From: "Java Poop" <[EMAIL PROTECTED]>
Subject: Re: Multi-byte character support.
Date: Mon, 26 Mar 2001 22:33:36 -0800
Message-ID: <[EMAIL PROTECTED]>
> What are the il8n limitations in 3.2.1?
For example, there are the following limitations:
1, The getParameter method in ServletRequest class returns non-Unicode
strings.
If a HTTP GET or POST request is done by a HTML form, servlet can't
know its charset because the request don't have its charset
information. The HTML 4.0 specification recommends to encode UTF-8
strings but most of systems don't do it.
Tomcat 3.2 treats parameters like their charset is ISO-8859-1. Other
servlet container may return Unicode strings by its own encoding
detection method but it is tricky and don't work under some situation.
For avoiding this problem, Servlet 2.3 API has setCharacterEncoding
method and Tomcat 4.0 based on it.
Because Tomcat 3.2 based on Servlet 2.2 API, we can't use this method.
But Costin will provide an extension module for Tomcat 3.3 (for
example, specify the default encoding per context).
2, A JSP content's encoding may not be recognized correctly.
This cause is a lack of a function to specify JSP content's encoding
(JSP output encoding is specified by contentType attribute of Page
directive).
JSP 1.2 has new pageEncoding attribute of Page directive. And a file
localization method, it is proposed by Arieh, may provide another
solution.
Kazuhiro Kazama ([EMAIL PROTECTED]) NTT Network Innovation Laboratories
