Contents | Prev | Next | Index Java Language Specification
Second Edition


CHAPTER 14

Blocks and Statements


The sequence of execution of a program is controlled by statements, which are executed for their effect and do not have values.

Some statements contain other statements as part of their structure; such other statements are substatements of the statement. We say that statement S immediately contains statement U if there is no statement T different from S and U such that S contains T and T contains U. In the same manner, some statements contain expressions (§15) as part of their structure.

The first section of this chapter discusses the distinction between normal and abrupt completion of statements (§14.1). Most of the remaining sections explain the various kinds of statements, describing in detail both their normal behavior and any special treatment of abrupt completion.

Blocks are explained first (§14.2), followed by local class declarations (§14.3) and local variable declaration statements (§14.4).

Next a grammatical maneuver that sidesteps the familiar "dangling else" problem (§14.5) is explained.

Statements that will be familiar to C and C++ programmers are the empty (§14.6), labeled (§14.7), expression (§14.8), if (§14.9), switch (§14.10), while (§14.11), do (§14.12), for (§14.13), break (§14.14), continue (§14.15), and return (§14.16) statements.

Unlike C and C++, the Java programming language has no goto statement. However, the break and continue statements are allowed to mention statement labels.

The Java programming language statements that are not in the C language are the throw (§14.17), synchronized (§14.18), and try (§14.19) statements.

The last section (§14.20) of this chapter addresses the requirement that every statement be reachable in a certain technical sense.

14.1 Normal and Abrupt Completion of Statements

Every statement has a normal mode of execution in which certain computational steps are carried out. The following sections describe the normal mode of execution for each kind of statement.

If all the steps are carried out as described, with no indication of abrupt completion, the statement is said to complete normally. However, certain events may prevent a statement from completing normally:

If such an event occurs, then execution of one or more statements may be terminated before all steps of their normal mode of execution have completed; such statements are said to complete abruptly.

An abrupt completion always has an associated reason, which is one of the following:

The terms "complete normally" and "complete abruptly" also apply to the evaluation of expressions (§15.6). The only reason an expression can complete abruptly is that an exception is thrown, because of either a throw with a given value (§14.17) or a run-time exception or error (§11, §15.6).

If a statement evaluates an expression, abrupt completion of the expression always causes the immediate abrupt completion of the statement, with the same reason. All succeeding steps in the normal mode of execution are not performed.

Unless otherwise specified in this chapter, abrupt completion of a substatement causes the immediate abrupt completion of the statement itself, with the same reason, and all succeeding steps in the normal mode of execution of the statement are not performed.

Unless otherwise specified, a statement completes normally if all expressions it evaluates and all substatements it executes complete normally.

14.2 Blocks

A block is a sequence of statements, local class declarations and local variable declaration statements within braces.

A block is executed by executing each of the local variable declaration statements and other statements in order from first to last (left to right). If all of these block statements complete normally, then the block completes normally. If any of these block statements complete abruptly for any reason, then the block completes abruptly for the same reason.

14.3 Local Class Declarations

A local class is a nested class (§8) that is not a member of any class and that has a name. All local classes are inner classes (§8.1.2). Every local class declaration statement is immediately contained by a block. Local class declaration statements may be intermixed freely with other kinds of statements in the block.

The scope of a local class declared in a block is the rest of the immediately enclosing block, including its own class declaration.

The name of a local class C may not be redeclared as a local class of the directly enclosing method, constructor, or initializer block within the scope of C, or a compile-time error occurs. However, a local class declaration may be shadowed (§6.3.1) anywhere inside a class declaration nested within the local class declaration's scope. A local class does not have a canonical name, nor does it have a fully qualified name.

It is a compile-time error if a local class declaration contains any one of the following access modifiers: public, protected, private, or static.

Here is an example that illustrates several aspects of the rules given above:

class Global {
	class Cyclic {}
	void foo() {
		new Cyclic(); // create a Global.Cyclic
		class Cyclic extends Cyclic{}; // circular definition
		{
			class Local{};
			{
				class Local{}; // compile-time error
			}
			class Local{}; // compile-time error
			class AnotherLocal {
				void bar() {
					class Local {}; // ok
				}
			}
		}
		class Local{}; // ok, not in scope of prior Local
}
The first statement of method foo creates an instance of the member class Global.Cyclic rather than an instance of the local class Cyclic, because the local class declaration is not yet in scope.

The fact that the scope of a local class encompasses its own declaration (not only its body) means that the definition of the local class Cyclic is indeed cyclic because it extends itself rather than Global.Cyclic. Consequently, the declaration of the local class Cyclic will be rejected at compile time.

