Customer Reviews

Higher-Order Perl: Transforming Programs with Programs

5 star:		(12)
4 star:		(1)
3 star:		(0)
2 star:		(1)
1 star:		(0)

 

 

As a programmer, your bookshelf is probably overflowing with books that did nothing to change the way you program...or think about programming.

You're going to need a completely different shelf for this book.

While discussing caching techniques in Chapter 3, Mark Jason Dominus points out how a large enough increase in power can change the fundamental way you think about a technology. And that's precisely what this entire book does for Perl.

It raids the deepest vaults and highest towers of Computer Science, and transforms the many arcane treasures it finds---recursion, iterators, filters, memoization, partitioning, numerical methods, higher-order functions, currying, cutsorting, grammar-based parsing, lazy evaluation, and constraint programming---into powerful and practical tools for real-world programming tasks: file system interactions, HTML processing, database access, web spidering, typesetting, mail processing, home finance, text outlining, and diagram generation.

Along the way it also scatters smaller (but equally invaluable) gems, like the elegant explanation of the difference between 'scope' and 'duration' in Chapter 3, or the careful exploration of how best to return error flags in Chapter 4. It even has practical tips for Perl evangelists.

Dominus presents even the most complex ideas in simple, comprehensible ways, but never compromises on the precision and attention to detail for which he is so widely and justly admired.

His writing is--as always--lucid, eloquent, witty, and compelling.

Aptly named, this truly is a Perl book of a higher order, and essential reading for every serious Perl programmer.

Many in the Perl community have been eagerly awaiting Higher Order Perl, and they will not be dissapointed. Not only is this a great Perl book, it's one of the best general computer science texts I've read in a long time. Dominus focuses on the functional, LISP-like aspects of Perl, breaking readers of the procedural habits they have developed writing Perl code. The book starts with a few simple examples of callbacks and closures, and quickly moves on to developing functions that dynamically manufacture and return other functions. These techniques are used to their fullest potential as Dominus shows us how to use dynamic iterators to eliminate recursion; an invaluable technique considering Perl's lack of tail call optimization. Further techniques include using iterators to transform other iterators (analagous to Perl's map function), currying, using linked lists to create "lazy" streams that produce their data upon request, and function memoization. Dominus also makes digressions into Perl internals, giving the reader a magnificent depth of understanding about how these techniques actually function under the hood.

In "Higher Order Perl" (or HOP as it's affectionately called in the Perl community), the renown Perl wizard Mark Jason Dominus (MJD) shows how to take Perl coding to the next level by applying advanced programming techniques from the domain of functional programming.

The book covers recursion (including methods to convert recursive code to iterative code), iterators, streams, memoization, currying, parsing, constraint programming and higher order functions (functions that take functions as arguments and/or return other functions). It is packed with great, sophisticated code which is explained very well and is a model for correct programming. The author takes an approach similar to Peter Norvig's PAIP - advanced coding techniques are presented, and then non-trivial programs are written to demonstrate these concepts.

The comparison with Lisp here is unavoidable, and MJD talks about Lisp in his preface. He claims that Perl shares 6 of the "7 features unique to Lisp" quoted from Norvig's PAIP, and that this basically means that most of what can be written in Lisp can be written in Perl in roughly the same manner. But as he himself admits in a later interview, the 7th "missing feature" of Lisp, namely its uniform syntax, is what *really* differentiates Lisp from the rest. Lisp's syntax allows a very clean handling of higher-order functions, list-processing, and most importantly macros. The contrast between MJD's own code in HOP and Norvig's PAIP code is the best example for this fundamental difference. Be MJD's code as clean and nice as it is (for Perl, anyway), it is nowhere near the sheer aesthetic appeal of Norvig's Lisp.

Still, Lisp is Lisp and Perl is Perl, and each has its respectable place in the world of programming. HOP is a great book to read, and I warmly recommend it to any intermediate+ Perl programmer. For people who have never programmed in Lisp or have never learned functional programming techniques, this book is a must - it will literally take your code to a higher level. For diehard fans of Lisp, this book demonstrates how to employ most of your favorite techniques in the most practical language out there (though the Perlish syntactic sugar will at times make your teeth grind).

I had originally planned to read Higher Order Perl and write my review immediately. When it became clear that I wasn't going to read HOP straight through, I figured I'd write a review when I was done. More and more, it looks like my reading of the last two chapters is being indefinitely delayed. I'm going to write what I think so far, while it's still more or less fresh in my mind.

Higher Order Perl was originally going to be given some fairly bland name, like "Handbook of Advancted Perl Techniques." This would have been a spot-on (but uninteresting) title. HOP provides the reader with explanation and demonstration of techniques for problem-solving that are often overlooked. The examples are complex and detailed, but not byzantine, and they're built up slowly, piece by piece, so that each line of code's meaning and significance are made clear.

The title "Higher Order Perl" refers to the book's most central technique, functional programming. While many programmers understand how to abstract a specific solution into a more general one, Dominus helps the reader learn to push the envelope, abstracing generic solutions into extremely generalized solutions that can be applied to seemingly-unrelated problems. This is frequently done by the construction of functions that build functions that build functions -- and so on, functions all the way down. Instead of solving the problem in base, earthly Perl, the programmer produces Perl elements of a higher order which, operating in harmony, become all things to all people.Well, I'm hyperbolizing, but I think it would be hard for me to over-emphasize the value of techniques like closures, iterators, and currying. They are, in part, what make Lisp so powerful, and the marriage of Lisp's power and Perl's expressivity is a happy one.As for the writing, it is good. The language is clear and the material is well-presented. One should be cautioned, though, that the book is dense. Dominus is constantly pressing onward, explaining new techniques or new ways to apply already-explained techniques. I found myself reading each page carefully and deliberately, only to turn back to it a few pages later, to be sure that I understood how the new material was relying on the old. It made the book a challenge to read, but it was a thoroughly enjoyable challenge. I never felt so frustrated by a bizarre idea that I gave up or so bored with an over-explained one that I skipped ahead.

Finally, while the techniques that Dominus presents are powerful and advanced, the required knowledge of Perl is not particularly great. Because he clearly explains the key Perl concepts that he uses (especially closures and associated scoping issues), any competent programmer with a working knowledge of Perl should be able to put the ideas in Higher Order Perl to work.

For serious Perl programmers, Higher Order Perl is a must-read book.

I'm mostly a Lisp hacker and don't do much with Perl, so I don't know why this book caught my eye, but I'm glad it did. It shows you how to use Perl for all kinds of powerful (and sometimes mind-blowing) tricks which are commonplace in functional languages, but which it never occurred to me to attempt in Perl. Like an earlier reviewer pointed out, this is far from the best book out there on functional programming; if that's solely what you're interested in, buy SICP. It is, however, by far the most interesting book I've ever seen on Perl. After I finish reading it, I'm going to give it away to someone who still thinks in Java.

Functional Programming is a programming paradigm that is very useful and must be part of the tool-arsenal of any software engineer and developer along with the well known Object Oriented Programming, the procedural programming approaches and the less known Aspect Oriented programming.

HOP gives an easy access to Functional Programming for Perl developers. The book is written in a personal manner and is fun to read and sometimes even funny. It is definitely not a dry factoid book.

The examples in the book are useful not only as they demonstrate nicely the need and power of functional programming but also due to their practicality for seasoned programmers.

I felt the influence of books such as "Structure and Interpretation of Computer Programs" by Abelson and the Sussmans" on HOP, and it was very nice and helpful to read MJD's recommendations for other good readings on the subject.

I got to reading HOP cover to cover, several times. I attempted to solve myself the problems raised in the book and found the process of doing that enjoying and rewarding. Reading on the commentary and approach suggested by MJD was in many ways enlightening. You can find some very amusing footnotes in the book that makes it yet more fun to read.

MJD gives a refreshing view on Perl programming, and shows that tasks that can be very complicated and hard, or computationally intense can be turned "inside out" and handled in a functional programming approach that simplifies understanding, simplifies the solution and also simplifies the implementation.

MJD presents a very useful introduction to functional programming using Perl using practical problems, and the nice thing is that even theoretical problems get practical usages when explained, which makes the reading fun, and worth the while for the non-theoretical people amongst the readers.

Though the book sometimes goes into detail or length too much for my taste, it does do a good job in giving the new view of the capabilities of a Perl programmer and the Perl programming language and how to use them.

I can testify that a lot of code that I got to see, write and maintain got a considerable "lift" and "upgrade" since me and other colleagues of mine read this book and used some of the wisdom that it offers.

There are many good and useful follow-ups to this book in the form of Perl modules on CPAN, articles and posts on the web and some book chapters. Look them up!

Also, if you feel that this functional programming approach is useful to you, you might want to probe further on this by reading Haskell, Lisp, Scheme (and other functional programming languages) literature. There are several "must read!" for these programming languages that will give you a very deep understanding and good skills in this programming paradigm. Also -- you will find, especially after reading HOP, that you now have the tools to easily and practically apply the functional programming principles that you learn in Functional Programming books to Perl and use it in your Perl code.

A few words on MDJ himself: I read a lot of what he wrote for Perl and used some of his modules and code that is available on the web. I had the pleasure of hosting him on the first YAPC in Israel and attend a 3 day course that he gave. He is a very smart person, a very nice person and an excellent teacher and a very good writer. If you don't already know MJD, this book is a very good way of getting access to some of the wisdom that this gut has to offer.

If you don't know which Perl books to buy and/or to read, know this: HOP is one of the books that you should read!

(It is funny that almost 3 years after I read this book for the first time and after reading some of the material before editing, as MJD offered in his private mailing list, I only now get to write a review on it. But still, it surely deserves a book review and a good mark too.)

In a book market awash with teach-yourself-in-x-time-units, cookbooks, and API tutorials, Higher-Order Perl is ambitious. It attempts to bring functional techniques to Perl, leaning heavily on two language features: closures, and functions as first-class entities. To enjoy this book you need to be completely comfortable with Perl's syntax, there's lots of code to read and digest, and if you can't follow it, you're going to get lost very quickly, because there are lots of new concepts going to be thrown your way. There's no outlandish symbol table hacking, almost no object orientation, not even much use of modules, but it is in a very different style to most of the Perl you're used to.

