[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: XQuery - The truth comes out!
- From: Steve Muench <Steve.Muench@oracle.com>
- To: Mike.Champion@SoftwareAG-USA.com
- Date: Sun, 25 Feb 2001 17:10:15 -0800
| Existing practice seems like an RDBMS world in which SQL has nothing but a
| SELECT ... WHERE clause
| What am I missing here?
GROUP BY, HAVING, subselects in arbitrary places where
"tuple-lists" are expected", etc.
Some RDBMS's go well beyond SELECT and WHERE, including object
extensions that let you map type structures and instances to
SQL99 object types, with rich, nested structures, and provide
convenient path-like syntax for querying these with the performance
expected of a relational database. They also allow adhoc
construction of instance of object types in queries, so construction
of typed-structures is not a foreign notion. These object/relational
extensions map naturally to data-oriented XML, and especially well
to XML Schema. Some ORDBMS's offer built-in XML integration on top
of these object/relational features today.
The path expression language that SQL99-compliant ORDBMS's use today is a
SQL99-flavored one and not an XPath path syntax, but various
SQL database vendors are pursuing a SQL/XML extension in the ANSI committee
that may prove attractive to folks with a deep investment in SQL databases
or a SQL background who want to work with XML querying. It's still early
days for this SQLX effort, though, as is clear from the effort's fledgling
website at http://www.sqlx.org so there's not much yet to offer information-wise
about what a SQLX standard might eventually look like, except the somewhat
obvious conclusion that it would be an extension of SQL.
| there has to be room for evolution/markets to select the best practices.
Steve Muench, Lead XML Evangelist & Consulting Product Manager
BC4J & XSQL Servlet Development Teams, Oracle Rep to XSL WG
Author "Building Oracle XML Applications", O'Reilly