Since local class names cannot be redeclared within the same method (or constructor or initializer, as the case may be), the second and third declarations of Local result in compile-time errors. However, Local can be redeclared in the context of another, more deeply nested, class such as AnotherLocal.

The fourth and last declaration of Local is legal, since it occurs outside the scope of any prior declaration of Local.

14.4 Local Variable Declaration Statements

A local variable declaration statement declares one or more local variable names.

The following are repeated from §8.3 to make the presentation here clearer:

Every local variable declaration statement is immediately contained by a block. Local variable declaration statements may be intermixed freely with other kinds of statements in the block.

A local variable declaration can also appear in the header of a for statement (§14.13). In this case it is executed in the same manner as if it were part of a local variable declaration statement.

14.4.1 Local Variable Declarators and Types

Each declarator in a local variable declaration declares one local variable, whose name is the Identifier that appears in the declarator.

If the optional keyword final appears at the start of the declarator, the variable being declared is a final variable(§4.5.4).

The type of the variable is denoted by the Type that appears in the local variable declaration, followed by any bracket pairs that follow the Identifier in the declarator.

Thus, the local variable declaration:

int a, b[], c[][];
is equivalent to the series of declarations:

int a;
int[] b;
int[][] c;
Brackets are allowed in declarators as a nod to the tradition of C and C++. The general rule, however, also means that the local variable declaration:

float[][] f[][], g[][][], h[];													// Yechh!
is equivalent to the series of declarations:

float[][][][] f;
float[][][][][] g;
float[][][] h;
We do not recommend such "mixed notation" for array declarations.

A local variable of type float always contains a value that is an element of the float value set (§4.2.3); similarly, a local variable of type double always contains a value that is an element of the double value set. It is not permitted for a local variable of type float to contain an element of the float-extended-exponent value set that is not also an element of the float value set, nor for a local variable of type double to contain an element of the double-extended-exponent value set that is not also an element of the double value set.

14.4.2 Scope of Local Variable Declarations

The scope of a local variable declaration in a block (§14.4.2) is the rest of the block in which the declaration appears, starting with its own initializer (§14.4) and including any further declarators to the right in the local variable declaration statement.

The name of a local variable v may not be redeclared as a local variable of the directly enclosing method, constructor or initializer block within the scope of v, or a compile-time error occurs. The name of a local variable v may not be redeclared as an exception parameter of a catch clause in a try statement of the directly enclosing method, constructor or initializer block within the scope of v, or a compile-time error occurs. However, a local variable of a method or initializer block may be shadowed (§6.3.1) anywhere inside a class declaration nested within the scope of the local variable.

A local variable cannot be referred to using a qualified name (§6.6), only a simple name.

The example:

class Test {
	static int x;
	public static void main(String[] args) {
		int x = x;
	}
}
causes a compile-time error because the initialization of x is within the scope of the declaration of x as a local variable, and the local x does not yet have a value and cannot be used.

The following program does compile:

class Test {
	static int x;
	public static void main(String[] args) {
		int x = (x=2)*2;
		System.out.println(x);
	}
}
because the local variable x is definitely assigned (§16) before it is used. It prints:

4

Here is another example:

class Test { public static void main(String[] args) { System.out.print("2+1="); int two = 2, three = two + 1; System.out.println(three); } }

which compiles correctly and produces the output:

2+1=3
The initializer for three can correctly refer to the variable two declared in an earlier declarator, and the method invocation in the next line can correctly refer to the variable three declared earlier in the block.

The scope of a local variable declared in a for statement is the rest of the for statement, including its own initializer.

If a declaration of an identifier as a local variable of the same method, constructor, or initializer block appears within the scope of a parameter or local variable of the same name, a compile-time error occurs.

Thus the following example does not compile:

class Test {
	public static void main(String[] args) {
		int i;
		for (int i = 0; i < 10; i++)
			System.out.println(i);
	}
}
This restriction helps to detect some otherwise very obscure bugs. A similar restriction on shadowing of members by local variables was judged impractical, because the addition of a member in a superclass could cause subclasses to have to rename local variables. Related considerations make restrictions on shadowing of local variables by members of nested classes, or on shadowing of local variables by local variables declared within nested classes unattractive as well. Hence, the following example compiles without error:

class Test {
	public static void main(String[] args) {
		int i;
		class Local {
			{
				for (int i = 0; i < 10; i++)
				System.out.println(i);
			}
		}
		new Local();
	}
}
On the other hand, local variables with the same name may be declared in two separate blocks or for statements neither of which contains the other. Thus:

class Test {
	public static void main(String[] args) {
		for (int i = 0; i < 10; i++)
			System.out.print(i + " ");
		for (int i = 10; i > 0; i--)
			System.out.print(i + " ");
		System.out.println();
	}
}
compiles without error and, when executed, produces the output:

0 1 2 3 4 5 6 7 8 9 10 9 8 7 6 5 4 3 2 1

14.4.3 Shadowing of Names by Local Variables

If a name declared as a local variable is already declared as a field name, then that outer declaration is shadowed (§6.3.1) throughout the scope of the local variable. Similarly, if a name is already declared as a variable or parameter name, then that outer declaration is shadowed throughout the scope of the local variable (provided that the shadowing does not cause a compile-time error under the rules of §14.4.2). The shadowed name can sometimes be accessed using an appropriately qualified name.

