Gregory Todd Williams

Version 1.041 of RDF::Query hit CPAN a few days ago.

Notable changes since the last announced release (1.038):

• Fixes to file:// URIs allowing their use when running on Windows.
• Updates to allow running on perl 5.6:
  • The SPARQL grammar now gracefully degrades Unicode support when running under perl 5.6.x (which was having trouble with the Unicode token rules).
  • Removed unwanted ‘+’ signs on stringified bigints produced by the SQL compiler when running under perl 5.6.x.
• Removed unneccesary forced dependencies (now listed as recommended modules):
  • Digest::SHA1 (used for jena function support).
  • Geo::Distance (used for distance-based filtering test).
• RDF::Core and RDF::Redland bridge classes now throw an error when passed a bad model object.
• Added support for BOUND() filters in SQL compiler.
• SQL compiler now produces valid SQL when query variable names are SQL reserved keywords.
• Removed code that tried to do JIT optimization of queries (this functionality was split out as a peephole optimizer).
• Fixed bug that prevented running queries against multiple models.
• Fixed SPARQL parsing bug for blank nodes as statement objects.
• Fixed bug in RDF::Core bridge class where isa_resource() was returning true for blank nodes.

Thanks to Nigam Shah for reporting the Windows and perl 5.6 bugs and being willing to help me test code until I fixed them, and to Yaron Koren for motivating some ongoing work at improving the documentation and examples.