在Java中编码可变长度的utf8字节数组

Question

Actually I am in a situation where I need to read a string which is in utf8 format but its chars use variable-length encoding so I have problem encoding them to string and I get weird chars when printing it, the chars seem to be in Korean and the is the code I used but had no result:

public static String byteToUTF8(byte[] bytes) {
    try {
        return (new String(bytes, "UTF-8"));

    } catch (UnsupportedEncodingException e) {
        e.printStackTrace();
    }
    Charset UTF8_CHARSET = Charset.forName("UTF-8");
    return new String(bytes, UTF8_CHARSET);
}

Also I used UTF-16 and got a bit better results, however it was giving me strange chars yet and according to doc provided above I should use utf8.

Thanks in advance for helping.

EDIT:

Base64 value: S0QtOTI2IEdHMDA2AAAAAA==\\n

Answer 1

Bluetooth name display issue:

If you check Bluetooth adapter setName(), you will get that

https://developer.android.com/reference/android/bluetooth/BluetoothAdapter.html#setName

Valid Bluetooth names are a maximum of 248 bytes using UTF-8 encoding, although many remote devices can only display the first 40 characters, and some may be limited to just 20.

Android Supported Versions:

If you check the link https://stackoverflow.com/a/7989085/2293534 , you will get the list of android supported version.

Supported and Non supported locales are given in the table:

-----------------------------------------------------------------------------------------------------
             | DEC Korean | Korean EUC | ISO-2022-KR | KSC5601/cp949 | UCS-2/UTF-16 | UCS-4 | UTF-8 |
-----------------------------------------------------------------------------------------------------
 DEC Korean  |      -     |      Y     |     N       |      Y        |        Y     |   Y   |   Y   |
-----------------------------------------------------------------------------------------------------
 Korean EUC  |      Y     |      -     |     Y       |      N        |        N     |   N   |   N   |
-----------------------------------------------------------------------------------------------------
 ISO-2022-KR |      N     |      Y     |     -       |      Y        |        N     |   N   |   N   |
-----------------------------------------------------------------------------------------------------
KSC5601/cp949|      Y     |      N     |     Y       |      -        |        Y     |   Y   |   Y   |
-----------------------------------------------------------------------------------------------------
 UCS-2/UTF-16|      Y     |      N     |     N       |      Y        |        -     |   Y   |   Y   |
-----------------------------------------------------------------------------------------------------
    UCS-4    |      Y     |      N     |     N       |      Y        |        Y     |   -   |   Y   |
-----------------------------------------------------------------------------------------------------
    UTF-8    |      Y     |      N     |     N       |      Y        |        Y     |   Y   |   -   |
-----------------------------------------------------------------------------------------------------

For solution,

Solution#1:

Michael has given a great example for conversion. For more you can check https://stackoverflow.com/a/40070761/2293534

When you call getBytes(), you are getting the raw bytes of the string encoded under your system's native character encoding (which may or may not be UTF-8). Then, you are treating those bytes as if they were encoded in UTF-8, which they might not be.

A more reliable approach would be to read the ko_KR-euc file into a Java String. Then, write out the Java String using UTF-8 encoding.

 InputStream in = ... Reader reader = new InputStreamReader(in, "ko_KR-euc"); // you can use specific korean locale here StringBuilder sb = new StringBuilder(); int read; while ((read = reader.read()) != -1){ sb.append((char)read); } reader.close(); String string = sb.toString(); OutputStream out = ... Writer writer = new OutputStreamWriter(out, "UTF-8"); writer.write(string); writer.close(); 

NB: You should, of course, use the correct encoding name

Solution#2:

Using StringUtils, you can do it https://stackoverflow.com/a/30170431/2293534

Solutions#3:

You can use Apache Commons IO for conversion. A very great example is given here: http://www.utdallas.edu/~lmorenoc/research/icse2015/commons-io-2.4/examples/toString_49.html

1 String resource;
2 //getClass().getResourceAsStream(resource) -> the <code>InputStream</code> to read from
3 //"UTF-8" -> the encoding to use, null means platform default
4 IOUtils.toString(getClass().getResourceAsStream(resource),"UTF-8");

Resource Links:

1. Korean Codesets and Codeset Conversion
2. Korean Localization
3. Changing the Default Locale
4. Byte Encodings and Strings

Answer 2

I suggest you use StringUtils per Apache libraries. I believe the necessary methods for your are documented here:

https://commons.apache.org/proper/commons-codec/apidocs/org/apache/commons/codec/binary/StringUtils.html