Hello all
I have a d2r-server (a joseki server providing a sparql endpoint using
d2rq, a library that translates sparql to sql) and noticed a strange effect
when issuing a certain query:
SELECT DISTINCT *
WHERE
{
SERVICE <http://0.0.0.0:2020/sparql> { ?document
<http://www.cipe.accamargo.org.br/ontologias/h2tc.owl#documentType>
"exam request" . }
SERVICE <http://0.0.0.0:2020/sparql> { ?document
<http://www.cipe.accamargo.org.br/ontologias/h2tc.owl#documentAuthor>
?author . }
}
[ This query is part of a bigger one that refers to several d2r-server
instances, and I know that I could join both SERVICEs in one, or even query
the service directly but that's not the point. ]
0.0.0.0:2020/sparql points to a d2r-server instance. This query was issued
not directly to this instance, but rather by a command line ARQ
2.10.1-snapshot over an empty graph.
The problem is, instead of getting proper URIs for ?author, for some
lines I'm getting those huge hex data, like
<687474703A2F2F7777772E636970652E616363616D6172676F2E6F72672E62722F6F6E746F6C6F676961732F683274632E6F776C237573657273554E495636363837353234353636>
Interestingly enough, converting consecutive pairs of those bytes to chars,
it turns into a URI!
http://www.cipe.accamargo.org.br/ontologias/h2tc.owl#usersUNIV6687524566
Which is a valid answer to ?author.
I have verified that a query without those SERVICE keywords works
perfectly, when issued directly to D2R-server. Or, if I join the two
SERVICEs in one, it also works ok.
*jena version (at D2R server)*
joseki-3.4.4
jena-arq-2.9.0-incubating
jena-core-2.7.0-incubating
jena-iri-9.9.0-incubating
Any help in order to get things straight will be appreciated.
Thanks,
--
diogo patrão
