Enabling Multilanguage Support With UTF-8 - the hack

There are several places where character encoding is used. They are:

1. Encoding which is used when JSP compiler reads files from disk
2. Encoding which is used when the web server writes the response to the socket
3. Encoding that is specified in the response header sent to the browser
4. Encoding which is used when the browser sends the request to the server.
   1. Query string data is encoded using percent signs, but after you decode it you need to specify character encoding to convert it to a java string
   2. Multipart form data is encoded using base64 encoding. The same, after decoding base64, character encoding is used

Out of the box, Magnolia uses ISO-8859-1 for most of these and I don't know any civilised way to switch to UTF-8. My solution is as follows:

• Replace org.apache.tomcat.util.http.Parameters class with the one attached. It sits inside $MAGNOLIA_HOME/author/server/tomcat-util.jar

Some browsers including my beloved Firefox 1.0 don't specify character encoding in the request header. Tomcat defaults to ISO-8859-1 when parsing the parameters. I changed the defaults to UTF-8 in that class. This patch should fix point 4.1

• Replace com.oreilly.servlet.multipart.MultipartParser class with the one attached. It sits inside $MAGNOLIA_HOME/author/webapps/magnolia/WEB-INF/lib/cos.jar

This should fix point 4.2

• Add parameter contentType="text/html;charset=UTF-8" to all @page directives inside your templates and inside all jsp files in the $MAGNOLIA_HOME/author/webapps/magnolia/admintemplates folder.

This should fix points 1, 2 and 3

I have attached the sources of the modified files as well. Places changed by me are marked by 'Mitek' word.

Enabling Multilanguage Support With UTF-8 or other encodings - a clean solution

1- configure Tomcat to use the correct encoding while compiling jsp files. By default Tomcat is already configured to use UTF8, you can switch to another encoding (e.g. UTF16) by changing $TOMCAT_HOME/conf/web.xml:

<servlet> 
  <servlet-name>jsp</servlet-name> 
  <servlet-class>org.apache.jasper.servlet.JspServlet</servlet-class> 
  <init-param> 
    <param-name>javaEncoding</param-name> 
    <param-value>UTF8</param-value> 
  </init-param> 
  ... 
</servlet>

For more details/throubleshooting you can see an interesting article here:
http://www.javaworld.com/javaworld/jw-04-2004/jw-0419-multibytes.html

2- you still have to force browser requests to contain the correct encoding. You can use a simple j2ee filter for this scope, without patching any tomcat/cos class (actually: this has been tested using jakarta commons-upload and not oreilly servlet... please report any problem with it).

Spring includes a filter in its distribution (org.springframework.web.filter.CharacterEncodingFilter), see http://www.springframework.org/docs/api/org/springframework/web/filter/CharacterEncodingFilter.html

To force requests to be UTF8 you only will need to add this to your web.xml (of course you will also need the appropriate jars/classes in your classpath):

<filter> 
  <filter-name>encodingFilter</filter-name> 
  <filter-class>org.springframework.web.filter.CharacterEncodingFilter</filter-class> 
  <init-param> 
    <param-name>encoding</param-name> 
    <param-value>UTF-8</param-value> 
  </init-param> 
  <init-param> 
    <param-name>forceEncoding</param-name> 
    <param-value>true</param-value> 
  </init-param> 
</filter>

In its previous incarnation on JspWiki, this page was last edited on Feb 9, 2007 10:12:01 AM by BorisKraft
Other known authors include :

• Mitek
• Ferris