DFTL: a Description Language for Transformation of Data Formats

Abstract

While conceptually simple, format transformation is not always easy, especially for end users of software systems. This paper presents a description language referred to as DFTL. DFTL is short for data format transformation language, which is a context-free grammar based language to take transformation between heterogeneous data. Employing XML as metalanguage, DFTL endeavors to describe a data format transformation model, which consists of logical structure of data and transformation rules associated with structural branches, in an actionable manner. When using DFTL, the transformation model of data in a data file is described by means of DFTL Schema, which is an XML Schema containing only a subset of the constructs available in full XML Schema description language, and augmented with special DFTL annotations. The DFTL Schema description sits in a logical separate file from the data itself.