Did that alrady. Here's the gory details and the conclusion is valid for 4 and 5. It seems the spec folks took care of the character encoding issue but forgot how to fix it for j_security_check. So the "short term" solution is probably a custom solution per platform. :( (Maybe google has the answer, haven't checked yet)


http://nagoya.apache.org/bugzilla/show_bug.cgi?id=21795


-Tim

Bill Barker wrote:
This is a really interesting issue.  I would hope that you would send a
message to [EMAIL PROTECTED] to hopefully get the expert-team
to clarify this before the 2.4 spec goes final.  Because the Servlet-2.2
spec lacks the request.setCharacterEncoding method, Tomcat 3.3 jumps through
a lot of hoops to try and guess the charset.  These were dropped in Tomcat
4+, since the request.setCharacterEncoding method was supposed to solve all
of these problems.  As you have pointed out, it is not possible to use this
in a Filter for the standard Form-auth config (for the simple reason that
Auth is called before Filters).  Therefore, the j_security_check target is
flying blind wrt charset.





--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to