Servlet receiving UTF-16 parameter

Question

I am attempting to build a servlet to receive a parameter encoded in UTF-16. At this stage, I am able to receive and process UTF-8 parameters but not UTF-16. I believe this is due to the request.getParameter() method not finding the encoded parameter name. My servlet code is:
request.setCharacterEncoding("UTF-16");
String xmlStr = request.getParameter("xml");When I send the encoded request I do:
System.out.println("Parameters:");
for (Enumeration e = request.getParameterNames() ; e.hasMoreElements() ;)
    String s = (String)e.nextElement();
    System.out.println(s);
     for (int i=0; i< s.length(); i++)
         System.out.println(s.codePointAt(i));
}and via the Tomcat stdout log get:
Parameters:
30720
27904
27648
which would indicate it received the parameter xml yet the request.getParameter("xml") returns null.
How can I retrieve the encoded parameter value?

Answer 1

I believe the previous problem was a combination of two issues:
1) The UTF-16 was being encoded in little endian, Java was reading it as big endian.
2) The JVM appears to append the character <fffd> to each query parameter and each value. This is defined in Unicode to be a 'Replacement Character' and since it is included as part of the parameter causes that parameter not to match that being searched for.
My new issue is this:
I can find and thus retrieve the xml parameter from the request, however, it appears to take no notice of the encoding specified. I have printed the bytes contained in the complete input string and verified they contain the null (00) character for the half width latin characters. However, once request.getParameter() is called, these characters are lost and the resultant string is invalid UTF-16.
So <xml> goes from x00 x3c, x00 x78, x00 x6d, x00 x6e, x00 x3e (10 bytes) to x3c, x78, x6d, x6e, x3e (5 bytes)
How can I correctly parse my request parameter as being UTF-16 encoded?