Evaluating Method Invocations

This explanation will use the following simplified example.

class A {

public void f() { System.out.println(" A-f"); }

public void h(A o) { System.out.print("A-h"); }

public void m(A o) { System.out.print("A-m"); }

}

class B extends A {

public void f() { System.out.println(" B-f"); }

public void h(A o) { System.out.print("B-h"); }

public void m(B o) { System.out.print("B-m"); }

}

public class MT2Test1 {

public static void main(String[] args) {

B b1 = new B();

B b2 = new B();

A a1 = new A();

A a2 = b2;

b1.m(a1);

a2.m(a2);

}

Algorithm for evaluating method invocations at compile-time:

· Determine the declared class of the variable on which the method is being called.

B b1 = new B1();

...

b1.m(a1); // The declared type of b1 is B.

· Start searching in that class for a method with the same name and arguments that match in number and declared type of the objects being passed in.

B b1 = new B1();

A a1 = new A1();

...

b1.m(a1); // Search in B for a method called m that takes

// as an argument a single A (or any superclass

// of A).

· Continue up the inheritance hierarchy looking for a matching method in the same manner as above. If not found, it is a compile-time error.

· Write down the method descriptor of the method that matches at compile-time. It will be used in the runtime evaluation. A method descriptor is the return type and signature of the method. Remember the signature is the name of the method and the type and order of the arguments.

void m(A) // write this down for use in runtime step

Algorithm for evaluating method invocations at runtime:

· Determine the type of the object to which the variable is bound. Your search will begin in this class.

· Look for a method with the exact descriptor you found earlier from the compile-time evaluation.

· If found, you’re done. If not, continue up the inheritance hierarchy. You’re guaranteed to find an implementation.

Walkthrough #1

Compile-time evaluation of b1.m(a1):

· Declared class of b1 is B. Begin search in this class.

· Declared class of a1 is A. Look for method called m in B that takes a single A (or Object) argument.

· Matching method not found in B so go to A. Found match!

· Write down the descriptor of the matching method. void m(A)

Runtime evaluation of b1.m(a1):

· The type of the object bound to b1 is B. Begin search in this class.

· Look for method in B with descriptor found in compile-time evaluation.

· Matching method (void m(A)) not found in B so go to A. Found match!

· A’s version of m is executed.

Walkthrough #2

Compile-time evaluation of a2.m(a2):

· Declared class of a2 is A. Begin search in this class.

· Declared class of a2 is A. Look for method called m in A that takes a single A (or Object) argument.

· Found match!

· Write down the descriptor of the matching method. void m(A)

Runtime evaluation of a2.m(a2):

· The type of the object bound to a2 is B. Begin search in this class.

· Look for method in B with descriptor found in compile-time evaluation.

· Matching method (void m(A)) not found in B so go to A. Found match!

· A’s version of m is executed.