Lists Home |
Date Index |
- To: "xml-Dev (E-mail)" <email@example.com>
- Subject: RE: [xml-dev] on the subject of schema crunching...
- From: "Paul Brown" <firstname.lastname@example.org>
- Date: Sun, 23 Jun 2002 18:34:03 -0400
- Cc: "Dare Obasanjo" <email@example.com>
- Thread-index: AcIa90SErMxbc3P6SYG8e0+v+arr6QAANMu9AAM1mnA=
- Thread-topic: on the subject of schema crunching...
> From: Dare Obasanjo
> A schema can only have a single targetNamespace while imports
> reference schemata from another namespace. I can't readily
> think of any work around to that fundamental restriction to
> how you can create a single unified schema.
The work-around would be to create one unified "module" for each target namespace. That's not such a big deal.
> Why exactly do you want to do this anyway?
Schemata with large data dictionaries are cumbersome, and examples of large data dictionaries are all around us: EDI, HL-7, etc. Most individual message types only use a small subset of the actual data dictionary. The individual message types -- once distilled -- are more portable, easier for humans and machines to process, etc.