Converting binary data to String

Question

If I have some binary data D And I convert it to string S. I expect than on converting it back to binary I will get D. But It's wrong.

public class A {
    public static void main(String[] args) throws IOException {
        final byte[] bytes = new byte[]{-114, 104, -35};// In hex: 8E 68 DD
        System.out.println(bytes.length);               //prints 3
        System.out.println(new String(bytes, "UTF-8").getBytes("UTF-8").length); //prints 7
    }
}

Why does this happens?

Are you trying to force arbitrary binary data into a string? Why? — Jesper
– Jesper, Commented Feb 4, 2012 at 20:22
(If you do have to hide binary in a string for some reason, you'd need to use an encoding that provides a one-to-one mapping between bytes and characters; ISO-8859-1 would be the obvious choice. UTF-8 has byte sequences that do not represent valid characters.) — bobince
– bobince, Commented Feb 5, 2012 at 9:29

John Ericksen · Accepted Answer · 2012-02-04 20:07:12Z

2

Converting between a byte array to a String and back again is not a one-to-one mapping operation. Reading the docs, the String implmentation uses the CharsetDecoder to convert the incoming byte array into unicode. The first and last bytes in your input byte array must not map to a valid unicode character, thus it replaces it with some replacement string.

answered Feb 4, 2012 at 20:07

John Ericksen

11.2k6 gold badges48 silver badges75 bronze badges

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

2 Comments

Stan Kurilin Over a year ago

Good point. But it seems strange. Why it should use some magic replacement string instead of throwing exception?

John Ericksen Over a year ago

I guess throwing an exception is possible from the CharsetDecoder when it encounters an unmappable character, but the default String implementation uses the less volatile option of a default error character. I bet you can use the CharsetDecoder yourself for more control over Byte[]<->String conversion.

David Wood · Accepted Answer · 2012-02-04 20:02:20Z

1

It's likely that the bytes you're converting to a string don't actually form a valid string. If java can't figure out what you mean by each byte, it will attempt to fix them. This means that when you convert back to the byte array, it won't be the same as when you started. If you try with a valid set of bytes, then you should be more successful.

answered Feb 4, 2012 at 20:02

David Wood

3851 gold badge4 silver badges9 bronze badges

2 Comments

Stan Kurilin Over a year ago

Yeah. But I except at least to get exception on this case.

DNA Over a year ago

The behaviour is configurable - you can ignore, replace or error. See docs.oracle.com/javase/1.5.0/docs/api/java/nio/charset/…, particularly the bit about the CodingErrorAction class

lxbndr · Accepted Answer · 2012-02-04 20:40:07Z

0

Your data can't be decoded into valid Unicode characters using UTF-8 encoding. Look at decoded string. It consists of 3 characters: 0xFFFD, 0x0068 and 0xFFFD. First and last are "�" - Unicode replacement characters. I think you need to choose other encoding. I.e. "CP866" produces valid string and converts back into same array.

answered Feb 4, 2012 at 20:40

lxbndr

2,2281 gold badge16 silver badges17 bronze badges

Collectives™ on Stack Overflow

Converting binary data to String

3 Answers 3

2 Comments

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related