250: Highlights

• Class hierarchies reflect subclass–superclass relations among classes.
• You have several reasons to arrange classes in hierarchies:
  • To parallel natural categories
  • To prevent avoidable duplication and to simplify maintenance
  • To avoid introducing bugs into previously debugged code
  • To use purchased code
• An object inherits member variables and member functions not only from the class to which it belongs, but also from all that class's superclasses.
• When a subclass–superclass chain contains multiple functions with the same name, the one closest to the object argument in the subclass–superclass chain is the one executed. All others are shadowed.
• When a subclass–superclass relation is direct, with no intervening classes, the subclass is called the derived class and the superclass is called the base class. The derived class is said to be derived from the base class.
• There are several types of derivations; so far, you know about only public derivations.
• If you want to create a class hierarchy, then draw a diagram that reflects natural categories, and populate the classes in that class hierarchy with member variables and member functions such that there is no needless duplication of a member variable or member function, and such that each member variable and member function is useful in every subclass.
• If you want to create a public derived-class–base-class relation, then augment the derived-class definition by instantiating the following pattern:

  class derived-class name : public base-class name { ... }