For example, the keyword this can be used to access a shadowed field x, using the form this.x. Indeed, this idiom typically appears in constructors (§8.8):

class Pair {
	Object first, second;
	public Pair(Object first, Object second) {
		this.first = first;
		this.second = second;
	}
}
In this example, the constructor takes parameters having the same names as the fields to be initialized. This is simpler than having to invent different names for the parameters and is not too confusing in this stylized context. In general, however, it is considered poor style to have local variables with the same names as fields.

14.4.4 Execution of Local Variable Declarations

A local variable declaration statement is an executable statement. Every time it is executed, the declarators are processed in order from left to right. If a declarator has an initialization expression, the expression is evaluated and its value is assigned to the variable. If a declarator does not have an initialization expression, then a Java compiler must prove, using exactly the algorithm given in §16, that every reference to the variable is necessarily preceded by execution of an assignment to the variable. If this is not the case, then a compile-time error occurs.

Each initialization (except the first) is executed only if the evaluation of the preceding initialization expression completes normally. Execution of the local variable declaration completes normally only if evaluation of the last initialization expression completes normally; if the local variable declaration contains no initialization expressions, then executing it always completes normally.

14.5 Statements

There are many kinds of statements in the Java programming language. Most correspond to statements in the C and C++ languages, but some are unique.

As in C and C++, the if statement of the Java programming language suffers from the so-called "dangling else problem," illustrated by this misleadingly formatted example:


if (door.isOpen())
	if (resident.isVisible())
		resident.greet("Hello!");
else door.bell.ring();	// A "dangling else"
The problem is that both the outer if statement and the inner if statement might conceivably own the else clause. In this example, one might surmise that the programmer intended the else clause to belong to the outer if statement. The Java programming language, like C and C++ and many programming languages before them, arbitrarily decree that an else clause belongs to the innermost if to which it might possibly belong. This rule is captured by the following grammar:

The following are repeated from §14.9 to make the presentation here clearer:

Statements are thus grammatically divided into two categories: those that might end in an if statement that has no else clause (a "short if statement") and those that definitely do not. Only statements that definitely do not end in a short if statement may appear as an immediate substatement before the keyword else in an if statement that does have an else clause.

This simple rule prevents the "dangling else" problem. The execution behavior of a statement with the "no short if" restriction is identical to the execution behavior of the same kind of statement without the "no short if" restriction; the distinction is drawn purely to resolve the syntactic difficulty.

14.6 The Empty Statement

An empty statement does nothing.

Execution of an empty statement always completes normally.

14.7 Labeled Statements

Statements may have label prefixes.

The Identifier is declared to be the label of the immediately contained Statement.

Unlike C and C++, the Java programming language has no goto statement; identifier statement labels are used with break (§14.14) or continue (§14.15) statements appearing anywhere within the labeled statement.

The scope of a label declared by a labeled statement is the statement immediately enclosed by the labeled statement.

Let l be a label, and let m be the immediately enclosing method, constructor, instance initializer or static initializer. It is a compile-time error if l shadows (§6.3.1) the declaration of another label immediately enclosed in m.

There is no restriction against using the same identifier as a label and as the name of a package, class, interface, method, field, parameter, or local variable. Use of an identifier to label a statement does not obscure (§6.3.2) a package, class, interface, method, field, parameter, or local variable with the same name. Use of an identifier as a class, interface, method, field, local variable or as the parameter of an exception handler (§14.19) does not obscure a statement label with the same name.

A labeled statement is executed by executing the immediately contained Statement. If the statement is labeled by an Identifier and the contained Statement completes abruptly because of a break with the same Identifier, then the labeled statement completes normally. In all other cases of abrupt completion of the Statement, the labeled statement completes abruptly for the same reason.

14.8 Expression Statements

Certain kinds of expressions may be used as statements by following them with semicolons:

An expression statement is executed by evaluating the expression; if the expression has a value, the value is discarded. Execution of the expression statement completes normally if and only if evaluation of the expression completes normally.

