Like a sentient creature,
a program must manipulate its world and make choices during execution.
In Java you manipulate objects and
data using operators, and you make choices with execution control statements.
Java was inherited from C++, so most of these statements and operators will be
familiar to C and C++ programmers. Java has also added some improvements and
simplifications.
An operator takes one or more
arguments and produces a new value. The arguments are in a different form than
ordinary method calls, but the effect is the same. You should be reasonably
comfortable with the general concept of operators from your previous programming
experience. Addition (+), subtraction and unary minus (-),
multiplication (*), division (/) and assignment (=) all
work much the same in any programming language.
All operators produce a value from
their operands. In addition, an operator can change the value of an operand.
This is called a side effect. The most common use
for operators that modify their operands is to generate the side effect, but you
should keep in mind that the value produced is available for your use just as in
operators without side effects.
Almost all operators work only with
primitives. The exceptions are ‘=’, ‘==’
and ‘!=’, which work with all objects (and are a point of
confusion for objects). In addition, the String class supports
‘+’ and
‘+=’.
Operator precedence defines how an
expression evaluates when several operators are present. Java has specific rules
that determine the order of evaluation. The easiest one to remember is that
multiplication and division happen before addition and subtraction. Programmers
often forget the other precedence rules, so you should use parentheses to make
the order of evaluation explicit. For example:
A = X + Y - 2/2 + Z;
has a very different meaning from
the same statement with a particular grouping of parentheses:
A = X + (Y - 2)/(2 + Z);
Assignment is performed with the
operator =. It means “take the value of the right-hand side (often called
the rvalue) and copy it into the left-hand side
(often called the lvalue). An rvalue is any
constant, variable or expression that can produce a value, but an lvalue must be
a distinct, named variable. (That is, there must be a physical space to store a
value.) For instance, you can assign a constant value to a variable (A =
4;), but you cannot assign anything to constant value – it cannot be
an lvalue. (You can’t say 4 = A;.)
Assignment of primitives is quite
straightforward. Since the primitive holds the actual value and not a handle to
an object, when you assign primitives you copy the contents from one place to
another. For example, if you say A = B for primitives, then the contents
of B is copied into A. If you then go on to modify A,
B is naturally unaffected by this modification. This is what you’ve
come to expect as a programmer for most situations.
When you
assign
objects, however, things change. Whenever you manipulate an object, what
you’re manipulating is the handle, so when you assign “from one
object to another” you’re actually copying a handle from one place
to another. This means that if you say C = D for objects, you end up with
both C and D pointing to the object that, originally, only
D pointed to. The following example will demonstrate this.
As an aside, the first thing you
see is a package statement for package c03,
indicating this book’s Chapter 3. The first code listing of each chapter
will contain a package statement like this to establish the chapter number for
the remaining code listings in that chapter. In Chapter 17, you’ll see
that as a result, all the listings in chapter 3 (except those that have
different package names) will be automatically placed in a subdirectory called
c03, Chapter 4’s listings will be in c04 and so on. All this
will happen via the CodePackager.java program shown in Chapter 17, and in
Chapter 5 the concept of packages will be fully explained. What you need to
recognize at this point is that, for this book, lines of code of the form
package c03 are used just to establish the chapter subdirectory for the
listings in the chapter.
In order to run the program, you
must ensure that the
classpath
contains the root directory where you installed the source code for this book.
(From this directory, you’ll see the subdirectories c02,
c03, c04, etc.)
For later versions of Java (1.1.4
and on), when your main( ) is inside a file with a package
statement, you must give the full package name before the program name in
order to run the program. In this case, the command line is:
java c03.Assignment
Keep this in mind any time
you’re running a program that’s in a
package.
Here’s the
example:
//: Assignment.java // Assignment with objects is a bit tricky package c03; class Number { int i; } public class Assignment { public static void main(String[] args) { Number n1 = new Number(); Number n2 = new Number(); n1.i = 9; n2.i = 47; System.out.println("1: n1.i: " + n1.i + ", n2.i: " + n2.i); n1 = n2; System.out.println("2: n1.i: " + n1.i + ", n2.i: " + n2.i); n1.i = 27; System.out.println("3: n1.i: " + n1.i + ", n2.i: " + n2.i); } } ///:~
The Number class is simple,
and two instances of it (n1 and n2) are created within
main( ). The i value within each Number is given a
different value, and then n2 is assigned to n1, and n1 is
changed. In many programming languages you would expect n1 and n2
to be independent at all times, but because you’ve assigned a handle
here’s the output you’ll see:
1: n1.i: 9, n2.i: 47 2: n1.i: 47, n2.i: 47 3: n1.i: 27, n2.i: 27
Changing the n1 object
appears to change the n2 object as well! This is because both n1
and n2 contain the same handle, which is pointing to the same object.
(The original handle that was in n1 that pointed to the object holding a
value of 9 was overwritten during the assignment and effectively lost; its
object will be cleaned up by the garbage collector.)
This phenomenon is often called
aliasing and it’s a
fundamental way that Java works with objects. But what if you don’t want
aliasing to occur in this case? You could forego the assignment and
say:
n1.i = n2.i;
This retains the two separate
objects instead of tossing one and tying n1 and n2 to the same
object, but you’ll soon realize that manipulating the fields within
objects is messy and goes against good object-oriented design principles. This
is a non-trivial topic, so it is left for Chapter 12, which is devoted to
aliasing. In the meantime, you should keep in mind that assignment for objects
can add surprises.
Aliasing will also occur when you
pass an object into a method:
//: PassObject.java // Passing objects to methods can be a bit tricky class Letter { char c; } public class PassObject { static void f(Letter y) { y.c = 'z'; } public static void main(String[] args) { Letter x = new Letter(); x.c = 'a'; System.out.println("1: x.c: " + x.c); f(x); System.out.println("2: x.c: " + x.c); } } ///:~
In many programming languages, the
method f( ) would appear to be making a copy of its argument
Letter y inside the scope of the method. But once again a handle is being
passed so the line
y.c = 'z';
is actually changing the object
outside of f( ). The output shows this:
1: x.c: a 2: x.c: z
Aliasing and its solution is a
complex issue and, although you must wait until Chapter 12 for all the answers,
you should be aware of it at this point so you can watch for
pitfalls.
The basic mathematical operators
are the same as the ones available in most programming languages: addition
(+), subtraction
(-), division (/),
multiplication (*) and
modulus (%, produces the remainder from integer
division). Integer division truncates, rather than rounds, the
result.
Java also uses a shorthand notation
to perform an operation and an assignment at the same time. This is denoted by
an operator followed by an equal sign, and is consistent with all the operators
in the language (whenever it makes sense). For example, to add 4 to the variable
x and assign the result to x, use: x += 4;.
