Hi there, I'm in the process of fleshing out a SPARQL driver for an RDF library, and I want it to be able to handle JSON, because its parser is by far the fastest for the language in question. My strategy was to supply an Accept: header which preferred JSON responses over others, but I ran into a few issues:
1) There appear to be at least three different JSON arrangements for CONSTRUCT responses, *plus* application/sparql-results+json. This is problematic because the driver does not, nor will it contain a bona fide SPARQL parser, and will thus have to rely on heuristics to tell whether or not to expect a variable bindings response versus a graph. Virtuoso (6.1.8) seems to prefer Talis-style JSON graph responses; if I tell it to use JSON-LD, it will respond with the graph embedded into application/sparql-results+json. 2) Virtuoso (6.1.8) does not seem to respect the q-values set in the Accept: header much anyway, but it could just be that I didn't find a set of values that tipped the scales against the source quality values set for the given content types. My question is: is there a way to tweak the source quality values for the different SPARQL response types? *Without* having to make product-specific concessions in the driver? My goal is to make the driver as generic as possible. Thanks, -- Dorian Taylor http://doriantaylor.com/ ------------------------------------------------------------------------------ _______________________________________________ Virtuoso-users mailing list Virtuoso-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtuoso-users
