datestamp in OAI records
|Reported by:||kjdon||Owned by:||kjdon|
As reported by Mariana Pichinini, 18-12-09.
I explain: where among the metadata reported is a DC.date tag, the OAI server treats this as a datestamp for the register. In common practice, this DC.date metadata means the publishing date for a resource, and since this is a internal resource, it usually is not formatted ISO 8601 compliant; in consequence, when OAI server try to read it, it reports a validation fail (I.e. with Open Archives or validators alike)
I don't know the reason for the OAI server to take this DC.date as a datestamp; AFAIK, the most reliable date to take in such a way is the modified date for a source document, which you can take from the 'lastmodified' metadate, present in doc.xml. This is the date that's got everywhere there is no DC.date tag, and I think this is the only one tag that that date should be read from.
When the collection do not implement Dublin Core in their metadata, there is no such trouble. So this mislead is not always raising...