Java Programming Tutorial
Object-oriented Programming (OOP) Basics
Suppose that you want to assemble your own PC, you go to a hardware store and pick up a motherboard, a processor, some RAMs, a hard disk, a casing, a power supply, and put them together.В You turn on the power, and the PC runs.В You need not worry whether the CPU is 1-core or 6-core; the motherboard is a 4-layer or 6-layer; the hard disk has 4 plates or 6 plates, 3 inches or 5 inches in diameter; the RAM is made in Japan or Korea, and so on. You simply put the hardware components together and expect the machine to run.В Of course, you have to make sure that you have the correct interfaces, i.e., you pick an IDE hard disk rather than a SCSI hard disk, if your motherboard supports only IDE; you have to select RAMs with the correct speed rating, and so on.В Nevertheless, it is not difficult to set up a machine from hardware components.
Similarly, a car is assembled from parts and components, such as chassis, doors, engine, wheels, brake and transmission. The components are reusable, e.g., a wheel can be used in many cars (of the same specifications).
Hardware, such as computers and cars, are assembled from parts, which are reusable hardware components.
How about software? В Can you "assemble" a software application by picking a routine here, a routine there, and expect the program to run?В The answer is obviously NO!В Unlike hardware, it is very difficult to "assemble" an application from software components.В Since the advent of computer 70 years ago, we have written tons and tons of programs and routines.В However, for each new application, we have to re-invent the wheels and write the program from scratch!
Why re-invent the wheels? Why re-writing codes? Can you write better codes than those codes written by the experts?
Traditional Procedural-Oriented languages
Traditional procedural-oriented programming languages (such as C, Fortran, Cobol and Pascal) suffer some notable drawbacks in creating reusable software components:
- The procedural-oriented programs are made up of functions. Functions are less reusable. It is very difficult to copy a function from one program and reuse in another program because the function is likely to reference the global variables and other functions. In other words, functions are not well-encapsulated as a self-contained reusable unit.
- The procedural languages are not suitable of high-level abstraction for solving real life problems. For example, C programs uses constructs such as if-else, for-loop, array, method, pointer, which are low-level and hard to abstract real problems such as a Customer Relationship Management (CRM) system or a computer soccer game.
The traditional procedural-languages separate the data structures (variables) and algorithms (functions).
In the early 1970s, the US Department of Defense (DoD) commissioned a task force to investigate why its IT budget always went out of control; but without much to show for. The findings are:
- 80% of the budget went to the software (with the remaining 20% to the hardware).
- More than 80% of the software budget went to maintenance (only the remaining 20% for new software development).
- Hardware components could be applied to various products, and their integrity normally did not affect other products. (Hardware can share and reuse! Hardware faults are isolated!)
- Software procedures were often non-sharable and not reusable. Software faults could affect other programs running in computers.
The task force proposed to make software behave like hardware OBJECT. Subsequently, DoD replaces over 450 computer languages, which were then used to build DoD systems, with an object-oriented language called Ada.
Object-Oriented Programming Languages
Object-oriented programming (OOP) languages are designed to overcome these problems.
- The basic unit of OOP is a class, which encapsulates both the static properties and dynamic operations within a "box", and specifies the public interface for using these boxes. Since classes are well-encapsulated, it is easier to reuse these classes. In other words, OOP combines the data structures and algorithms of a software entity inside the same box.
- OOP languages permit higher level of abstraction for solving real-life problems. The traditional procedural language (such as C and Pascal) forces you to think in terms of the structure of the computer (e.g. memory bits and bytes, array, decision, loop) rather than thinking in terms of the problem you are trying to solve. The OOP languages (such as Java, C++ and C#) let you think in the problem space, and use software objects to represent and abstract entities of the problem space to solve the problem.
As an example, suppose you wish to write a computer soccer games (which I consider as a complex application). It is quite difficult to model the game in procedural-oriented languages. But using OOP languages, you can easily model the program accordingly to the "real things" appear in the soccer games.
- Player: attributes include name, number, location in the field, and etc; operations include run, jump, kick-the-ball, and etc.
