- From: Jeni Tennison <jeni@jenitennison.com>
- Date: Thu, 17 Jan 2002 11:15:02 +0000
- To: ht@cogsci.ed.ac.uk ((Henry S. Thompson))
- CC: Khaled Noaman <knoaman@ca.ibm.com>, xmlschema-dev@w3.org
Hi Henry, >> Yes. When Schema A or Schema B get used, the schema validator >> constructs, in memory, one big schema that includes everything from >> Schema A, Schema B and Schema C. All the components can reference each >> other across physical file boundaries. > > Actually, not necessarily 'Yes'. Unless I'm confused, the target > namespace of schema document C must be different from that of schema > document A. So to allow references in schema document C to > components defined in schema document A, schema document C must > import schema document A's target namespace. Now I'm confused :) If you use Schema A then I thought you got a schema information item consisting of the components from: Schema A Schema B (through the import from Schema A) Schema C (through the include from Schema B) If you use Schema B then I thought you got a schema information item consisting of the components from: Schema B Schema C (through the include from Schema B) Schema A (through the import from Schema B) I thought that what was important when resolving QName references was not the physical document, but rather the schema components of the schema information item during validation. But it wouldn't surprise me if I've missed something in the Rec? Cheers, Jeni --- Jeni Tennison http://www.jenitennison.com/
Received on Thursday, 17 January 2002 06:15:05 UTC