在 Java 开关中声明和初始化变量

我有一个关于 Java 交换机的疯狂问题。

int key = 2;


switch (key) {
case 1:
int value = 1;
break;
case 2:
value = 2;
System.out.println(value);
break;
default:
break;
}

场景1-当 key为2时，它成功地将值打印为2。
场景2-当我要在 case 2:中注释 value = 2时，它会发出嘎嘎的声音说 局部变量值可能尚未初始化。

问题:

场景1: 如果执行流不到 case 1:(当 key = 2时) ，那么它如何知道值变量的类型为 int？

场景2: 如果编译器知道值变量的类型为 int，那么它必须访问 case 1:中的 int value = 1;表达式。(声明及初始化)。那么为什么它嘎嘎作响，当我要在 case 2:中注释 value = 2，说 局部变量值可能尚未初始化。

22670 次浏览
小开

From http://www.coderanch.com/t/447381/java-programmer-SCJP/certification/variable-initialization-within-case-block

Declarations are processed at compile time and do not depend on the execution flow of your code. Since value is declared within the local scope of the switch block, it is useable anywhere in that block from the point of its declaration.

小开

The variable has been declared (as an int), but not initialized (assigned an initial value). Think of the line:

int value = 1;

As:

int value;
value = 1;

The int value part tells the compiler at compile time that you have a variable called value which is an int. The value = 1 part initializes it, but that happens at run-time, and doesn't happen at all if that branch of the switch isn't entered.

小开
最佳答案

Switch statements are odd in terms of scoping, basically. From section 6.3 of the JLS:

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

In your case, case 2 is in the same block as case 1 and appears after it, even though case 1 will never execute... so the local variable is in scope and available for writing despite you logically never "executing" the declaration. (A declaration isn't really "executable" although initialization is.)

If you comment out the value = 2; assignment, the compiler still knows which variable you're referring to, but you won't have gone through any execution path which assigns it a value, which is why you get an error as you would when you try to read any other not-definitely-assigned local variable.

I would strongly recommend you not to use local variables declared in other cases - it leads to highly confusing code, as you've seen. When I introduce local variables in switch statements (which I try to do rarely - cases should be very short, ideally) I usually prefer to introduce a new scope:

case 1: {
int value = 1;
...
break;
}
case 2: {
int value = 2;
...
break;
}

I believe this is clearer.

小开

With the integration of JEP 325: Switch Expressions (Preview) in JDK-12 early access builds. There are certain changes that could be seen from Jon's answer -

  1. Local Variable Scope - The local variables in the switch cases can now be local to the case itself instead of the entire switch block. An example (similar to what Jon had attempted syntactically as well) considering the Day enum class for further explanation :

    public enum Day {
MONDAY, TUESDAY, WEDNESDAY, THURSDAY, FRIDAY, SATURDAY, SUNDAY
}


// some another method implementation
Day day = Day.valueOf(scanner.next());
switch (day) {
case MONDAY,TUESDAY -> {
var temp = "mon-tue";
System.out.println(temp);
}
case WEDNESDAY,THURSDAY -> {
var temp = Date.from(Instant.now()); // same variable name 'temp'
System.out.println(temp);
}
default ->{
var temp = 0.04; // different types as well (not mandatory ofcourse)
System.out.println(temp);
}
}

  2. Switch Expressions - If the intent is to assign a value to a variable and then make use of it, once can make use of the switch expressions. e.g.

    private static void useSwitchExpression() {
int key = 2;
int value = switch (key) {
case 1 ->  1;
case 2 -> 2;
default -> {break 0;}
};
System.out.println("value = " + value); // prints 'value = 2'
}
小开

This Explanation might help.

    int id=1;


switch(id){
default:
boolean b= false; // all switch scope going down, because there is no scope tag


case 1:
b = false;
case 2:{
//String b= "test"; you can't declare scope here. because it's in the scope @top
b=true; // b is still accessible
}
case 3:{
boolean c= true; // case c scope only
b=true; // case 3 scope is whole switch
}
case 4:{
boolean c= false; // case 4 scope only
}
}
小开

Java spec:

https://docs.oracle.com/javase/specs/jls/se12/html/jls-14.html#jls-14.11

The case of abrupt completion because of a break with a label is handled by the general rule for labeled statements (§14.7).

https://docs.oracle.com/javase/specs/jls/se12/html/jls-14.html#jls-14.7

Labeled statements:

LabeledStatement: Identifier : Statement

LabeledStatementNoShortIf: Identifier : StatementNoShortIf

Unlike C and C++, the Java programming language has no goto statement; identifier statement labels are used with break (§14.15) or continue (§14.16) statements appearing anywhere within the labeled statement.

The scope of a label of a labeled statement is the immediately contained Statement.

In other words, case 1, case 2 are labels within the switch statement. break and continue statements can be applied to labels.

Because labels share the scope of the statement, all variables defined within labels share the scope of the switch statement.


提交答案