This example shows the use of the
mathematical operators:
//: MathOps.java // Demonstrates the mathematical operators import java.util.*; public class MathOps { // Create a shorthand to save typing: static void prt(String s) { System.out.println(s); } // shorthand to print a string and an int: static void pInt(String s, int i) { prt(s + " = " + i); } // shorthand to print a string and a float: static void pFlt(String s, float f) { prt(s + " = " + f); } public static void main(String[] args) { // Create a random number generator, // seeds with current time by default: Random rand = new Random(); int i, j, k; // '%' limits maximum value to 99: j = rand.nextInt() % 100; k = rand.nextInt() % 100; pInt("j",j); pInt("k",k); i = j + k; pInt("j + k", i); i = j - k; pInt("j - k", i); i = k / j; pInt("k / j", i); i = k * j; pInt("k * j", i); i = k % j; pInt("k % j", i); j %= k; pInt("j %= k", j); // Floating-point number tests: float u,v,w; // applies to doubles, too v = rand.nextFloat(); w = rand.nextFloat(); pFlt("v", v); pFlt("w", w); u = v + w; pFlt("v + w", u); u = v - w; pFlt("v - w", u); u = v * w; pFlt("v * w", u); u = v / w; pFlt("v / w", u); // the following also works for // char, byte, short, int, long, // and double: u += v; pFlt("u += v", u); u -= v; pFlt("u -= v", u); u *= v; pFlt("u *= v", u); u /= v; pFlt("u /= v", u); } } ///:~
The first thing you will see are
some shorthand methods for printing: the prt( ) method prints a
String, the pInt( ) prints a String followed by an
int and the pFlt( ) prints a String followed by a
float. Of course, they all ultimately end up using
System.out.println( ).
To generate numbers, the program
first creates a Random object. Because no arguments are passed during
creation, Java uses the current time as a seed for the random number generator.
The program generates a number of different types of random numbers with the
Random object simply by calling different methods:
nextInt( ), nextLong( ), nextFloat( ) or
nextDouble( ).
The modulus operator, when used
with the result of the random number generator, limits the result to an upper
bound of the operand minus one (99 in this case).
The unary minus
(-) and unary plus
(+) are the same operators as
binary minus and plus. The compiler figures out which use is intended by the way
you write the expression. For instance, the statement
x = -a;
has an obvious meaning. The
compiler is able to figure out:
x = a * -b;
but the reader might get confused,
so it is more clear to say:
x = a * (-b);
The unary minus produces the
negative of the value. Unary plus provides symmetry with unary minus, although
it doesn’t do
much.
Java, like C, is full of shortcuts.
Shortcuts can make code much easier to type, and either easier or harder to
read.
Two of the nicer shortcuts are the
increment and decrement operators
(often referred to as the auto-increment and
auto-decrement operators). The decrement operator is
-- and means “decrease by one unit.” The increment operator
is ++ and means “increase by one unit.” If A is an
int, for example, the expression ++A is equivalent to (A = A +
1). Increment and decrement operators produce the value of the variable as a
result.
There are two versions of each type
of operator, often called the prefix and postfix versions. Pre-increment means
the ++ operator appears before the variable or expression, and
post-increment means the ++ operator appears after the variable or
expression. Similarly, pre-decrement means the -- operator appears before
the variable or expression, and post-decrement means the -- operator
appears after the variable or expression. For pre-increment and pre-decrement,
(i.e., ++A or --A), the operation is performed and the value is
produced. For post-increment and post-decrement (i.e. A++ or A--),
the value is produced, then the operation is performed. As an
example:
//: AutoInc.java // Demonstrates the ++ and -- operators public class AutoInc { public static void main(String[] args) { int i = 1; prt("i : " + i); prt("++i : " + ++i); // Pre-increment prt("i++ : " + i++); // Post-increment prt("i : " + i); prt("--i : " + --i); // Pre-decrement prt("i-- : " + i--); // Post-decrement prt("i : " + i); } static void prt(String s) { System.out.println(s); } } ///:~
The output for this program
is:
i : 1 ++i : 2 i++ : 2 i : 3 --i : 2 i-- : 2 i : 1
You can see that for the prefix
form you get the value after the operation has been performed, but with the
postfix form you get the value before the operation is performed. These are the
only operators (other than those involving assignment) that have side effects.
(That is, they change the operand rather than using just its
value.)
The increment operator is one
explanation for the name C++, implying “one step beyond C.” In an
early Java speech, Bill Joy (one of the creators), said
that “Java=C++--“ (C plus plus minus minus), suggesting that Java is
C++ with the unnecessary hard parts removed and therefore a much simpler
language. As you progress in this book you’ll see that many parts are
simpler, and yet Java isn’t that much easier than
C++.
Relational operators generate a
boolean result. They evaluate the relationship between the values of the
operands. A relational expression produces true if the relationship is
true, and false if the relationship is untrue. The relational operators
are less than (<), greater than
(>), less than or equal to
(<=), greater than or equal to
(>=), equivalent (==) and not
equivalent
(!=).
Equivalence and nonequivalence works with all built-in data types, but the other
comparisons won’t work with type boolean.
The relational operators ==
and != also work with all objects, but their meaning often confuses the
first-time Java programmer. Here’s an example:
//: Equivalence.java public class Equivalence { public static void main(String[] args) { Integer n1 = new Integer(47); Integer n2 = new Integer(47); System.out.println(n1 == n2); System.out.println(n1 != n2); } } ///:~
The expression
System.out.println(n1 == n2) will print out the result of the
boolean comparison within it. Surely the output should be true and
then false, since both Integer objects are the same. But
while the contents of the objects are the same, the
handles are not the same and the
operators == and != compare object handles. So the output is
actually false and then true. Naturally, this surprises people at
first.
What if you want to compare the
actual contents of an object for equivalence? You must use the special method
equals( ) that exists
for all objects (not primitives, which work fine with
== and !=). Here’s how it’s used:
//: EqualsMethod.java public class EqualsMethod { public static void main(String[] args) { Integer n1 = new Integer(47); Integer n2 = new Integer(47); System.out.println(n1.equals(n2)); } } ///:~
The result will be true, as
you would expect. Ah, but it’s not as simple as that. If you create your
own class, like this:
//: EqualsMethod2.java class Value { int i; } public class EqualsMethod2 { public static void main(String[] args) { Value v1 = new Value(); Value v2 = new Value(); v1.i = v2.i = 100; System.out.println(v1.equals(v2)); } } ///:~
you’re back to square one:
the result is false. This is because the default behavior of
equals( ) is to compare handles. So unless you override
equals( ) in your new class you won’t get the desired
behavior. Unfortunately, you won’t learn about overriding until Chapter 7,
but being aware of the way equals( ) behaves might save you some
grief in the meantime.
Most of the Java library classes
implement equals( ) so that it compares the contents of objects
instead of their
handles.
The logical operators AND
(&&), OR
(||) and NOT (!) produce a
boolean value of true or false
based on the logical relationship
of its arguments. This example uses the relational and logical
operators:
//: Bool.java // Relational and logical operators import java.util.*; public class Bool { public static void main(String[] args) { Random rand = new Random(); int i = rand.nextInt() % 100; int j = rand.nextInt() % 100; prt("i = " + i); prt("j = " + j); prt("i > j is " + (i > j)); prt("i < j is " + (i < j)); prt("i >= j is " + (i >= j)); prt("i <= j is " + (i <= j)); prt("i == j is " + (i == j)); prt("i != j is " + (i != j)); // Treating an int as a boolean is // not legal Java //! prt("i && j is " + (i && j)); //! prt("i || j is " + (i || j)); //! prt("!i is " + !i); prt("(i < 10) && (j < 10) is " + ((i < 10) && (j < 10)) ); prt("(i < 10) || (j < 10) is " + ((i < 10) || (j < 10)) ); } static void prt(String s) { System.out.println(s); } } ///:~
You can apply AND, OR, or NOT to
boolean values only. You can’t use a non-boolean as if it
were a boolean in a logical expression as you can
in C and C++. You can see the failed attempts at doing this commented out with a
//! comment marker. The subsequent expressions, however, produce
boolean values using relational comparisons, then use logical operations
on the results.
