Michiel; One suggestion is to refactor your query a bit. Any update on SPO can
have strange effects, so I'd suggest using FILTER to ensure that the update
operations occur only on the intended triples:
delete {
?s ?p ?o .
}
insert {
?s_new ?p ?o_new .
}
where {
?s ?p ?o .
BIND(IRI(REPLACE(str(?s), "mySource", "mySourceX")) AS ?s_new)
BIND(IRI(REPLACE(str(?o), "myTarget", "myTargetY")) AS ?o_new)
FILTER (isIRI(?o) && CONTAINS(str(?s), "mySource") && CONTAINS(str(?o),
"myTarget"))
FILTER (?s != <http://marklogic.com/todel3>)
}
Note also there isn't a need to test whether the subject is an IRI – it must be
in RDF.
Try that and the original triple should get deleted.
— Scott
Scott Henninger
Senior Consultant, NAC Central
MarkLogic Corporation
[email protected]<[email protected]>
www.marklogic.com<http://www.marklogic.com/>
Hi,
I have a question of the following SPARQL query:
-----------------------------------------------------
delete {
?s ?p ?o .
}
insert {
?s_new ?p ?o_new .
}
where {
?s ?p ?o .
bind (if(isIRI(?s),IRI(replace(str(?s),"mySource","mySourceX")),?s) as
?s_new) .
bind (if(isIRI(?o),IRI(replace(str(?o),"myTarget","myTargetY")),?o) as
?o_new) .
}
-----------------------------------------------------
and the following triple:
<mySource> <myRelation> <myTarget>
If I execute this query in MarkLogic, the result is 2 triples:
<mySource> <myRelation> <myTarget>
<mySourceX> <myRelation> <myTargetY>
where I expected only one:
<mySourceX> <myRelation> <myTargetY>
Kind regards,
Michiel Vanhoutte
Wolters Kluwer Belgium
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://developer.marklogic.com/pipermail/general/attachments/20160309/60f9f381/attachment.html
_______________________________________________
General mailing list
[email protected]
Manage your subscription at:
http://developer.marklogic.com/mailman/listinfo/general
