Child pages
  • Hydra and Fedora 4 Federated Content
Skip to end of metadata
Go to start of metadata
At this time we are leaning away from federation as a solution, so this the content of this page is historical. At this point we are working towards redirecting to services for interactive mediation for file retrieval. See HydraDAM2 asynchronous storage interactions.

This is a scratch pad page to explore using Fedora 4 federated content in the context of a Hydra/ActiveFedora Rails application

Problem to solve

If you have a federated content via a projection, your Fedora repo knows about that content and has nodes for all of the files and directories. It appears as being "in Fedora" even though it isn't. However, Hydra/ActiveFedora didn't put any of those nodes there, so therefore they aren't Solrized. They may be in Solr somewhere depending on the local Fedora architecture, but that's not the same as saying they were Solrized by ActiveFedora specific to a Hydra app's content models. They probably also lack key properties and/or content nodes required by a Hydra apps functionality.

Given that, it should be possible to walk the federated content and Solrize things in a way that would be usable (in a read-only sense) by some Hydra app in question.  This may also involve establishing additional relations between the content's nodes and real Fedora 4 objects, or maybe creating proxy nodes to tie things together.


  • Create a new blank Hydra app using ActiveFedora 9 for Fedora 4
  • Follow the latest Dive Into Hydra Tutorial and create the simple book model with RDF desc metadata
  • Create a simple directory tree representing books, properties, and TEI XML
  • Configure Fedora 4, preferably within Jetty, to project over book XML directory

Experimental scenarios:

  • Can extra properties be set on external book items and be Solrized to look like native book objects in the Hydra app?
  • Can external book digital objects nodes be associated as files to native book objects in the Hydra app?
  • Can ActiveFedora be employed to help Solrize the federated content (vs. custom asynchronous indexing code?)
  • Could a directory (nt:folder) hold properties to represent a book, and make Hydra understand that the child digital nodes are the files?
  • Would Fedora4 attachments be helpful? (would be like using Fedora3 datastreams)
  • No labels