Querying XML with locator semantics

Peter Fankhauser
German National Research Center for Information Technology
Integrated Publication and Information Systems Institute
Dolivostraße 15, D-64293 Darmstadt, Germany


The prevailing data model for semi-structured data is XML. XQL is a query language for XML that supports flexible navigation and filtering on possibly irregular document trees by means of simple path expressions. I present an approach to extend XQL with support for joins and for more expressive extraction. To this end, locators are introduced, which allow to identify document nodes by paths. On this basis, a locator algebra (or L'Algebra for short) can be defined, which supports matching, combination, and restructuring of XML documents without excessive generation of new nodes. L'Algebra preserves document order, and - by virtue of the locator semantics - also preservers regular or irregular structure and navigation history. I define the semantics of the main operators of L'Algebra, introduce its axioms, and describe a mapping from XQL expressions to L'Algebra.


Peter Fankhauser has studied computer science at the Technical University of Vienna. His PHD was on knowledge based methods for the integration of heterogeneous database schemas. Since 1989 he is affiliated with the Institute for Integrated Publication and Information Systems of the GMD (German National Research Center for Information Technology) in Darmstadt, Germany. He currently works on interoperability of information systems.