Learning Python is a well-written book by an experienced Python trainer that has served the Python community well since the first edition was published, back in 1999. Now at its fourth edition, this book by Mark Lutz arguably continues to be Python's bible. This article reviews the fourth edition by discussing the target audience, scope, content, and pedagogical features. It concludes by a recommendation.

Audience

This book is intended as an introductory text to programmers new to Python.  Although people with no programming experience are not discouraged from reading it, they are warned that time is mostly spent teaching Python, not programming fundamentals. I agree with this picture, though from my own experience as well as those of others, the book is equally valuable to more experienced Python programmers both as a pseudo-reference, and as an introduction to more advanced topics. The critical point here is that the book does not make assumptions about educational or vocational experiences and provides many examples, which renders the book approachable by a large audience.

Scope

Both Python 2.6 and 3.x are covered in this edition. However, the latest 3.x line is considered the reference from which variations in 2.6 are discussed when appropriate. This approach is logical; the new Python 3.x presents a major change to the language, but is not sufficiently dominant to warrant exclusive treatment.

This book discusses the Python language and excludes the Python standard and non-standard libraries. The latter are discussed in other places, including Lutz's own Programming Python, which stands at its third edition at the time of writing of this article. I find this separation necessary because of size considerations and, in fact, this division did not exist in the first edition of the book! However, one topic does not seem to fit in the language/libraries division, and that is packaging and deployment.

I will argue that there aren't many (if any) books that discuss packaging and deployment of Python programs. I will also argue that this topic should be included in the book being reviewed here since it is essential to real Python programming. Since Lutz discusses the Python run-time environment, I do not think it would detract from the book's coherence to include a single chapter on packaging.  It is possible that the proliferation of various packaging and deployment options such as distutils, setuptools, pip, buildout, virtualenv, paver, fabric and others, is the reason for this exclusion. Or it could be that these tools are in a state of major flux that any text will be quickly outdated.  If size is the reason for this exclusion, maybe Lutz or someone else can publish a "Packaging and Deploying Python" as a separate volume.

Content

The book starts by making a case for the use of Python. Both the features of the language and its prominent users are discussed to build credibility. Then, the run-time environment is discussed: how to run programs in various ways on various operating systems and language interpreters.

Types and statements, which are at the core of any language, are discussed next. Notably, there is an excellent discussion on the topic of iterators and generators (both are also covered in a later chapter).  Functions, modules and classes are then introduced. The text also includes a discussion of general object-oriented programming (OOP) principles which I find to be invaluable, as it brings the topic of classes to life.

Exceptions are introduced and discussed in detail. The placement here is appropriate since exceptions are now objects in Python so classes had to be discussed first. This chapter should prove to be especially useful for people migrating from other languages that do not have simple, yet effective, exception-handling constructs.

Finally, four advanced topics are covered: decorators, unicode, managed attributes, and meta-classes. I find the first two to be absolutely necessary for almost any system nowadays, even the smaller ones! The latter two are not as ubiquitous, but should be useful to more experienced programmers.

I should mention here that the discussion of the topics mentioned above doesn't stop at the basics, but provides comprehensive coverage. This is also the case with the discussion of topics such as dynamic typing, inheritance order, iterators, generators, comprehensions, and functional programming, among many others. There is even an interlude on documentation and the pydoc library.

Pedagogical Features

Like many programming texts, Learning Python uses small programming examples (appropriately executed in the Python interactive shell). These small examples hope to capture the essence of the topic at hand, and it does that well within the limitations of a small-scale context. This fourth edition adds a new chapter on classes (Chapter 27) that contains a more realistic code example presented in a tutorial format.

In addition to examples, each chapter ends with a summary of the chapter's content as well as a quiz. The quiz is immediately followed by its answers for easy reference. I have to admit that I do not use any of these two features, so I will not be able to comment on their efficacy.

Conclusion

Like many O'Reilly books, this is a well-written, coherent, and beautifully type-set book. I highly recommend it to anyone who wants to, or already does, program using python. It should help the novice in their transition to an excellent programming language, or otherwise, make an already familiar environment more powerful in the hands of veterans.

About the reviewer: Ahmed Al-Saadi is a Software Analyst who works for a Montreal Python house. He wrote his first lines of code on a Sinclair ZX Spectrum+, though unfortunately not in Python at the time.