We're looking at using Omeka as a light-weight public site for materials
we're already managing in a Fedora repository. I've tried the
FedoraConnector plugin, and it looks like it do what we need at the item
and datastream level. I'm now looking at managing the ingest as a batch,
and I thought I'd check in and see what others are doing or thinking about
I'd like to feed in a list of Fedora PIDs and have the FedoraConnector
create an item for each one, import the DC metadata, and attach the
necessary datastreams (which would be the same for every item). To do the
item level work (create, import, attach) in a single operation looks like a
straight-forward enhancement to FedoraConnector: in
DatastreamsController.php I'll refactor importAction() and
insertdatastreamAction() to separate the grabbing of values from a post
from the insertion of those values into the db, to allow for feeding those
values in from a batch process.
At the batch level, I think the ideal solution would be to generalize the
CSVImport plugin's batch management system (with the ability to interrupt
and roll back a batch etc.) to handle generic batch ingests, with
sub-plugins for CSV, Fedora, etc. I haven't started investigating to see
how feasible that would be.
Finally, we'll be interested in adding some new functionality: during an
import, detect whether an item already exists (based on the PID), and if so
update it rather than create a new item. This would allow us to sync
changes from Fedora to Omeka.
This is all still at the investigation stage, and it's still to be
determined whether we can invest the necessary effort to make this happen.
Is anyone else thinking (or, dare I hope, working) along these lines? Any
advice from those who know the internals of these plugins?
Digital Initiatives Technology Librarian
Information Technology Services
4-30 Cameron Library
University of Alberta
Canada T6G 2J8