We've been looking here lately at exporting data out of Omeka.
Specifically, we are OAI harvesting a set of 40,000+ records from CDP Heritage West/DC.Builder into Omeka. However, we want to use the same data we've harvested as part of a CDP dark archive migration, where we would use the DC as source metadata to be transformed to MODS wrapped in METS and re-connected with master files in the METS . Basically we need either a csv export that we can map to an xml template or flat xml files we can transform by applying and xsl.
Looking at how Omeka stores the element data it's harvested, a simple csv export looks not so simple at all - matching element ids and record ids and data values....
We see the new version of Omeka including a set of xml exports, including DC, but at the record/item view - not in batch...
You may find of interest a yet to be released plugin, Export (svn). It generates a snapshot of your entire Omeka repository, including all items and their DC metadata. The output is an XML file that conforms to the Omeka schema. You could transform this directly to MODS. I'm not sure why the plugin isn't released, but I'll look into it.
You could also write an Omeka plugin that iterates the items while writing MODS to a file.
if you could look why the plug-in wasn't released, that would be appreciated.
Perhaps you could put me in touch with developers and we could move this thing ahead?
Apologies for suggesting yet another alternative that is as-yet unreleased, but there's another solution that may fit your needs.
The unreleased version of the OAI-PMH Repository plugin can export directly to MODS for OAI-PMH harvesters. The repository has the benefit of having lower server and configuration requirements, as well.
I'm the developer of both the repository and the Export plugins. Both of these plugins are in the "testing and polish" stage, so you should be able to expect an update on them fairly soon.
May we offer ourselves here at ADR as b/testers for the Export plugin?
We are well familiar with (old) OAI-PMH plugin and welcome direct MODS harvest.
Our current challenge is to extract the DC records from Omeka repository, for later batch processing, however...