encoding messy code

[xu cheng]

hi all:
I configure an app with solr to index documents
and there are some Chinese content in the documents
and I've configure the apache tomcat URIEncoding to be utf-8
and I use the program curl to sent the documents in xml format
however , when I query the documents, all the Chinese content becomes messy
code. It've cost me a lot of time.
does anyone has some idea about it?
thanks
by the way , the tomcat edition is 6.0.20,while the solr is 1.4