java4python 3.0 documentation

Classes

«  Lists and Maps   ::   Contents

Classes

Defining Classes in Java

You have already seen how to define classes in Java. Its unavoidable for even the simplest of programs. In this section we will look at how we define classes to create our own data types. Lets start by creating a fraction class to extend the set of numeric data types provided by our language. The requirements for this new data type are as follows:

  • Given a numerator and a denominator create a new Fraction.
  • When a fraction is printed it should be simplified.
  • Two fractions can be added or subtracted
  • Two fractions can be multiplied or divided
  • Two fractions can be compared
  • A fraction and an integer can be added together.
  • Given a list of Fractions that list should be sortable by the default sorting function.

Here is a mostly complete implementation of a Fraction class in Python that we will refer to throughout this section:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
class Fraction:

    def __init__(self,top,bottom):

        self.num = top        #the numerator is on top
        self.den = bottom     #the denominator is on the bottom


    def __repr__(self):
        if self.num > self.den:
            retWhole = self.num / self.den
            retNum = self.num - (retWhole * self.den)
            return str(retWhole) + " " + str(retNum)+"/"+str(self.den)
        else:
            return str(self.num)+"/"+str(self.den)

    def show(self):
        print self.num,"/",self.den

    def __add__(self,other):
        # convert to a fraction
        other = self.toFract(other)

        newnum = self.num*other.den + self.den*other.num
        newden = self.den * other.den

        common = gcd(newnum,newden)

        return Fraction(newnum/common,newden/common)

    def __radd__(self,leftNum):
        other = self.toFract(leftNum)            
        newnum = self.num*other.den + self.den*other.num
        newden = self.den * other.den

        common = gcd(newnum,newden)

        return Fraction(newnum/common,newden/common)

    def __cmp__(self,other):

        num1 = self.num*other.den 
        num2 = self.den*other.num

        if num1 < num2:
           return -1
        else:
           if num1 == num2:
              return 0
           else:
              return 1

    def toFract(self,n):
        if isinstance(n,int):
            other = Fraction(n,1)
        elif isinstance(n, float):
            wholePart = int(n)
            fracPart = n - wholePart
            # convert to 100ths???
            fracNum = int(fracPart * 100)
            newNum = wholePart * 100 + fracNum
            other = Fraction(newNum,100)
        elif isinstance(n,Fraction):
            other = n
        else:
            print "Error: cannot add a fraction to a ", type(n)
            return None
        return other

#gcd is a helper function for Fraction

def gcd(m,n):
    while m%n != 0:
        oldm = m
        oldn = n

        m = oldn
        n = oldm%oldn

    return n

The instance variables (data members) we will need for our fraction class are the numerator and denominator. Of course in Python we can add instance variables to a class at any time by simply assigning a value to objectReferenc.variableName In Java all data members must be declared up front.

The declarations of instance variables can come at the beginning of the class definition or the end. Cay Horstman, Author of the Core Java books puts the declarations at the end of the class. I like them at the very beginning so you see the variables that are declared before you begin looking at the code that uses them. With that in mind the first part of the Fraction class definition is as follows:

public class Fraction {
    private int numerator;
    private int denominator;

Notice that we have declared the numerator and denominator to be private. This means that the compiler will generate an error if another method tries to write code like the following:

Fraction f = new Fraction(1,2);
int y = f.numerator * 10;

Direct access to private members is not allowed from outside the class. Therefore, if we do not include a way to update the numerator or denominator, then our fractions will be immutable. For private fields, we typically provide access to outsiders using getter methods.

1
2
3
4
5
6
7
public Integer getNumerator() {
    return numerator;
}

public Integer getDenominator() {
    return denominator;
}

If we want to allow the numerator and denominator to change, then we would have a mutable fraction. One way to do this would be to make the fields public. However, it is more common in Java to provide setter methods.

