Proceedings

Abstract:

XSLT 2.0 provides a wide range of new features, many of which make light work of tasks that are notoriously difficult in XSLT 1.0, such as grouping and string manipulation. This paper attempts to show how these facilities not only make coding easier, but will also extend the scope of the language making it possible to tackle problems that were quite outside the range of XSLT 1.0.

The paper shows case study of a multi-phase transformation taking data from a legacy ASCII-based interchange format, to XML based on a standardized vocabulary. The transformations illustrate the power of new features including regular expression handling, grouping, recursive functions, and schema-aware processing.

The conclusion of the paper is that these new facilities - notably regular expression handling and grouping - take XSLT into new territory, where languages such as Perl previously reigned supreme. XSLT 1.0 works best where all the structure in a document is already identified by markup. XSLT 2.0 will also be able to handle many situations where the structure is implicit in the text, or in markup designed for presentation purposes rather than to capture the information semantics. It thus becomes a powerful tool for "up-conversion" applications. These facilities work will in conjunction with schema-aware processing, where the aim of the exercise is to create XML that conforms to a target schema.