[
https://issues.apache.org/jira/browse/FELIX-4416?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13919065#comment-13919065
]
Uwe Barthel commented on FELIX-4416:
------------------------------------
Properties are always defined in ISO-8859-1.
See http://docs.oracle.com/javase/7/docs/api/java/util/Properties.html.
For use another Characterset instead of ISO-8859-1, the properties files has to
prepare with native2ascii tool provided by JDK.
> CFG file encoding
> -----------------
>
> Key: FELIX-4416
> URL: https://issues.apache.org/jira/browse/FELIX-4416
> Project: Felix
> Issue Type: Bug
> Components: File Install
> Affects Versions: fileinstall-3.2.6
> Environment: Karaf 2.3.3
> Reporter: Alexandre Cartapanis
>
> FileInstall always use ISO-8859-1 for reading cfg file, even if the CFG file
> is not encoded in ISO-8859-1.
> So when the CFG file contains non-ascii characters, those characters are
> "transformed" during parsing. Typically my file contains the line
> "error.message.210=Identifiant d'entité cible invalide" (without the "") but
> after FileInstall parsing, my string is transformed into "Identifiant
> d'entité cible invalide" (still without the "").
> The org.apache.felix.utils.properties.Properties class has a default encoding
> defined to ISO-8859-1,which is always used whatever the actual file encoding
> is.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
