When dealing with rdf:XMLLiteral's which includes HTML entities such as ’ does preference go to:
asked 10 Jun '10, 14:43
To maximise portability, and that includes people round-tripping the data through different syntaxes and/or stores I'd recommend that you use the numeric forms in RDF/XML (and appropriate escaping in other syntaxes).
answered 10 Jun '10, 15:48
XML Literals are supposed to be in exclusive canonical form. If I read correctly that means you ought to do the following:
So for ’ you should enter '’', the character itself.
Serialisations other than rdf/xml may impose further restrictions. For example in N-Triples you must encode everything beyond US-ASCII, so <span>’</span> would be written as <span>\u2019</span>.
answered 10 Jun '10, 21:09