[hibernate-issues] [Hibernate-JIRA] Commented: (HSEARCH-309) PurgeAllLuceneWork duplicates in work queue

Sanne Grinovero (JIRA) noreply at atlassian.com
Wed Nov 26 07:04:16 EST 2008


    [ http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-309?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=31822#action_31822 ] 

Sanne Grinovero commented on HSEARCH-309:
-----------------------------------------

I am more worried about the "subclasses" variable containing the class itself too;
IMHO the purgeAll is the only one behaving out-of-spec but is innocent...
we should either rename the "subclasses" variable or remove the class from
it and fix the rest of the code.

I imagine more code going to be written getting confused by the current naming,
or some other more critical bug we don't know about yet.

Are you sure you want to remove that? it's looking ok, we need to remove subclasses too,
but I don't know what you did about the "polymorphic query stuff".

> PurgeAllLuceneWork duplicates in work queue
> -------------------------------------------
>
>                 Key: HSEARCH-309
>                 URL: http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-309
>             Project: Hibernate Search
>          Issue Type: Bug
>          Components: engine
>    Affects Versions: 3.1.0.CR1
>            Reporter: Sanne Grinovero
>             Fix For: 3.1.0
>
>
> when using purgeAll( typeA ); you get two works to purge typeA in the queue, +1 for each subclass.
> The problem is caused by the presence of each type in the set of it's subtypes in DocumentBuilderContainedEntity.subclasses 
> I tried having the typeA removed from the subclasses set, but this breaks many other features (junit told).
> It appears some other code is relying on the fact that this set has to contain the super class too; this looks strange?
> so the easy fix would be to change the current purgeAll() to not purge on the root entity too, as doing it on subclasses is enough;
> but it actually feels to me this is the only correct method, and we should check other usages.
> I've seen this affect only purgeAll but could be more critical, so priority must be investigated.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        



More information about the hibernate-issues mailing list