[jira] [Updated] (JCR-3968) Joins do not evaluate correctly

["Pascal Knüppel (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/JCR-3968?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Pascal Knüppel updated JCR-3968:
--------------------------------
    Attachment: jcrDataStructure.png

here is the datastructure that I am using

> Joins do not evaluate correctly
> -------------------------------
>
>                 Key: JCR-3968
>                 URL: https://issues.apache.org/jira/browse/JCR-3968
>             Project: Jackrabbit Content Repository
>          Issue Type: Bug
>          Components: query
>    Affects Versions: 2.12.1
>            Reporter: Pascal Knüppel
>         Attachments: jcrDataStructure.png
>
>
> I got a tree simple structure that represents my data and the webdav-interface is showing that my structure is correctly inserted into the repository. (I will append a graphstructure after the post)
> My problem now is that I want to query all my ${id}-nodes that match certain conditions. If I execute the following query I get several results:
> {noformat}
> SELECT * 
> FROM [nt:unstructured] AS receipt
> {noformat}
> results in my testversion in:
> {noformat}
> /
> /301234562/201604/999999996/front
> /301234562/201604/999999999/back
> /301234562
> /301234562/201604/999999996/back
> /301234562/201604
> /301234562/201604/999999999/front
> /301234562/201604/999999996
> /301234562/201604/999999999
> /301234561/201604/999999996/front
> /301234561/201604/999999999/back
> /301234561/201604/999999996/back
> /301234561
> /301234561/201604
> /301234561/201604/999999999/front
> /301234561/201604/999999996
> /301234561/201604/999999999
> {noformat}
> but if I now execute the following join condition I get an empty set of results
> {noformat}
> SELECT *
>  FROM [nt:unstructured] AS receipt
> INNER JOIN [nt:unstructured] AS child
> ON ISSAMENODE(child, receipt)
> {noformat}
> from JSR 283 I got the following:
> {noformat}
> Let *L x R* be the Cartesian product of *L* and *R* as a set of (m + n)-tuples
> {noformat}
> so this join should result in a set of 2-tuples containing exactly the same elements as the first query, or am I missing something obvious again?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)