[
https://issues.apache.org/jira/browse/JENA-1187?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15309050#comment-15309050
]
Eetu Mäkelä commented on JENA-1187:
-----------------------------------
Okay, so the answer to my primary question seems to be that I was relying on a
bug, where the actual linear order that the query was being run by leaked out
instead of staying inside the pure bottom-up SPARQL algebra interpretation.
(Fixing which is a bit of a shame for me, because it was really useful to be
able to use unioned blocks just as alternates 'inside' a graph pattern, and now
I'll be forced to copy a lot of VALUES and binds and filters to each of parts
of the unions.)
There's still something funny going on though too, as based on the generative
interpretation, shouldn't
{noformat}
SELECT *
WHERE
{ { BIND("nonexistant" AS ?cl)
{ BIND(?cl AS ?cl2)
?c a ?cl
}
UNION
{ BIND(?cl AS ?cl2)
?c a ?cl2
}
}
}
{noformat}
still produce all rdf:type statements inside a dataset from the second part of
the union? Because it doesn't, but results in an empty result set (where your
reduced example or my other example does produce them).
> Wrong results/performance regression when using BIND and graph pattern groups
> -----------------------------------------------------------------------------
>
> Key: JENA-1187
> URL: https://issues.apache.org/jira/browse/JENA-1187
> Project: Apache Jena
> Issue Type: Bug
> Components: ARQ
> Affects Versions: Jena 3.1.0, Fuseki 2.4.0
> Reporter: Eetu Mäkelä
>
> I've been using SPARQL queries with BINDs and/or VALUESs that precede UNION
> blocks. These used to work efficiently (at least in Jena 2 times) with the
> bound value being bound also in the subpattern, but no longer do. In
> addition, at certain times, they produce nonsensical results. For example,
> the below returns {{rdf:type}} statements in the dataset without any regard
> to ?cl:
> {code}
> SELECT * {
> {
> BIND("nonexistant" AS ?cl)
> {
> BIND(?cl AS ?cl2)
> ?c a ?cl2 .
> } UNION {
> BIND(?cl AS ?cl2)
> ?c a ?cl2 .
> }
> }
> }
> LIMIT 10
> {code}
> while if you change _just one_ of the subpatterns in the union to directly
> refer to ?cl, it returns an empty result set:
> {code}
> SELECT * {
> {
> BIND("nonexistant" AS ?cl)
> {
> BIND(?cl AS ?cl2)
> ?c a ?cl .
> } UNION {
> BIND(?cl AS ?cl2)
> ?c a ?cl2 .
> }
> }
> }
> LIMIT 10
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
