Roland Hedberg wrote: > http://www.ldap.com/1/schema/ldapv3.owl#0.9.2342.19200300 > > Now, rdflib barks on this and says it can not split the uri. > > Looking at the code I find the reason to be that the leading character > in "0.9.2342.19200300" does not belong to one of the > NAME_START_CATEGORIES. In fact none of the characters does. > > So are rdflib correct and therefor the ontology faulty ? Or ?
I'm not 100% sure. I actually saw this same behavior (that the fragment part of the URI must not start with a digit) in the Jena API today (that's a Java RDF API), so I'm inclined to think it's not allowed by the specs. However, I can't find that in the specs. First stop is RFC 2396, which defines URIs, says it's allowed http://www.ietf.org/rfc/rfc2396.txt (look for the fragments definition) Next stop was the RDF spec: http://www.w3.org/TR/rdf-concepts/#dfn-URI-reference http://www.w3.org/TR/rdf-concepts/#section-fragID Actually, they mostly refer to RFC 2396, and add some constraints (about absolute vs relative, Unicode formatted, and no escaping), but nothing about the fragment syntax. Searching on Google for "fragment RDF syntax site:w3.org" gives a lot of interesting stuff (mostly discussion that a URI does not contain a fragment, but a URI-Reference does -- I learn every day). So nothing definitive, sorry. Maybe you have more luck searching. Regards, Freek
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Dev mailing list Dev@rdflib.net http://rdflib.net/mailman/listinfo/dev
