Question

UML Diagram : https://i.stack.imgur.com/ajsVQ.png

My OOP teacher said that we could implement something like in the above diagram in C++ (for a lab assignment). She said we could do something like this so we wouldn’t need two List classes (one for Books and one for Clients).

I started implementing it and i’m stuck on getFromPosition(int) (it is supposed to return the object at certain position in the list). What I am trying to do is to save a Book object in a List object (this works fine) and i would later like to be able to return/read/modify these saved Book objects.It returns the object , but then I (probably) should make a cast of some sort so I could use the methods of the Book. Is this even possible?
This is the code for the method:

   Obj List::getFromPosition(int i){
            return list[i];
        }

I hope I gave enough info (this is my first question ).
Thanks in advance!

Answer 1

Since you’re returning an Obj by value, I’m guessing internally you hold a collection of Obj objects, and not pointers.

Which is wrong. You run into object slicing.

Firstly, you should keep pointers or smart pointers internally, and return an Obj* or a SmartPtr<Obj>.

Second, whether you need the cast or not depends.

If you have a common method between your deriving classes, it should probably go in the base class. If not, you probably need a cast.

Here’s how I’d do it rough cut:

class List
{
   std::vector<Obj*> _objects;
public:
   Obj* getFromPosition(int i)
   {
      return _objects[i];
   }
};

//...
List list;
//...
Obj* obj = list.getFromPosition(0);

obj->commonMethod();   //common method is virtual and a member of Obj
(dynamic_cast<Book*>(obj))->bookMethod(); //book method is only available for Book class