Jetty does not adhere to UTF-8 spec and uses CESU-8 for supplementary characters

Jetty does not parse URIs correctly. It interprets them as CESU-8 instead of UTF-8. This is an issue when
using Unicode supplementary characters beyond \uFFFF. What's happening is that it accepts these supplementary characters as 6 bytes (3 bytes per surrogate code unit). To be using UTF_8, it needs to accept the 4 byte form.

Here is some example code that replicates the issue.

http://svn.apache.org/repos/asf/lucene/dev/trunk/solr/example/exampledocs/test_utf8.sh

Here is an example report of the issue:

http://www.mail-archive.com/solr-user@lucene.apache.org/msg43399.html

Parts of Solr has worked around this issue, but it would be better if Jetty conformed to the UTF-8 spec. Lots of web browsers and software out there do support CESU-8.