified name of an instance variable looks like this:

objectReference.variableName 

The qualified name of a method looks like this:

objectReference.methodName(argumentList) 
 or 
objectReference.methodName() 

The garbage collector automatically cleans up unused objects. An object is unused if the program holds no more references to it. You can explicitly drop a reference by setting the variable holding the reference to null.

Questions and Exercises: Objects

Some of the following questions and exercises refer to the Point and Rectangle classes. Use the classes shown on page 123 and page 124, respectively.

Questions

1:	What's wrong with the following program? public class SomethingIsWrong { public static void main(String[] args) { Rectangle myRect; myRect.width = 40; myRect.height = 50; System.out.println("myRect's area is " + myRect.area()); } }
2:	The following code creates one Point object and one Rectangle object. How many references to those objects exist after the code executes? Is either object eligible for garbage collection? ... Point point = new Point(2,4); Rectangle rectangle = new Rectangle(point, 20, 20); point = null; ...
3:	How does a program destroy an object that it creates?

Exercises

1:	Fix the program called SomethingIsWrong shown in question 1.
2:	Given the following class, write some code that creates an instance of the class, initializes its two member variables, and then displays the value of each member variable. public class NumberHolder { public int anInt; public float aFloat; }

Answers

You can find answers to these Questions and Exercises online:

http://java.sun.com/docs/books/tutorial/java/data/QandE/objects-answers.html