One output listing looked like
this:
i = 85 j = 4 i > j is true i < j is false i >= j is true i <= j is false i == j is false i != j is true (i < 10) && (j < 10) is false (i < 10) || (j < 10) is true
Note that a boolean value is
automatically converted to an appropriate text form if it’s used where a
String is expected.
You can replace the definition for
int in the above program with any other primitive data type except
boolean. Be aware, however, that the comparison of floating-point numbers
is very strict. A number that is the tiniest fraction different from another
number is still “not equal.” A number that is the tiniest bit above
zero is still nonzero.
When dealing with
logical operators you run into a
phenomenon called “short circuiting.” This means that the expression
will be evaluated only until the truth or falsehood of the entire expression can
be unambiguously determined. As a result, all the parts of a logical expression
might not be evaluated. Here’s an example that demonstrates
short-circuiting:
//: ShortCircuit.java // Demonstrates short-circuiting behavior // with logical operators. public class ShortCircuit { static boolean test1(int val) { System.out.println("test1(" + val + ")"); System.out.println("result: " + (val < 1)); return val < 1; } static boolean test2(int val) { System.out.println("test2(" + val + ")"); System.out.println("result: " + (val < 2)); return val < 2; } static boolean test3(int val) { System.out.println("test3(" + val + ")"); System.out.println("result: " + (val < 3)); return val < 3; } public static void main(String[] args) { if(test1(0) && test2(2) && test3(2)) System.out.println("expression is true"); else System.out.println("expression is false"); } } ///:~
Each test performs a comparison
against the argument and returns true or false. It also prints information to
show you that it’s being called. The tests are used in the
expression:
if(test1(0) && test2(2) && test3(2))
You might naturally think that all
three tests would be executed, but the output shows otherwise:
test1(0) result: true test2(2) result: false expression is false
The first test produced a
true result, so the expression evaluation continues. However, the second
test produced a false result. Since this means that the whole expression
must be false, why continue evaluating the rest of the expression? It
could be expensive. The reason for short-circuiting, in fact, is precisely that;
you can get a potential performance increase if all the parts of a logical
expression do not need to be
evaluated.
The bitwise operators allow you to
manipulate individual bits in an integral primitive data type. Bitwise operators
perform boolean algebra on the corresponding bits in the
two arguments to produce the result.
The bitwise operators come from
C’s low-level orientation; you were often manipulating hardware directly
and had to set the bits in hardware registers. Java was originally designed to
be embedded in TV set-top boxes, so this low-level
orientation still made sense. However, you probably won’t use the bitwise
operators much.
The bitwise AND operator
(&) produces a one in the
output bit if both input bits are one; otherwise it produces a zero. The bitwise
OR operator (|) produces a one in
the output bit if either input bit is a one and produces a zero only if both
input bits are zero. The bitwise, EXCLUSIVE OR, or XOR
(^),
produces a one in the output bit if one or the other input bit is a one, but not
both. The bitwise NOT (~, also called the ones complement
operator) is a unary
operator; it takes only one
argument. (All other bitwise operators are binary
operators.) Bitwise NOT produces
the opposite of the input bit – a one if the input bit is zero, a zero if
the input bit is one.
The bitwise operators and logical
operators use the same characters, so it is helpful to have a mnemonic device to
help you remember the meanings: since bits are “small,” there is
only one character in the bitwise operators.
Bitwise operators can be combined
with the = sign to unite the operation and assignment:
&=, |= and
^= are all legitimate. (Since ~ is a unary operator it
cannot be combined with the = sign.)
The boolean type is treated
as a one-bit value so it is somewhat different. You can perform a bitwise AND,
OR and XOR, but you can’t perform a bitwise NOT (presumably to prevent
confusion with the logical NOT). For booleans the bitwise operators have
the same effect as the logical operators except that they do not short circuit.
Also, the bitwise operators on booleans gives you a XOR logical operator
that is not included under the list of “logical” operators.
You’re prevented from using booleans in shift expressions, which is
described
next.
The shift operators also manipulate
bits. They can be used solely with primitive, integral types. The left-shift
operator (<<)
produces the operand to the left of the operator shifted to the left by the
number of bits specified after the operator (inserting zeroes at the lower-order
bits). The signed right-shift operator
(>>) produces the
operand to the left of the operator shifted to the right by the number of bits
specified after the operator. The signed right shift >> uses
sign extension: if the value is positive, zeroes are inserted at the
higher-order bits; if the value is negative, ones are inserted at the
higher-order bits. Java has also added the unsigned right shift >>>,
which uses zero extension: regardless of the sign, zeroes are
inserted at the higher-order bits. This operator does not exist in C or
C++.
If you shift a char,
byte, or short, it will be promoted to int before the shift
takes place, and the result will be an int. Only the five low-order bits
of the right-hand side will be used. This prevents you from shifting more than
the number of bits in an int. If you’re operating on a long,
long will be the result. Only the six low-order bits of the right-hand
side will be used so you can’t shift more than the number of bits in a
long. There is a problem, however, with the unsigned right shift. If you
use it with byte or short you might not get the correct results.
(It’s broken in Java 1.0 and Java
1.1.) These are promoted to int and right shifted,
but the zero extension does not occur, so you get -1 in those
cases. The following example can be used to test your
implementation:
//: URShift.java // Test of unsigned right shift public class URShift { public static void main(String[] args) { int i = -1; i >>>= 10; System.out.println(i); long l = -1; l >>>= 10; System.out.println(l); short s = -1; s >>>= 10; System.out.println(s); byte b = -1; b >>>= 10; System.out.println(b); } } ///:~
Shifts can be combined with the
equal sign (<<= or >>= or
>>>=). The lvalue
is replaced by the lvalue shifted by the rvalue.
Here’s an example that
demonstrates the use of all the operators involving bits:
//: BitManipulation.java // Using the bitwise operators import java.util.*; public class BitManipulation { public static void main(String[] args) { Random rand = new Random(); int i = rand.nextInt(); int j = rand.nextInt(); pBinInt("-1", -1); pBinInt("+1", +1); int maxpos = 2147483647; pBinInt("maxpos", maxpos); int maxneg = -2147483648; pBinInt("maxneg", maxneg); pBinInt("i", i); pBinInt("~i", ~i); pBinInt("-i", -i); pBinInt("j", j); pBinInt("i & j", i & j); pBinInt("i | j", i | j); pBinInt("i ^ j", i ^ j); pBinInt("i << 5", i << 5); pBinInt("i >> 5", i >> 5); pBinInt("(~i) >> 5", (~i) >> 5); pBinInt("i >>> 5", i >>> 5); pBinInt("(~i) >>> 5", (~i) >>> 5); long l = rand.nextLong(); long m = rand.nextLong(); pBinLong("-1L", -1L); pBinLong("+1L", +1L); long ll = 9223372036854775807L; pBinLong("maxpos", ll); long lln = -9223372036854775808L; pBinLong("maxneg", lln); pBinLong("l", l); pBinLong("~l", ~l); pBinLong("-l", -l); pBinLong("m", m); pBinLong("l & m", l & m); pBinLong("l | m", l | m); pBinLong("l ^ m", l ^ m); pBinLong("l << 5", l << 5); pBinLong("l >> 5", l >> 5); pBinLong("(~l) >> 5", (~l) >> 5); pBinLong("l >>> 5", l >>> 5); pBinLong("(~l) >>> 5", (~l) >>> 5); } static void pBinInt(String s, int i) { System.out.println( s + ", int: " + i + ", binary: "); System.out.print(" "); for(int j = 31; j >=0; j--) if(((1 << j) & i) != 0) System.out.print("1"); else System.out.print("0"); System.out.println(); } static void pBinLong(String s, long l) { System.out.println( s + ", long: " + l + ", binary: "); System.out.print(" "); for(int i = 63; i >=0; i--) if(((1L << i) & l) != 0) System.out.print("1"); else System.out.print("0"); System.out.println(); } } ///:~
The two methods at the end,
pBinInt( ) and pBinLong( ) take an int or a
long, respectively, and print it out in binary format along with a
descriptive string. You can ignore the implementation of these for
now.
