A simple question about converting byte array to unicode string

[Genom]

Hi,

I have a bytearray which is coding for a unicode text ("BlaBla"). So my byte array is 12 bytes long (there are 6 characters in the text and 2 bytes for each character).

And here is the byte array:
66, 0, 108, 0, 97, 0, 66, 0, 108, 0, 97, 0

Everything seems to be normal but I can't convert this byte array back to "BlaBla" and show it in an text area. Instead of "BlaBla" I get some chinese like characters (䈀氀愀䈀氀愀BB䈀氀愀䈀氀愀).

I have tried so far:
String Message=new String(packageBytes,0,12, "UTF-16");
String Message=new String(packageBytes,0,12, "US-ASCII");
String Message=new String(packageBytes,0,12, "UTF-16BE");
String Message=new String(packageBytes,0,12, "UTF-16LE");

I have tried a lot of things and googled a lot but can't find a solution :confused:?

[javaforum$]

Try this -

String msg = new String(new byte[]{66, 0, 108, 0, 97, 0, 66, 0, 108, 0, 97, 0},0,12,"US-ASCII");
System.out.println(msg);

[JosAH]

Try this -

String msg = new String(new byte[]{66, 0, 108, 0, 97, 0, 66, 0, 108, 0, 97, 0},0,12,"US-ASCII");
System.out.println(msg);

That is incorrect; print this instead:

System.out.println(Arrays.toString(msg.toCharArray())); 
System.out.println(msg.length());

... and you'll see; the encoding is UTF-16, low byte first.

kind regards,

Jos

[Genom]

I can't put the output here because of the squares representing "0" bytes. 0's are being recognized as a character. Actually "66" and "0" together have to be a unicode character.
I get something like this as output:

[B, *, l, *, a, * ...]
12

*... is actually a square. Maybe I have to explain it better:

I am writing a java client which binds over a socket to a vb.net written server. Programs written with .Net together can exchange strings over network without a problem. But If I send the string to a java client, java client doesn't have a readString() method like .Net. So I decided to send data in byte array and convert it to string in java client. But then illegal characters are being a problem "ş" is then replaced with 2 other symbols in java side. So I decided to send data as unicode. Data is being delivered with 2 headers:

1 byte as "1" or "2" telling to other side what type of data is coming (same connection is going to be used for file transfer as well. 1 for text 2 for file chunks).

1 byte as length of data (in bytes) to be read (as x).

x bytes of data.

				InputStream is = socket.getInputStream();
				int packageType;
				int packageLen;			
				
				packageType=is.read();
				
				if (packageType==1) {
				packageLen=is.read();
				System.out.println("Package Length: " + packageLen);
				byte[] packageBytes= new byte[packageLen];
				is.read(packageBytes, 0, packageLen);
					int i;
				for(i=0;i<packageLen;i++){
					System.out.println(packageBytes[i]);//just to verify bytes are stored in my byte array
				}
					String Message=new String(packageBytes,0,12, "UTF-16");
					System.out.println(Message);
					ta.append(Message); //ta is a text area. Result is "B" (byte=66) so only the first character is written. I think that is because of the character with byte=0.
				}

I really don't get why String(packageBytes,0,12, "UTF-16"); can create a string according to my byte array?

[JosAH]

Check the CharSet class for the available encodings; it is a low byte first (little endian) UTF-16 encoding.

kind regards,

Jos

[Genom]

"UTF-16LE" as charset worked thanks a lot!!!! D But I have tried that as well. Interestingly it didn't work I think I forgot something. Thanks again...!

[JosAH]

"UTF-16LE" as charset worked thanks a lot!!!! D But I have tried that as well. Interestingly it didn't work I think I forgot something. Thanks again...!

Good to hear that your problem is solved now; I recognize UTF-16 low byte first when I see it ;-) Just using UTF-16 is platform dependent when there are no two 'byte order' bytes at the start of the byte sequence (i.e. 0xFF 0xFE or 0xFE 0xFF)

kind regards,

Jos