On Jun 1, 1:18 pm, Anees <anees.hai...@gmail.com> wrote:
Dear Fellows:
I have following scenario:
class A {
public:
void me_print () { cout << "A" << endl; }
static void print() { cout << "static A" << endl; }
};
class B {
};
class C : public A, public B {
public:
void foo() { PRINT; }
};
class D: public B {
public:
void foo() { PRINT; }
};
int _tmain(int argc, _TCHAR* argv[])
{
C c;
D d;
c.foo();
d.foo();
return 0;
}
I want to define a macro (#define) such that when instance of A's
derived class (i.e. C) call PRINT; it gets A's member method me_print
(), and when instance of NON A's derived class (i.e. D) call PRINT, it
should be resolved to A's static subroutine print().
You don't need a macro for that, you can do that in pure C++ (for
clarity, I changed class to struct and removed all public keywords):
struct A {
void me_print () { std::cout << "A\n"; }
static void print() { std::cout << "static A\n"; }
};
void do_print(A* a) { a->me_print(); }
void do_print(void*) { A::print(); }
struct B {};
struct C : A, B {
void foo() { do_print(this); }
};
struct D : B {
void foo() { do_print(this); }
};
int main()
{
C c;
D d;
c.foo();
d.foo();
}
The idea here is that there are two overloads of do_print(). One takes
A*, the other one void*. The first one is chosen when you pass A* or
its publicly derived sub-class pointer. The void* overload is chosen
for all other pointers.
--
Max- Hide quoted text -
- Show quoted text -