You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Speed; we've measured cases of SPARQL queries where Oxrdflib runs ~2 orders of magnitude faster than the default RDFlib.Store implementation.
Cons:
Oxrdflib has some troubles with namespaces ... specifically, any prefixes used in a namespace must be explicitly added to each SPARQL query
binding variables do not appear to be working
result rows from a query differ from the default impl when there are OPTIONAL clauses
Oxrdflib uses setup.py to register its RDFlib plugins and it's not clear that this is working properly
We've been able to isolate these issues with a minimum Python code + RDF + SPARQL example, and I'll be opening an issue for Oxrdflib which is linked to this issue.
The SQLAlchemy issues are a different matter. Ostensibly, our new support for specifying a store="foo" in the KnowledgeGraph constructor similarly applies here. The exceptions listed seem to be in the SQLAlchemy store's support for required methods, specifically in the support for contexts. Perhaps they didn't quite get that far?
FWIW, we've got a new RDFlib.Store plugin in development which is based on NumPy / cuNumeric and it does implement the contexts support.
Perhaps this is a Google Colab-only issue?
And oxrdflib:
The text was updated successfully, but these errors were encountered: