[Contents] [Previous] [Next]

Objects

Objects are software entities made of data (nouns) and definitions of actions that can be performed on the data (verbs). Together, the data and actions associated with objects allow them to simulate the characteristics and behavior of almost any real-world entity, including concrete items, such as computers, people, and physical places, and abstract concepts, such as geometric shapes and data structures.

Data abstraction

Much like the cells that make up living organisms, to which they are often compared, objects hide their internal details from other objects. Just as cells communicate with each other via chemical messages that the cell membrane recognizes and allows to pass through to the inside of the cell, objects communicate with each other by means of messages (evoked by calling member functions associated with each object). If one object needs to trigger some particular action on the data within another object, it simply sends that object the appropriate message. The message's sender doesn't need to know anything about the recipient or its data other than the names of the member functions to which it responds.

This capacity for data abstraction--the organization of data into small, independent objects that can communicate with each other--helps software developers break complex programming problems into many smaller, simpler problems. Data abstraction is one of the primary benefits of OOP. By learning how to use the member functions that control an object, a programmer can concentrate on what the object does rather than on how it does it.

Encapsulation

The enforcement of the data abstraction provided by objects is called encapsulation. Just as a cell membrane protects the cell's interior from all external chemical signals except those the membrane recognizes, encapsulation prevents activities in the rest of the program from accidentally damaging an object but allows other objects to interact with that object by calling its member functions.

Figure 7 shows how encapsulation works. The left side of the figure shows the appearance of a hypothetical window object of type TWindow, and the right side shows the implementation of that object. The implementation shows some of the members that the TWindow object provides. These include four data members that hold the coordinates for the left side, right side, top, and bottom of the window and several member functions that operate on the window. The member functions define the actions that a TWindow object can respond to. The Open function opens the window, the Close function closes it when the user clicks the box in the upper-left corner, and the Resize function adjusts the size of the window when, for example, the user drags its lower-right corner.


NOTE Although the TWindow objects in Figure 7 and later figures reflect the Taligent convention of beginning with "T" for "Type," they present a highly simplified view of window programming and are not real CommonPoint objects. In reality, the details of window implementation depend on the development system, the role of the window, and so on.

The Resize member function in Figure 7 has arguments that specify the window's coordinates. The only way to change the actual coordinates from outside the object is to call the Resize member function; the program can't do anything to the data members directly. The data in any object can be changed only by that object's member functions.

Encapsulation protects an object's data from accidental or inappropriate changes, which in this case could mean that the values of the coordinates might not match the way the window actually looks, and text or other data inside the window might run over its edges. Because there is only one way to change a window's coordinates--by using the Resize member function--the drawing of the window and its location on screen are guaranteed to match. Consistency
is ensured.

TWindow includes several other member functions. The Zoom function is triggered when a user clicks the box in the window's upper-right corner to enlarge the window so it fills up the whole screen or clicks the box again to shrink it to its original size. The Zoom function in turn calls the Resize function, specifying coordinates that are inset slightly from the sides of the screen and keeping track of the original coordinates so it can restore them when the user clicks the Zoom box again.

Three member functions take care of actually drawing the window shown in Figure 7. The DrawSelf member function can be called to redraw the entire window and its contents. The DrawSelf function in turn calls DrawFrame and DrawContents. DrawFrame draws the window's frame and controls, and DrawContents draws its contents, adjusting automatically for the window's current size.


[Contents] [Previous] [Next]

Addison-Wesley Publishing Company, Copyright©1995 by Sean Cotter and Taligent,Inc. All rights reserved.