The default setting for PRESERVE_BNODE_IDS in RDF4J is false, but currently the Jelly parser always assumes true. This makes loading multiple files with blank nodes potentially destructive, as blank node identifiers may get randomly merged.
We should change it to false by default, and also allow reverting it to true.
I suspect we may have the same issue in Jena, check that.
We will also need follow-up changes in jelly-cli, I think?
The default setting for
PRESERVE_BNODE_IDSin RDF4J is false, but currently the Jelly parser always assumes true. This makes loading multiple files with blank nodes potentially destructive, as blank node identifiers may get randomly merged.We should change it to false by default, and also allow reverting it to true.
I suspect we may have the same issue in Jena, check that.
We will also need follow-up changes in jelly-cli, I think?