Mailing-List: contact commits-help@openejb.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@openejb.apache.org
Message-ID: <194331005.1252670938253.JavaMail.jira@brutus>
Date: Fri, 11 Sep 2009 05:08:58 -0700 (PDT)
From: "Taner Mansur (JIRA)" <jira@apache.org>
To: commits@openejb.apache.org
Subject: [jira] Created: (OPENEJB-1071) Turkish Locale Problem
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

Turkish Locale Problem
----------------------

                 Key: OPENEJB-1071
                 URL: https://issues.apache.org/jira/browse/OPENEJB-1071
             Project: OpenEJB
          Issue Type: Bug
    Affects Versions: 3.1.1
         Environment: Windows environments (Windows XP and Vista are tested=
) with Turkish Locale settings
            Reporter: Taner Mansur
            Priority: Minor


There is a bug in string transformation when writing to stdout, which conve=
rts uppercase characters to lowercase. It seems that the conversion is made=
 with a String.toLowerCase() function call which takes no arguments and use=
s the default system locale to determine local lowercase letters. In Turkis=
h, the lowercase equivalent of 'I' (U+0049) character is not 'i' (U+0069). =
It is '=C4=B1' (U+0131). So the following output is shown on the console wh=
ile OpenEJB initializes:

DEBUG - Using default 'openejb.nobanner=3Dtrue'
Apache OpenEJB 3.1.1    build: 20090530-06:18
http://openejb.apache.org/

*****

INFO - Beginning load: D:\ext\openejb\lib\xml-resolver-1.2.jar
DEBUG - Using default 'openejb.tempclassloader.skip=3Dnone'  Possible value=
s are: none, annotat=C4=B1ons, enums
INFO - Configuring enterprise application: classpath.ear
WARN - No application-client.xml found assuming annotations present: classp=
ath.ear, module: xml-resolver-1.2.jar
DEBUG - Using default 'openejb.descriptors.output=3Dfalse'
DEBUG - Using default 'openejb.validation.output.level=3Dmed=C4=B1um'.  Pos=
sible values are: terse, med=C4=B1um, verbose

If you can't see the lowercase 'I': it is like 'med(U+0131)um' in the outpu=
t which should be medium.=20
I believe this issue is closely related with the confirmed BCEL bug: 38787

https://issues.apache.org/bugzilla/show_bug.cgi?id=3D38787

It seems the BCEL function now in jre6/rt.jar : com/sun/org/apache/bcel/int=
ernal/util/InstructionFinder.compilePattern() (Line 149) still has

  pattern.toLowerCase();

--=20
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.