Overload-Problem

Question

Hey,
I just try to get through the Programmers Exam and found an error while overloading, that I can't understand.
Here is the code:
class A
public void test1(int a) { System.out.println("test1, integer");}
public void test1(float a) { System.out.println("test1, float");}
public void test2(int a) { System.out.println("test2, integer");}
public class OverloadTest extends A{
public OverloadTest() {
public static void main(String[] args)
int a=5;
OverloadTest ow=new OverloadTest();
ow.test1(a);
ow.test2(a); // Error at compile time
public void test2(float a) { System.out.println("test2, float");}
This produces following error at compile-time:
OverloadTest.java [19:1] reference to test2 is ambiguous, both method test2(int) in A and method test2(float) in OverloadTest match
ow.test2(a); // Error at compile time
I have the same overloaded methods test1 and test2. In the first case this is done in class A alone, in the second case in class A and OverloadTest.
What is the problem here ??
Why does the integer variable 'a' match both test2 methods ? ( I know that there is an implizit cast possible from int to float, but why does the compiler try this when he finds a perfect matching method in test2(int) ?)
The compiler does not have any problems with test1, which should be the same.
Anyone having an idear ?
Best regards
Markus

Answer 1

Approximately because each requires the same number of steps to get to it.
ow.test1(a) => A.test2(int a)
...Convert ow to A
ow.test1(a) => OW.test2(float a)
...Convert a to float
The precise answer requires that you break down which method would be used by going through the JLS step by step.
This thread provides an example which is fairly close to yours and the break down that I did of it. (The thread is probably cached so give it time to load.)
http://forum.java.sun.com/thread.jsp?forum=31&thread=159240
Let me emphasize that going through the break down is very interesting. However, it is not something that I am going to repeat more than once, so you will need to proceed with your own analysis.