Google translate latin characters (�����) problem

Question

Hi All,
I took a job where I am asked to write a J2ME translator which uses a back-end servlet to get the translation from google translate. Well it is almost done, however I get translation differences depending, I think, on which platform, tomcat, or java vm is used to run the servlet. I cannot really figure out why this happens.
result on my notebook (xp sp2, tomcat 5.5.20, jdk 1.5.0_10):
*** Request ***
Request Path: /translate_t
Request Query: hl=en&ie=UTF8&text=how+much+is+this&langpair=en|it
User-Agent: Jakarta Commons-HttpClient/3.0.1
Host: www.google.com
*** Response ***
Status Line: HTTP/1.1 200 OK
Content-Type: text/html; charset=ISO-8859-1
Content-Language: en
Server: TWS/0.9
Cache-Control: private, x-gzip-ok=""
Transfer-Encoding: chunked
Date: Wed, 25 Jul 2007 21:55:31 GMT
en_US
HERE I GET THE RIGHT TRANSLATION : quanto � questo
result on the provider's server (not sure but maybe linux, tomcat 5.5.xx, jdk 1.5.0_xx, where xx means versions older then mine):
*** Request ***
Request Path: /translate_t
Request Query: hl=en&ie=UTF8&text=how+much+is+this&langpair=en|it
User-Agent: Jakarta Commons-HttpClient/3.0.1
Host: www.google.com
*** Response ***
Status Line: HTTP/1.1 200 OK
Content-Type: text/html; charset=ISO-8859-1
Content-Language: en
Server: TWS/0.9
Cache-Control: private, x-gzip-ok=""
Transfer-Encoding: chunked
Date: Wed, 25 Jul 2007 21:58:48 GMT
en_US
HERE I GET A WRONG TRANSLATION : quanto ��� questo
Has someone out there ever faced and solved such or similar problems?
Thanks,
Regards,
Silvio.

Answer 1

Hi hunter,
thx for your reply.
This is what I do :
String url="http://www.google.com/translate_t?hl=en&ie="+charset+"&text="+text+"&langpair="+googlelang;
HttpClient client = new HttpClient();
client.getHttpConnectionManager().getParams().setConnectionTimeout(5000);
HttpMethod method = null;
method = new GetMethod(url);
method.setFollowRedirects(true);
String responseBody = null;
try{
client.executeMethod(method);
// gets the response from google as an html page
BufferedReader in = new BufferedReader(new InputStreamReader(method.getResponseBodyAsStream()));
String input = null;
StringBuffer sb = new StringBuffer("");
while ((input = in.readLine())!=null) {
     sb.append(input);
// here I have the html page; already here I have the wrong or right translation, quite strange no?
responseBody = sb.toString();
} catch (Exception e) {
e.printStackTrace();
After simply I parse the html extracting the translated string.
Any other suggestions?
Thx again, silvio