A practitioners approach to normalizing XQuery expressions

Abstract

XQuery becomes a standard of the XML query language. Just like in SQL, XQuery allows nested expressions. To optimize XQuery processing, a lot of research has been done on normalization, i.e., transforming nested expressions to equivalent unnested ones. Previous normalization rules are classified into two categories - source-level and algebra-level - depending on whether a construct is specified by using a query language or an algebraic expression. In implementation point of view, we contend that the source-level rule is preferable to the algebra-level rule because algebras used for normalization are hard to be directly exploited in a typical DBMS. However, a complete set of source-level rules is yet to be developed. In this paper, we propose source-level rules for normalizing XQuery expressions and present an implementation mechanism. We show that our rules are correct and complete according to the nesting types classified by Kim. Our mechanism is easily implementable since it adapts the well-known Query Graph Model (QGM) representation. We have successfully implemented this mechanism into our XML DBMS named Odysseus/XML.