A minor version update of the dotNetRDF library is now available for download.
This release makes two enhancements requested by library users.
The DefaultDocumentLoader static class has been updated to use the more modern
.NET HttpClient stack under the hood, and now exposes properties to allow control over the maximum processed response size and maximum number
of redirects to follow when loading JSON-LD documents from the Web.
The QueryBuilder class now applies the default configured SPARQL query optimisers to the generated SparqlQuery. Whether or not to apply
query optimisation, and which optimisers to apply can now also be controlled through optional arguments to the Build method.
This release also introduces a new experimental feature - the PullQueryProcessor. This is a SPARQL query processor implementation written to use an asynchronous enumeration approach to processing SPARQL queries. The goal is to provide a SPARQL query engine which can work on larger data than Leviathan by avoiding creating large intermediate result sets while processing a SPARQL query. At this stage the implementation is complete and passes the SPARQL 1.1 query conformance tests (excepting a couple of URI normalisation tests). However in performance testing, the PullQueryProcessor is a couple of orders of magnitude slower than the default Leviathan query processor. There will be further work on this engine to attempt to improve performance, but at this point the PullQueryProcessor is being treated as experimental and should not be used in production code.
dotNetRDF 3.3.0 can be found here on NuGet and also on GitHub.