Recently, I’ve been trying to pick up Perl again. I learned it back around version 4.0, before PHP even existed. After a few years of disuse, you lose the ability to communicate in a language – real or artificial. (Nihongo ga wakarimasen – I can’t speak Japanese [anymore].)

Of course, with a language that is actually maturing, like Perl and PHP, there’s new stuff to learn that’s happened since you’ve been away. Combine that with the experience you gain – I’m just starting to venture into the brave new world of object-oriented design, for example – and catching up can be hard.

Since I’m fairly fluent in both PHP and Perl, I thought I’d share some of my insights for anybody who’s thinking of getting started in web programming, and feeling confused as to where to start.

Many people don’t consider this, but programming languages tend to be stronger in some tasks than others. Because each language is designed differently, syntax, functionality, and style tend to form around an expected task. Take PHP and Perl, for example. It is widely believed that PHP is the best language for building web pages, and Perl is best for text processing. I think this is mostly correct.

PHP was designed from the ground up to be an Internet language – in fact, the official meaning of the acronym is “PHP Hypertext Preprocessor.” The language is unusual in that large blocks of text can be passed through the interpreter without being executed. This could potentially eliminate a lot of extra print functions the Perl folks have to type.

The problem with this that many programmers cite is that this ease of use discourages the conscientious separation of presentation and business logic that makes a huge difference in big projects. Solutions have been developed to address this, including a semi-official template library called Smarty, but all of these solutions require the support of the programmer, who may be unable or unwilling to cooperate.

Perl is recognized as one of the most sophisticated text-processing programming languages ever invented. Its regular expression support is unparalleled. In fact, Perl is one of the very few languages were regular expressions are an operator (=~) rather than a function (ereg and 13 (!) others in PHP). Because this is what the language is designed for, it’s heavily optimized.

Perl also excels at being a general-purpose language. CPAN, a repository of Perl modules, is one of the largest language-specific code repositories in the world. PHP’s PEAR is laughable in comparison. As of the date this was blogged, the catalog consisted of over 5,000 modules comprising 1,936 MB of space, served on 227 mirrors scattered around the world. If you can imagine it, somebody’s likely wrote it already. It’s the core of many unusual systems, including chat rooms, virtual communities, and even home automation.

Perl has downsides, though. One of the main problems is its very odd syntax. With most C-style languages, for example, a function includes the parameters in the declaration. In Perl, you have to use a special array – @_. (I don’t know why they chose that – maybe because it’s short.)

Even common concepts like global variables have strange implementations. For example, Perl has the concept of packages, which can hold different variables with the same name. For example, I could have $main::test and $blah::test, and the value of $test would be different depending on what package I was using. Also, the local function is deprecated in favor of my, because local was implemented differently at an earlier point.

So, what would I choose? It’s likely going to be both for the foreseeable future. I think for web programming, I’m going to stick with PHP, because it’s very easy to write and easy to find support for. Perl is a very powerful language, and I’m starting to write a number of maintenance scripts with it. It’s very different, but I would feel confident in saying that it’s probably the most powerful language in use today, other than C.