You’ll note the use of
System.out.print( ) instead of System.out.println( ).
The print( ) method does not put out a new line, so it allows you to
output a line in pieces.
As well as demonstrating the effect
of all the bitwise operators for int and long, this example also
shows the minimum, maximum, +1 and -1 values for int and long so
you can see what they look like. Note that the high bit represents the sign: 0
means positive and 1 means negative. The output for the int portion looks
like this:
-1, int: -1, binary: 11111111111111111111111111111111 +1, int: 1, binary: 00000000000000000000000000000001 maxpos, int: 2147483647, binary: 01111111111111111111111111111111 maxneg, int: -2147483648, binary: 10000000000000000000000000000000 i, int: 59081716, binary: 00000011100001011000001111110100 ~i, int: -59081717, binary: 11111100011110100111110000001011 -i, int: -59081716, binary: 11111100011110100111110000001100 j, int: 198850956, binary: 00001011110110100011100110001100 i & j, int: 58720644, binary: 00000011100000000000000110000100 i | j, int: 199212028, binary: 00001011110111111011101111111100 i ^ j, int: 140491384, binary: 00001000010111111011101001111000 i << 5, int: 1890614912, binary: 01110000101100000111111010000000 i >> 5, int: 1846303, binary: 00000000000111000010110000011111 (~i) >> 5, int: -1846304, binary: 11111111111000111101001111100000 i >>> 5, int: 1846303, binary: 00000000000111000010110000011111 (~i) >>> 5, int: 132371424, binary: 00000111111000111101001111100000
This operator is unusual because it
has three operands. It is truly an operator because it produces a value, unlike
the ordinary if-else statement that you’ll see in the next section of this
chapter. The expression is of the form
boolean-exp ? value0 :
value1
If boolean-exp
evaluates to true, value0 is evaluated and its result becomes the
value produced by the operator. If boolean-exp is false,
value1 is evaluated and its result becomes the value produced by the
operator.
Of course, you could use an
ordinary if-else statement (described later), but the ternary operator is
much terser. Although C prides itself on being a terse language, and the ternary
operator might have been introduced partly for efficiency, you should be
somewhat wary of using it on an everyday basis – it’s easy to
produce unreadable code.
The conditional operator can be
used for its side effects or for the value it produces, but in general you want
the value since that’s what makes the operator distinct from the
if-else. Here’s an example:
static int ternary(int i) { return i < 10 ? i * 100 : i * 10; }
You can see that this code is more
compact than what you’d need to write without the ternary
operator:
static int alternative(int i) { if (i < 10) return i * 100; return i * 10; }
The second form is easier to
understand, and doesn’t require a lot more typing. So be sure to ponder
your reasons when choosing the ternary
operator.
The comma is used in C and C++ not
only as a separator in function argument lists, but also as an operator for
sequential evaluation. The sole place that the comma operator is used in
Java is in for loops, which will be described later in this
chapter.
There’s one special usage of
an operator in Java: the + operator can be used to
concatenate strings, as you’ve already seen. It
seems a natural use of the + even though it doesn’t fit with the
traditional way that + is used. This capability seemed like a good idea
in C++, so operator
overloading was added to C++ to allow the C++ programmer to add meanings to
almost any operator. Unfortunately, operator overloading combined with some of
the other restrictions in C++ turns out to be a fairly complicated feature for
programmers to design into their classes. Although operator overloading would
have been much simpler to implement in Java than it was in C++, this feature was
still considered too complex, so Java programmers cannot implement their own
overloaded operators as C++ programmers can.
The use of the String + has
some interesting behavior. If an expression begins with a String, then
all operands that follow must be Strings:
int x = 0, y = 1, z = 2; String sString = "x, y, z "; System.out.println(sString + x + y + z);
Here, the Java compiler will
convert x, y, and z into their String
representations instead of adding them together first. However, if you
say:
System.out.println(x + sString);
earlier versions of Java will
signal an error. (Later versions, however, will turn x into a
String.) So if you’re putting together a String (using an
earlier version of Java) with addition, make sure the first element is a
String (or a quoted sequence of characters, which the compiler recognizes
as a
String).
One of the pitfalls when using
operators is trying to get away without parentheses when you are even the least
bit uncertain about how an expression will evaluate. This is still true in
Java.
An extremely common error in C and
C++ looks like this:
while(x = y) { // .... }
The programmer was trying to test
for equivalence (==) rather than do an assignment. In C and C++ the
result of this assignment will always be true if y is nonzero, and
you’ll probably get an infinite loop. In Java, the result of this
expression is not a boolean, and the compiler expects a boolean
and won’t convert from an int, so it will conveniently give you a
compile-time error and catch the problem before you ever try to run the program.
So the pitfall never happens in Java. (The only time you
won’t get a compile-time error is when x and y are
boolean, in which case x = y is a legal expression, and in the
above case, probably an error.)
A similar problem in C and C++ is
using bitwise AND and OR instead of logical. Bitwise AND and OR use one of the
characters (& or |) while logical AND and OR use two
(&& and ||). Just as with = and ==,
it’s easy to type just one character instead of
two.
In Java, the compiler again prevents this because it won’t let you
cavalierly use one type where it doesn’t
belong.
The word cast is used in the
sense of “casting into a mold.” Java will automatically change one
type of data into another when appropriate. For instance, if you assign an
integral value to a floating-point variable, the compiler will automatically
convert the int to a float. Casting allows you to make this type
conversion explicit, or to force it when it wouldn’t normally
happen.
To perform a cast, put the desired
data type (including all modifiers) inside parentheses to the left of any value.
Here’s an example:
void casts() { int i = 200; long l = (long)i; long l2 = (long)200; }
As you can see, it’s possible
to perform a cast on a numeric value as well as on a variable. In both casts
shown here, however, the cast is superfluous, since the compiler will
automatically promote an int value to a long when necessary. You
can still put a cast in to make a point or to make your code more clear. In
other situations, a cast is essential just to get the code to
compile.
In C and C++, casting can cause
some headaches. In Java, casting is safe, with the exception that when you
perform a so-called narrowing
conversion (that is, when you go from a data type that can hold more
information to one that doesn’t hold as much) you run the risk of losing
information. Here the compiler forces you to do a cast, in effect saying
“this can be a dangerous thing to do – if you want me to do it
anyway you must make the cast explicit.” With a
widening conversion an
explicit cast is not needed because the new type will more than hold the
information from the old type so that no information is ever
lost.
Java allows you to cast any
primitive type to any other primitive type, except for
boolean, which doesn’t allow any casting at
all. Class types do not allow casting. To convert one to the other there must be
special methods. (String is a special case, and you’ll find out
later in the book that objects can be cast within a family of types; an
Oak can be cast to a Tree and vice-versa, but not to a foreign
type such as a Rock.)
Ordinarily when you insert a
literal value into a program the compiler knows exactly what type to make it.