Unlike C and C++, the Java programming language allows only certain forms of expressions to be used as expression statements. Note that the Java programming language does not allow a "cast to void"-void is not a type-so the traditional C trick of writing an expression statement such as:

(void) ... ;			// incorrect!
does not work. On the other hand, the language allows all the most useful kinds of expressions in expressions statements, and it does not require a method invocation used as an expression statement to invoke a void method, so such a trick is almost never needed. If a trick is needed, either an assignment statement (§15.26) or a local variable declaration statement (§14.4) can be used instead.

14.9 The if Statement

The if statement allows conditional execution of a statement or a conditional choice of two statements, executing one or the other but not both.

The Expression must have type boolean, or a compile-time error occurs.

14.9.1 The if-then Statement

An if-then statement is executed by first evaluating the Expression. If evaluation of the Expression completes abruptly for some reason, the if-then statement completes abruptly for the same reason. Otherwise, execution continues by making a choice based on the resulting value:

14.9.2 The if-then-else Statement

An if-then-else statement is executed by first evaluating the Expression. If evaluation of the Expression completes abruptly for some reason, then the if-then-else statement completes abruptly for the same reason. Otherwise, execution continues by making a choice based on the resulting value:

14.10 The switch Statement

The switch statement transfers control to one of several statements depending on the value of an expression.

The type of the Expression must be char, byte, short, or int, or a compile-time error occurs.

The body of a switch statement is known as a switch block. Any statement immediately contained by the switch block may be labeled with one or more case or default labels. These labels are said to be associated with the switch statement, as are the values of the constant expressions (§15.28) in the case labels.

All of the following must be true, or a compile-time error will result:

for (i = 0; i < n; ++i) foo();
where n is known to be positive. A trick known as Duff's device can be used in C or C++ to unroll the loop, but this is not valid code in the Java programming language:

int q = (n+7)/8;
switch (n%8) {
case 0:		do {	foo();		// Great C hack, Tom,
case 7:			foo();		// but it's not valid here.
case 6:			foo();
case 5:			foo();
case 4:			foo();
case 3:			foo();
case 2:			foo();
case 1			foo();
		} while (--q >= 0);
}
Fortunately, this trick does not seem to be widely known or used. Moreover, it is less needed nowadays; this sort of code transformation is properly in the province of state-of-the-art optimizing compilers.

When the switch statement is executed, first the Expression is evaluated. If evaluation of the Expression completes abruptly for some reason, the switch statement completes abruptly for the same reason. Otherwise, execution continues by comparing the value of the Expression with each case constant. Then there is a choice:

If any statement immediately contained by the Block body of the switch statement completes abruptly, it is handled as follows:

As in C and C++, execution of statements in a switch block "falls through labels."

For example, the program:

class Toomany {
	static void howMany(int k) {
		switch (k) {
		case 1:			System.out.print("one ");
		case 2:			System.out.print("too ");
		case 3:			System.out.println("many");
		}
	}
	public static void main(String[] args) {
		howMany(3);
		howMany(2);
		howMany(1);
	}
}
contains a switch block in which the code for each case falls through into the code for the next case. As a result, the program prints:

many
too many
one too many
If code is not to fall through case to case in this manner, then break statements should be used, as in this example:

class Twomany {
	static void howMany(int k) {
		switch (k) {
		case 1:			System.out.println("one");
					break;					// exit the switch
		case 2:			System.out.println("two");
					break;					// exit the switch
		case 3:			System.out.println("many");
					break;					// not needed, but good style
		}
	}
	public static void main(String[] args) {
		howMany(1);
		howMany(2);
		howMany(3);
	}
}
This program prints:

one
two
many

14.11 The while Statement

The while statement executes an Expression and a Statement repeatedly until the value of the Expression is false.

The Expression must have type boolean, or a compile-time error occurs.

A while statement is executed by first evaluating the Expression. If evaluation of the Expression completes abruptly for some reason, the while statement completes abruptly for the same reason. Otherwise, execution continues by making a choice based on the resulting value:

If the value of the Expression is false the first time it is evaluated, then the Statement is not executed.

14.11.1 Abrupt Completion

Abrupt completion of the contained Statement is handled in the following manner:

14.12 The do Statement

The do statement executes a Statement and an Expression repeatedly until the value of the Expression is false.

The Expression must have type boolean, or a compile-time error occurs.

A do statement is executed by first executing the Statement. Then there is a choice:

Executing a do statement always executes the contained Statement at least once.

14.12.1 Abrupt Completion

Abrupt completion of the contained Statement is handled in the following manner:

14.12.2 Example of do statement

The following code is one possible implementation of the toHexString method of class Integer:

public static String toHexString(int i) {
	StringBuffer buf = new StringBuffer(8);
	do {
		buf.append(Character.forDigit(i & 0xF, 16));
		i >>>= 4;
	} while (i != 0);
	return buf.reverse().toString();
}
Because at least one digit must be generated, the do statement is an appropriate control structure.