1
2
3
4
5
6
7
public void setNumerator(Integer numerator) {
    this.numerator = numerator;
}

public void setDenominator(Integer denominator) {
    this.denominator = denominator;
}

Writing a constructor

Once you have identified the instance variables for you class the next thing to consider is the constructor. In Java, constructors have the same name as the class and are declared public. They are declared without a return type. So any function that is named the same as the class and has no return type is a constructor. Our constructor will take two parameters the numerator and the denominator.

public Fraction(int top, int bottom) {
    num = top;
    den = bottom;
}

There are a couple of important things to notice here. First, you will notice that the constructor does not have a self parameter. You will also notice that we can simply refer to the instance variables by name without the self prefix, because they have already been declared. This allows the Java compiler to do the work of dereferencing the current Java object. Java does provide a special variable called this that works like the self variable. In Java, this is typically only used when it is needed to differentiate between a parameter or local variable and an instance variable. For example this alternate definition of the the Fraction constructor uses this to differentiate between parameters and instance variables.

public Fraction(int num, int den) {
    this.num = num;
    this.den = den;
}

Methods or Member Functions

Now we come to one of the major differences between Java and Python. The Python class definition used the special methods for addition, and comparison that have the effect of redefining how the standard operators behave. In Java there is no operator overloading. So we will have to write member functions to do addition, subtraction, multiplication, and division. Lets begin with addition.

1
2
3
4
5
6
7
8
9
public Fraction add(Fraction other) {
    int newNum, newDen, common;

    newNum = other.denominator*this.numerator +
                             this.denominator*other.numerator;
    newDen = this.denominator * other.denominator;
    common = gcd(newNum,newDen);
    return new Fraction(newNum/common, newDen/common );
}

First you will notice that the add member function is declared as public Fraction The public part means that any other method may call the add method. The Fraction part means that add will return a fraction as its result.

Second, you will notice that on line two all of the local variables used in the function are declared. In this case there are three local variables: newNum, newDen, and common. It is a good idea for you to get in the habit of declaring your local variables at the beginning of your function. This declaration section provides a simple road map for the function in terms of the data that will be used. The listing above also makes use of the this variable, you may find it useful to use this until you are comfortable with abandoning your Pythonic ideas about self.

Declaring your variables at the top is not a requirement, it is just a recommended practice for you. Java only requires that you declare your variables before they are used. The following version of Fraction is also legal Java, but may be somewhat less readable.

1
2
3
4
5
6
7
public Fraction add(Fraction other) {
    int newNum = other.denominator*numerator +
                             denominator*other.numerator;
    int newDen = denominator * other.denominator;
    int common = gcd(newNum,newDen);
    return new Fraction(newNum/common, newDen/common );
}

The addition takes place by multiplying each numerator by the opposite denominator before adding. This procedure ensures that we are adding two fractions with common denominators. Using this approach the denominator is computed by multiplying the two denominators. The greatest common divisor function is used to find a common divisor to simplify the numerator and denominator in the result.

Finally on line 8 a new fraction is returned as the result of the computation. The value that is returned by the return statement must match the value that is specified as part of the declaration. So, in this case the return value on line 8 must match the declared value on line 1.

Method Signatures and Overloading

Our specification for this project said that we need to be able to add a Fraction to an int. In Python we can do this by checking the type of the parameter using the isinstance function at runtime. Recall that isinstance(1,int) returns True to indicate that 1 is indeed an instance of the int class. See lines 22 and 53—68 of the Python version of the Fraction class to see how our Python implementation fulfills this requirement.

In Java we can do runtime type checking, but the compiler will not allow us to pass an int to the add function since the parameter has been declared to be a Fraction. The way that we solve this problem is by writing another add method with a different set of parameters. In Java this practice is legal and common we call this practice overloading.

This idea of overloading raises a very important difference between Python and Java. In Python a method is known by its name only. In Java a method is known by its signature. The signature of a method includes its name, and the types of all of its parameters. The name and the types of the parameters are enough information for the Java compiler to decide which method to call at runtime.

To solve the problem of adding an int and a Fraction in Java we will overload both the constructor and the add function. We will overload the constructor so that if it only receives a single int it will convert the int into a Fraction. We will also overload the add method so that if it receives an int as a parameter it first construct a Fraction from that integer and then add the two Fractions together. The new methods that accomplish this task are as follows:

1
2
3
4
5
6
7
8
public Fraction(int num) {
    this.numerator = num;
    this.denominator = 1;
}

public Fraction add(int other) {
    return add(new Fraction(other));
}

Notice that the overloading approach can provide us with a certain elegance to our code. Rather than utilizing if statements to check the types of parameters we just overload functions ahead of time which allows us to call the method we want and allow the compiler to make the decisions for us. This way of thinking about programming takes some practice.

