Chapter 6. Getting Relational: Database Schema Design

In which my car leaves me stranded on the highway, and we finally discuss the integration of XML and SQL!

A database schema is a collection of database objects that are designed together to fulfill a specific function. The most commonly known database object is a database table, in which information is stored in columns and rows (see the section Structured Query Language (SQL) in Chapter 2), but other commonly used database objects include sequences (incremental number generators), stored procedures (sometimes called by database-specific terms such as Oracle's PL/SQL), and triggers (stored procedures that “fire off” before, during, or after a specific event, like the insertion or deletion of data).

In this chapter, we discuss how to put together a database schema that best complements the storage of data in XML form. We start with database schema basics and then delve into more detailed examples. The underlying principle presented in this chapter is that of partial decomposition, a method for integrating XML with relational databases by which the entire XML instance is stored as a whole (possibly in the database itself), and selected data from the XML instances is extracted and placed in relational tables to speed access to data and to maintain data integrity.