Register additional serializer plugins for SPARQL mime types. by darrengarvey · Pull Request #987 · RDFLib/rdflib (original) (raw)

@darrengarvey
This looks great, thanks.

Just one thing we'd like to discuss before merging.

Do we really need two different entries for application/sparql-results+xml; charset=UTF-8 and application/sparql-results+xml? Seems this is an opportunity to make this lookup more concise.

I understand the content-type header can and sometimes does include the charset (or other additional parameters), do you expect the user will pass that directly into the serializer? Should there be a mimetype sanitizer that splits removes anything after a semicolon when matching the mimetype to serializer (or parser)?