Our full Fraction class to this point would look like the following. You may want to try to compile and run the short test program provided just to see what happens.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
public class Fraction {

    private int numerator;
    private int denominator;

    public Fraction(int num, int den) {
        this.numerator = num;
        this.denominator = den;
    }

    public Fraction(int num) {
        this.numerator = num;
        this.denominator = 1;
    }

    public Fraction add(Fraction other) {
        int newNum, newDen, common;

        newNum = other.denominator*this.numerator + this.denominator*other.numerator;
        newDen = this.denominator * other.denominator;
        common = gcd(newNum,newDen);
        return new Fraction(newNum/common, newDen/common );
    }

    public Fraction add(int other) {
        return add(new Fraction(other));
    }

    private static int gcd(int m, int n) {
        while (m % n != 0) {
            int oldm = m;
            int oldn = n;
            m = oldn;
            n = oldm%oldn;
        }
        return n;
    }

    public static void main(String[] args) {
        Fraction f1 = new Fraction(1,2);
        Fraction f2 = new Fraction(2,3);

        System.out.println(f1.mul(f2));
        System.out.println(f1.add(1));
    }

}

Inheritance

If you ran the program above you probably noticed that the output is not very satisfying. Chances are your output looked something like this:

Fraction@7b11a3ac
Fraction@6c22c95b

The reason is that we have not yet provided a friendly string representation for our Fraction objects. The truth is that, just like in Python, whenever an object is printed by the println method it must be converted to string format. In Python you can control how that looks by writing an __str__ method for your class. If you do not then you will get the default, which looked something like the above.

The Object Class

In Java, the equivalent of __str__ is the toString method. Every object in Java already has a toString method defined for it because every class in Java automatically inherits from the Object class. The object class provides default implementations for the following functions (among others).

  • equals
  • getClass
  • toString

To make our output nicer we will implement the toString method for the Fraction class. A simple version of the method is provided below.

public String toString() {
    return numerator + "/" + denominator;
}

The other important class for us to implement from the list of methods inherited from Object is the equals method. When two objects are compared in Java using the == operator they are tested to see if they are exactly the same object, that is do the two objects occupy the same exact space in the computers memory. This is the default behavior of the equals method provided by Object. The equals method allows us to decide if two objects are equal by looking at their instance variables. However it is important to remember that since Java does not have operator overloading if you want to use your equals method you must call it directly. Therefore once you write your own equals method:

object1 == object2

is NOT the same as

object1.equals(object2)

Here is the equals method we would like to write for the Fraction class:

public boolean equals(Fraction other) {
    int num1 = this.numerator * other.denominator;
    int num2 = this.denominator * other.numerator;
    return num1 == num2;
}

However, this is not quite correct, since equals must work for null and for objects of classes other than Fraction. The proper implementation of equals include a bit of boilerplate before the code above.

1
2
3
4
5
6
7
8
9
public boolean equals(Object that) {
    if (that == null) return false;
    if (this.getClass() != that.getClass()) return false;
    Fraction other = (Fraction) that;

    int num1 = this.numerator * other.denominator;
    int num2 = this.denominator * other.numerator;
    return num1 == num2;
}

One important thing to remember about equals is that it only checks to see if two objects are equal it does not have any notion of less than or greater than. We’ll see more about that shortly.

Abstract Classes and Methods

If we want to make our Fraction class behave like Integer, Double, and the other numeric classes in Java, we need to make a couple of additional modifications to the class. The first thing we will do is plug Fraction into the Java class hierarchy at the same place as Integer and its siblings. If you look at the documentation for Integer you will see that Integer’s parent class is Number. Number is an abstract class that specifies several methods that all of its children must implement. In Java an abstract class is more than just a placeholder for common functions. In Java an abstract class has the power to specify certain functions that all of its children must implement. You can trace this power back to the strong typing nature of Java.

The that makes the Fraction class a child of Number is as follows:

public class Fraction extends Number {
    ...
}

The keyword extends tells the compiler that the class Fraction extends, or adds new functionality to the Number class. A child class always extends its parent.

