From david.nickerson at nus.edu.sg Fri Jan 5 04:21:10 2007 From: david.nickerson at nus.edu.sg (David Nickerson) Date: Fri Jan 5 04:25:53 2007 Subject: [redland-dev] librdf_model_get_arcs_out assertion failure Message-ID: <459DD236.7070805@nus.edu.sg> Hi all, I am using the librdf_model_get_arcs_out method (from C code) to iterate through all the arcs out of a given node in an RDF model. Whenever I call this method I get the following printed to stderr rdf_node.c:677: (librdf_new_node_from_node) assertion failed: object pointer of type librdf_node is NULL. I get this for both the Redland 1.0.5 I built myself and the FC5 binary packages (redland 1.0.5-1, rasqal 0.9.13-1, and raptor 1.4.13-1) from http://download.librdf.org/binaries/redhat/fc5/. I'm also getting the same behaviour when using rdfproc with the arcs-out command. Despite the assertion failure everything seems to work fine and I'm getting all the results (arcs) I expect from a given node - again in both my code and by using rdfproc. Has anyone else seen this behaviour? or have I missed something obvious? I'm new to redland (and RDF in general) and some pointers to where to start looking would be much appreciated. Thanks, David. -- David Nickerson, PhD Research Fellow Division of Bioengineering Faculty of Engineering National University of Singapore Email: david.nickerson@nus.edu.sg From Jeff.Dutton at stratus.com Fri Jan 12 04:07:53 2007 From: Jeff.Dutton at stratus.com (Dutton, Jeff) Date: Fri Jan 12 04:13:22 2007 Subject: [redland-dev] NULL string for librdf_query_results_to_string() on empty results? Message-ID: <1C68BCE03F80CD46A821B5B9C5F2163E020E7685@EXNA.corp.stratus.com> I am trying to implement a SPARQL endpoint webservice to a Redland database. When I do a query that finds no results, the stringification of the query results is NULL (undef in Perl). My understanding of the required XML output(http://www.w3.org/TR/rdf-sparql-XMLres/) for the results of an empty (but not malformed or otherwise errant) query should list the section of selected variables and an empty element. My simple SELECT * WHERE { ?s ?p ?o } query isn't getting any errors when I put some data in the database, but when there is no data (or when the WHERE clause doesn't match any data I get this bogus string, with no error or warning from Redland. I am using Redland v 1.0.4 with Rasqal 0.9.13. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.usefulinc.com/pipermail/redland-dev/attachments/20070111/ddca9a25/attachment.htm From Jagdev.Bhogal at uce.ac.uk Tue Jan 16 11:01:39 2007 From: Jagdev.Bhogal at uce.ac.uk (Jagdev Bhogal) Date: Tue Jan 16 11:08:15 2007 Subject: [redland-dev] rdql question Message-ID: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A49@exchangea.staff.uce.ac.uk> I am using rdfprint.c to parse an rdf file and output the set of resulting triples into a file (eg tripleout) I then want to use rdql to query the triples. In the synopsis given on librdf.org/rasqal/librasqal.html, query string has been set to "select * from http://example.org/data.rdf" How can I set up the query string so that it selects the data from my file of triples "tripleout"? Can I replace http... with a file name? Also which query language is better supported is it SPARQL or RDQL? I hope you can help with the above questions. Thank you Mrs Jagdev Bhogal Senior Lecturer in Database Systems Dept. of Computing University of Central England Feeney Building Perry Barr Birmingham B42 2SU Tel: 0121 331 7658 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.usefulinc.com/pipermail/redland-dev/attachments/20070116/a10ae872/attachment.htm From mendezster at gmail.com Wed Jan 24 02:07:30 2007 From: mendezster at gmail.com (=?ISO-8859-1?Q?Ricardo_J._M=E9ndez?=) Date: Wed Jan 24 02:08:10 2007 Subject: [redland-dev] Querying execute debug information Message-ID: <286cf0820701231807m5107dfcfoabf074fe88ed4431@mail.gmail.com> Hi, I've just begun experimenting with redland, modifying the original example.py provided in the bindings. Whenever I step through the results in for result in q.execute(model): print "{" for k in result: print " "+k+" = "+str(result[k]) print "}" the library prints out trace information like rasqal_engine.c:1319:rasqal_engine_do_step: Graph pattern 0 returned 1 rasqal_engine.c:1347:rasqal_engine_do_step: Got match rasqal_engine.c:1613:rasqal_engine_get_next_result: Returned step is got match rasqal_engine.c:1620:rasqal_engine_get_next_result: Solution binds 2 values rasqal_engine.c:1621:rasqal_engine_get_next_result: New bindings 2 rasqal_engine.c:1634:rasqal_engine_get_next_result: Ending with step got match and graph pattern 0 How can I turn this debug information off? RDF.debug(0) doesn't do the trick. Thanks in advance, Ricardo J. M?ndez http://ricardo.strangevistas.net/ From dave at dajobe.org Wed Jan 24 02:33:44 2007 From: dave at dajobe.org (Dave Beckett) Date: Wed Jan 24 02:34:19 2007 Subject: [redland-dev] librdf_model_get_arcs_out assertion failure In-Reply-To: <459DD236.7070805@nus.edu.sg> References: <459DD236.7070805@nus.edu.sg> Message-ID: <45B6C588.5050500@dajobe.org> David Nickerson wrote: > Hi all, > > I am using the librdf_model_get_arcs_out method (from C code) to iterate > through all the arcs out of a given node in an RDF model. Whenever I > call this method I get the following printed to stderr > > rdf_node.c:677: (librdf_new_node_from_node) assertion failed: object > pointer of type librdf_node is NULL. > > I get this for both the Redland 1.0.5 I built myself and the FC5 binary > packages (redland 1.0.5-1, rasqal 0.9.13-1, and raptor 1.4.13-1) from > http://download.librdf.org/binaries/redhat/fc5/. I'm also getting the > same behaviour when using rdfproc with the arcs-out command. > > Despite the assertion failure everything seems to work fine and I'm > getting all the results (arcs) I expect from a given node - again in > both my code and by using rdfproc. > > Has anyone else seen this behaviour? or have I missed something obvious? > I'm new to redland (and RDF in general) and some pointers to where to > start looking would be much appreciated. Yeah it's a bug with failing to allow NULLs, here's the patch: Index: rdf_storage.c =================================================================== --- rdf_storage.c (revision 11806) +++ rdf_storage.c (working copy) @@ -943,8 +943,10 @@ } - node1=librdf_new_node_from_node(node1); - node2=librdf_new_node_from_node(node2); + if(node1) + node1=librdf_new_node_from_node(node1); + if(node2) + node2=librdf_new_node_from_node(node2); switch(want) { case LIBRDF_STATEMENT_SUBJECT: Dave From dave at dajobe.org Wed Jan 24 03:20:36 2007 From: dave at dajobe.org (Dave Beckett) Date: Wed Jan 24 03:21:13 2007 Subject: [redland-dev] NULL string for librdf_query_results_to_string() on empty results? In-Reply-To: <1C68BCE03F80CD46A821B5B9C5F2163E020E7685@EXNA.corp.stratus.com> References: <1C68BCE03F80CD46A821B5B9C5F2163E020E7685@EXNA.corp.stratus.com> Message-ID: <45B6D084.5010000@dajobe.org> Dutton, Jeff wrote: > I am trying to implement a SPARQL endpoint webservice to a Redland > database. When I do a query that finds no results, the stringification > of the query results is NULL (undef in Perl). My understanding of the > required XML output(http://www.w3.org/TR/rdf-sparql-XMLres/) for the > results of an empty (but not malformed or otherwise errant) query should > list the section of selected variables and an empty > element. > > My simple SELECT * WHERE { ?s ?p ?o } query isn't getting any errors > when I put some data in the database, but when there is no data (or when > the WHERE clause doesn't match any data I get this bogus string, with no > error or warning from Redland. > > I am using Redland v 1.0.4 with Rasqal 0.9.13. You are correct, this is a bug. It's a 1 line fix (this is redland 1.0.4) --- rasqal/src/rasqal_query_results.c.orig 2006-04-29 23:38:19.000000000 -0700 +++ rasqal/src/rasqal_query_results.c 2007-01-23 19:18:31.000000000 -0800 @@ -737,7 +737,7 @@ raptor_uri *format_uri, raptor_uri *base_uri) { - if(!results || results->failed || results->finished) + if(!results || results->failed) return 1; /* Dave From dave at dajobe.org Wed Jan 24 04:13:58 2007 From: dave at dajobe.org (Dave Beckett) Date: Wed Jan 24 04:14:31 2007 Subject: [redland-dev] rdql question In-Reply-To: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A49@exchangea.staff.uce.ac.uk> References: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A49@exchangea.staff.uce.ac.uk> Message-ID: <45B6DD06.9020408@dajobe.org> Jagdev Bhogal wrote: > I am using rdfprint.c to parse an rdf file and output the set of > resulting triples into a file (eg tripleout) > > I then want to use rdql to query the triples. > > In the synopsis given on librdf.org/rasqal/librasqal.html, query string > has been set to ?select * from http://example.org/data.rdf? > > How can I set up the query string so that it selects the data from my > file of triples ?tripleout?? Can I replace http? with a file name? Yes, use a URI like file:tripleout or file://absolute/path/to/tripleout > Also which query language is better supported is it SPARQL or RDQL? RDQL is completely implemented. SPARQL is still a work in progress since it's a more complex query language, but it is the future :) Dave From mendezster at gmail.com Wed Jan 24 06:13:55 2007 From: mendezster at gmail.com (=?ISO-8859-1?Q?Ricardo_J._M=E9ndez?