regarding overloading-Dan Chisholm's Mock Exam question

[debasmita pattnayak]

Hi Ranchers,

Can anyone explain me more clearly the following code:

  class A{} class B extends A{} class C extends B{static void m(A a,A a1){System.out.println("A,A");}
static void m(A a, B b){System.out.println("A,B");} 
static void m(B b,A a){System.out.println("B,A");}
static void m(B b,B b1){System.out.println("B,B");}
static void m(A a,C c){System.out.println("A,C");} 
public static void main(String args[]){
 C c = new C(); m(c,c);
}
}

options:

a) Prints: A,A
b) Prints: A,B
c) Prints: B,A
d) Prints: B,B
e) Prints: A,C
f) Compile-time error
g) Run-time error
h) none of the above


I marked the answer as " d". but to my surprise the answer is "f".
Can somebody explain me the answer in a better way.

[chandra sekhar.k]

Hi debasmita,
One thumb rule that you can remember is "Super class references can hold child class objects where as child class reference cannot hold superclass object".

next in your question,
all each and every method m() defined ,can handle m(c,c).
because all are super classes of c.

next
"method call to the methodbody linkage will be done at the time of compile time only"
so you are putting the compiler in confusion ,like which method to link to your m(c,c).

so..the above situation leads to ambiguity...thats why compile time error...

i hope you got it....

[Chandra Bhatt]

class A{} class B extends A{} class C extends B{static void m(A a,A a1){System.out.println("A,A");}
static void m(A a, B b){System.out.println("A,B");} 
static void m(B b,A a){System.out.println("B,A");}
static void m(B b,B b1){System.out.println("B,B");}
static void m(A a,C c){System.out.println("A,C");} 
public static void main(String args[]){
C c = new C(); m(c,c);
}
}

There is method call ambiguity that's why compiler error. Let's see how:
Compiler has to select most specific method to satisfy your method call.
Method selection should be so that others method also pass the test
without any bother.
Now lets check, whether m(B,B) is appropriate, (think in a way that if
m(B,B) does not exist and any one method exist individually.
#1- m(A a, B b)will do the job
#2- m(B b,A a) will also do the job
#3- m(A a,C c), will NOT pass the method call

Because test #3 fails, your method selection is not most OK.

In this way we see that none of the method is more appropriate than the
other.

How was that technique?

Thanks,

[yogesh srinivasan]

but still i'm confused on how this ambiguity is to be identified with respect to the method call

[Manfred Klug]

Originally posted by yogesh srinivasan:
but still i'm confused on how this ambiguity is to be identified with respect to the method call

Have a look at the following discussion.