v1.6.0

v1.6.0 Release Notes

See the
release and closed tickets on the
v1.6.0 milestone on Github.

Required changes to existing data:

• A project is now required to have at least one description, so potentially a description will need
  to be added to those projects that don't have one.

New features:

General:

• Added a /health endpoint
• KnoraService waits on startup for a triplestore before trying to load the ontologies

Gravsearch enhancements:

• Accept queries in POST requests (@github#650).
• Allow a Gravsearch query to specify the IRI of the main resource (@github#871) (by allowing BIND).
• Allow lang to be used with !=.
• A UNION or OPTIONAL can now be nested in an OPTIONAL (@github#882).
• Gravsearch now does type inference (@github#884).
• The Knora API v2 complex schema can now be used in Gravsearch, making it possible to search
  for list nodes (@github#899).

Admin API:

• Make project description required (@github#875).

Conversion to TEI:

• Conversion of standard standoff entities to TEI
• Custom conversion of project specific standoff entities and metadata to TEI

Sipi integration:

• The Knora specific Sipi configuration and scripts can now be found under the sipi/ directory (@github#404).
• Documentation on how Sipi can be started changed (@github#404).

Bugfixes:

• Allow a class or property definition to have more than one object for rdf:type (@github#885).
• Exclude list values from v2 fulltext search (@github#906).

Gravsearch fixes:

• Allow the lang function to be used in a comparison inside AND/OR (@github#846).
• Fix the processing of resources with multiple incoming links that use the same property (@github#878).
• Fix the parsing of a FILTER inside an OPTIONAL (@github#879).
• Require the match function to be the top-level expression in a FILTER.