Hi Rob, I'm not subscribed to any other RDF mailing list myself, but feel free to forward this issue and example N-Triples to any mailing list that you think might benefit.
Thanks, Frank. "Rob Vesse" <[email protected]> wrote on 03/22/2011 06:26:09 AM: > [image removed] > > RE: Problem with "Turtle" output? > > Rob Vesse > > to: > > jena-users > > 03/22/2011 06:27 AM > > Cc: > > Frank Budinsky > > Please respond to jena-users > > Hi Frank > > Interesting bug, have you considered publicising this more widely e.g. via > the [email protected] list since I suspect many other libraries may > encounter the same issue. > > You could just provide the sample input as NTriples and then > developers/users can check whether their libraries have the issue and take > appropriate steps to get it fixed elsewhere. > > I have already reproduced and fixed this issue in my own library's Turtle > Writer so I'll echo Andy's sentiment in saying thanks for a clear test case. > Took me only a minute to code up the equivalent in my library and confirm > the problem. > > Rob > > > -----Original Message----- > > From: Andy Seaborne [mailto:[email protected]] > > Sent: 22 March 2011 08:22 > > To: [email protected] > > Subject: Re: Problem with "Turtle" output? > > > > > Is this a bug > > > > Yes - I've raised JENA-55 with your example code as the test case. > > > > Pretty printing Turtle requires a prepass over the data to find lists > > that can't written prettily in the (...) form. It should exclude lists > > where one or more of the cons-cells making up the list has a URI. This > > is a very unusual case; it takes some effort to create a list not > > comprised of blank node cons-cells (there are the Jena API operations > > for lists). In syntax, you have to write the list out long form - no > > short form syntax enables this in Turtle). > > > > Thanks, and thanks for a clear test case, > > > > Andy > > > > https://issues.apache.org/jira/browse/JENA-55 > > > > > > On 21/03/11 22:08, Frank Budinsky wrote: > > > > > > > > > Hi, > > > > > > I created an rdf:list with named list entries, but when I call > > Model.write > > > ("Turtle") it seems to ignore the names and generates the simple "( > > ... )" > > > list syntax. Here's a test program that illustrates the problem: > > > > > > Model model = ModelFactory.createDefaultModel(); > > > Resource data1 = model.createResource(); > > > > > data1.addProperty(model.createProperty("http://property";), > > > "test1"); > > > Resource data2 = model.createResource(); > > > > > data2.addProperty(model.createProperty("http://property";), > > > "test2"); > > > > > > Resource listEntry1 = > > model.createResource("http://test/1";); > > > model.add(listEntry1, RDF.first, data1); > > > model.add(listEntry1, RDF.rest, RDF.nil); > > > > > > Resource listEntry2 = > > model.createResource("http://test/2";); > > > model.add(listEntry2, RDF.first, data2); > > > model.add(listEntry2, RDF.rest, listEntry1); > > > > > > Resource root = model.createResource("http://root";); > > > root.addProperty(model.createProperty("http://list";), > > > listEntry2); > > > > > > model.write(System.out, "N-TRIPLE"); > > > model.write(System.out, "Turtle"); > > > > > > The output looks like this: > > > > > > _:AX2dX96d4dd4X3aX12eda6ae213X3aXX2dX7fff<http://property> "test1" . > > > <http://root> <http://list> <http://test/2> . > > > <http://test/2> <http://www.w3.org/1999/02/22-rdf-syntax-ns#rest> > > > <http://test/1> . > > > <http://test/2> <http://www.w3.org/1999/02/22-rdf-syntax-ns#first> > > > _:AX2dX96d4dd4X3aX12eda6ae213X3aXX2dX7ffe . > > > _:AX2dX96d4dd4X3aX12eda6ae213X3aXX2dX7ffe<http://property> "test2" . > > > <http://test/1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#rest> > > > <http://www.w3.org/1999/02/22-rdf-syntax-ns#nil> . > > > <http://test/1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#first> > > > _:AX2dX96d4dd4X3aX12eda6ae213X3aXX2dX7fff . > > > <http://root> > > > <http://list> ([<http://property> "test2" > > > ] [<http://property> "test1" > > > ]) . > > > > > > The "N-TRIPLE" output is as expected, but the "Turtle" output has > > discarded > > > the entry URIs. > > > > > > Is this a bug? > > > > > > Thanks, > > > Frank >
