> If their structure isn't so regular, using SAX to extract the records and > running them individually through XSLT might be a good approach.
This seems like a good approach. The XML that we have to read into the database will potentially have some irregularity. But it might be worth a little bit of extra work to code the extraction process manually. After thinking it over, it seems like using SAX directly won't really be that much extra work and it's sure to provide better performance. -ishmael.
