English could not be shown correctly in respond of HTTP

Question

Hi, everyone, I am writing a class that it could connect to http://babelfish.altavista.com/ and send out a "Chinese-simp to English" translation request. And then it would print out the respond from the server which is in HTML language. Everything works alright so far except the display of the translation result. The results, including all the HTML tags and translation results, supposed to be all in English. However, only the English translation result doesn't appear correctly. There is no problem at all if I enter the same Chinese string in the program into babelfish website and translate it into English.
I also tried to convert the charset of the outputstream, which would send the request to the server, into UTF-8 since the webpage was encoded in UTF-8. Correspondingly, I also use the method toChinese() to convert the Chinese string from GBk to UTF-8 before sending it out. I tred many charsets (ASCII, UTF-8, UTF-16, GBK, ISO8859_1), but none of them work correctly.
I have already successfully translate French to English by setting the both of charsets of outputstream and inputstream to ISO8859_1.
I don't know what is wrong with it. I wish someone could point it out for me. Here is the code of this program:
import java.io.*;
import java.net.*;
import java.util.*;
import javax.swing.text.html.*;
import javax.swing.text.html.HTMLEditorKit.*;
public class FormProcessor {
public static String doPost(URL url, Properties nameValuePairs) throws IOException{
URLConnection connection = url.openConnection();
connection.setDoOutput(true);
PrintWriter out = new PrintWriter(new OutputStreamWriter(connection.getOutputStream(), "UTF8"));
Enumeration enume = nameValuePairs.keys();
while(enume.hasMoreElements()){
String name = (String)enume.nextElement();
String value = nameValuePairs.getProperty(name);
out.print(name);
out.print('=');
if(name.equals("trtext")) // debug if-statement
value = toChinese(value);
out.print(value);
if(enume.hasMoreElements())
out.print('&');
out.close();
BufferedReader in;
StringBuffer response = new StringBuffer();
try{
in = new BufferedReader(new InputStreamReader(connection.getInputStream(), "UTF8"));
catch(IOException e){
if(!(connection instanceof HttpURLConnection)) throw e;
InputStream err = ((HttpURLConnection)connection).getErrorStream();
if(err == null) throw e;
in = new BufferedReader(new InputStreamReader(err));
String line;
while((line = in.readLine()) != null)
response.append(new String(line.getBytes(), "ISO8859_1") + "\n");
in.close();
return response.toString();
public static String toChinese(String strvalue){
try{
if(strvalue==null)
return null;
else{
strvalue = new String(strvalue.getBytes("GBK"), "UTF8");
return strvalue;
}catch(Exception e){
return null;
public static void main(String[] args){
try{
Properties props = new Properties();
URL url = new URL("http://babelfish.altavista.com/tr");
props.put("URL", "http://babelfish.altavista.com/tr");
props.put("trtext", "中文测试");
props.put("lp", "zh_en");
System.out.println(props.getProperty("trtext"));
String r = doPost(url, props);
System.out.println(r);
catch(IOException e){
e.printStackTrace();
}Please help me solve this problem. Thank you very much!!!

Answer 1

The central thing to bear in mind is that Java characters should always be unicode and character encodings are only to be used when the characters are translated to and from bytes. So when you see the
new String(otherString.getBytes("xxx"), "yyy")
construction you can be pretty sure things are going wrong.
Assuming bablefish is sending it's response in UTF-8 (which is logical) then your UTF-8 input reader will suffice to do the job. No further transliteration should be required except, possibly, if you output the results to a file. In that case you simply use and OutputStreamWriter with the appropriate encoding.