uima-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Klügl (JIRA) <...@uima.apache.org>
Subject [jira] [Commented] (UIMA-6057) Avoid falsely switching classloader
Date Tue, 09 Jul 2019 06:46:00 GMT

    [ https://issues.apache.org/jira/browse/UIMA-6057?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16880980#comment-16880980
] 

Peter Klügl commented on UIMA-6057:
-----------------------------------

Thanks Marshall, we will test it.

Some answers/comments to your previous comment:

-> 1
I have not thought about that the CAS will be unlocked. I would be good to prevent that, but
IMHO that a different level of issue. You cannot really prevent an annotator to reset the
CAS during process() (a colleague of me did that once) and it rather won't be by accident
if the annotator explicitly calls cas.reset().

-> 2
In our pipelines all components but the PEARs have the same merged type system. I think only
the result specification of AE1 should count.

-> 3 
I do not understand this issue.

-> 4
I forgot about that. What about the time of an aggregate AE? Is it the sum of the delegated
AEs?
 For me, it would be ok the count it twice, but that is of course not true for everyone else.

-> 5

-> 6
I do not see the problem. In any case the creator of AE1 and AE2 need to take care that the
combination works.




> Avoid falsely switching classloader
> -----------------------------------
>
>                 Key: UIMA-6057
>                 URL: https://issues.apache.org/jira/browse/UIMA-6057
>             Project: UIMA
>          Issue Type: Bug
>          Components: Core Java Framework
>            Reporter: Matthias Koch
>            Priority: Major
>         Attachments: UIMA-6057.diff, classloadertest.zip
>
>
> In some cases the classloader is switched back, although it hasn't be switched before
processing.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message