Most importantly, some of these classes (such as Ball and Audience ) can be reused in another application, e.g., computer basketball game, with little or no modification.
Benefits of OOP
The procedural-oriented languages focus on procedures, with function as the basic unit. You need to first figure out all the functions and then think about how to represent data.
The object-oriented languages focus on components that the user perceives, with objects as the basic unit. You figure out all the objects by putting all the data and operations that describe the user’s interaction with the data.
Object-Oriented technology has many benefits:
- Ease in software design as you could think in the problem space rather than the machine’s bits and bytes. You are dealing with high-level concepts and abstractions. Ease in design leads to more productive software development.
- Ease in software maintenance: object-oriented software are easier to understand, therefore easier to test, debug, and maintain.
- Reusable software: you don’t need to keep re-inventing the wheels and re-write the same functions for different situations. The fastest and safest way of developing a new application is to reuse existing codes – fully tested and proven codes.
OOP in Java
Class & Instances
In Java, a class is a definition of objects of the same kind. In other words, a class is a blueprint, template, or prototype that defines and describes the static attributes and dynamic behaviors common to all objects of the same kind.
An instance is a realization of a particular item of a class. In other words, an instance is an instantiation of a class. All the instances of a class have similar properties, as described in the class definition. For example, you can define a class called " Student " and create three instances of the class " Student " for " Peter ", " Paul " and " Pauline ".
The term "object" usually refers to instance. But it is often used loosely, and may refer to a class or an instance.
A Class is a 3-Compartment Box Encapsulating Data and Operations
A class can be visualized as a three-compartment box, as illustrated:
- Name (or identity): identifies the class.
- Variables (or attribute, state, field): contains the static attributes of the class.
- Methods (or behaviors, function, operation): contains the dynamic behaviors of the class.
In other words, a class encapsulates the static attributes (data) and dynamic behaviors (operations that operate on the data) in a box.
The followings figure shows a few examples of classes:
The following figure shows two instances of the class Student , identified as " paul " and " peter ".
Unified Modeling Language (UML) Class and Instance Diagrams: The above class diagrams are drawn according to the UML notations. A class is represented as a 3-compartment box, containing name, variables, and methods, respectively. Class name is shown in bold and centralized. An instance is also represented as a 3-compartment box, with instance name shown as instanceName:Classname and underlined.
- A class is a programmer-defined, abstract, self-contained, reusable software entity that mimics a real-world thing.
- A class is a 3-compartment box containing the name, variables and the methods.
- A class encapsulates the data structures (in variables) and algorithms (in methods). The values of the variables constitute its state. The methods constitute its behaviors.
- An instance is an instantiation (or realization) of a particular item of a class.
Class Definition in Java
In Java, we use the keyword class to define a class. For examples:
The syntax for class definition in Java is:
We shall explain the access control modifier, such as public and private , later.
Class Naming Convention: A class name shall be a noun or a noun phrase made up of several words. All the words shall be initial-capitalized (camel-case). Use a singular noun for class name. Choose a meaningful and self-descriptive classname. For examples, SoccerPlayer , HttpProxyServer , FileInputStream , PrintStream and SocketFactory .
Creating Instances of a Class
To create an instance of a class, you have to:
- Declare an instance identifier (instance name) of a particular class.
- Construct the instance (i.e., allocate storage for the instance and initialize the instance) using the " new " operator.
For examples, suppose that we have a class called Circle , we can create instances of Circle as follows:
When an instance is declared but not constructed, it holds a special value called null .
Dot ( . ) Operator
The variables and methods belonging to a class are formally called member variables and member methods. To reference a member variable or method, you must:
- First identify the instance you are interested in, and then,
- Use the dot operator ( . ) to reference the desired member variable or method.
For example, suppose that we have a class called Circle , with two member variables ( radius and color ) and two member methods ( getRadius() and getArea() ). We have created three instances of the class Circle , namely, c1 , c2 and c3 . To invoke the method getArea() , you must first identity the instance of interest, says c2 , then use the dot operator, in the form of c2.getArea() .
