XPERANTO: Middleware for Publishing Object-Relational Data as XML Documents

The eXtended Markup Language (XML) is quickly emerging as the universal format for publishing and exchanging data on the World Wide Web. As a result, data sources, including object-relational databases, are now faced with a new class of users and applications; customers and programs that would like to deal directly with XML data rather than being forced to deal with the data source’s particular (e.g., objectrelational) schema and query language. The goal of the XPERANTO middleware project is to support this new class of users and applications. XPERANTO does this by providing query-able “XML Views” over the underlying objectrelational database. Users can then query and (re)structure XML data using an XML query language, without having to deal with the underlying SQL tables and query language. The XPERANTO system translates XML-based queries into SQL requests, receives the tabular query results, converts them into XML, and then returns XML documents to the system’s users and applications. 1. Motivation and Introduction XML [2][4] is emerging as the standard for publishing and exchanging data on the World Wide Web. As a result, many “e-businesses” want to publish their existing data as XML documents so that their business partners can process them. For example, an online wholesale bookstore may want to make its inventory available in XML form so that retail merchants can process this information over the Internet. It is also important to provide query capability over such XML data because, for instance, a retail merchant may want to check whether the wholesale bookstore has a particular book in stock (this is a selection query over the inventory XML document). The goal of the XPERANTO project [3] at the IBM Almaden Research Center is to publish object-relational data as XML documents. XPERANTO operates as a middleware on top of existing (object-)relational database systems and provides query-able “XML Views” over them. In doing so, it shields users of XPERANTO from the underlying database and language (SQL) and allows them to create and query XML data using an an XML query language (currently XML-QL [5]). Thus, businesses that have large amounts of important business data stored in existing (object-)relational databases can use XPERANTO to make this data available on the Internet in query-able XML form. XPERANTO differs from other similar systems that we are aware of [6] in the following ways. First, XPERANTO is based on a “pure XML” philosophy – users and developers do no need to know or learn SQL. *Currently at Propel, 2350 Mission College Blvd., Santa Clara, CA 95054. + Also at the University of Wisconsin, Madison, WI 53706. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the VLDB copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Very Large Data Base Endowment. To copy otherwise, or to republish, requires a fee and/or special permission from the Endowment. Proceedings of the 26th International Conference on Very Large Databases, Cairo, Egypt, 2000 Second, XPERANTO attempts to push as much processing down to the relational engine as possible for more efficient query execution [7]. Third, XPERANTO handles object-relational structures in addition to flat relational structures. Finally, XPERANTO enables the seamless querying of relational data (such as column values) and meta-data (such as column names). For the remainder of this demo proposal, we outline the architecture of the XPERANTO system and show it can efficiently support query-able XML views by doing smart XML query rewrites and by harnessing the sophisticated processing capability of the underlying (object-)relational engine. 2. The XPERANTO High Level Architecture As mentioned earlier, XPERANTO is a middleware system that works on top of existing (object-)relational database systems. Figure 1 shows the high level architecture of the XPERANTO system. XPERANTO starts by providing a default virtual view of a given (object-)relational database (step 1). XML application developers can then create more complex or specialized views (which are also virtual) on top of this default view (step 2) by using an XML query language (currently XML-QL). These tailored views can then be made available to other applications and/or user groups, who can in turn create more complex view over the existing views (step 3). The views can then be queried using the same XML query language used to define them (step 4) and XPERANTO returns the result XML document (step 5). XPERANTO clients use an XML-based RPC facility – Simple Object Access Protocol [1] – to communicate with the XPERANTO middleware for creating and querying XML views and retrieving the XML results. Note that XPERANTO takes a "pure XML" approach to the problem of publishing (object-)relational data as XML i.e., developers interact with XPERANTO using only XML (and an XML query language); they do not have to use or learn SQL. An interesting and novel aspect of XPERANTO is that it provides a uniform framework that allows users to use XML as a way to seamlessly query over both relational data and relational metadata. This power stems from the fact that the XPERANTO default XML view of a relational database does not distinguish between data (such as column values) and metadata (such as column names). In this sense, XPERANTO provides users and applications a more powerful query capability than SQL. 3. XPERANTO Internals We now describe the internals of the XPERANTO system and show how it supports query-able XML views over relational sources. As Figure 2 shows, when an user or applications poses an XML-QL query over an XML view, it is first parsed and translated into an internal representation called XQGM (Xml Query Graph Model). The representation is designed to be language neutral (i.e., not tied to the XML-QL query language) so that XPERANTO can be easily modified to support the standard XML query language when one becomes available. Figure 1: XPERANTO Architecture (1) (Object-)Relational Database System XPERANTO