This chapter presents the state of the art approaches for storing and retrieving the XML documents from relational databases. Approaches are classified into schema-based mapping and schemaless-based mapping. It also discusses the solutions which are included in Database Management Systems such as SQL Server, Oracle, and DB2. The discussion addresses the issues of: rebuilding XML from RDBMS approaches, comparison of mapping approaches, and their advantages and disadvantages. The chapter concludes with the issues addressed.
TopApproaches For Storing And Querying Xml
A number of different techniques for storing XML documents in a RDB have been established. These techniques can be divided into two groups: the schemaless-centric technique and the schema–centric technique (Dweib et al., 2008). The first one makes use of XML document structure to manage the mapping process (Tatarinov et al., 2002;Dweib et al., 2008;Soltan and Rahgozar, 2006; Zhang and Tompa, 2004b;Jiang et al., 2002;Yoshikawa et al., 2001). The second one depends on schema information to develop a relational schema for XML documents (Fujimoto et al., 2005; Shanmugasundaram et al., 1999; Amer-Yahia et al., 2004; Atay et al., 2007b; Xing et al., 2007b; Knudsen et al., 2005; Lee et al., 2006).
The aim of mapping XML documents into relational database is to make use of the capabilities of the relational database which are: indexes, triggers, data integrity, security, multi-user access, and query optimization by SQL query language. In the meanwhile XML technology is trying to gain the above-mentioned capabilities, developed for RDBs, and efficiently store, retrieve, and rebuild XML data from RDBs.
The studies that address the problem of mapping XML document into RDB take care of the above issues, and attempt to translate users’ XML queries, either XPath expression (Berglund et al., 2007) or W3C’s recommendation XQuery expression (Boag et al., 2007), into SQL queries (Oracle, n. a.). XQuery gives power to the translation method since XQuery comprises XPath, and it is recommended by W3C, while XPath is not. The translation method should also consider its ability to rebuild, the stored XML document without losing information, and retrieve it in an acceptable time. Many studies have tried to address translation and restore constructing labelling methods. Labelling methods aim to reserve nodes order, parent-child and ancestor-descendant relationships, and document structure(Tatarinov et al., 2002;Chung and Jesurajaiah, 2005;Soltan and Rahgozar, 2006;Li and Moon, 2001;O’Neil et al., 2004;Wu et al., 2004;Kobayashi et al., 2005).