If, like me, most of the programming languages you're familiar with are more closely related to C than Lisp, this is going to be a challenging read. It's going to take you a long time to get through this book, and it'll probably require re-reading (more than once) to fully get it all. This is a book bursting at the seams with ideas, beginning with recursion, and then onto caching, iterators, streams and currying. The last two chapters show how to apply these techniques to parsing and declarative programming, and they feel like mini books in themselves. Applications of the code range from classics like Tower of Hanoi, Fibonacci sequences and the Newton Raphson method, to more practical material including databases, tied files, and directory walking.

Part of the reason I enjoyed this book so much may be because I have no formal training in Computer Science, and of course, the likes of SICP and The Little Schemer contain similar concepts. But for someone without functional programming experience it's great to have it in Perl. Lispniks, MLers and Haskellites may well be familiar with these ideas already, but they should consider this an enticing 'gateway book' for Perl programmers. And Introduction to Functional Programming using Haskell just isn't as well-written as HOP. There's no hiding from the fact that if you want to fully appreciate the material here, you're going to have to work hard at it, but Mark Jason Dominus does a great job at stopping things from feeling like a textbook, and knows how to inject some wit into the proceedings without being distracting. Full marks to Morgan Kaufmann for the layout and overall production quality, too.

I keep The Camel, Perl Cookbook and Perl Best Practices on my desk when programming Perl. HOP is not going to be joining them (yet), it's not that sort of a book. Instead it's a very different intellectual pleasure and easily one of my favourite and most important Perl books (only PBP edges it out due to the latter's everyday practicality). If you're serious about Perl, you need to read this book. If you're serious about programming, ditto. Hell, this might just be a good enough reason to learn Perl if you don't already (admittedly, the chances of such a person reading this far are small).

HOP has set the bar very high. Python and Ruby authors, please step up!

I am NOT a SW guy - I am a perl autodidact HW engineer. I have written some pretty BIG "scripts" that get the job done - but a fringe "benefit" to me from the demise of the previous company I worked for is that I DON'T HAVE to maintain those scripts - a big bowl of spaghetti :(
I tried several times to learn the OO side of Perl but this stuff was w-a-a-a-a-y out there for my scull :(

HOP just made it for me! I wish it was available 5 years ago so I wouldn't waste my time writing the aforementioned spaghetti! Study it! It will twist your brains to the point of pain but stick with it - well worthed - you won't look at scripting/programming the same old way ever again!

vess

Non-Lisp programmers, I think, fundamentally misunderstand the appeal of Lisp, and why it is superior to other programming languages. Lisp is not superior because it offers higher order functions or any of the other things, it is superior because it does not get in the way: Lisp meets problems, rather than forcing problems to meet it.

This book serves two purposes: on the one hand it helps Perl programmers to move the center of gravity of their solutions further away from Perl, and a bit closer to their problems (quite a bit closer, actually), and on the other it provides a good introduction to Perl for Lisp programmers, and for that it gets an effortless five stars.

I suspect that it is maybe more successful for the second purpose than the first: a relaxed read for someone who knows what a closure is, but for someone who does not know what a closure is, most likely not a relaxed read. Further, anyone who masters the contents may find that his newly refined coding style is somewhat caviary to the general, something that would be a problem in any IT/ORG that I have experienced.

Small quibbles: A Pratt Precedence parser might have fitted the discussion better than the recursive descent parser that is developed, and a deeper discussion of the ways that Perl nevertheless falls short of Lisp, instead of just a passing reference to Norvig's list, would actually have made the arguement stronger - for instance you cannot add a macro facility similar to that of Lisp to Perl, and you cannot add the general continuations available in Scheme (or the limited continuations that Graham, e.g., shows how to add to Common Lisp).

In short, Higher Order Perl makes a good case for the the aesthetics of haiku, but one should understand that (in spite of what my cousin, Leontia Flynn, thinks) it is haiku written with a typewriter, not with a caligrapher's brush.

P.S. One minor repeated irritation for me was the way the author, in spite of everything, insisted on emphasising how input functions should read input files one line at a time. This is a bizarre mindset holdover from the seventies, that is inconsistent with the general argument of the book: if you have a computer with hundreds of megabytes, or gigabytes, of core, then the efficient way to read a text file is not line by line, but in one go.

This book is excellent. Unless you're a programming genius as well as a Perl genius, this book will teach a a lot of cool Perl tricks, and a lot of cool programming techniques which can be applied to all sorts of thorny problems and even implemented in other languages (although, some of the techniques might be a real pain to write in one of the "real" programming languages i.e. C/C++/C#/Java).
I am much more of a Python programmer and I tend to avoid Perl if I can, but reading this book makes me want to write Perl code again.
OK, maybe that feeling doesn't last for long, but it reminds why Python (and Ruby!) owe so much to Perl and the Perl community.