[
https://issues.apache.org/jira/browse/OAK-2660?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14946861#comment-14946861
]
Davide Giannella commented on OAK-2660:
---------------------------------------
Gave a run at the test case with the features of OAK-1617 enabled and
if we force the computed UNION query it works.
In OAK-1617 the query executed, between the ORIGINAL and the
OPTIMISED/UNION depends on their cost. Therefore we're not sure what
version will be executed.
Two possible solutions:
- we come up with a specific oak query syntax that will force the
execution of the optimised query.
- we try to solve this problem at the root.
> wrong resultset multiple ORs, lucene index, full-text
> -----------------------------------------------------
>
> Key: OAK-2660
> URL: https://issues.apache.org/jira/browse/OAK-2660
> Project: Jackrabbit Oak
> Issue Type: Bug
> Components: query
> Affects Versions: 1.1.7
> Reporter: Davide Giannella
> Assignee: Davide Giannella
> Fix For: 1.3.8
>
>
> When executing a query like
> {code}
> SELECT *
> FROM [nt:unstructured] AS c
> WHERE ( c.[name] = 'yes'
> OR CONTAINS(c.[surname], 'yes')
> OR CONTAINS(c.[description], 'yes') )
> AND ISDESCENDANTNODE(c, '/content')
> ORDER BY added DESC
> {code}
> and a lucene property index is serving all the properties: {{name, surname,
> description, added}} the full index is returned and no extra condition is
> applied.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
