Ruby Cookbook, 2nd Edition

Chapter 11. Reflection and Metaprogramming

In a dynamic language like Ruby, few pieces are static. Classes can grow new methods and lose the ones they had before. Methods can be defined manually, or automatically with well-written code.

Probably the most interesting aspect of the Ruby programming philosophy is its use of reflection and metaprogramming to save the programmer from having to write repetitive code. In this chapter, we will teach you the ways and the joys of these techniques.

Reflection lets you treat classes and methods as objects. With reflection you can see which methods you can call on an object (Recipes 11.2 and 11.3). You can grab one of its methods as an object (Recipe 11.4), and call it or pass it in to another method as a code block. You can get references to the class an object implements and the modules it includes, and print out its inheritance structure (Recipe 11.1). Reflection is especially useful when you’re interactively examining an unfamiliar object or class structure.

Metaprogramming is to programming as programming is to doing a task by hand. If you need to sort a file of a hundred lines, you don’t open it up in a text editor and start shuffling the lines; you write a program to do the sort. By the same token, if you need to give a Ruby class a hundred similar methods, you shouldn’t just start writing the methods one at a time. You should write Ruby code that defines the methods for you (Recipe 11.10). Or you should make your class capable of intercepting ...