Subject:Large xquery dataset. Author:Steve Cox Date:28 Nov 2005 09:30 AM
I was experimenting with what was effectively a left outer join of two datasets with about 15,000 records each - sharing a key.
I used an xquery for the join similar to:
for $row in /Contact1/row,
$row1 in /Contact2/row
where $row/ACCOUNTNO = $row1/ACCOUNTNO
and pulled out most of the fields for each row (10-15 fields per dataset).
Doing this on a database system where the key is indexed takes 10-15 seconds, using xquery within stylusstudio took nearly 3 hours on a 1GB 3GHz Pentium4. I was wondering if there is any way to speed this up - maybe defining keys in each dataset?
Subject:Large xquery dataset. Author:(Deleted User) Date:28 Nov 2005 10:03 AM
These types of optimizations depend entirely on which xquery processor you are using. The Stylus Studio xquery processor has been built as a tool to design, test, and debug xquery programs. Users are normally expected to deploy the application on a different xquery processor which has been designed to be efficient in a production environment.
For you example, the DDXQ xquery processor would optimize this query quite nicely. We suggest you view the video tutorial at: