start page | rating of books | rating of authors | reviews | copyrights
Programming PHPProgramming PHPSearch this book
PreviousNext

Chapter 6. Objects

Contents:

Terminology
Creating an Object
Accessing Properties and Methods
Declaring a Class
Introspection
Serialization

Object-oriented programming (OOP) opens the door to cleaner designs, easier maintenance, and greater code reuse. Such is the proven value of OOP that few today would dare to introduce a language that wasn't object-oriented. PHP supports many useful features of OOP, and this chapter shows you how to use them.

OOP acknowledges the fundamental connection between data and the code that works on that data, and it lets you design and implement programs around that connection. For example, a bulletin-board system usually keeps track of many users. In a procedural programming language, each user would be a data structure, and there would probably be a set of functions that work with users' data structures (create the new users, get their information, etc.). In an object-oriented programming language, each user would be an object—a data structure with attached code. The data and the code are still there, but they're treated as an inseparable unit.

In this hypothetical bulletin-board design, objects can represent not just users, but also messages and threads. A user object has a username and password for that user, and code to identify all the messages by that author. A message object knows which thread it belongs to and has code to post a new message, reply to an existing message, and display messages. A thread object is a collection of message objects, and it has code to display a thread index. This is only one way of dividing the necessary functionality into objects, though. For instance, in an alternate design, the code to post a new message lives in the user object, not the message object. Designing object-oriented systems is a complex topic, and many books have been written on it. The good news is that however you design your system, you can implement it in PHP.

The object as union of code and data is the modular unit for application development and code reuse. This chapter shows you how to define, create, and use objects in PHP. It covers basic OO concepts as well as advanced topics such as introspection and serialization.

6.1. Terminology

Every object-oriented language seems to have a different set of terminology for the same old concepts. This section describes the terms that PHP uses, but be warned that in other languages these terms may have different meanings.

Let's return to the example of the users of a bulletin board. You need to keep track of the same information for each user, and the same functions can be called on each user's data structure. When you design the program, you decide the fields for each user and come up with the functions. In OOP terms, you're designing the user class. A class is a template for building objects.

An object is an instance of a class. In this case, it's an actual user data structure with attached code. Objects and classes are a bit like values and data types. There's only one integer data type, but there are many possible integers. Similarly, your program defines only one user class but can create many different (or identical) users from it.

The data associated with an object are called its properties . The functions associated with an object are called its methods . When you define a class, you define the names of its properties and give the code for its methods.

Debugging and maintenance of programs is much easier if you use encapsulation. This is the idea that a class provides certain methods (the interface) to the code that uses its objects, so the outside code does not directly access the data structures of those objects. Debugging is thus easier because you know where to look for bugs—the only code that changes an object's data structures is in the class—and maintenance is easier because you can swap out implementations of a class without changing the code that uses the class, as long as you maintain the same interface.

Any nontrivial object-oriented design probably involves inheritance. This is a way of defining a new class by saying that it's like an existing class, but with certain new or changed properties and methods. The old class is called the superclass (or base class), and the new class is called the subclass (or derived class). Inheritance is a form of code reuse—the base-class code is reused instead of being copied and pasted into the new class. Any improvements or modifications to the base class are automatically passed on to the derived class.

PreviousHomeNext
5.10. Using ArraysBook Index6.2. Creating an Object
Library Navigation Links

Copyright © 2003 O'Reilly & Associates. All rights reserved.