Function Pointers in Classes

[Stephen Hewitt]

Firstly, your approach using void (*display)(char*); doesn't work as it points to a global or static class member function, not a member function. One way to to do it is the following, assuming the function to be called is always display:

1. Define an interface that contains the function:

   class ICallable
   {
   public:
   virtual void display(char *msg) = 0;
   };

   Note the function is a pure virtual function.

2. Derive the classes that have the function to be called from it:

   class A : public ICallable
   {
   public:
   virtual void display(char *msg)
   {
   // A's impelentation here...
   }
   };

3. Your pointer now looks like the following:

   ICallable *pPointer;

4. The call through the pointer like this:

   pPointer->display("Hello world!");

Steve

[ParagPatel]

What is your requirement? Is it ok if you have static "display" function in class A?

Parag Patel Sr. Software Eng, Varaha Systems

[Divyang Mithaiwala]

What you have done is shown that you want some common operation in display method of class A & class B. So, for that here you can take one base class which define display method. And after that you can inherit your 2 classes A & B from that base class. Both have access of display method.

---

Do not trust a computer... Always check what computer is doing regards, Divyang Mithaiwala Software Engineer

[Stuart Dootson]

I would implement this using Boost.Function[^] and Boost.Bind[^]: In class B, declare display like this:

boost::function<void(char*)> display;

in A, set b.display like this:

B b;
b.display = boost::bind(&A::display, this);

in B, call it like this:

if (display) display(whatever string you want to display

The if (display) bit is to protect you against the case that nothing's been assigned to B::display. Here's a complete, buildable and runnable example:

#include <iostream>
#include <boost/bind.hpp>
#include <boost/function.hpp>

class B
{
public:
boost::function<void(char*)> display;
void b_test() { if (display) display("test"); }
};

class A
{
public:
A() { b.display = boost::bind(&A::display, this, _1); }
void a_test() { b.b_test(); }
void display(char* s) { std::cout << s << std::endl; }
B b;
};

int main()
{
A a;
a.a_test();
}

Java, Basic, who cares - it's all a bunch of tree-hugging hippy cr*p

[led mike]

Stuart Dootson wrote:

I would implement this using Boost.Function[^] and Boost.Bind[^]:

The Visual C++ 2008 Feature Pack has implemented TR1[^] Although, I would prefer using interfaces as Stephen Hewitt posted.

[CPallini]

Why do you want to follow a very plain-C path? You know, C++ is a fully featured OOP launguage. :)

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler. -- Alfonso the Wise, 13th Century King of Castile.
This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong. -- Iain Clarke
[My articles]

[Stuart Dootson]

led mike wrote:

The Visual C++ 2008 Feature Pack has implemented TR1[^]

Caught me - I've not got used to having TR1 yet. Also, TR1 doesn't have Boost.Bind in it :-( (or Boost.Lambda, for that matter - even more :-()

led mike wrote:

Although, I would prefer using interfaces as Stephen Hewitt posted.

Fair enough - each approach has its own benefits and drawbacks.

Java, Basic, who cares - it's all a bunch of tree-hugging hippy cr*p

[led mike]

Stuart Dootson wrote:

Also, TR1 doesn't have Boost.Bind in it Frown (or Boost.Lambda, for that matter - even more Frown)

Well no, there is no Boost namespace but those features are in TR1.

[Stuart Dootson]

So it is (Boost.Bind equivalent, that is)! Well, you learn something new every day! So, come VS2010, when C++0x lambdas are available, I'll be well happy :-)

Java, Basic, who cares - it's all a bunch of tree-hugging hippy cr*p

[led mike]

Stuart Dootson wrote:

lambdas are available

Yeah, I suppose that's a good thing, we'll see.

[CPallini]

Stuart Dootson wrote:

Well, you learn something new every day!

Get in times Stuart! ;P of course I knew nothing about... :rolleyes:

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler. -- Alfonso the Wise, 13th Century King of Castile.
This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong. -- Iain Clarke
[My articles]

[Arsalan Malik]

You are right, thanks, perhaps I am thinking too much in C :-D

ARSALAN MALIK

[Arsalan Malik]

Thanks a lot for the hint, perhaps I was thinking too much in C :-D

ARSALAN MALIK