- From: Damian Steer <pldms@mac.com>
- Date: Tue, 17 Nov 2009 12:43:47 +0000
- To: Richard Cyganiak <richard@cyganiak.de>
- CC: Sebastian Hellmann <hellmann@informatik.uni-leipzig.de>, semantic-web@w3.org, Michael Martin <martin@informatik.uni-leipzig.de>, Christopher Jona Sahnwaldt <christopher@sahnwaldt.de>, Matthias Weidl <matthias.weidl@googlemail.com>, Anja Jentzsch <anja@anjeve.de>, Robert Isele <robertisele@gmail.com>
Richard Cyganiak wrote: > Just throwing in one comment: I *believe* (but am not 100% sure) that > XML element names may contain characters outside of the US-ASCII range, > such as “ä”. So, <dbpedia:längengrad> might actually be a valid XML > element. This is true for both xml 1.0 and 1.1, see [1] and [2]. IIRC the exclusions are mostly obvious, non-printable things once you're beyond ascii. > I have no idea if this also applies to RDF/XML. The > relationships between URIs, IRIs, RDF, XML, UTF-8 etc are incredibly > complex... But it might be worth trying if a URI such as > <http://de.dbpedia.org/property/längengrad> is actually somehow allowed > *in the RDF data model*, and how it would be serialized in the different > RDF surface syntaxes. Things will probably go pear shaped once the IRI has been converted to a URI, particularly comparisons. I doubt any libraries are routinely normalising incoming IRIs. Damian [1] <http://www.w3.org/TR/xml-names/#NT-NCNameStartChar> [2] <http://www.w3.org/TR/xml11/#IDAKUDS>
Received on Tuesday, 17 November 2009 12:43:23 UTC