Scoring of a rep:similar search is not as expected
--------------------------------------------------
Key: JCR-1945
URL: https://issues.apache.org/jira/browse/JCR-1945
Project: Jackrabbit Content Repository
Issue Type: Bug
Components: jackrabbit-core, query
Reporter: Alexander Klimetschek
Priority: Minor
A normal rep:similar query, such as
//element(*, nt:hierarchyNode)[rep:similar(jcr:content,
'/content/some/path/jcr:content')]
will return a properly ordered result, with the most similar nodes at the top.
However, adding an order by @jcr:score
//element(*, nt:hierarchyNode)[rep:similar(jcr:content,
'/content/some/path/jcr:content')] order by @jcr:score desc
will return a list with not-so-well-matching nodes at the top. I think one
would expect the score value to depend on the similarity, ie. the more similar,
the higher the score, so the result should look just as the one without an
order by.
(tested only on a Jackrabbit 1.4.6)
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
