Encapsulating Functionality


Computer science authors and professors often discuss the concepts of encapsulation, abstraction, and information hiding. These concepts are fundamental to programming, and they are required in order to manage complex systems. By encapsulating information into separate components, programmers can reason about those subsystems alone. Instead of understanding and mentally grasping the whole system at once, coders can partition systems into coherent sections. Unfortunately, encapsulation, abstraction, and information hiding are usually discussed with respect to objects and object-oriented programming (as you can see in the links above). But encapsulation is equally important in the lower-level tools of functions and methods.

Fundamentally, functions are designed to encapsulate related functionality. Because the functionality and information are abstracted away in a function, the programmer can avoid thinking about those lower-level details when they are reasoning about the calling code. If the function is well designed, coders can also modify it more easily because they will not need to think about any external code when doing so.

Here is an oversimplified example of a function with poor encapsulation:

// There are three configuration files:
// custom.config.xml, default.config.xml, and core.config.xml
public List<File> getConfigFiles() {
    ArrayList<File> configFiles = new ArrayList<File>();
    configFiles.add(new File("default.config.xml"));
    configFiles.add(new File("core.config.xml"));
    return Collections.unmodifiableList(configFiles);
}

Because the getConfigFiles() method requires the developer to know which of the three configuration files it returns, the method has poor encapsulation. When the developer wants to use this method, they must think about the internals. Then they must obtain the custom.config.xml separately. Therefore the method does not properly encapsulate the functionality of getting the configuration files.

In order to fix this, the method must return all three configuration files:

// There are three configuration files:
// custom.config.xml, default.config.xml, and core.config.xml
public List<File> getConfigFiles() {
    ArrayList<File> configFiles = new ArrayList<File>();
    configFiles.add(new File("custom.config.xml"));
    configFiles.add(new File("default.config.xml"));
    configFiles.add(new File("core.config.xml"));
    return Collections.unmodifiableList(configFiles);
}

Now the details are hidden or abstracted within the method. When the developer calls getConfigFiles(), he or she does not need to know each configuration file that the method obtains by name. The reverse is also true. The developer only needs to consider the configuration filenames when modifying the getConfigFiles() method. By ensuring that getConfigFiles() truly encapsulates its named functionality, the developer has made the whole program easier to understand.

For further reading on constructing excellent functions, check out Chapter 7 of Steve McConnell’s excellent book, Code Complete—“High-Quality Routines” (pp. 161-186).

Comments

Related Posts

Step Through Your Code

Java Testing Tip #2:

Java Testing Tip #1:

Programming Terms Cheat Sheet

Technical Writing