This is just sample data for Oregon Digital's Hydra application.
Data files (images) are in full_res; sample data is in various .xml and .nt files.
We can probably extend and reuse the mapping script in bracero2nt.py later to generate our data on a collection-by-collection basis.