Sometimes, however, the type is ambiguous. When this happens you must guide the
compiler by adding some extra information in the form of characters associated
with the literal value. The following code shows these
characters:
//: Literals.java class Literals { char c = 0xffff; // max char hex value byte b = 0x7f; // max byte hex value short s = 0x7fff; // max short hex value int i1 = 0x2f; // Hexadecimal (lowercase) int i2 = 0X2F; // Hexadecimal (uppercase) int i3 = 0177; // Octal (leading zero) // Hex and Oct also work with long. long n1 = 200L; // long suffix long n2 = 200l; // long suffix long n3 = 200; //! long l6(200); // not allowed float f1 = 1; float f2 = 1F; // float suffix float f3 = 1f; // float suffix float f4 = 1e-45f; // 10 to the power float f5 = 1e+9f; // float suffix double d1 = 1d; // double suffix double d2 = 1D; // double suffix double d3 = 47e47d; // 10 to the power } ///:~
Hexadecimal
(base 16), which works with all the integral data types,
is denoted by a leading 0x or 0X followed by 0–9 and
a–f either in upper or lower case. If you try to initialize a variable
with a value bigger than it can hold (regardless of the numerical form of the
value), the compiler will give you an error message. Notice in the above code
the maximum possible hexadecimal values for char, byte, and
short. If you exceed these, the compiler will automatically make the
value an int and tell you that you need a narrowing cast for the
assignment. You’ll know you’ve stepped over the
line.
Octal
(base 8) is denoted by a leading zero in the number and
digits from 0-7. There is no literal representation for
binary numbers in C, C++ or
Java.
A trailing character after a
literal value establishes its type. Upper or lowercase L means
long, upper or lowercase
F means float and
upper or lowercase D means
double.
Exponents
use a notation that I’ve always found rather dismaying: 1.39 e-47f.
In science and engineering, ‘e’ refers to the base of
natural logarithms, approximately
2.718. (A more precise double value is available in Java as
Math.E.) This is used in exponentiation expressions such as 1.39 x
e-47, which means 1.39 x 2.718-47. However, when
FORTRAN was invented they decided that e would
naturally mean “ten to the power,” which is an odd decision because
FORTRAN was designed for science and engineering and one would think its
designers would be sensitive about introducing such an
ambiguity.[16]
At any rate, this custom was followed in C, C++ and now Java. So if you’re
used to thinking in terms of e as the base of natural logarithms, you
must do a mental translation when you see an expression such as 1.39
e-47f in Java; it means 1.39 x 10-47.
Note that you don’t need to
use the trailing character when the compiler can figure out the appropriate
type. With
long n3 = 200;
there’s no ambiguity, so an
L after the 200 would be superfluous. However, with
float f4 = 1e-47f; // 10 to the power
the compiler normally takes
exponential numbers as doubles, so without the trailing f it will give
you an error telling you that you must use a cast to convert double to
float.
You’ll discover that if you
perform any mathematical or bitwise operations on primitive data types that are
smaller than an int (that is, char, byte, or short),
those values will be promoted to int before
performing the operations, and the resulting value will be of type int.
So if you want to assign back into the smaller type, you must use a cast. (And,
since you’re assigning back into a smaller type, you might be losing
information.) In general, the largest data type in an expression is the one that
determines the size of the result of that expression; if you multiply a
float and a double, the result will be double; if you add
an int and a long, the result will be
long.
In C and C++, the
sizeof( ) operator satisfies a specific need:
it tells you the number of bytes allocated for data items. The most compelling
need for sizeof( ) in C and C++ is
portability. Different data types might be different
sizes on different machines, so the programmer must find out how big those types
are when performing operations that are sensitive to size. For example, one
computer might store integers in 32 bits, whereas another might store integers
as 16 bits. Programs could store larger values in integers on the first machine.
As you might imagine, portability is a huge headache for C and C++
programmers.
Java does not need a
sizeof( ) operator for this purpose because all the data types are
the same size on all machines. You do not need to think about portability on
this level – it is designed into the
language.
Upon hearing me complain about the
complexity of remembering operator
precedence during one of my seminars, a student suggested a mnemonic that is
simultaneously a commentary: “Ulcer Addicts Really Like C A
lot.”
Mnemonic |
Operator type |
Operators |
Ulcer |
Unary |
+ - ++ – [[
rest...]] |
Addicts |
Arithmetic (and
shift) |
* / % + - <<
>> |
Really |
Relational |
> < >= <= ==
!= |
Like |
Logical (and
bitwise) |
&& || & | ^
|
C |
Conditional
(ternary) |
A > B ? X : Y |
A Lot |
Assignment |
= (and compound assignment like
*=) |
Of course, with the shift and
bitwise operators distributed around the table it is not a perfect mnemonic, but
for non-bit operations it
works.
The following example shows which
primitive
data types can be used with particular operators. Basically, it is the same
example repeated over and over, but using different primitive data types. The
file will compile without error because the lines that would cause errors are
commented out with a //!.
//: AllOps.java // Tests all the operators on all the // primitive data types to show which // ones are accepted by the Java compiler. class AllOps { // To accept the results of a boolean test: void f(boolean b) {} void boolTest(boolean x, boolean y) { // Arithmetic operators: //! x = x * y; //! x = x / y; //! x = x % y; //! x = x + y; //! x = x - y; //! x++; //! x--; //! x = +y; //! x = -y; // Relational and logical: //! f(x > y); //! f(x >= y); //! f(x < y); //! f(x <= y); f(x == y); f(x != y); f(!y); x = x && y; x = x || y; // Bitwise operators: //! x = ~y; x = x & y; x = x | y; x = x ^ y; //! x = x << 1; //! x = x >> 1; //! x = x >>> 1; // Compound assignment: //! x += y; //! x -= y; //! x *= y; //! x /= y; //! x %= y; //! x <<= 1; //! x >>= 1; //! x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! char c = (char)x; //! byte B = (byte)x; //! short s = (short)x; //! int i = (int)x; //! long l = (long)x; //! float f = (float)x; //! double d = (double)x; } void charTest(char x, char y) { // Arithmetic operators: x = (char)(x * y); x = (char)(x / y); x = (char)(x % y); x = (char)(x + y); x = (char)(x - y); x++; x--; x = (char)+y; x = (char)-y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: x= (char)~y; x = (char)(x & y); x = (char)(x | y); x = (char)(x ^ y); x = (char)(x << 1); x = (char)(x >> 1); x = (char)(x >>> 1); // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; x <<= 1; x >>= 1; x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! boolean b = (boolean)x; byte B = (byte)x; short s = (short)x; int i = (int)x; long l = (long)x; float f = (float)x; double d = (double)x; } void byteTest(byte x, byte y) { // Arithmetic operators: x = (byte)(x* y); x = (byte)(x / y); x = (byte)(x % y); x = (byte)(x + y); x = (byte)(x - y); x++; x--; x = (byte)+ y; x = (byte)- y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: x = (byte)~y; x = (byte)(x & y); x = (byte)(x | y); x = (byte)(x ^ y); x = (byte)(x << 1); x = (byte)(x >> 1); x = (byte)(x >>> 1); // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; x <<= 1; x >>= 1; x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; short s = (short)x; int i = (int)x; long l = (long)x; float f = (float)x; double d = (double)x; } void shortTest(short x, short y) { // Arithmetic operators: x = (short)(x * y); x = (short)(x / y); x = (short)(x % y); x = (short)(x + y); x = (short)(x - y); x++; x--; x = (short)+y; x = (short)-y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: x = (short)~y; x = (short)(x & y); x = (short)(x | y); x = (short)(x ^ y); x = (short)(x << 1); x = (short)(x >> 1); x = (short)(x >>> 1); // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; x <<= 1; x >>= 1; x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; byte B = (byte)x; int i = (int)x; long l = (long)x; float f = (float)x; double d = (double)x; } void intTest(int x, int y) { // Arithmetic operators: x = x * y; x = x / y; x = x % y; x = x + y; x = x - y; x++; x--; x = +y; x = -y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: x = ~y; x = x & y; x = x | y; x = x ^ y; x = x << 1; x = x >> 1; x = x >>> 1; // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; x <<= 1; x >>= 1; x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; byte B = (byte)x; short s = (short)x; long l = (long)x; float f = (float)x; double d = (double)x; } void longTest(long x, long y) { // Arithmetic operators: x = x * y; x = x / y; x = x % y; x = x + y; x = x - y; x++; x--; x = +y; x = -y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: x = ~y; x = x & y; x = x | y; x = x ^ y; x = x << 1; x = x >> 1; x = x >>> 1; // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; x <<= 1; x >>= 1; x >>>= 1; x &= y; x ^= y; x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; byte B = (byte)x; short s = (short)x; int i = (int)x; float f = (float)x; double d = (double)x; } void floatTest(float x, float y) { // Arithmetic operators: x = x * y; x = x / y; x = x % y; x = x + y; x = x - y; x++; x--; x = +y; x = -y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: //! x = ~y; //! x = x & y; //! x = x | y; //! x = x ^ y; //! x = x << 1; //! x = x >> 1; //! x = x >>> 1; // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; //! x <<= 1; //! x >>= 1; //! x >>>= 1; //! x &= y; //! x ^= y; //! x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; byte B = (byte)x; short s = (short)x; int i = (int)x; long l = (long)x; double d = (double)x; } void doubleTest(double x, double y) { // Arithmetic operators: x = x * y; x = x / y; x = x % y; x = x + y; x = x - y; x++; x--; x = +y; x = -y; // Relational and logical: f(x > y); f(x >= y); f(x < y); f(x <= y); f(x == y); f(x != y); //! f(!x); //! f(x && y); //! f(x || y); // Bitwise operators: //! x = ~y; //! x = x & y; //! x = x | y; //! x = x ^ y; //! x = x << 1; //! x = x >> 1; //! x = x >>> 1; // Compound assignment: x += y; x -= y; x *= y; x /= y; x %= y; //! x <<= 1; //! x >>= 1; //! x >>>= 1; //! x &= y; //! x ^= y; //! x |= y; // Casting: //! boolean b = (boolean)x; char c = (char)x; byte B = (byte)x; short s = (short)x; int i = (int)x; long l = (long)x; float f = (float)x; } } ///:~
Note that
boolean is quite limited. You can assign to it the
values true and false, and you can test it for truth or falsehood,
but you cannot add booleans or perform any other type of operation on
them.
In char, byte, and
short you can see the effect of promotion with the
arithmetic operators. Each arithmetic operation on any of those types results in
an int result, which must be explicitly cast back to the original type (a
narrowing conversion that might lose information) to assign back to that type.
With int values, however, you do not need to cast, because everything is
already an int. Don’t be lulled into thinking everything is safe,
though. If you multiply two ints that are big enough, you’ll
overflow the result. The following example demonstrates
this:
//: Overflow.java // Surprise! Java lets you overflow. public class Overflow { public static void main(String[] args) { int big = 0x7fffffff; // max int value prt("big = " + big); int bigger = big * 4; prt("bigger = " + bigger); } static void prt(String s) { System.out.println(s); } } ///:~
The output of this
is:
big = 2147483647 bigger = -4
and you get no errors or warnings
from the compiler, and no exceptions at run-time. Java is good, but it’s
not that good.
Compound assignments do not
require casts for char, byte, or short, even though they
are performing promotions that have the same results as the direct arithmetic
operations. On the other hand, the lack of the cast certainly simplifies the
code.
You can see that, with the
exception of boolean, any primitive type can be
cast to any other primitive type. Again, you must be aware of the effect of a
narrowing conversion when casting to a smaller type, otherwise you might
unknowingly lose information during the
cast.
Java uses all of C’s
execution control statements, so if you’ve programmed with C or C++ then
most of what you see will be familiar. Most procedural programming languages
have some kind of control statements, and there is often overlap among
languages. In Java, the keywords include if-else, while,
do-while, for, and a selection statement called switch.
Java does not, however, support the much-maligned goto (which can still
be the most expedient way to solve certain types of problems). You can still do
a goto-like jump, but it is much more constrained than a typical
goto.
All conditional statements use the
truth or falsehood of a conditional expression to determine the execution path.
An example of a conditional expression is A == B. This uses the
conditional operator == to see if the value of A is equivalent to
the value of B. The expression returns true or false. Any
of the relational operators you’ve seen earlier in this chapter can be
used to produce a conditional statement. Note that Java doesn’t allow you
to use a number as a boolean, even though it’s allowed in C and C++
(where truth is nonzero and falsehood is zero). If you want to use a
non-boolean in a boolean test, such as if(a), you must
first convert it to a boolean value using a conditional expression, such
as if(a !=
0).
The if-else statement is
probably the most basic way to control program flow. The else is
optional, so you can use if in two forms:
if(Boolean-expression) statement
or
if(Boolean-expression) statement else statement
The conditional must produce a
Boolean result. The statement means either a simple statement terminated
by a semicolon or a compound statement, which is a group of simple statements
enclosed in braces. Anytime the word “statement” is used, it
always implies that the statement can be simple or compound.
As an example of if-else,
here is a test( ) method that will tell you whether a guess is
above, below, or equivalent to a target number:
static int test(int testval) { int result = 0; if(testval > target) result = -1; else if(testval < target) result = +1; else result = 0; // match return result; }
It is conventional to indent the
body of a control flow statement so the reader might easily determine where it
begins and ends.
The return keyword has two
purposes: it specifies what value a method will return (if it doesn’t have
a void return value) and it causes that value to be returned immediately.
The test( ) method above can be rewritten to take advantage of
this:
static int test2(int testval) { if(testval > target) return -1; if(testval < target) return +1; return 0; // match }
while, do-while and
for control looping and are sometimes classified as iteration
statements. A statement repeats until the controlling
Boolean-expression evaluates to false. The form for a while
loop is
while(Boolean-expression) statement
The Boolean-expression is
evaluated once at the beginning of the loop and again before each further
iteration of the statement.
Here’s a simple example that
generates random numbers until a particular condition is met:
//: WhileTest.java // Demonstrates the while loop public class WhileTest { public static void main(String[] args) { double r = 0; while(r < 0.99d) { r = Math.random(); System.out.println(r); } } } ///:~
This uses the static method
random( ) in the Math library, which generates a double
value between 0 and 1. (It includes 0, but not 1.) The conditional
expression for the while says “keep doing this loop until the
number is 0.99 or greater.” Each time you run this program you’ll
get a different-sized list of
numbers.
The form for do-while
is
do statement while(Boolean-expression);
The sole difference between
while and do-while is that the statement of the do-while
always executes at least once, even if the expression evaluates to false the
first time. In a while, if the conditional is false the first time the
statement never executes. In practice, do-while is less common than
while.
A for loop performs
initialization before the first iteration. Then it performs conditional testing
and, at the end of each iteration, some form of “stepping.” The form
of the for loop is:
for(initialization; Boolean-expression; step) statement
Any of the expressions
initialization, Boolean-expression or step can be empty.