14.13 The for Statement

The for statement executes some initialization code, then executes an Expression, a Statement, and some update code repeatedly until the value of the Expression is false.

The Expression must have type boolean, or a compile-time error occurs.

14.13.1 Initialization of for statement

A for statement is executed by first executing the ForInit code:

If the ForInit code is a local variable declaration, it is executed as if it were a local variable declaration statement (§14.4) appearing in a block. The scope of a local variable declared in the ForInit part of a for statement (§14.13) includes all of the following:

If execution of the local variable declaration completes abruptly for any reason, the for statement completes abruptly for the same reason.

14.13.2 Iteration of for statement

Next, a for iteration step is performed, as follows:

If the value of the Expression is false the first time it is evaluated, then the Statement is not executed.

If the Expression is not present, then the only way a for statement can complete normally is by use of a break statement.

14.13.3 Abrupt Completion of for statement

Abrupt completion of the contained Statement is handled in the following manner:

14.14 The break Statement

A break statement transfers control out of an enclosing statement.

A break statement with no label attempts to transfer control to the innermost enclosing switch, while, do, or for statement of the immediately enclosing method or initializer block; this statement, which is called the break target, then immediately completes normally.

To be precise, a break statement with no label always completes abruptly, the reason being a break with no label. If no switch, while, do, or for statement encloses the break statement, a compile-time error occurs.

A break statement with label Identifier attempts to transfer control to the enclosing labeled statement (§14.7) that has the same Identifier as its label; this statement, which is called the break target, then immediately completes normally. In this case, the break target need not be a while, do, for, or switch statement. A break statement must refer to a label within the immediately enclosing method or initializer block. There are no non-local jumps.

To be precise, a break statement with label Identifier always completes abruptly, the reason being a break with label Identifier. If no labeled statement with Identifier as its label encloses the break statement, a compile-time error occurs.

It can be seen, then, that a break statement always completes abruptly.

The preceding descriptions say "attempts to transfer control" rather than just "transfers control" because if there are any try statements (§14.19) within the break target whose try blocks contain the break statement, then any finally clauses of those try statements are executed, in order, innermost to outermost, before control is transferred to the break target. Abrupt completion of a finally clause can disrupt the transfer of control initiated by a break statement.

In the following example, a mathematical graph is represented by an array of arrays. A graph consists of a set of nodes and a set of edges; each edge is an arrow that points from some node to some other node, or from a node to itself. In this example it is assumed that there are no redundant edges; that is, for any two nodes P and Q, where Q may be the same as P, there is at most one edge from P to Q. Nodes are represented by integers, and there is an edge from node i to node edges[i][j] for every i and j for which the array reference edges[i][j] does not throw an IndexOutOfBoundsException.

The task of the method loseEdges, given integers i and j, is to construct a new graph by copying a given graph but omitting the edge from node i to node j, if any, and the edge from node j to node i, if any:

class Graph {
	int edges[][];
	public Graph(int[][] edges) { this.edges = edges; }
	public Graph loseEdges(int i, int j) {
		int n = edges.length;
		int[][] newedges = new int[n][];
		for (int k = 0; k < n; ++k) {
			edgelist: {
				int z;
				search: {
					if (k == i) {
						for (z = 0; z < edges[k].length; ++z)
							if (edges[k][z] == j)
								break search;
					} else if (k == j) {
						for (z = 0; z < edges[k].length; ++z)
							if (edges[k][z] == i)
								break search;
					}
					// No edge to be deleted; share this list.
					newedges[k] = edges[k];
					break edgelist;
				} //search
				// Copy the list, omitting the edge at position z.
				int m = edges[k].length - 1;
				int ne[] = new int[m];
				System.arraycopy(edges[k], 0, ne, 0, z);
				System.arraycopy(edges[k], z+1, ne, z, m-z);
				newedges[k] = ne;
			} //edgelist
		}
		return new Graph(newedges);
	}
}
Note the use of two statement labels, edgelist and search, and the use of break statements. This allows the code that copies a list, omitting one edge, to be shared between two separate tests, the test for an edge from node i to node j, and the test for an edge from node j to node i.

14.15 The continue Statement

A continue statement may occur only in a while, do, or for statement; statements of these three kinds are called iteration statements. Control passes to the loop-continuation point of an iteration statement.

A continue statement with no label attempts to transfer control to the innermost enclosing while, do, or for statement of the immediately enclosing method or initializer block; this statement, which is called the continue target, then immediately ends the current iteration and begins a new one.

