R2RML mappings for the MusicBrainz schema on an entity-by-entity basis.
These can be run on the MusicBrainz server using ultrawrap, for which a script is provided (dump.sh).
You must set an environment variable ULTRAWRAP_HOME.
Running musicbrainz-r2rml/dump.sh entity (where entity is artist, track, etc.) runs the appropriate set of mappings (e.g. mappings/artist.ttl) to produce output in the form of NTriples (e.g. output/artist.nt).
A virtual machine is available (for use with VirtualBox, VMware, etc.) with a replicated MusicBrainz database.
Note that the file musicbrainz_compile_config.properties must reflect your DB name:
musicbrainz_dbis the default for a snapshotmusicbrainz_db_slaveis the default for a replicated database
Please report any issues on our Jira tracker, under the LINKB project.