Bug 31084

Summary: fmt charset problem when data come from database (mysql 4.0.x)
Product: Taglibs Reporter: Lorenzo Sicilia <arnor>
Component: Standard TaglibAssignee: Tomcat Developers Mailing List <dev>
Severity: normal    
Priority: P3    
Version: unspecified   
Target Milestone: ---   
Hardware: PC   
OS: Linux   

Description Lorenzo Sicilia 2004-09-07 08:37:34 UTC
in version box there isn't

I use jstl and jvm sun 1.4.2 on my linux box. 

I have this small snippet:

<%@ taglib prefix="c"   uri="/WEB-INF/tld/c.tld" %>
<%@ taglib prefix="fmt" uri="/WEB-INF/tld/fmt.tld" %>
<%@ taglib prefix="sql" uri="/WEB-INF/tld/sql.tld" %>

<html xmlns="http://www.w3.org/1999/xhtml">
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
<title>bug fmt</title>

<fmt:message key="catalogue.detail" /><br/>
<sql:query var="languageList">
    SELECT id_language,name FROM language
 <c:forEach items="${languageList.rows}" var="languageRow">
    <c:out value="${languageRow.name}" />

and this table with mysql 4.0.x

CREATE TABLE `language` (
  `id_language` int(11) unsigned NOT NULL auto_increment,
  `name` varchar(50) NOT NULL default '',
  `code` varchar(4) NOT NULL default '',
  PRIMARY KEY  (`id_language`)

INSERT INTO `language` VALUES (1, 'italiano', 'it');
INSERT INTO `language` VALUES (2, 'fran├žaise', 'fr');
INSERT INTO `language` VALUES (3, 'english', 'en');

in my web.xml I use:

When I use a fmt tag the output from db it is wrong.
I get:
italiano fran?aise english

instead of:
 italiano fran├žaise english

If I remove in web xml the fmt context param and I add in my page haeder:
<fmt:setBundle basename="localtext" />
// bug fmt:setBundle. We need reset contentType
response.setContentType("text/html; charset=ISO-8859-1");
it work fine
Comment 1 Henri Yandell 2007-12-07 05:47:36 UTC
I was unable to recreate this back when I dug into it. I suspect more so because
I didn't manage to put together the right test case than because I'm sure it's
not a bug.
Comment 2 Henri Yandell 2009-07-10 00:45:30 UTC
That said (back in 2007) - I don't see myself being able to create a better test case to show the problem. Resolving as I couldn't get it to exhibit. If it is a problem in 1.1.x of Standard (or trunk), please reopen with an attachment demonstrating the issue.