M.F. Cowlishaw
IBM Systems Journal
Although the Extensible Markup Language (XML) has gained in popularity and has resulted in the creation of powerful software for authoring, transforming, and querying XML-based business data, much information remains in non-XML form. In this paper we introduce an approach to virtualize data resources and thus enable applications to access both XML and non-XML sources. We describe the architectural components that enable virtual XML - a toolbox that includes a cursor model, an XML-view mechanism such as the view created with the Data Format Description Language (DFDL), and XML processing languages. We illustrate the applicability of virtual XML through a number of use cases in various environments. We discuss the products that we expect from vendors and the open-source community and the way enterprises can plan to take advantage of virtual XML developments. Finally, we outline future research directions that include a vision of virtual XML that covers large-scale structures such as entire file systems, databases, or even the World Wide Web. © Copyright 2006 by International Business Machines Corporation.
M.F. Cowlishaw
IBM Systems Journal
Oliver Bodemer
IBM J. Res. Dev
Michael Ray, Yves C. Martin
Proceedings of SPIE - The International Society for Optical Engineering
György E. Révész
Theoretical Computer Science