Calling getArea() without identifying the instance is meaningless, as the radius is unknown (there could be many instances of Circle – each maintaining its own radius). Furthermore, c1.getArea() and c2.getArea() are likely to produce different results.
In general, suppose there is a class called AClass with a member variable called aVariable and a member method called aMethod() . An instance called anInstance is constructed for AClass . You use anInstance.aVariable and anInstance.aMethod() .
A member variable has a name (or identifier) and a type; and holds a value of that particular type (as descried in the earlier chapter).
Variable Naming Convention: A variable name shall be a noun or a noun phrase made up of several words. The first word is in lowercase and the rest of the words are initial-capitalized (camel-case), e.g., fontSize , roomNumber , xMax , yMin and xTopLeft .
The formal syntax for variable definition in Java is:
A method (as described in the earlier chapter):
- receives arguments from the caller,
- performs the operations defined in the method body, and
- returns a piece of result (or void ) to the caller.
The syntax for method declaration in Java is as follows:
Method Naming Convention: A method name shall be a verb, or a verb phrase made up of several words. The first word is in lowercase and the rest of the words are initial-capitalized (camel-case). For example, getArea() , setRadius() , getParameterValues() , hasNext() .
Variable name vs. Method name vs. Class name : A variable name is a noun, denoting an attribute; while a method name is a verb, denoting an action. They have the same naming convention (the first word in lowercase and the rest are initial-capitalized). Nevertheless, you can easily distinguish them from the context. Methods take arguments in parentheses (possibly zero arguments with empty parentheses), but variables do not. In this writing, methods are denoted with a pair of parentheses, e.g., println() , getArea() for clarity.
On the other hand, class name is a noun beginning with uppercase.
Putting them Together: An OOP Example
A class called Circle is defined as shown in the class diagram. It contains two private member variables: radius (of type double ) and color (of type String ); and three public member methods: getRadius() , getColor() , and getArea() .
Three instances of Circle s, called c1 , c2 , and c3 , shall be constructed with their respective data members, as shown in the instance diagrams.
The source codes for Circle.java is as follows:
Compile " Circle.java " into " Circle.class ".
Notice that the Circle class does not have a main() method. Hence, it is NOT a standalone program and you cannot run the Circle class by itself. The Circle class is meant to be a building block – to be used in other programs.
We shall now write another class called TestCircle , which uses the Circle class. The TestCircle class has a main() method and can be executed.
Compile TestCircle.java into TestCircle.class .
Run the TestCircle and study the output:
A constructor looks like a special method that has the same method name as the class name. In the above Circle class, we define three overloaded versions of constructor Circle(. ) . A constructor is used to construct and initialize all the member variables. To construct a new instance of a class, you need to use a special " new " operator followed by a call to one of the constructors. For example,
A constructor is different from an ordinary method in the following aspects:
- The name of the constructor method is the same as the class name. By classname’s convention, it begins with an uppercase (instead of lowercase for ordinary methods).
- Constructor has no return type. It implicitly returns void . No return statement is allowed inside the constructor’s body.
- Constructor can only be invoked via the " new " operator. It can only be used once to initialize the instance constructed. Once an instance is constructed, you cannot call the constructor anymore.
- Constructors are not inherited (to be explained later).
Default Constructor : A constructor with no parameter is called the default constructor. It initializes the member variables to their default value. For example, the Circle() in the above example initialize member variables radius and color to their default value.
Method Overloading (Revisit)
Method overloading means that the same method name can have different implementations (versions). However, the different implementations must be distinguishable by their parameter list (either the number of parameters, or the type of parameters, or their order).
Example: The method average() has 3 versions, with different parameter lists. The caller can invoke the chosen version by supplying the matching arguments.
Overloading Circle Class’ Constructor
Constructor, like an ordinary method, can also be overloaded. The above Circle class has three overloaded versions of constructors differentiated by their parameter list, as followed:
Depending on the actual argument list used when invoking the method, the matching constructor will be invoked. If your argument list does not match any one of the methods, you will get a compilation error.
public vs. private – Access Control Modifiers
An access control modifier can be used to control the visibility of a class, or a member variable or a member method within a class. We begin with the following two access control modifiers:
- public : The class/variable/method is accessible and available to all the other objects in the system.
