Why can’t a derived class call protected member function in this code?

by

Even though access control in C++ works on per-class basis (as opposed to per-instance basis), protected access specifier has some peculiarities.

The language specification wants to ensure that you are accessing a protected member of some base subobject that belongs to the derived class. You are not supposed to be able access protected members of some unrelated independent objects of base type. In particular, you cannot access protected members of freestanding objects of base type. You are only allowed to access protected members of base objects that are embedded into derived objects as base subobjects.

For this reason, you have to access protected members through pointer->member syntax, reference.member or object.member syntax, where the pointer/reference/object refers to the derived class.

This means that in your example, protected member somethingProtected() is not accessible through Base objects, Base * pointers or Base & references, but it is accessible through Derived objects, Derived * pointers and Derived & references. Your plain somethingProtected() access is allowed, since it is just a shorthand for this->somethingProtected() where this is of type Derived *.

b.somethingProtected() violates the above requirements.

Note that in accordance with the above rules in

void Derived::somethingDerived()
{
    Base *b = this;
    b->somethingProtected();    // ERROR
    this->somethingProtected(); // OK
}

the first call will also fail while the second one will compile, even though both are trying to access the same entity.

More Related Contents:

  1. Which access specifiers will not affect a friend function?
  2. Friend function strange behaviour
  3. Why should I prefer to use member initialization lists?
  4. Function pointer to member function
  5. When should you use a class vs a struct in C++?
  6. Thou shalt not inherit from std::vector
  7. Why should the “PIMPL” idiom be used? [duplicate]
  8. When should you use ‘friend’ in C++?
  9. What is the difference between private and protected members of C++ classes?
  10. Why are unnamed namespaces used and what are their benefits?
  11. What are the differences between struct and class in C++?
  12. Why should I avoid multiple inheritance in C++?
  13. When should I use C++ private inheritance?
  14. What are Mixins (as a concept)
  15. How do you create a static class in C++?
  16. Why should I use the “using” keyword to access my base class method?
  17. Declaring multiple object pointers on one line causes compiler error
  18. Implementing multiple interfaces in c++
  19. map set/get requests into C++ class/structure changes
  20. Will an ’empty’ constructor or destructor do the same thing as the generated one?
  21. Diamond inheritance (C++)
  22. overloading base class method in derived class
  23. Best method for storing this pointer for use in WndProc
  24. Should I use public or private variables?
  25. How to initialize the reference member variable of a class?
  26. Object oriented programming in Haskell
  27. How to create class objects dynamically?
  28. What is the purpose of forward declaration?
  29. Why should I prefer to use member initializer lists?
  30. What does slicing mean in C++?

Leave a Comment