Andy,
The data in question is this: Arāḑ Muḩtallah Apparently the "h" is a latin U+0068, but it is combined with a U+0327 code to get the curve under the "h". My understanding is that the Unicode Normal Form C prefers one character (a combination character) instead of the two codes for the one character. Regardless of the data, I shouldn't expect a NullPointerException, I would expect a warning. The only way I found the Unicode warning was through stepping through the code in my debugger to figure out what was causing the problem. Thanks, Elli ________________________________ From: Andy Seaborne <[email protected]> To: [email protected] Sent: Thursday, July 19, 2012 1:01 PM Subject: Re: NullPointerException when writing Unicode data (switch to the users list) On 19/07/12 17:53, Elli Schwarz wrote: > Hello, > > I am attempting to write a graph stored in Fuseki out as RDF/XML, and I get a > NullPointerException from line 199 of LangRDFXML. It looks like the variable > errorHandler is null. > > There is actually a warning that "... {W131} String not in Unicode Normal > Form C: ..." that is coming from Jena's XMLHandler, but instead of this being > propagated back as a warning it is throwing a NullPointerException. > > It seems that it isn't a fatal error, so no exception at all should be > thrown, just a warning should be logged, so I'm guessing this is a bug? > > I'm using Jena 2.7.2, ARQ 2.9.2, and I'm connecting to a Fuseki 0.2.3 back > end (this error occurs when I do ds.getModel(modelName) where ds is a Fuseki > DataAccessor. > > Thank you! > -Elli > What does the data look like? Andy
