What is information hiding in c?

Answer 1

Information hiding is employed by object designers to make certain aspects of an object inaccessible to end-users. This is achieved via the access specifiers: public, protected and private.

Private members are accessible to the class itself (including other objects of the same class type) as well as friends of the class type. Protected members are similar to private members but are also accessible to classes that are derived from the class. Public members are fully accessible.

Typically, an object's member variables will be private to the class of object, with public accessors (getters) and mutators (setters) which act as gatekeepers, ensuring the member variables (and therefore the object itself) remain in a valid state at all times. Since objects of the same class type must behave in exactly the same way (as defined by their class), they have privileged access to each other's private members. Friends of the class have similar privileges which is often said to undermine encapsulation. However this is nonsense. Friends extend a class interface, nothing more. You -- the class designer -- are ultimately responsible for ensuring friends of the class do not undermine the encapsulation.

Methods that operate upon the member variables will generally be made public, however any methods that are used internally, solely by the class itself, should remain private to the class.

Protected access is intended for those classes that are intentionally designed to be used as base classes. However, protected access should be limited only to those members that a derived class actually requires access to. While it may be tempting to replace all private accessors to protected, this reduces the effectiveness of the data hiding mechanism. A better solution is to provide protected methods that only derived classes can call.

Remember that all class members are private by default. It is up to you to decide what is made available outside of the class, but as a general rule, less is more. The less you expose, the easier it will be for end-users to use your class.

As an example, imagine a complex class that represents a complex piece of machinery that normally requires the user to complete a check list before invoking a series of start up procedures. In the real world that may well be the case, but in class design, all of that responsibility can be fully encapsulated by the class itself, as private members and methods. The only public member is the start() method, so all the user has to do is hit the start button and the class itself decides if it is good to go. If so, the machine starts, otherwise an error message is displayed. Everything else is completely hidden from the user.