Name: skT88420 Date: 05/19/99


A URL is encoded using java.net.URLEncode which encodes the URL
using the client OS's character encoding. One the server side,
the java.net.URLDecode parses the URL and makes a unicode
character out of each byte. If the client is using a Japanese OS
with the SJIS encoding and types in the Japanese word setsume,
the java.net.URLEncode converts URL word setsume to %90%D8%96%BE.
On the server java.net.URLDecode parses the URL and makes a
four unicode character string out of the four bytes (not a two
character string).

There are two problems.
1) The URL is being encoded using the OS's default encoding but
that encoding is not being sent to the server So, the server side
does not know how to decode the string properly.

2) The java.net.URLDecode class is decoding the string
as if every byte is a character. It is not handling double or
triple byte encoding (it must take into consideration the encoding
of the URL
(Review ID: 83263)
======================================================================