Catálogo de publicaciones - libros

As I mentioned in the book’s introduction, Ruby is my language of choice. It is the tool I instinctively reach for first when solving system administration problems. To recap, it has principally achieved this place in my affections by • Making it simple to write nicely readable code • Emphasizing convention over configuration, so a little effort goes a long way • Offering seamless mechanisms for interfacing with C libraries • Having syntax, extension, and execution conventions that very often make coding the “right” way and the “quick” way the same thing • Adopting object-oriented principles so completely that extremely powerful metaprogramming techniques become readily available

Much of this book is devoted to writing “proper” code in nice little scripts and demonstrating how much easier Ruby makes this process. Even so, there will always be times when you just want to remove the commas from something or quickly rotate a log. Thus this relatively short chapter sits in deference to the needs of the ten-second script.

This chapter covers both the analysis and the optimization of script performance. We will warm up with some scenarios in which a good script can outpace a traditional pipeline of C applications. With this musing undertaken, we will proceed to tour the available analysis tools and apply them to a sample script. Finally, we will look at some algorithmic and linguistic optimization strategies.

So far, you’ll have acquired an impression of the basics of Ruby—its flavor, if you will. Forme, what really sets it apart is the way in which it facilitates metaprogramming . Of course, this is a somewhat nebulous statement if one hasn’t properly defined what class of programming is worthy of such a prefix (see the sidebar “Meta ⋯ What?”).

One of the most brilliant design choices found in UNIX-style operating systems is the notion that “everything is just a file.” In other words, if your programming language can do very basic reading/writing of files and a little string handling, you can drive the configuration and environment of every bit of software you care about.

Data, data, everywhere, but not a byte to eat. How we store, organize, and access data can determine the future success of software systems and thus the organizations that rely on them so heavily. Choosing poorly in these respects can doom us to a long, agonizing information death as complexity, inefficiency, and competing constraints spiral out of control.

Ruby’s value in a corporate environment is not to be underestimated. In this chapter, we’ll cover how to perform some of the most common parsing tasks found in daily enterprise life, manipulating delimited values and XML documents with ease. The second half of the chapter is devoted to the vital requirement of being able to interact with business information systems over LDAP, XML-RPC, and SOAP, and using the increasingly popular REST principals.

Making systems talk to one another is a fundamental task for system administrators. Often, we are called upon to provide abstraction, analysis, monitoring, and even translation of network protocols and services. Thus a grasp of how Ruby augments our network toolkit is vitally important.

Networks are full of metadata that is useful to techies and management alike. So often this data sits and rots because nobody has the time or the inclination to bring it together and use it effectively.

Perl has CPAN (www.cpan.org), JavaScript has JSAN (www.openjsan.org), and Python has a darkened crypt (http://py.vaults.ca). Ruby also has a unified packaging system for supplementing the evailable cloud of libraries on a system: RubyGems . Each individual bundle of functionality under this scheme is called a gem and contains, as you will see, all of the code and metainformation pertinent to given library.