Skip to content

Home

Jump to bottom
Jeremy Nelson edited this page Feb 5, 2017 · 8 revisions

DPLA Service Hub Documentation

This project provides the source code and tools for converting different metadata vocabularies and file formats into BIBFRAME 2.0 linked data stored in a RDF Triplestore. From this RDF Triplestore, we generate ResourceSync feed for consumption by DP.LA.

Installing

Currently the only way to get the DPLA Service Hub software up and running is to build the project from source. The steps to do so are outlined in the README.md.

Running

The DPLA Service Hub requires a running instance of the Docker image jermnelson/semantic-server-core that bundles a Blazegraph RDF Triplesstore and Fedora Commons Digital Repository. The easiest way to run this image with Blazegraph running on port 9999 and Fedora on port 8080 is:

docker run -p 8080:8080 -p 9999:9999 jermnelson/semantic-server-core:dev

Command-line Ingester

To ingest input metadata into BIBFRAME Linked Data, a command-line script profile.py is called with different required arguments and optional parameters that either directly posts the resulting RDF into a triplestore or you can optionally output to a RDF Turtle file. There are two commands, add_record and add_batch, that are available in profile.py.

add_record

The add_record command is for a single record ingestion. This command takes a RDF ttl Rule file called a profile, a metadata input file, and either outputs to an RDF BIBFRAME turtle file or to the RDF triplestore defined in the application's configuration file located in the instance directory. The Profile RDF file should be located in the custom directory. The additional parameters for add_record are:

--ingest_type [dc|csv|mods|ptfs] Ingester type, calls and configures the different metadata ingester. The available options are dc (Dublin Core), csv (Comma- separated value), mods (MODS metadata vocabulary), ptfs (Custom XML format).

--item_iri Optional IRI for Item. You may already have a pre-existing URL you want to use as the BIBFRAME IRI for the Item, the profile will use this instead of having the ingester generate a new IRI for the incoming record.

--in_file Full path including the metadata input file name to be transformed into BIBFRAME RDF. This is an optional parameter but if missing then, the --at_url MUST be present.

--at_url A URL to the metadata you want to ingest, This is an optional parameter but if missing then, the --in_file MUST be present.

--out_file Full path and file name that saves the output of the ingester to a RDF Turtle file. This is an optional parameter and if missing the ingester will submit the new linked data to the triplestore defined in the configuration file.

Custom Jobs

Clone this wiki locally