To be precise, such a continue statement always completes abruptly, the reason being a continue with no label. If no while, do, or for statement of the immediately enclosing method or initializer block encloses the continue statement, a compile-time error occurs.

A continue statement with label Identifier attempts to transfer control to the enclosing labeled statement (§14.7) that has the same Identifier as its label; that statement, which is called the continue target, then immediately ends the current iteration and begins a new one. The continue target must be a while, do, or for statement or a compile-time error occurs. A continue statement must refer to a label within the immediately enclosing method or initializer block. There are no non-local jumps.

More precisely, a continue statement with label Identifier always completes abruptly, the reason being a continue with label Identifier. If no labeled statement with Identifier as its label contains the continue statement, a compile-time error occurs.

It can be seen, then, that a continue statement always completes abruptly.

See the descriptions of the while statement (§14.11), do statement (§14.12), and for statement (§14.13) for a discussion of the handling of abrupt termination because of continue.

The preceding descriptions say "attempts to transfer control" rather than just "transfers control" because if there are any try statements (§14.19) within the continue target whose try blocks contain the continue statement, then any finally clauses of those try statements are executed, in order, innermost to outermost, before control is transferred to the continue target. Abrupt completion of a finally clause can disrupt the transfer of control initiated by a continue statement.

In the Graph example in the preceding section, one of the break statements is used to finish execution of the entire body of the outermost for loop. This break can be replaced by a continue if the for loop itself is labeled:

class Graph {
	. . .
	public Graph loseEdges(int i, int j) {
		int n = edges.length;
		int[][] newedges = new int[n][];
		edgelists: for (int k = 0; k < n; ++k) {
			int z;
			search: {
				if (k == i) {
					. . .
				} else if (k == j) {
					. . .
				}
				newedges[k] = edges[k];
				continue edgelists;
			} // search
			. . .
		} // edgelists
		return new Graph(newedges);
	}
}
Which to use, if either, is largely a matter of programming style.

14.16 The return Statement

A return statement returns control to the invoker of a method (§8.4, §15.12) or constructor (§8.8, §15.9).

A return statement with no Expression must be contained in the body of a method that is declared, using the keyword void, not to return any value (§8.4), or in the body of a constructor (§8.8). A compile-time error occurs if a return statement appears within an instance initializer or a static initializer (§8.7). A return statement with no Expression attempts to transfer control to the invoker of the method or constructor that contains it.

To be precise, a return statement with no Expression always completes abruptly, the reason being a return with no value.

A return statement with an Expression must be contained in a method declaration that is declared to return a value (§8.4) or a compile-time error occurs. The Expression must denote a variable or value of some type T, or a compile-time error occurs. The type T must be assignable (§5.2) to the declared result type of the method, or a compile-time error occurs.

A return statement with an Expression attempts to transfer control to the invoker of the method that contains it; the value of the Expression becomes the value of the method invocation. More precisely, execution of such a return statement first evaluates the Expression. If the evaluation of the Expression completes abruptly for some reason, then the return statement completes abruptly for that reason. If evaluation of the Expression completes normally, producing a value V, then the return statement completes abruptly, the reason being a return with value V. If the expression is of type float and is not FP-strict (§15.4), then the value may be an element of either the float value set or the float-extended-exponent value set (§4.2.3). If the expression is of type double and is not FP-strict, then the value may be an element of either the double value set or the double-extended-exponent value set.

It can be seen, then, that a return statement always completes abruptly.

The preceding descriptions say "attempts to transfer control" rather than just "transfers control" because if there are any try statements (§14.19) within the method or constructor whose try blocks contain the return statement, then any finally clauses of those try statements will be executed, in order, innermost to outermost, before control is transferred to the invoker of the method or constructor. Abrupt completion of a finally clause can disrupt the transfer of control initiated by a return statement.

14.17 The throw Statement

A throw statement causes an exception (§11) to be thrown. The result is an immediate transfer of control (§11.3) that may exit multiple statements and multiple constructor, instance initializer, static initializer and field initializer evaluations, and method invocations until a try statement (§14.19) is found that catches the thrown value. If no such try statement is found, then execution of the thread (§17) that executed the throw is terminated (§11.3) after invocation of the uncaughtException method for the thread group to which the thread belongs.

The Expression in a throw statement must denote a variable or value of a reference type which is assignable (§5.2) to the type Throwable, or a compile-time error occurs. Moreover, at least one of the following three conditions must be true, or a compile-time error occurs:

A throw statement first evaluates the Expression. If the evaluation of the Expression completes abruptly for some reason, then the throw completes abruptly for that reason. If evaluation of the Expression completes normally, producing a non-null value V, then the throw statement completes abruptly, the reason being a throw with value V. If evaluation of the Expression completes normally, producing a null value, then an instance V' of class NullPointerException is created and thrown instead of null. The throw statement then completes abruptly, the reason being a throw with value V'.