- private : The class/variable/method is accessible and available within this class only.
For example, in the above Circle definition, the member variable radius is declared private . As the result, radius is accessible inside the Circle class, but NOT in the TestCircle class. In other words, you cannot use " c1.radius " to refer to c1 ‘s radius in TestCircle .
- Try inserting the statement " System.out.println(c1.radius) " in TestCircle and observe the error message.
- Try changing radius to public in the Circle class, and re-run the above statement.
On the other hand, the method getRadius() is declared public in the Circle class. Hence, it can be invoked in the TestCircle class.
UML Notation: In UML class diagram, public members are denoted with a " + "; while private members with a " – ".
More access control modifiers will be discussed later.
Information Hiding and Encapsulation
A class encapsulates the name, static attributes and dynamic behaviors into a "3-compartment box". Once a class is defined, you can seal up the "box" and put the "box" on the shelve for others to use and reuse. Anyone can pick up the "box" and use it in their application. This cannot be done in the traditional procedural-oriented language like C, as the static attributes (or variables) are scattered over the entire program and header files. You cannot "cut" out a portion of C program, plug into another program and expect the program to run without extensive changes.
Member variables of a class are typically hidden from the outside word (i.e., the other classes), with private access control modifier. Access to the member variables are provided via public assessor methods, e.g., getRadius() and getColor() .
This follows the principle of information hiding. That is, objects communicate with each others using well-defined interfaces (public methods). Objects are not allowed to know the implementation details of others. The implementation details are hidden or encapsulated within the class. Information hiding facilitates reuse of the class.
Rule of Thumb: Do not make any variable public , unless you have a good reason.
The public Getters and Setters for private Variables
To allow other classes to read the value of a private variable says xxx , we provide a get method (or getter or accessor method) called getXxx() . A get method needs not expose the data in raw format. It can process the data and limit the view of the data others will see. The getters shall not modify the variable.
To allow other classes to modify the value of a private variable says xxx , we provide a set method (or setter or mutator method) called setXxx() . A set method could provide data validation (such as range checking), or transform the raw data into the internal representation.
For example, in our Circle class, the variables radius and color are declared private . That is to say, they are only accessible within the Circle class and not visible in any other classes, such as the TestCircle class. You cannot access the private variables radius and color from the TestCircle class directly – via says c1.radius or c1.color . The Circle class provides two public accessor methods, namely, getRadius() and getColor() . These methods are declared public . The class TestCircle can invoke these public accessor methods to retrieve the radius and color of a Circle object, via says c1.getRadius() and c1.getColor() .
There is no way you can change the radius or color of a Circle object, after it is constructed in the TestCircle class. You cannot issue statements such as c1.radius = 5.0 to change the radius of instance c1 , as radius is declared as private in the Circle class and is not visible to other classes including TestCircle .
If the designer of the Circle class permits the change the radius and color after a Circle object is constructed, he has to provide the appropriate set methods (or setters or mutator methods), e.g.,
With proper implementation of information hiding, the designer of a class has full control of what the user of the class can and cannot do.
Keyword " this "
You can use keyword " this " to refer to this instance inside a class definition.
One of the main usage of keyword this is to resolve ambiguity.
In the above codes, there are two identifiers called radius – a member variable of the class and the method’s argument. This causes naming conflict. To avoid the naming conflict, you could name the method’s argument r instead of radius . However, radius is more approximate and meaningful in this context. Java provides a keyword called this to resolve this naming conflict. " this.radius " refers to the member variable; while " radius " resolves to the method’s argument.
Using the keyword " this ", the constructor, getter and setter methods for a private variable called xxx of type T are as follows:
For a boolean variable xxx , the getter shall be named isXxx() or hasXxx() , which is more meaningful than getXxx() . The setter remains setXxx() .
