jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jukka Zitting (JIRA)" <j...@apache.org>
Subject [jira] Updated: (JCR-1463) TCK: testRestore assumes too much about OPV=VERSION
Date Mon, 21 Apr 2008 23:29:23 GMT

     [ https://issues.apache.org/jira/browse/JCR-1463?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Jukka Zitting updated JCR-1463:

      Component/s:     (was: jackrabbit-core)
    Fix Version/s:     (was: core 1.4.3)
          Summary: TCK: testRestore assumes too much about OPV=VERSION  (was: Core: Failing
version tests (testRestore....))

Agreed with Toby, reverted the 1.4 merge in revision 650308.

I dug a bit deeper on this. It seems that the root of the problem is that the test case assumes
that restore() will restore an OPV=VERSION child node to the specific version it was in when
the parent node was checked in. This is not (always) true in Jackrabbit and not required by
the JSR 170 spec.

The spec merely says the following about such a child node (c):

    If the workspace does not have an instance of C then one is
    restored from C's version history. The workspace in which the
    restore is being performed will determine which particular version
    of C will be restored. This determination depends on the
    configuration of the workspace and is outside the scope of this

Jackrabbit by default uses the DateVersionSelector class that often makes it look like the
exact child version information is kept around in the frozen node, but that is not the case
as evidenced by the reported same-millisecond versions causing the DateVersionSelector to
select an "unexpected" child node version. Note that this behaviour is perfectly acceptable,
and there is no reason why child node could not for example always be restored to the latest
available version, regardless of the version of the parent node.

Also, if the versioned child node already exists in the workspace, restoring the parent node
to some version should not (or at least may not) affect the version of the child node.

Re-categorizing this as a TCK issue. We might also want to revert the changes to jackrabbit-core
in trunk.

> TCK: testRestore assumes too much about OPV=VERSION
> ---------------------------------------------------
>                 Key: JCR-1463
>                 URL: https://issues.apache.org/jira/browse/JCR-1463
>             Project: Jackrabbit
>          Issue Type: Bug
>          Components: jackrabbit-jcr-tests
>            Reporter: angela
>            Assignee: Tobias Bocanegra
>         Attachments: TEST-org.apache.jackrabbit.core.integration.JCRAPITest.xml
> i run into failing restore test in the core project while trying to reproduce JCR-1293.
> since that i had similar (or the same) tests failing more than once. therefore i guess
this is really caused by a bug in the restore code or in the corresponding tests and not an
> the next time i get it, i will attach the surefire report.
> regards
> angela

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message