=) Date: Wed Jan 24 06:14:38 2007 Subject: [redland-dev] Re: Querying execute debug information In-Reply-To: <286cf0820701231807m5107dfcfoabf074fe88ed4431@mail.gmail.com> References: <286cf0820701231807m5107dfcfoabf074fe88ed4431@mail.gmail.com> Message-ID: <286cf0820701232213saa0e29elc957b81d8142cecf@mail.gmail.com> Hello again, Bad form replying to yourself, but I've figured it out using the source - the executables were built with --enable-maintainer-mode. Does anybody have any objections to moving logging to log4cxx in the future ( http://logging.apache.org/log4cxx/ )? I thought about doing it now, but I think it might be better to wait until 0.9.8 is released. Ricardo J. M?ndez http://ricardo.strangevistas.net/ From mendezster at gmail.com Thu Jan 25 02:13:14 2007 From: mendezster at gmail.com (=?ISO-8859-1?Q?Ricardo_J._M=E9ndez?=) Date: Thu Jan 25 02:14:01 2007 Subject: [redland-dev] Default serializer not saving rdf:about Message-ID: <286cf0820701241813t755c0674x673fbe31afcc7afe@mail.gmail.com> Hi, I've run into an interesting problem with Redland. I'm using raptor-1.4.13 redland-1.0.5 rasqal-0.9.13 When attempting serialization of a model using the default serializator, as serializer=RDF.Serializer() serializer.set_namespace("dc", RDF.Uri("http://xmlns.com/foaf/0.1/")) serializer.set_namespace("rdf", RDF.Uri("http://www.w3.org/1999/02/22-rdf-syntax-ns#")) serializer.serialize_model_to_file(out_file, model) the resulting output does not contain the rdf:about tags for any of the nodes. Instead, they look like The Name However, pretty much any other serializer includes it, like The Name if it's rdfxml or rdfxml-abbrev, or as an element if it's ntriples. Two questions: 1) Is this behavior a bug? 2) Which one is the default serializer? I've serialized data using RDF.Serializer(), RDF.Serializer("rdfxml"), RDF.Serializer("rdfxml-abbrev") and RDF.Serializer("ntriples"), and while the default one matches rdfxml-abbrev the closest, the output isn't exactly the same. I've got some test code if anyone considers it necessary. Thanks in advance, Ricardo J. M?ndez http://ricardo.strangevistas.net/ From Jagdev.Bhogal at uce.ac.uk Thu Jan 25 14:58:34 2007 From: Jagdev.Bhogal at uce.ac.uk (Jagdev Bhogal) Date: Thu Jan 25 14:59:58 2007 Subject: [redland-dev] FW: segmentation fault References: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A53@exchangea.staff.uce.ac.uk> Message-ID: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A54@exchangea.staff.uce.ac.uk> SGkNCldoZW4gSSBydW4gdGhlIGZvbGxvd2luZyBjb2RlIEkgZ2V0IGEgc2VnbWVudGF0aW9uIGZh dWx0LiBJZiB0aGVyZSBpcyBhbnlvbmUgd2hvIHVuZGVyc3RhbmRzIHdoeSBJIGFtIGdldHRpbmcg dGhpcyBlcnJvciBhbmQgaG93IHRvIHJlc29sdmUgaXQgSSB3b3VsZCBiZSB2ZXJ5IGdyYXRlZnVs Lg0KIA0KVGhlIGNvZGUgaXMgYXMgZm9sbG93cyAoYW5kIHNlZW1zIHRvIGNyYXNoIGF0IHJhc3Fh bF9xdWVyeV9wcmVwYXJlKQ0KIA0KICNpbmNsdWRlIDxzdGRpby5oPg0KI2luY2x1ZGUgPHJhcHRv ci5oPg0KI2luY2x1ZGUgPHJhc3FhbC5oPg0KbWFpbigpDQp7DQogIGludCBpOw0KICByYXNxYWxf cXVlcnlfcmVzdWx0cyAqcmVzdWx0czsNCiAgdW5zaWduZWQgY2hhciAqdXJpX3N0cmluZzsNCiAg cmFzcWFsX3F1ZXJ5ICpycT1yYXNxYWxfbmV3X3F1ZXJ5KCJyZHFsIixOVUxMKTsNCiAgcmFwdG9y X2lvc3RyZWFtICpyYXB0b3Jfb3V0cHV0Ow0KICBjb25zdCBjaGFyICpxdWVyeV9zdHJpbmc9InNl bGVjdCAqIGZyb20gPGZpbGU6Ly9leHBvcnQvQ0lTUi1EL2phZ2Rldi9va2FwaS9wcm9ncy9uZXdz dHJpcGxlcz4iOw0KICByYXNxYWxfaW5pdCgpOw0KICByYXB0b3JfaW5pdCgpOw0KICByYXNxYWxf cXVlcnlfcHJlcGFyZShycSxxdWVyeV9zdHJpbmcsIE5VTEwpOw0KICByZXN1bHRzPXJhc3FhbF9x dWVyeV9leGVjdXRlKHJxKTsNCiAgd2hpbGUoIXJhc3FhbF9xdWVyeV9yZXN1bHRzX2ZpbmlzaGVk KHJlc3VsdHMpKSB7DQogICAgICAgIGZvcihpPTA7aTxyYXNxYWxfcXVlcnlfcmVzdWx0c19nZXRf YmluZGluZ3NfY291bnQocmVzdWx0cyk7aSsrKSB7DQogICAgICAgY29uc3QgY2hhciAqbmFtZT1y YXNxYWxfcXVlcnlfcmVzdWx0c19nZXRfYmluZGluZ19uYW1lKHJlc3VsdHMsaSk7DQogICAgICAg ICAgICAgICAgcmFzcWFsX2xpdGVyYWwgKnZhbHVlPXJhc3FhbF9xdWVyeV9yZXN1bHRzX2dldF9i aW5kaW5nX3ZhbHVlKHJlc3VsdHMsaSk7DQogICAgICAgICAgICAgICAgZnByaW50ZihzdGRlcnIs ImNvdW50IFxuIik7DQogICAgICAgICAgICAgICAgcmFzcWFsX3F1ZXJ5X3Jlc3VsdHNfd3JpdGUo cmFwdG9yX291dHB1dCxyZXN1bHRzLE5VTEwsTlVMTCk7DQogICAgICAgIH0NCiAgICAgICAgcmFz cWFsX3F1ZXJ5X3Jlc3VsdHNfbmV4dChyZXN1bHRzKTsNCiAgfQ0KICByYXNxYWxfZnJlZV9xdWVy eV9yZXN1bHRzKHJlc3VsdHMpOw0KICByYXNxYWxfZnJlZV9xdWVyeShycSk7DQogIHJhc3FhbF9m aW5pc2goKTsNCiAgcmFwdG9yX2ZpbmlzaCgpOw0KfQ0KIA0KQXMgdGhlIGRpcmVjdG9yeSBsaXN0 aW5nIHNob3dzLCB0aGUgZmlsZSBuZXdzdHJpcGxlcyBkb2VzIGV4aXN0Og0KIA0KL2V4cG9ydC9D SVNSLUQvamFnZGV2L29rYXBpL3Byb2dzJSBscyAtbA0KdG90YWwgMzQ4NTYNCg0KLXJ3eHJ3eHJ3 eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDQ4MzY2IERlYyAxNSAxNDoxMCBuZXdzLnJkZg0K LXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDk3ODY5IERlYyAxNSAxNDoxNyBu ZXdzdHJpcGxlcw0KLXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDQ4MzY2IERl YyAxNSAxNDowNSBuZXdzLnhtbA0KLXJ3eHJ3eHIteCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAg IDE2MTg1IEphbiAyNSAxMzowOCBxdWVyeXJkZg0KLXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2 MDc3ICAgICAgICAxMDg1IEphbiAyNSAxMzowNCBxdWVyeXJkZi5jDQogDQpJZiBhbnlvbmUgY2Fu IHBvaW50IG1lIGluIHRoZSByaWdodCBkaXJlY3Rpb24gSSB3b3VsZCBiZSB2ZXJ5IGdyYXRlZnVs Lg0KIA0KVGhhbmtzDQogDQpKYWdkZXYgQmhvZ2FsDQoNCg0KLS0tLS0tLS0tLS0tLS0gbmV4dCBw YXJ0IC0tLS0tLS0tLS0tLS0tCkFuIEhUTUwgYXR0YWNobWVudCB3YXMgc2NydWJiZWQuLi4KVVJM OiBodHRwOi8vbGlzdHMudXNlZnVsaW5jLmNvbS9waXBlcm1haWwvcmVkbGFuZC1kZXYvYXR0YWNo bWVudHMvMjAwNzAxMjUvM2YzYzAwOTkvYXR0YWNobWVudC5odG0K From Jagdev.Bhogal at uce.ac.uk Thu Jan 25 14:55:52 2007 From: Jagdev.Bhogal at uce.ac.uk (Jagdev Bhogal) Date: Thu Jan 25 15:17:21 2007 Subject: [redland-dev] segmentation fault Message-ID: <294B4B3243E76C4BA4FF7F54003B3BE1014A6A53@exchangea.staff.uce.ac.uk> SGkNCldoZW4gSSBydW4gdGhlIGZvbGxvd2luZyBjb2RlIEkgZ2V0IGEgc2VnbWVudGF0aW9uIGZh dWx0LiBJZiB0aGVyZSBpcyBhbnlvbmUgd2hvIHVuZGVyc3RhbmRzIHdoeSBJIGFtIGdldHRpbmcg dGhpcyBlcnJvciBhbmQgaG93IHRvIHJlc29sdmUgaXQgSSB3b3VsZCBiZSB2ZXJ5IGdyYXRlZnVs Lg0KIA0KVGhlIGNvZGUgaXMgYXMgZm9sbG93cyAoYW5kIHNlZW1zIHRvIGNyYXNoIGF0IHJhc3Fh bF9xdWVyeV9wcmVwYXJlKQ0KIA0KICNpbmNsdWRlIDxzdGRpby5oPg0KI2luY2x1ZGUgPHJhcHRv ci5oPg0KI2luY2x1ZGUgPHJhc3FhbC5oPg0KbWFpbigpDQp7DQogIGludCBpOw0KICByYXNxYWxf cXVlcnlfcmVzdWx0cyAqcmVzdWx0czsNCiAgdW5zaWduZWQgY2hhciAqdXJpX3N0cmluZzsNCiAg cmFzcWFsX3F1ZXJ5ICpycT1yYXNxYWxfbmV3X3F1ZXJ5KCJyZHFsIixOVUxMKTsNCiAgcmFwdG9y X2lvc3RyZWFtICpyYXB0b3Jfb3V0cHV0Ow0KICBjb25zdCBjaGFyICpxdWVyeV9zdHJpbmc9InNl