xml-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mike Dierken <m...@DataChannel.com>
Subject RE: [vote] A native XML database project under Apache
Date Mon, 29 Oct 2001 17:28:36 GMT

> -----Original Message-----
> From: Robert Stuart [mailto:Robert.Stuart@NFER-NELSON.co.uk]
> Relational database theory derives from set theory.  (EF Codd 
> circa 1980)
> A tree structure may be viewed as a set of branches each 
> containing a set
> branches and leaves.
> 	The root is simply a branch with no parent branch.
> 	A leaf is simply a branch with no child branch.
> Thus, we are mapping one set-of-sets onto another set-of-sets.
Each traversal of a relationship is an expensive operation. If you look at
an XML document as a graph of nodes (ala Groves, etc.) the number of
traversals to transform the data is very high. If each relationship is
expensive, then XML document processing will be very expensive.

It would be better to store the XML as a blob, and extract XPath-style node
identifiers for a searchable index (via SAX events as you store the blob). 

Mike Dierken
"In theory, there is no difference between theory and practice. In practice
there is."

View raw message