Summary

Dynamic languages provide their own benefits when working with XML. Their ability to rapidly prototype and iteratively develop a solution make writing routines for reading or writing XML faster than their static cousins like Java or C#. When your compilation cycle is reduced to the time it takes to save the file you are working on, you can be more productive with your code. In addition, as each of these languages has been available for quite some time, many libraries have been created for working with common XML formats. Many even provide a variety of techniques for working with XML: from tree-based syntaxes that work like the W3C DOM, through stream-based interfaces like SAX, all the way to object-based syntaxes that make XML work like the native objects of the language.

Perl, through its powerful implementation of regular expressions, is almost synonymous with text processing. As XML is simply text, it should come as no surprise that Perl works quite well with XML. Whether it is through a low level API, such as XML::Simple, more capable APIs (such as XML::Parser), or through a library for a specific syntax (such as the SOAP or SVG libraries), Perl provides a capable, if slightly opaque, language for processing XML.

The breadth of libraries for Python truly expresses a common motto of Python, "Batteries included." If you need simply to manipulate XML for a configuration file, or other simple format, you can make use of the powerful built-in functionality. If you need to read or write a specific format of XML, it is almost guaranteed that there is a library out there to help you. This, combined with the expressiveness of the Python language makes it an ideal choice for developing with XML.

Although Ruby does not have native support for a wide variety of XML formats, its native XML handling with REXML provides a rapid means of reading and writing just about any format. Support for creating and calling Web services and other common syntaxes, such as RSS, XML-RPC and others is a nice added bonus. While the total number of libraries is generally less than for Perl or Python, those that are integrate so easily into the nature of the language that it sometimes seems like they were always part of the language.