More on " this "
- this.varName refers to varName of this instance; this.methodName(. ) invokes methodName(. ) of this instance.
- In a constructor, we can use this(. ) to call another constructor of this class.
- Inside a method, we can use the statement " return this " to return this instance to the caller.
Every well-designed Java class should have a public method called toString() that returns a string description of this instance. You can invoke the toString() method explicitly by calling anInstanceName.toString() , or implicitly via println() or String concatenation operator ‘+’ . That is, running println(anInstance) invokes the toString() method of that instance implicitly.
For example, include the following toString() method in our Circle class:
In your TestCircle class, you can get a description of a Circle instance via:
The signature of toString() is:
Constants ( final )
Constants are variables defined with the modifier final . A final variable can only be assigned once and its value cannot be modified once assigned. For example,
Constant Naming Convention: A constant name is a noun, or a noun phrase made up of several words. All words are in uppercase separated by underscores ‘ _ ‘, for examples, X_REFERENCE , MAX_INTEGER and MIN_VALUE .
- A final primitive variable cannot be re-assigned a new value.
- A final instance cannot be re-assigned a new object.
- A final class cannot be sub-classed (or extended).
- A final method cannot be overridden.
Putting Them Together in the Revised Circle Class
We shall include constructors, getters, setters, toString() , and use the keyword " this ". The class diagram for the final Circle class is as follows:
A Test Driver for the Circle Class
The expected outputs are:
More Examples on Classes
EG. 1: The Account Class
A class called Account , which models a bank account, is designed as shown in the class diagram. It contains the following members:
- Two private instance variables: accountNumber ( int ), and balance ( double ) which maintains the current account balance.
- Constructors (overloaded).
- Getters and Setters for the private instance variables. There is no setter for accountNumber as it is not designed to be changed.
- public methods credit() and debit() , which adds/subtracts the given amount to/from the balance, respectively.
- A toString() , which returns " A/C no:xxx, Balance=$xxx.xx ", with balance rounded to two decimal places.
Write the Account class and a test driver to test all the public methods.
The Account Class ( Account.java )
A Test Driver for the Account Class ( TestAccount.java )
The expected outputs are:
EG. 2: The Date class
A Date class models a calendar date with day, month and year, is designed as shown in the class diagram. It contains the following members:
- 3 private instance variables day , month , and year .
- Constructors, public getters and setters for the private instance variables.
- A method setDate() , which sets the day , month and year .
- A toString() , which returns " DD/MM/YYYY ", with leading zero for DD and MM if applicable.
Write the Date class and a test driver to test all the public methods. No Input validations are required for day , month , and year .
The Date Class ( Date.java )
A Test Driver for the Date Class ( TestDate.java )
The expected outputs are:
EG. 3: The Time class
A class called Time , which models a time instance with hour, minute and second, is designed as shown in the class diagram. It contains the following members:
- 3 private instance variables hour , minute , and second .
- Constructors, getters and setters.
- A method setTime() to set hour , minute and second .
- A toString() that returns " hh:mm:ss " with leading zero if applicable.
- A method nextSecond() that advances this instance by one second. It returns this instance to support chaining (cascading) operations, e.g., t1.nextSecond().nextSecond() . Take note that the nextSecond() of 23:59:59 is 00:00:00 .
Write the Time class and a test driver to test all the public methods. No input validations are required.
The Time Class ( Time.java )
A Test Driver ( TestTime.java )
The expected outputs are:
EG. 4: The Time class with Input Validation
In this example, we shall validate the inputs to ensure that 0≤hour≤23 , 0≤minute≤59 , and 0≤second≤59 . We re-write our Time class as follows. Take note that all the validations are done in the setters. All other methods (such as constructors and setTime() invoke the setters to perform input validations – so as to avoid duplication of codes.
EG. 5 (Advanced): The Time Class with Input Validation via Exception Handling
In the previous example, we print a error message and set the variable to 0, if the input is invalid. This is less than perfect. The proper way to handle invalid inputs is via the so-called exception handling mechanism.
