Abstract :
This paper describes a declarative and unifying approach, based on XQuery and XML Schema, to modeling and accessing the variety of data source types found in a typical enterprise. These include relational, Web service, function-based (a.k.a. servicebased), and file-based data sources. The approach that we detail here is based on introspection of data source meta-information and generation of metadata artifacts that conform to a common model and that provide a uniform framework in which the data sources become available to an XQuery-based query processor. We explain how our approach addresses various aspects of data sources, including data source connectivity, data typing, integrity constraints, access control, and the need for performant access. We also explain why XQuery and XML Schema can serve as an excellent vehicle for data and metadata integration. The approach described in this paper is used in BEA´s AquaLogic Data Services Platform product and serves as the substrate for its data service modeling concepts.