The methods we must implement if Fraction is going to be a child of Number are:

  • longValue
  • intValue
  • floatValue
  • doubleValue

This really isn’t much work for us to implement these functions as all we have to do is some conversion of our own and some division. The implementation of these methods is as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
public double doubleValue() {
    return ((double) numerator) / ((double) denominator);
}

public float floatValue() {
    return ((float) numerator) / ((float) denominator);
}

public int intValue() {
    return numerator / denominator;
}

public long longValue() {
    return ((long) numerator) / ((long) denominator);
}

By having the Fraction class extend the Number class we can now pass a Fraction to any Java function that specifies it can receive a Number as one of its parameters. For example many Java user interface methods accept any object that is a subclass of Number as a parameter. In Java the class hierarchy and the IS-A relationships are very important. Whereas in Python you can pass any kind of object as a parameter to any function the strong typing of Java makes sure that you only pass an object as a parameter that is of the type specified in the function call or one of its children. So, in this case when you see a parameter of type Number its important to remember that an Integer is-a Number and a Double is-a Number and a Fraction is-a Number.

However, and this is a big however, it is also important to remember that if you specify Number as the type on a particular parameter then the Java compiler will only let you use the methods of a Number. In this case longValue, intValue, floatValue, and doubleValue.

Lets suppose you define a method in some class as follows:

public void test(Number a, Number b) {
    a.add(b);
}

The Java compiler would give an error because add is not a defined method of the Number class. Even if you called the add method and passed two Fractions as parameters.

Interfaces

Lets turn our attention to making a list of fractions sortable by the standard Java sorting method Collections.sort. In Python all we would need to do is implement the __cmp__ method. But in Java we cannot be that informal. In Java Things that are sortable must be Comparable. Your first thought might be that Comparable is Superclass of Number. That would be a good thought but it would not be correct. Java only supports single inheritance, that is, a class can have only one parent. Although it would be possible to add an additional Layer to the class hierarchy it would also complicate things dramatically. Because Not only are Numbers comparable, but Strings are also Comparable as would many other types. For example we might have a Student class and we want to be able to sort Students by their gpa. But Student already extends the class Person for which we have no natural comparison function.

Java’s answer to this problem is the Interface mechanism. Interfaces are like a combination of Inheritance and contracts all rolled into one. An interface is a specification that says any object that claims it implements this interface must provide the following methods. It sounds a little bit like an abstract class, however it is outside the inheritance mechanism. You can never create an instance of Comparable. Many objects, however, do implement the Comparable interface. What does the Comparable interface specify?

The Comparable interface says that any object that claims to be Comparable must implement the compareTo method.

int compareTo(T o)

The following is the documentation for the compareTo method as specified by the Comparable interface.

Compares this object with the specified object for order. Returns a negative integer, zero, or a positive integer as this object is less than, equal to, or greater than the specified object. The implementor must ensure sgn(x.compareTo(y)) == -sgn(y.compareTo(x)) for all x and y. (This implies that x.compareTo(y) must throw an exception iff y.compareTo(x) throws an exception.)

The implementor must also ensure that the relation is transitive: (x.compareTo(y)>0 && y.compareTo(z)>0) implies x.compareTo(z)>0.

Finally, the implementor must ensure that x.compareTo(y)==0 implies that sgn(x.compareTo(z)) == sgn(y.compareTo(z)), for all z.

It is strongly recommended, but not strictly required that (x.compareTo(y)==0) == (x.equals(y)). Generally speaking, any class that implements the Comparable interface and violates this condition should clearly indicate this fact. The recommended language is “Note: this class has a natural ordering that is inconsistent with equals.”

In the foregoing description, the notation sgn(expression) designates the mathematical signum function, which is defined to return one of -1, 0, or 1 according to whether the value of expression is negative, zero or positive.

To make our Fraction class Comparable we must modify the class declaration line as follows:

public class Fraction extends Number implements Comparable<Fraction> {
    ...
}

The specification Comparable<Fraction> makes it clear that Fraction is only comparable with another Fraction. The compareTo method could be implemented as follows:

public int compareTo(Fraction other) {
    int num1 = this.numerator * other.denominator;
    int num2 = this.denominator * other.numerator;
    return num1 - num2;
}

Unlike the equals method, compareTo can assume that it’s argument has type Fraction. In addition, compareTo need not return a value if other is null; in this case, the function will throw a NullPointerException.