The revised Time.java that uses exception handling mechanism is as follows:
What to do if an invalid hour , minute or second was given as input? Print an error message? Terminate the program Abruptly? Continue operation by setting the parameter to its default? This is a really hard decision and there is no perfect solution that suits all situations.
In Java, instead of printing an error message, you can throw an so-called Exception object (such as IllegalArgumentException ) to the caller, and let the caller handles the exception gracefully. For example,
The caller can use the try-catch construct to handle the exception gracefully. For example,
The statements in the try -clause will be executed. If all the statements in the try -clause are successful, the catch -clause is ignored, and execution continues to the next statement after try-catch . However, if one of the statement in the try -clause throws an exception (in this case, an IllegalArgumentException ), the rest of try -clause will be skipped, and the execution will be transferred to the catch -clause. The program always continues to the next statement after the try-catch (instead of abruptly terminated).
A Test Driver Class for the Time Class ( TestTime.java )
Without the proper try-catch , the " Time t2 " will abruptly terminate the program, i.e., the rest of the program will not be run (try it by un-commenting the statement). With proper try-catch handling, the program can continue its operation (i.e., graceful handling of exception).
EG. 6: The Point class
A Point class models a 2D point at (x,y) , as shown in the class diagram. It contains the following members:
- 2 private instance variables x and y , which maintain the location of the point.
- Constructors, getters and setters.
- A method setXY() , which sets the x and y of the point; and a method getXY() , which returns the x and y in a 2-element int array.
- A toString() , which returns " (x,y) ".
- 3 versions of overloaded distance() :
- distance(int x, int y) returns the distance from this instance to the given point at (x,y) .
- distance(Point another) returns the distance from this instance to the given Point instance (called another ).
- distance() returns the distance from this instance to (0,0) .
The Point Class ( Point.java )
A Test Driver ( TestPoint.java )
EG. 7: The Ball class
A Ball class models a moving ball, is designed as shown in the class diagram. It contains the following members:
- 4 private variables x , y , xStep , yStep , which maintain the position of the ball and the displacement per move step.
- Constructors, getters and setters.
- Method setXY() and setXYStep() , which sets the position and step size of the ball; and getXY() and getXYSpeed() .
- A toString() , which returns " Ball@(x,y),speed=(xStep,yStep) ".
- A method move() , which increases x and y by xStep and yStep respectively; and returns this instance to support chaining operation.
The Ball Class ( Ball.java )
A Test Driver ( TestBall.java )
Try: To support bouncing ball within a rectangular boundary, add a variable called radius , and methods reflectHorizontal() and reflectVertical() .
EG. 8: The Student Class
Suppose that our application requires us to model students. A student has a name and an address. We are required to keep track of the courses taken by each student, together with the grades (between 0 and 100) for each of the courses. A student shall not take more than 30 courses for the entire program. We are required to print all course grades, and also the overall average grade.
We can design the Student class as shown in the class diagram. It contains the following members:
- private instance variables name ( String ), address ( String ), numCourses ( int ), course ( String ) and grades ( int ). The numCourses keeps track of the number of courses taken by this student so far. The courses and grades are two parallel arrays, storing the courses taken (e.g., <"IM101", "IM102", "IM103">) and their respective grades (e.g. <89, 56, 98>).
- A constructor that constructs an instance with the given name and Address . It also constructs the courses and grades arrays and set the numCourses to 0.
- Getters for name and address ; setter for address . No setter is defined for name as it is not designed to be changed.
- A toString() , which prints " name(address) ".
- A method addCourseGrade(course, grade) , which appends the given course and grade into the courses and grades arrays, respectively; and increments numCourses .
- A method printGrades() , which prints " name course1:grade1, course2:grade2. ".
- A method getAverageGrade() , which returns the average grade of all the courses taken.
The Student Class ( Student.java )
A Test Driver for the Student Class ( TestStudent.java )
The expected outputs are:
Notes : We used arrays In this example, which has several limitations. Arrays need to be pre-allocated with a fixed-length. Furthermore, we need two parallel arrays to keep track of two entities. There are advanced data structures that could represent these data better and more efficiently.