Unicode text through socket in java

Question

I am facing a tiny issue (I believe) in socket programming. When sending text from non-English languages, I get garbled results. After a lot of researching on google, I made some corrections. I changed getBytes() to getBytes("UTF-8") and tried to send some Arabic text.

When connecting sockets locally, it works fine. I see the arabic text I expected. But when testing from online, the results display strange/garbled characters.

Here is the text I tried:

"مرحبا" (this is the arab text of "hello") which displayed to me as "Ù…Ø±ØØ¨Ø§"

Please help me in resolving this issue.

Post your server write and client read code

Miserable Variable
– Miserable Variable

2012-02-21 20:39:05 +00:00
Commented Feb 21, 2012 at 20:39 — Miserable Variable
– Miserable Variable, Commented Feb 21, 2012 at 20:39

tchrist · Accepted Answer · 2012-02-21 21:47:21Z

5

This is some Java code I had lying around that’s used for setting the stream encodings on a pair of byte streams, but you could do the same with a singleton, at least assuming you’re using TCP stream sockets not UDP datagrams.

    Process
    slave_process = Runtime.getRuntime().exec("cmdname -opts cmdargs");

 OutputStream
 __bytes_into_his_stdin  = slave_process.getOutputStream();

 OutputStreamWriter
   chars_into_his_stdin  = new OutputStreamWriter(
                             __bytes_into_his_stdin,
         /* DO NOT OMIT! */  Charset.forName("UTF-8").newEncoder()
                         );

 InputStream
 __bytes_from_his_stdout = slave_process.getInputStream();

 InputStreamReader
   chars_from_his_stdout = new InputStreamReader(
                             __bytes_from_his_stdout,
         /* DO NOT OMIT! */  Charset.forName("UTF-8").newDecoder()
                         );

 InputStream
 __bytes_from_his_stderr = slave_process.getErrorStream();

 InputStreamReader
   chars_from_his_stderr = new InputStreamReader(
                             __bytes_from_his_stderr,
         /* DO NOT OMIT! */  Charset.forName("UTF-8").newDecoder()
                         );

answered Feb 21, 2012 at 21:47

tchrist

80.7k31 gold badges135 silver badges186 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Dinesh Ravichandran Over a year ago

yes man "InputStreamReader & OutputStreamWriter" did the work well for me... thank you for your contribution...

lxbndr · Accepted Answer · 2012-02-21 20:42:11Z

4

Perhaps, you forgot to specify encoding on string creation.

byte[] utf8bytes = yourString.getBytes("UTF-8");       // encoding
String otherString = new String(utf8bytes, "UTF-8");   // decoding

answered Feb 21, 2012 at 20:42

lxbndr

2,2281 gold badge16 silver badges17 bronze badges

7 Comments

tchrist Over a year ago

Why wouldn’t you just wrap the stream with the correct decoder? Manual decoding data to or from a stream that is all in one encoding is surely the wrong approach, don’t you think?

Dinesh Ravichandran Over a year ago

Well @mersenne i used the same like what you said... with that above two line which u said i successfully sent unicode text to socket listener in server... and after server gets it it sends to other client listeners where i get the same annoying text..

lxbndr Over a year ago

@tchrist: What is the correct decoder then? It is good to advice architectural changes, but it requires more detailed examples and some knowledge about source code.

tchrist Over a year ago

@Mersenne The rule-of-thumb is that streams in a single encoding should always have their raw bytes implicitly converted into characters by the right call to InputStreamReader(InputStream in, CharsetDecoder dec) or OutputStreamWriter(OutputStream out, CharsetEncoder enc). That way you deal only in Java characters; it’s a cleaner abstraction less prone to error and to repeating yourself. Calling getBytes is an antipattern for those scenarios.

lxbndr Over a year ago

@tchrist: that's right, that's right. I just can't be sure that it is all about simple text data. Sources are pending. Also, if we throw away old code completely and replace it with new logic (with better logic, of course), author will never know what was wrong with his original sources.

|

Shinzul · Accepted Answer · 2012-02-21 20:40:18Z

0

I think the easiest way to solve this would be to use a Serialized object that has a String container with your arabic text inside it.

Don't write the bytes directly, instead use:

ObjectOutputStream oos = yourSocket.getOutputStream();
oos.writeObject(yourContainer);

Then on the receiving end, do this:

if (receivedObject instanceof YourContainer) {
    // get out arabic string
}

answered Feb 21, 2012 at 20:40

Shinzul

2962 silver badges11 bronze badges

Comments

obeid salem · Accepted Answer · 2022-04-01 18:34:59Z

0

If anyone still trying to solve this :

in your Socket response:

HTTP/1.1 200 OK\r\n
Content-Type: text/html; charset=utf8\r\n\r\n

Just don't forget the Content-Type with charset set to utf8 it should work with Arabic letters.

answered Apr 1, 2022 at 18:34

obeid salem

1452 silver badges14 bronze badges

Collectives™ on Stack Overflow

Unicode text through socket in java

4 Answers 4

1 Comment

7 Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

1 Comment

7 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related