It can be seen, then, that a throw statement always completes abruptly.

If there are any enclosing try statements (§14.19) whose try blocks contain the throw statement, then any finally clauses of those try statements are executed as control is transferred outward, until the thrown value is caught. Note that abrupt completion of a finally clause can disrupt the transfer of control initiated by a throw statement.

If a throw statement is contained in a method declaration, but its value is not caught by some try statement that contains it, then the invocation of the method completes abruptly because of the throw.

If a throw statement is contained in a constructor declaration, but its value is not caught by some try statement that contains it, then the class instance creation expression that invoked the constructor will complete abruptly because of the throw.

If a throw statement is contained in a static initializer (§8.7), then a compile-time check ensures that either its value is always an unchecked exception or its value is always caught by some try statement that contains it. If at run-time, despite this check, the value is not caught by some try statement that contains the throw statement, then the value is rethrown if it is an instance of class Error or one of its subclasses; otherwise, it is wrapped in an ExceptionInInitializerError object, which is then thrown (§12.4.2).

If a throw statement is contained in an instance initializer (§8.6), then a compile-time check ensures that either its value is always an unchecked exception or its value is always caught by some try statement that contains it, or the type of the thrown exception (or one of its superclasses) occurs in the throws clause of every constructor of the class.

By convention, user-declared throwable types should usually be declared to be subclasses of class Exception, which is a subclass of class Throwable (§11.5).

14.18 The synchronized Statement

A synchronized statement acquires a mutual-exclusion lock (§17.13) on behalf of the executing thread, executes a block, then releases the lock. While the executing thread owns the lock, no other thread may acquire the lock.

The type of Expression must be a reference type, or a compile-time error occurs.

A synchronized statement is executed by first evaluating the Expression.

If evaluation of the Expression completes abruptly for some reason, then the synchronized statement completes abruptly for the same reason.

Otherwise, if the value of the Expression is null, a NullPointerException is thrown.

Otherwise, let the non-null value of the Expression be V. The executing thread locks the lock associated with V. Then the Block is executed. If execution of the Block completes normally, then the lock is unlocked and the synchronized statement completes normally. If execution of the Block completes abruptly for any reason, then the lock is unlocked and the synchronized statement then completes abruptly for the same reason.

Acquiring the lock associated with an object does not of itself prevent other threads from accessing fields of the object or invoking unsynchronized methods on the object. Other threads can also use synchronized methods or the synchronized statement in a conventional manner to achieve mutual exclusion.

The locks acquired by synchronized statements are the same as the locks that are acquired implicitly by synchronized methods; see §8.4.3.6. A single thread may hold a lock more than once.

The example:

class Test {
	public static void main(String[] args) {
		Test t = new Test();
		synchronized(t) {
			synchronized(t) {
				System.out.println("made it!");
			}
		}
	}
}
prints:

made it!
This example would deadlock if a single thread were not permitted to lock a lock more than once.

14.19 The try statement

A try statement executes a block. If a value is thrown and the try statement has one or more catch clauses that can catch it, then control will be transferred to the first such catch clause. If the try statement has a finally clause, then another block of code is executed, no matter whether the try block completes normally or abruptly, and no matter whether a catch clause is first given control.

The following is repeated from §8.4.1 to make the presentation here clearer:

The following is repeated from §8.3 to make the presentation here clearer:

The Block immediately after the keyword try is called the try block of the try statement. The Block immediately after the keyword finally is called the finally block of the try statement.

A try statement may have catch clauses (also called exception handlers). A catch clause must have exactly one parameter (which is called an exception parameter); the declared type of the exception parameter must be the class Throwable or a subclass of Throwable, or a compile-time error occurs. The scope of the parameter variable is the Block of the catch clause.

An exception parameter of a catch clause must not have the same name as a local variable or parameter of the method or initializer block immediately enclosing the catch clause, or a compile-time error occurs.

The scope of a parameter of an exception handler that is declared in a catch clause of a try statement (§14.19) is the entire block associated with the catch.

Within the Block of the catch clause, the name of the parameter may not be redeclared as a local variable of the directly enclosing method or initializer block, nor may it be redeclared as an exception parameter of a catch clause in a try statement of the directly enclosing method or initializer block, or a compile-time error occurs. However, an exception parameter may be shadowed (§6.3.1) anywhere inside a class declaration nested within the Block of the catch clause.

It is a compile-time error if an exception parameter that is declared final is assigned to within the body of the catch clause.

