Metaprogramming lets you program more expressively. This makes your code easier to write and easier to maintain and extend. Learn both the hows and whys of metaprogramming Ruby from Dave Thomas, one of the most experienced Ruby programmers in the western world.

Buy Now

  • Download and watch DRM-free episodes when and where it's convenient for you
  • You get Theora Ogg, iPod/iPhone 3, iPad/iPhone 4, and QuickTime formats.

All the episodes in this series have been released.

 

About this Screencast

Initially, metaprogramming Ruby can seem really difficult. There are all these unfamiliar terms: singletons, self, metaclasses, and so on. And things never seem to work quite right the first time—Hmm… maybe it’ll work if I change class_eval to instance_eval...

Well, the good news is that it really isn’t that complex. You just have to understand the underlying model, and everything falls into place. Dave Thomas has been digging into that model for the last ten years, and has a fresh (and simple) way of looking at it. Now you’ll be able to see why metaprogramming works, and how to do it for yourself.

It doesn’t matter if you’ve been programming Ruby for a month or for five years. We bet that you’ll learn plenty new in each of these episodes, as Dave digs into both the how and the why of the Ruby object model and metaprogramming.

  • Dave talks about the series in this free preview.
    (3 mins, QuickTime | Ogg)
  • Watch some free extracts from Episodes 2 and 3.
    (9 mins, QuickTime | Ogg)

Contents and Extracts

  • Episode 1: Objects and Classes (29 minutes)

    Just what is a Ruby object, and how can Ruby move you away from class-oriented development and back to object-oriented development? Learn about:

    • the internals of objects,
    • how classes really aren’t anything special,
    • what self does, and the two ways it can get changed
    • the method call mantra: “one to the right, then up,”
    • singleton methods and ghost classes,
    • why class methods don’t exist,
    • how classes get their names,
    • how the concept of the current class interacts with def
  • Episode 2: Sharing Behavior (39 minutes)

    One of the primary goals of good design is to put the right behavior in the right place without duplication. We’ll see three ways Ruby excels at this:

    • Using prototype-based programming, where you can create hierarchies of objects with controlled sharing of both state and behavior, and without a class definition in sight.
    • using inheritance (but you can only watch this section after you’ve watched the included Public Safety announcement on why inheritance is normally a bad idea). We’ll also look at what happens when you do class << object (and why it’s nothing whatsoever to do with inheritance).
    • using modules and mixins, the sweet spot of Ruby programming. How to using include and extend, and how to create modules that mix in both instance and class methods.

    Episodes 1 and 2 are the foundation for the rest of the screencast.

  • Episode 3: Dynamic Code (31 minutes)

    Metaprogramming is sometimes defined as “writing code that writes code.” Here we’ll see just how to do that in Ruby.

    • Blocks and the two kinds of block-objects
    • How bindings capture execution context
    • How blocks can act as closures, and why that’s vital to metaprogramming
    • Writing methods that define other methods using nested defs
    • Using define_method
    • Writing your own accessor methods
  • Episode 4: instance_eval and class_eval (29 minutes)

    Two of the workhorse methods of metaprogramming are instance_eval and class_eval. They allow you to execute chunks of code dynamically. Here we’ll see how to use them.

    • How instance_eval and class_eval differ
    • When to use one versus the other
    • Breaking down barriers
    • Creating methods on the fly, but without using closures
    • Defining stuff in classes given a class object
    • Writing DSLs in a block structure
  • Episode 5: Nine Examples of Metaprogramming (36 minutes)

    Here’s where all the theory and practice comes together! We’ll take a simple problem and find nine different ways of attacking it using metaprogramming techniques.

    • Adding behavior directly inside the class
    • Using subclassing to add behavior
    • Subclassing with a generator using Class.new
    • Using a ghost class
    • Ghost class with a generator using class_eval
    • Rewrite the method with alias_method
    • Rewrite using a module
    • Rewrite using bind
    • Writing a DSL in a block
  • Episode 6: Some Hook Methods (35 minutes)

    Ruby hook methods are a way for your application to hook itself into the execution of the Ruby interpreter. Using hook methods is crucial for some kinds of metaprogramming, and they can make your code more flexible. In this episode, we’ll see how to use two powerful hook methods: inherited and const_missing.

    • Overriding hook methods to intercept and deal with certain Ruby interpreter events during the lifetime of your application
    • Decoupling code using inherited to keep track of subclasses
    • Implementing enumerated types with const_missing
    • Using const_missing to autoload classes based on the names of files
    • Applying const_missing in both global and localized situations
    • How to chain an overridden hook method to its original hook method
    • Using const_set to define constants
    • Two practical uses for hook methods
  • Episode 7: More Hook Methods (53 minutes)

    We’ll pick up where we left off in the last episode by looking at two more Ruby hook methods: included and method_added. But we’ll also take it a step further. We’ll use these hook methods to develop a metaprogramming library that traces the execution of a Ruby program. Along the way we’ll see all the various subtle (and important!) things you need to think about when you’re trying to write a general-purpose metaprogramming library.

    • Using included to intercept when a module is included in a class, and use it to set up another hook method in a different context
    • Using method_added to track when a new method is defined on a class, and trace the method’s execution
    • Refactoring the tracing to support blocks
    • Using method objects to bypass naming issues
    • Adding tracing to methods that have already been added
    • Suppressing tracing for certain methods
    • Using Thread.current to define thread-local variables
    • Differences between Ruby 1.8 and 1.9
    • Practical examples (and corner cases) of metaprogramming

Audience

These screencasts assume you have some basic Ruby knowledge.

About the Author

Dave Thomas has been using Ruby since 1999, and has been teaching it around the world since 2002. He’s the author of Programming Ruby, Agile Web Development with Rails, The Pragmatic Programmer, and four other titles. He was one of the authors of The Agile Manifesto. But mostly he’s a Ruby programmer who still loves the language, and loves telling other people how cool it is.

Upcoming Author Events

Comments and Reviews

  • Holy cats those first three episodes are fantastic! Your quality efforts really show—these screencasts are vastly superior to any of the others in my growing collection (both from a content and fit-and-polish standpoint). While I kinda sorta almost understood metaprogramming in Ruby before, these presentations really helped to hammer it home.

    —Rex Walters
  • Joon You said:

    I’m extremely impressed with the quality of the video as well as the contents. All I can say is that you should try them out for yourself, especially at $5 per episode, there’s really nothing to think about.

  • Antonio Cangiano said:

    Five stars, and I’m certainly looking forward to future episodes… the $5 price was a steal.

  • These are really awesome. I particularly liked the exercises you included in the third screencast; they really helped me to solidify some of the concepts.

    —Zach Langley