[
https://issues.apache.org/jira/browse/JENA-1280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16458038#comment-16458038
]
ASF subversion and git services commented on JENA-1280:
-------------------------------------------------------
Commit 1bd5fdcf3ac0694e0bbabd2f8e63deb92d3c2bff in jena's branch
refs/heads/master from [~andy.seaborne]
[ https://git-wip-us.apache.org/repos/asf?p=jena.git;h=1bd5fdc ]
JENA-1534: Test for filter-only variables in EXISTS.
Tighten the classifer, assumes proper use of (sequence)
Relates to JENA-1167, JENA-1280.
> JoinClassifer is too cautious with EXISTS variables.
> ----------------------------------------------------
>
> Key: JENA-1280
> URL: https://issues.apache.org/jira/browse/JENA-1280
> Project: Apache Jena
> Issue Type: Bug
> Components: ARQ
> Affects Versions: Jena 3.1.1
> Reporter: Andy Seaborne
> Assignee: Andy Seaborne
> Priority: Major
> Fix For: Jena 3.2.0
>
>
> This query
> {noformat}
> SELECT * {
> :s :p ?g # ?g
> GRAPH ?g { # ?g
> :s1 :p1 ?x
> FILTER EXISTS { ?s ?p ?o }
> }
> }
> {noformat}
> is not executed as a {{(sequence)}} because there are variables in the
> {{FILTER}} that appear to be not bound: {{?s}} etc in {{EXISTS}}. If it were
> a function, this would be correct but {{EXISTS}} is different.
> Variables only in an {{EXISTS}} BGP patterns should not contribute to the
> unsafe variables of a {{FILTER}}.
> This query:
> {noformat}
> SELECT * {
> ?s :p ?g # :s -> ?s
> GRAPH ?g {
> :s1 :p1 ?x
> FILTER EXISTS { ?s ?p ?o }
> }
> }
> {noformat}
> is not safe to allow {{?s}} to substitute. This is correctly detected in
> Jena 3.1.0.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