Exception parameters cannot be referred to using qualified names (§6.6), only by simple names.

Exception handlers are considered in left-to-right order: the earliest possible catch clause accepts the exception, receiving as its actual argument the thrown exception object.

A finally clause ensures that the finally block is executed after the try block and any catch block that might be executed, no matter how control leaves the try block or catch block.

Handling of the finally block is rather complex, so the two cases of a try statement with and without a finally block are described separately.

14.19.1 Execution of try-catch

A try statement without a finally block is executed by first executing the try block. Then there is a choice:

class BlewIt extends Exception {
	BlewIt() { }
	BlewIt(String s) { super(s); }
}
class Test {
	static void blowUp() throws BlewIt { throw new BlewIt(); }
	public static void main(String[] args) {
		try {
			blowUp();
		} catch (RuntimeException r) {
			System.out.println("RuntimeException:" + r);
		} catch (BlewIt b) {
			System.out.println("BlewIt");
		}
	}
}
the exception BlewIt is thrown by the method blowUp. The try-catch statement in the body of main has two catch clauses. The run-time type of the exception is BlewIt which is not assignable to a variable of type RuntimeException, but is assignable to a variable of type BlewIt, so the output of the example is:

BlewIt

14.19.2 Execution of try-catch-finally

A try statement with a finally block is executed by first executing the try block. Then there is a choice:

class BlewIt extends Exception {
	BlewIt() { }
	BlewIt(String s) { super(s); }
}
class Test {
	static void blowUp() throws BlewIt {
		throw new NullPointerException();
	}
	public static void main(String[] args) {
		try {
			blowUp();
		} catch (BlewIt b) {
			System.out.println("BlewIt");
		} finally {
			System.out.println("Uncaught Exception");
		}
	}
}
produces the output:

Uncaught Exception
java.lang.NullPointerException
	at Test.blowUp(Test.java:7)
	at Test.main(Test.java:11)
The NullPointerException (which is a kind of RuntimeException) that is thrown by method blowUp is not caught by the try statement in main, because a NullPointerException is not assignable to a variable of type BlewIt. This causes the finally clause to execute, after which the thread executing main, which is the only thread of the test program, terminates because of an uncaught exception, which typically results in printing the exception name and a simple backtrace.

14.20 Unreachable Statements

It is a compile-time error if a statement cannot be executed because it is unreachable. Every Java compiler must carry out the conservative flow analysis specified here to make sure all statements are reachable.

This section is devoted to a precise explanation of the word "reachable." The idea is that there must be some possible execution path from the beginning of the constructor, method, instance initializer or static initializer that contains the statement to the statement itself. The analysis takes into account the structure of statements. Except for the special treatment of while, do, and for statements whose condition expression has the constant value true, the values of expressions are not taken into account in the flow analysis.

For example, a Java compiler will accept the code:

{
	int n = 5;
	while (n > 7) k = 2;
}
even though the value of n is known at compile time and in principle it can be known at compile time that the assignment to k can never be executed.

A Java compiler must operate according to the rules laid out in this section.

The rules in this section define two technical terms:

The definitions here allow a statement to complete normally only if it is reachable.

To shorten the description of the rules, the customary abbreviation "iff" is used to mean "if and only if."

The rules are as follows:

The then-statement is reachable iff the if-then statement is reachable and the condition expression is not a constant expression whose value is false.

The actual rules for the if statement are as follows:

while (false) { x=3; }
because the statement x=3; is not reachable; but the superficially similar case:

if (false) { x=3; }
does not result in a compile-time error. An optimizing compiler may realize that the statement x=3; will never be executed and may choose to omit the code for that statement from the generated class file, but the statement x=3; is not regarded as "unreachable" in the technical sense specified here.

The rationale for this differing treatment is to allow programmers to define "flag variables" such as:

static final boolean DEBUG = false;
and then write code such as:

if (DEBUG) { x=3; }
The idea is that it should be possible to change the value of DEBUG from false to true or from true to false and then compile the code correctly with no other changes to the program text.

This ability to "conditionally compile" has a significant impact on, and relationship to, binary compatibility (§13). If a set of classes that use such a "flag" variable are compiled and conditional code is omitted, it does not suffice later to distribute just a new version of the class or interface that contains the definition of the flag. A change to the value of a flag is, therefore, not binary compatible with preexisting binaries (§13.4.8). (There are other reasons for such incompatibility as well, such as the use of constants in case labels in switch statements; see §13.4.8.)


Contents | Prev | Next | Index Java Language Specification
Second Edition
Copyright © 2000 Sun Microsystems, Inc. All rights reserved
Please send any comments or corrections to jls@java.sun.com