bGVjdCAqIGZyb20gPGZpbGU6Ly9leHBvcnQvQ0lTUi1EL2phZ2Rldi9va2FwaS9wcm9ncy9uZXdz dHJpcGxlcz4iOw0KICByYXNxYWxfaW5pdCgpOw0KICByYXB0b3JfaW5pdCgpOw0KICByYXNxYWxf cXVlcnlfcHJlcGFyZShycSxxdWVyeV9zdHJpbmcsIE5VTEwpOw0KICByZXN1bHRzPXJhc3FhbF9x dWVyeV9leGVjdXRlKHJxKTsNCiAgd2hpbGUoIXJhc3FhbF9xdWVyeV9yZXN1bHRzX2ZpbmlzaGVk KHJlc3VsdHMpKSB7DQogICAgICAgIGZvcihpPTA7aTxyYXNxYWxfcXVlcnlfcmVzdWx0c19nZXRf YmluZGluZ3NfY291bnQocmVzdWx0cyk7aSsrKSB7DQogICAgICAgY29uc3QgY2hhciAqbmFtZT1y YXNxYWxfcXVlcnlfcmVzdWx0c19nZXRfYmluZGluZ19uYW1lKHJlc3VsdHMsaSk7DQogICAgICAg ICAgICAgICAgcmFzcWFsX2xpdGVyYWwgKnZhbHVlPXJhc3FhbF9xdWVyeV9yZXN1bHRzX2dldF9i aW5kaW5nX3ZhbHVlKHJlc3VsdHMsaSk7DQogICAgICAgICAgICAgICAgZnByaW50ZihzdGRlcnIs ImNvdW50IFxuIik7DQogICAgICAgICAgICAgICAgcmFzcWFsX3F1ZXJ5X3Jlc3VsdHNfd3JpdGUo cmFwdG9yX291dHB1dCxyZXN1bHRzLE5VTEwsTlVMTCk7DQogICAgICAgIH0NCiAgICAgICAgcmFz cWFsX3F1ZXJ5X3Jlc3VsdHNfbmV4dChyZXN1bHRzKTsNCiAgfQ0KICByYXNxYWxfZnJlZV9xdWVy eV9yZXN1bHRzKHJlc3VsdHMpOw0KICByYXNxYWxfZnJlZV9xdWVyeShycSk7DQogIHJhc3FhbF9m aW5pc2goKTsNCiAgcmFwdG9yX2ZpbmlzaCgpOw0KfQ0KIA0KQXMgdGhlIGRpcmVjdG9yeSBsaXN0 aW5nIHNob3dzLCB0aGUgZmlsZSBuZXdzdHJpcGxlcyBkb2VzIGV4aXN0Og0KIA0KL2V4cG9ydC9D SVNSLUQvamFnZGV2L29rYXBpL3Byb2dzJSBscyAtbA0KdG90YWwgMzQ4NTYNCg0KLXJ3eHJ3eHJ3 eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDQ4MzY2IERlYyAxNSAxNDoxMCBuZXdzLnJkZg0K LXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDk3ODY5IERlYyAxNSAxNDoxNyBu ZXdzdHJpcGxlcw0KLXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAgIDQ4MzY2IERl YyAxNSAxNDowNSBuZXdzLnhtbA0KLXJ3eHJ3eHIteCAgICAxIGFkODkzICAgIDI2MDc3ICAgICAg IDE2MTg1IEphbiAyNSAxMzowOCBxdWVyeXJkZg0KLXJ3eHJ3eHJ3eCAgICAxIGFkODkzICAgIDI2 MDc3ICAgICAgICAxMDg1IEphbiAyNSAxMzowNCBxdWVyeXJkZi5jDQogDQpJZiBhbnlvbmUgY2Fu IHBvaW50IG1lIGluIHRoZSByaWdodCBkaXJlY3Rpb24gSSB3b3VsZCBiZSB2ZXJ5IGdyYXRlZnVs Lg0KIA0KVGhhbmtzDQogDQpKYWdkZXYgQmhvZ2FsDQoNCg0KLS0tLS0tLS0tLS0tLS0gbmV4dCBw YXJ0IC0tLS0tLS0tLS0tLS0tCkFuIEhUTUwgYXR0YWNobWVudCB3YXMgc2NydWJiZWQuLi4KVVJM OiBodHRwOi8vbGlzdHMudXNlZnVsaW5jLmNvbS9waXBlcm1haWwvcmVkbGFuZC1kZXYvYXR0YWNo bWVudHMvMjAwNzAxMjUvNzAzNjRiODIvYXR0YWNobWVudC5odG0K From dave at dajobe.org Mon Jan 29 00:44:24 2007 From: dave at dajobe.org (Dave Beckett) Date: Mon Jan 29 00:45:48 2007 Subject: [redland-dev] Re: Querying execute debug information In-Reply-To: <286cf0820701232213saa0e29elc957b81d8142cecf@mail.gmail.com> References: <286cf0820701231807m5107dfcfoabf074fe88ed4431@mail.gmail.com> <286cf0820701232213saa0e29elc957b81d8142cecf@mail.gmail.com> Message-ID: <45BD4368.9060903@dajobe.org> Ricardo J. M?ndez wrote: > Hello again, > > Bad form replying to yourself, but I've figured it out using the > source - the executables were built with --enable-maintainer-mode. > > Does anybody have any objections to moving logging to log4cxx in the > future ( http://logging.apache.org/log4cxx/ )? I thought about doing > it now, but I think it might be better to wait until 0.9.8 is > released. I wouldn't want to keep that compiled in all the time. Also, it's C++ and rasqal (raptor, redland) are all in C. Dave From dave at dajobe.org Mon Jan 29 00:51:19 2007 From: dave at dajobe.org (Dave Beckett) Date: Mon Jan 29 00:52:39 2007 Subject: [redland-dev] Default serializer not saving rdf:about In-Reply-To: <286cf0820701241813t755c0674x673fbe31afcc7afe@mail.gmail.com> References: <286cf0820701241813t755c0674x673fbe31afcc7afe@mail.gmail.com> Message-ID: <45BD4507.8050705@dajobe.org> Ricardo J. M?ndez wrote: > Hi, > > I've run into an interesting problem with Redland. I'm using > > raptor-1.4.13 > redland-1.0.5 > rasqal-0.9.13 > > When attempting serialization of a model using the default serializator, as > > serializer=RDF.Serializer() > serializer.set_namespace("dc", RDF.Uri("http://xmlns.com/foaf/0.1/")) > serializer.set_namespace("rdf", > RDF.Uri("http://www.w3.org/1999/02/22-rdf-syntax-ns#")) > serializer.serialize_model_to_file(out_file, model) > > the resulting output does not contain the rdf:about tags for any of > the nodes. Instead, they look like > > > The Name That's the relative URI handling. The serializers by default try to generate the shortest relative URI if they can, or are allowed to (not in ntriples, it's all absolute). "" is the document URI, in this case , the filename for 'out_file' in your example, something like file://path/to/file/ > However, pretty much any other serializer includes it, like > > > The Name > > if it's rdfxml or rdfxml-abbrev, or as an element if it's ntriples. > > Two questions: > > 1) Is this behavior a bug? Feature. Actually you can turn this off with a serializer feature called relativeURIs Write relative URIs wherever possible in serializing. (from rapper -f help) Feature URI: http://feature.librdf.org/raptor-relativeURIs > 2) Which one is the default serializer? I've serialized data using > RDF.Serializer(), RDF.Serializer("rdfxml"), > RDF.Serializer("rdfxml-abbrev") and RDF.Serializer("ntriples"), and > while the default one matches rdfxml-abbrev the closest, the output > isn't exactly the same. > > I've got some test code if anyone considers it necessary. The default varies, it's usually RDF/XML (rdfxml) although RDF/XML Abbrev (rdfxml-abbrev) is a friendlier choice for people. again: rapper -h shows the names Dave From benno at iri.columbia.edu Mon Jan 29 20:04:56 2007 From: benno at iri.columbia.edu (Benno Blumenthal) Date: Mon Jan 29 20:06:22 2007 Subject: [redland-dev] ruby bindings and typed literals In-Reply-To: <20060927163108.GC7781@eyaloren.org> References: <451A3E34.9040700@cs.pdx.edu> <20060927163108.GC7781@eyaloren.org> Message-ID: <45BE5368.9010705@iri.columbia.edu> Hello All, I am interested in modifying ActiveRDF so that typed literals are handled properly. In the Redland Ruby bindings, there is a literal subclass of node which has a lang and value properties. And it is possible to define a literal node with the datatype property set correctly. But I cannot see how to get the datatype property back out with the ruby bindings. So what happens to numeric/date/XML literals in the ruby bindings? Are they already converted to Ruby datatypes, or am I missing something? -- Dr. M. Benno Blumenthal benno@iri.columbia.edu International Research Institute for climate and society The Earth Institute at Columbia University Lamont Campus, Palisades NY 10964-8000 (845) 680-4450