I’m trying to use the auto-generated XSD schema at /cgi/schema in 3.2 to assist with some integration work involving SWORD and EPrints XML packages.  The schema is making it difficult to validate the XML packages we are using.

·         The encoding attribute on the data element is not allowed by the schema even though it is supplied by the XML with files embedded export plugin.

·         All sub-fields of a compound type are required by the schema whereas, to my knowledge, EPrints elsewhere doesn’t support this notion explicitly (except by custom validation)

·         The schema doesn’t allow Namedset fields to be blank or ‘UNSPECIFIED’ etc even though this may be a valid value depending on the workflow configuration.

I’ve attached a patch against 3.2 that addresses these three issues. It looks like it should work in 3.3 also.

The patch for the latter two issues weakens the constraints imposed by the schema but makes the schema applicable in a greater number of what I would consider normal scenarios.

With the patch I can validate the simple EPrints XML documents/packages we are using in our current work successfully but if I export a real eprint and validate the output there are numerous validity errors reported, which seem to relate to issues and indexcodes files. I haven’t got the time to look at these at the moment.

Cheers

Mark

Mark Gregson | Applications and Development Team Leader
Library eServices | Queensland University of Technology
Level 3 | R Block | Kelvin Grove Campus | GPO Box 2434 | Brisbane 4001
Phone: +61 7 3138 3782 | Web: http://eprints.qut.edu.au/
ABN: 83 791 724 622
CRICOS No: 00213J