The expression is tested before each iteration, and as soon as it evaluates to
false execution will continue at the line following the for
statement. At the end of each loop, the step executes.
for loops are usually used
for “counting” tasks:
//: ListCharacters.java // Demonstrates "for" loop by listing // all the ASCII characters. public class ListCharacters { public static void main(String[] args) { for( char c = 0; c < 128; c++) if (c != 26 ) // ANSI Clear screen System.out.println( "value: " + (int)c + " character: " + c); } } ///:~
Note that the variable c is
defined at the point where it is used, inside the control expression of the
for loop, rather than at the beginning of the block denoted by the open
curly brace. The scope of c is the expression controlled by the
for.
Traditional procedural languages
like C require that all variables be defined at the
beginning of a block so when the compiler creates a block it can allocate space
for those variables. In Java and C++ you can spread your variable declarations
throughout the block, defining them at the point that you need them. This allows
a more natural coding style and makes code easier to
understand.
You can define multiple variables
within a for statement, but they must be of the same
type:
for(int i = 0, j = 1; i < 10 && j != 11; i++, j++) /* body of for loop */;
The int definition in the
for statement covers both i and j. The ability to
define variables in the control expression is limited to the for loop.
You cannot use this approach with any of the other selection or iteration
statements.
Earlier in this chapter I stated
that the comma operator
(not the comma separator, which is used to separate function arguments)
has only one use in Java: in the control expression of a for loop. In
both the initialization and step portions of the control expression you can have
a number of statements separated by commas, and those statements will be
evaluated sequentially. The previous bit of code uses this ability. Here’s
another example:
//: CommaOperator.java public class CommaOperator { public static void main(String[] args) { for(int i = 1, j = i + 10; i < 5; i++, j = i * 2) { System.out.println("i= " + i + " j= " + j); } } } ///:~
Here’s the
output:
i= 1 j= 11 i= 2 j= 4 i= 3 j= 6 i= 4 j= 8
You can see that in both the
initialization and step portions the statements are evaluated in sequential
order. Also, the initialization portion can have any number of definitions of
one type.
Inside the body of any of the
iteration statements you can also control the flow of the loop by using
break and continue. break quits the loop without executing
the rest of the statements in the loop. continue stops the execution of
the current iteration and goes back to the beginning of the loop to begin a new
iteration.
This program shows examples of
break and continue within for and while
loops:
//: BreakAndContinue.java // Demonstrates break and continue keywords public class BreakAndContinue { public static void main(String[] args) { for(int i = 0; i < 100; i++) { if(i == 74) break; // Out of for loop if(i % 9 != 0) continue; // Next iteration System.out.println(i); } int i = 0; // An "infinite loop": while(true) { i++; int j = i * 27; if(j == 1269) break; // Out of loop if(i % 10 != 0) continue; // Top of loop System.out.println(i); } } } ///:~
In the for loop the value of
i never gets to 100 because the break statement breaks out of the
loop when i is 74. Normally, you’d use a break like this
only if you didn’t know when the terminating condition was going to occur.
The continue statement causes execution to go back to the top of the
iteration loop (thus incrementing i) whenever i is not evenly
divisible by 9. When it is, the value is printed.
The second portion shows an
“infinite loop” that would, in theory, continue forever. However,
inside the loop there is a break statement that will break out of the
loop. In addition, you’ll see that the continue moves back to the
top of the loop without completing the remainder. (Thus printing happens only
when the value of i is divisible by 9.) The output is:
0 9 18 27 36 45 54 63 72 10 20 30 40
The value 0 is printed because 0 %
9 produces 0.
A second form of the infinite loop
is for(;;). The compiler treats both while(true) and
for(;;) in the same way so whichever one you use is a matter of
programming taste.
The goto
keyword has been present in programming languages from the beginning.
Indeed, goto was the genesis of program control in assembly language:
“if condition A, then jump here, otherwise jump there.” If you read
the assembly code that is ultimately generated by virtually any compiler,
you’ll see that program control contains many jumps. However, goto
jumps at the source-code level, and that’s what brought it into
disrepute. If a program will always jump from one point to another, isn’t
there some way to reorganize the code so the flow of control is not so jumpy?
goto fell into true disfavor with the publication of the famous
“Goto considered harmful” paper by Edsger Dijkstra, and since then
goto-bashing has been a popular sport, with advocates of the cast-out keyword
scurrying for cover.
As is typical in situations like
this, the middle ground is the most fruitful. The problem is not the use of
goto but the overuse of goto, and in rare situations goto
is the best way to structure control flow.
Although goto is a reserved
word in Java, it is not used in the language; Java has no goto. However,
it does have something that looks a bit like a jump tied in with the
break and continue keywords. It’s not a jump but rather a
way to break out of an iteration statement.
The reason it’s often thrown in with
discussions of goto is because it uses the same mechanism: a
label.
label1:
The only place a label is
useful in Java is right before an iteration statement. And that means
right before – it does no good to put any other statement between
the label and the iteration. And the sole reason to put a label before an
iteration is if you’re going to nest another iteration or a switch inside
it. That’s because the break and
continue keywords will normally interrupt only the
current loop, but when used with a label they’ll interrupt the loops up to
where the label exists:
label1: outer-iteration { inner-iteration { //... break; // 1 //... continue; // 2 //... continue label1; // 3 //... break label1; // 4 } }
In case 1, the break breaks
out of the inner iteration and you end up in the outer iteration. In case 2, the
continue moves back to the beginning of the inner iteration. But in case
3, the continue label1 breaks out of the inner iteration and the
outer iteration, all the way back to label1. Then it does in fact
continue the iteration, but starting at the outer iteration. In case 4, the
break label1 also breaks all the way out to label1, but it does
not re-enter the iteration. It actually does break out of both
iterations.
Here is an example using for
loops:
//: LabeledFor.java // Java’s "labeled for loop" public class LabeledFor { public static void main(String[] args) { int i = 0; outer: // Can't have statements here for(; true ;) { // infinite loop inner: // Can't have statements here for(; i < 10; i++) { prt("i = " + i); if(i == 2) { prt("continue"); continue; } if(i == 3) { prt("break"); i++; // Otherwise i never // gets incremented. break; } if(i == 7) { prt("continue outer"); i++; // Otherwise i never // gets incremented. continue outer; } if(i == 8) { prt("break outer"); break outer; } for(int k = 0; k < 5; k++) { if(k == 3) { prt("continue inner"); continue inner; } } } } // Can't break or continue // to labels here } static void prt(String s) { System.out.println(s); } } ///:~
This uses the prt( )
method that has been defined in the other examples.
Note that break breaks out
of the for loop, and that the increment-expression doesn’t occur
until the end of the pass through the for loop. Since break skips
the increment expression, the increment is performed directly in the case of
i == 3. The continue outer statement in the case of I == 7
also goes to the top of the loop and also skips the increment, so it too is
incremented directly.
Here is the
output:
i = 0 continue inner i = 1 continue inner i = 2 continue i = 3 break i = 4 continue inner i = 5 continue inner i = 6 continue inner i = 7 continue outer i = 8 break outer
If not for the break outer
statement, there would be no way to get out of the outer loop from within an
inner loop, since break by itself can break out of only the innermost
loop. (The same is true for continue.)
Of course, in the cases where
breaking out of a loop will also exit the method, you can simply use a
return.
