One of the best-known and most common structures that produce maintainable programs is called Model/View/Controller (MVC). There are plenty of other good methodologies, but we’ll use MVC in this chapter because it’s particularly well-suited to database applications. It’s a very typical three-tier methodology.
MVC splits an application into the Model, the View, and the Controller. Each layer is an independent unit that performs a specific function. If you design the program correctly, you can—at least in theory—completely strip out one layer and replace it with something different without having to change the code in the other two layers.
The body of the application. Here, all objects that represent real-world “things” within the application are modeled. For example, a book, publisher, and author are represented in code on this layer. The Model is also responsible for making these objects persistent; that is, storing them so they can be retrieved after the program finishes. Therefore, all database interaction is performed by the Model.
The eyes, ears, and hands of the application. The View is responsible for presenting information to the user and collecting any user feedback. In a traditional desktop application, the View draws the screens and reads input from the keyboard and mouse. In a web-based application, the View generates the HTML viewed by the user’s browser and passes down any form data submitted by the user. All I/O that ...