XML to Relational Mapping

From: Brian <boneill_at_calpoly.edu>
Date: Tue, 11 Jun 2002 11:10:15 -0700
Message-ID: <3d063db9$1_6@news.teranews.com>

I have a fairly large complex xml document that needs to go into a db schema and I'm concerned about performace. If I go the standard mapping route of making container elements as tables and attributes/simple elements as columns, I will have about 50 tables w/ a *lot* of data.

Every element has a max string value, so I'm wondering if I should create a one column table, a CLOB, and then create 50 views based on that table, and maybe create a couple of tables for indexing purposes that mimic the access paths since the block density will not be too good. I've seen this done w/very large databases but don't know when the appropriate situation to apply this pattern.

Does anyone have any thought, experiences, or comments about either approaches? T/X.

Brian. Received on Tue Jun 11 2002 - 13:10:15 CDT