static inner class and static member of a class shares SAME NAME?

Question

How can the static inner class M and static member M [of class C] share the same name?

The following code generates "White" as output:

public class Amazed{

    public static void main(String[] args) {
        System.out.println(B.M.W);
    }
}

class B {
    public static class M {       
        static String W = "Black";
    }

    static C M = new C();
}

class C {
    String W = "White";
}

how the member object is accessed and not the static class member : W ["Black"]

if i want to access the member within static class M how to do that?

I like the method `meth` takes you nowhere. An very long while loop might be more appropriate though. — thatidiotguy, Jul 26 '13 at 17:09
@thatidiotguy: Or something that recurses until it blows the stack. — T.J. Crowder, Jul 26 '13 at 17:13
Thanks BMT. But i don't understand the specification of obscuring and reclassification... could someone help me with that. — Dineshkumar, Jul 26 '13 at 17:22
@Dineshkumar There's precedence in how accessing fields is done. — Sotirios Delimanolis, Jul 26 '13 at 17:23
well if you are amazed, read this http://stackoverflow.com/questions/17690135/what-is-the-meaning-of-this-java-code — ZhongYu, Jul 26 '13 at 23:18

Ted Hopp · Accepted Answer · 2013-07-26T17:37:15.927

Chapter 6 of the Java Language Specification (especially Section 6.5) spells out the gruesome details of how Java sorts out what a particular identifier means in a particular context. The rules are fairly complex, but, roughly speaking, Java has six name spaces:

Package names
type names
field (variable) names
method names
local variable names (including parameters)
labels

The same identifier can be used for entities in each of these name spaces. Note that type (class) names and field names live separately, which explains why your code is legal.

Inherited names in the same name space can also sometimes be shadowed or masked. Sometimes an identifier is ambiguous; it then needs to be qualified in some way (e.g., with a package name) or the compiler will complain.

Code obfuscators use this to great advantage, to the point where you can end up with a package named a.a and a class named a in package a (which also would be identified as a.a). Not to mention that Java key words like do and for are legal names in .class files (but not in Java source). It helps make reverse engineering a real bear.

score 1 · Answer 2 · answered Jul 26 '13 at 17:29

1

The variable obscures the type of the same name. If a name could be interpreted to be either a variable or a type, the variable is preferred.

You can avoid this by not giving them the same name.

From the Java Language Specification, section 6.4.2: Obscuring:

6.4.2. Obscuring

A simple name may occur in contexts where it may potentially be interpreted as the name of a variable, a type, or a package. In these situations, the rules of §6.5 specify that a variable will be chosen in preference to a type, and that a type will be chosen in preference to a package. Thus, it is may sometimes be impossible to refer to a visible type or package declaration via its simple name. We say that such a declaration is obscured.

answered Jul 26 '13 at 17:29

Andy Thomas

84,978
11
107
151

how it is chosen? whats the order of preference? – Dineshkumar Jul 26 '13 at 17:30
1

1) Variable 2) Type 3) Package name – Andy Thomas Jul 26 '13 at 17:33
does Type here means anything other than class? – Dineshkumar Jul 26 '13 at 17:33
It means any type, including classes and interfaces. Primitive types are keywords, so you couldn't use them as a variable name. – Andy Thomas Jul 26 '13 at 17:35

cyon · Answer 3 · 2014-03-04T20:09:21.863

In your particular case where M can be instantiated you could use the fact that statics can be accessed in a non-static way:

public class Amazed{
   @SuppressWarnings("static-access")
   public static void main(String[] args) {
       B.M val = new B.M();
       System.out.println(val.W);
   }
}

class B {
   public static class M {       
       static String W = "Black";
   }

   static C M = new C();
}

class C {
   String W = "White";
}

The above will print 'Black' because you disambiguated the identifier by referring to it through an instance of B.M. By no means is this a good idea in production code, of course, since static fields should not be accessed in a non-static way. Likewise, it requires you to create an instance of B.M.

score 0 · Answer 4 · answered Jul 26 '13 at 17:42

After everyone's help and little play i found that even without creating object for inner class we can access the member W "Black".

simply using this statement

M.W

But i can't figure out access level. [because its public?] even without public it worked. how's that?

static inner class and static member of a class shares SAME NAME?

4 Answers4