[
https://issues.apache.org/jira/browse/JENA-1570?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16534661#comment-16534661
]
Andy Seaborne commented on JENA-1570:
-------------------------------------
It is the current behaviour and has been the behaviour for a long time.
Converting data as it comes into the system, at the same point that values like
integers would be canonicalized, is the current way to get lang tag
canonicalization.
Any change the community would like to see needs to take into account all
viewpoints, including the silent majority of existing deployed applications,
and also the fact that RDF itself implies a non-BCP-47 definition of canonical
language tag.
Jena is a general framework for semantic web applications. That means it can
not do exactly what every app wants in every detail; the apps need to work with
the general facilities to build what they want.
> Inconsistent handling of language tags in GraphMem
> --------------------------------------------------
>
> Key: JENA-1570
> URL: https://issues.apache.org/jira/browse/JENA-1570
> Project: Apache Jena
> Issue Type: Bug
> Components: Core
> Affects Versions: Jena 3.3.0
> Reporter: Patrycja Rewerska
> Priority: Minor
> Attachments: GraphMemTest.java
>
>
> We've found that the implementation of the GraphMem graph handles language
> tags cases in an inconsistent way.
> There is the unit test attached which shows the issue.
> It seems that the contains method is case insensitive for language tags but
> the insert and delete methods are case sensitive.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
