Improve content type handling in SPARQL source loading

[aucampia]

This is not the best way of doing things, better to rather try content-type header first.

rdflib/rdflib/plugins/sparql/sparql.py

Lines 315 to 333 in ad56044

	def _load(graph, source):
	try:
	return graph.parse(source, format="turtle", **kwargs)
	except Exception:
	pass
	try:
	return graph.parse(source, format="xml", **kwargs)
	except Exception:
	pass
	try:
	return graph.parse(source, format="n3", **kwargs)
	except Exception:
	pass
	try:
	return graph.parse(source, format="nt", **kwargs)
	except Exception:
	raise Exception(
	"Could not load %s as either RDF/XML, N3 or NTriples" % source
	)