Here is a demonstration of labeled
break and continue statements with while
loops:
//: LabeledWhile.java // Java's "labeled while" loop public class LabeledWhile { public static void main(String[] args) { int i = 0; outer: while(true) { prt("Outer while loop"); while(true) { i++; prt("i = " + i); if(i == 1) { prt("continue"); continue; } if(i == 3) { prt("continue outer"); continue outer; } if(i == 5) { prt("break"); break; } if(i == 7) { prt("break outer"); break outer; } } } } static void prt(String s) { System.out.println(s); } } ///:~
The same rules hold true for
while:
The output of
this method makes it clear:
Outer while loop i = 1 continue i = 2 i = 3 continue outer Outer while loop i = 4 i = 5 break Outer while loop i = 6 i = 7 break outer
It’s important to remember
that the only reason to use labels in Java is when you have nested loops
and you want to break or continue through more than one nested
level.
In Dijkstra’s “goto
considered harmful” paper, what he specifically objected to was the
labels, not the goto. He observed that the number of bugs seems to increase with
the number of labels in a program. Labels and gotos make programs difficult to
analyze statically, since it introduces cycles in the program execution graph.
Note that Java labels don’t suffer from this problem, since they are
constrained in their placement and can’t be used to transfer control in an
ad hoc manner. It’s also interesting to note that this is a case where a
language feature is made more useful by restricting the power of the
statement.
The switch is sometimes
classified as a selection statement. The switch statement selects
from among pieces of code based on the value of an integral expression. Its form
is:
switch(integral-selector) { case integral-value1 : statement; break; case integral-value2 : statement; break; case integral-value3 : statement; break; case integral-value4 : statement; break; case integral-value5 : statement; break; // ... default: statement; }
Integral-selector is an
expression that produces an integral value. The switch compares the
result of integral-selector to each integral-value. If it finds a
match, the corresponding statement (simple or compound) executes. If no
match occurs, the default statement
executes.
You will notice in the above
definition that each case ends with a
break, which causes execution to jump to the end of the switch
body. This is the conventional way to build a switch statement, but the
break is optional. If it is missing, the code for the following case
statements execute until a break is encountered. Although you don’t
usually want this kind of behavior, it can be useful to an experienced
programmer. Note the last statement, for the default, doesn’t have
a break because the execution just falls through to where the
break would have taken it anyway. You could put a break at the end
of the default statement with no harm if you considered it important for
style’s sake.
The switch statement is a
clean way to implement multi-way selection (i.e., selecting from among a number
of different execution paths), but it requires a selector that evaluates to an
integral value such as int or char. If you want to use, for
example, a string or a floating-point number as a selector, it won’t work
in a switch statement. For non-integral types, you must use a series of
if statements.
Here’s an example that
creates letters randomly and determines whether they’re vowels or
consonants:
//: VowelsAndConsonants.java // Demonstrates the switch statement public class VowelsAndConsonants { public static void main(String[] args) { for(int i = 0; i < 100; i++) { char c = (char)(Math.random() * 26 + 'a'); System.out.print(c + ": "); switch(c) { case 'a': case 'e': case 'i': case 'o': case 'u': System.out.println("vowel"); break; case 'y': case 'w': System.out.println( "Sometimes a vowel"); break; default: System.out.println("consonant"); } } } } ///:~
Since Math.random( )
generates a value between 0 and 1, you need only multiply it by the upper bound
of the range of numbers you want to produce (26 for the letters in the alphabet)
and add an offset to establish the lower bound.
Although it appears you’re
switching on a character here, the switch statement is actually using the
integral value of the character. The singly-quoted characters in the case
statements also produce integral values that are used for
comparison.
Notice how the cases can be
“stacked” on top of each other to provide multiple matches for a
particular piece of code. You should also be aware that it’s essential to
put the break statement at the end of a particular case, otherwise
control will simply drop through and continue processing on the next
case.
The statement:
char c = (char)(Math.random() * 26 + 'a');
deserves a closer look.
Math.random( ) produces a double, so the value 26 is
converted to a double to perform the multiplication, which also produces
a double. This means that ‘a’ must be converted to a
double to perform the addition. The double result is turned back
into a char with a cast.
First, what does the cast to
char do? That is, if you have the value 29.7 and you cast it to a
char, is the resulting value 30 or 29? The answer to this can be seen in
this example:
//: CastingNumbers.java // What happens when you cast a float or double // to an integral value? public class CastingNumbers { public static void main(String[] args) { double above = 0.7, below = 0.4; System.out.println("above: " + above); System.out.println("below: " + below); System.out.println( "(int)above: " + (int)above); System.out.println( "(int)below: " + (int)below); System.out.println( "(char)('a' + above): " + (char)('a' + above)); System.out.println( "(char)('a' + below): " + (char)('a' + below)); } } ///:~
The output is:
above: 0.7 below: 0.4 (int)above: 0 (int)below: 0 (char)('a' + above): a (char)('a' + below): a
The second question has to do with
Math.random( ). Does
it produce a value from zero to one, inclusive or exclusive of the value
‘1’? In math lingo, is it (0,1), or [0,1], or (0,1] or [0,1)? (The
square bracket means “includes” whereas the parenthesis means
“doesn’t include.”) Again, a test program provides the
answer:
//: RandomBounds.java // Does Math.random() produce 0.0 and 1.0? public class RandomBounds { static void usage() { System.err.println("Usage: \n\t" + "RandomBounds lower\n\t" + "RandomBounds upper"); System.exit(1); } public static void main(String[] args) { if(args.length != 1) usage(); if(args[0].equals("lower")) { while(Math.random() != 0.0) ; // Keep trying System.out.println("Produced 0.0!"); } else if(args[0].equals("upper")) { while(Math.random() != 1.0) ; // Keep trying System.out.println("Produced 1.0!"); } else usage(); } } ///:~
To run the program, you type a
command line of either:
java RandomBounds lower
or
java RandomBounds upper
In both cases you are forced to
break out of the program manually, so it would appear that
Math.random( ) never produces either 0.0 or 1.0. But this is where
such an experiment can be deceiving. If you consider that there are
2128 different double fractions between 0 and 1, the likelihood of
reaching any one value experimentally might exceed the lifetime of one computer,
or even one experimenter. It turns out that 0.0 is included in the output
of Math.random( ). Or, in math lingo, it is
[0,1).
This chapter concludes the study of
fundamental features that appear in most programming languages: calculation,
operator precedence, type casting, and selection and iteration. Now you’re
ready to begin taking steps that move you closer to the world of object-oriented
programming. The next chapter will cover the important issues of initialization
and cleanup of objects, followed in the subsequent chapter by the essential
concept of implementation
hiding.
[16]
John Kirkham writes, “I started computing in 1962 using FORTRAN II on an
IBM 1620. At that time, and throughout the 1960s and into the 1970s, FORTRAN was
an all uppercase language. This probably started because many of the early input
devices were old teletype units that used 5 bit Baudot code, which had no
lowercase capability. The ‘E’ in the exponential notation was also
always upper case and was never confused with the natural logarithm base
‘e’, which is always lower case. The ‘E’ simply stood
for exponential, which was for the base of the number system used –
usually 10. At the time octal was also widely used by programmers. Although I
never saw it used, if I had seen an octal number in exponential notation I would
have considered it to be base 8. The first time I remember seeing an exponential
using a lower case ‘e’ was in the late 1970s and I also found it
confusing. The problem arose as lowercase crept into FORTRAN, not at its
beginning. We actually had functions to use if you really wanted to use the
natural logarithm base, but they were all uppercase.”