Static member variables

Suppose that you wanted to write a Student class so that the class could keep track of the number of students it had created. Although you could do this with a global counter variable that is an ugly solution. The right way to do it is to use a static variable. In Python we could do this as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
class Student:
    numStudents = 0

    def __init__(self, id, name):
        self.id = id
        self.name = name

        Student.numStudents = Student.numStudents + 1

def main():
    for i in range(10):
        s = Student(i,"Student-"+str(i))
    print 'The number of students is: ', Student.numStudents
>>> main()
The number of students is:  10
>>>

In Java we would write this same example using a static declaration.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
public class Student {
    public static int numStudents = 0;

        private int id;
        private String name;

        public Student(int id, String name) {
        this.id = id;
        this.name = name;

        numStudents = numStudents + 1;
    }

    public static void main(String[] args) {
        for(int i = 0; i < 10; i++) {
            Student s = new Student(i,"Student"+i.toString());
        }
        System.out.println("The number of students: "+Student.numStudents.toString());
    }
}

In this example notice that we create a static member variable by using the static modifier on the variable declaration. Once a variable has been declared static in Java it can be access from inside the class without prefixing the name of the class as we had to do in Python.

Static Methods

We have already discussed the most common static method of all, main. However in our Fraction class we also implemented a method to calculate the greatest common divisor for two fractions (gdc). There is no reason for this method to be a member method since it takes two int values as its parameters. Therefore we declare the method to be a static method of the class. Furthermore since we are only going to use this gcd method for our own purposes we can make it private.

1
2
3
4
5
6
7
8
9
private static int gcd(int m, int n) {
    while (m % n != 0) {
        int oldm = m;
        int oldn = n;
        m = oldn;
        n = oldm%oldn;
    }
    return n;
}

Full Implementation of the Fraction Class

A final version of the Fraction class that exercises all of the features we have discussed is as follows.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
import java.util.Arrays;

public class Fraction extends Number implements Comparable<Fraction> {
    private int numerator;
    private int denominator;
  
    public Fraction(int num, int den) {
        this.numerator = num;
        this.denominator = den;
    }
    public Fraction(int num) {
        this.numerator = num;
        this.denominator = 1;
    }

    public Integer getNumerator()                   { return numerator; }
    public Integer getDenominator()                 { return denominator; }
    public void setNumerator  (Integer numerator)   { this.numerator = numerator; }
    public void setDenominator(Integer denominator) { this.denominator = denominator; }

    public Fraction add(int other) {
        return add(new Fraction(other));
    }
    public Fraction add(Fraction other) {
        int newNum = other.denominator*this.numerator + this.denominator*other.numerator;
        int newDen = this.denominator * other.denominator;
        int common = gcd(newNum,newDen);
        return new Fraction(newNum/common, newDen/common );
    }
    private static int gcd(int m, int n) {
        while (m % n != 0) {
            int oldm = m;
            int oldn = n;
            m = oldn;
            n = oldm%oldn;
        }
        return n;
    }

    public double doubleValue() { return ((double) numerator) / ((double) denominator); } 
    public float  floatValue()  { return ((float) numerator) / ((float) denominator); } 
    public int    intValue()    { return numerator / denominator; } 
    public long   longValue()   { return ((long) numerator) / ((long) denominator); }
    public String toString()    { return numerator + "/" + denominator; }
    public boolean equals(Object that) {
        if (that == null) return false;
        if (this.getClass() != that.getClass()) return false;
        Fraction other = (Fraction) that;
            
        int num1 = this.numerator * other.denominator;
        int num2 = this.denominator * other.numerator;       
        return num1 == num2;
    }
    public int compareTo(Fraction other) {
        int num1 = this.numerator * other.denominator;
        int num2 = this.denominator * other.numerator;
        return num1 - num2;
    }

    // This is a unit test program to test the class.
    public static void main(String[] args) {
        Fraction f1 = new Fraction(1,2);
        Fraction f2 = new Fraction(2,3);
        Fraction f3 = new Fraction(1,4);

        System.out.println(f1.add(1));
        System.out.println(f1.intValue());
        System.out.println(f1.doubleValue());

        Fraction[] myFracs = { f1, f2, f3 };
        Arrays.sort(myFracs);

        for(Fraction f : myFracs) {
            System.out.println(f);
        }
    }
}

«  Lists and Maps   ::   Contents