How could I name devices using alljoyn not with the default dbus kind of name but using any UTF8 string?

Question

Please take for granted that I am using at the most just one alljoyn service in each device

From what I have got so far this is how you locate other devices:

1. You set a well known prefix which is the name of the service and everybody knows about
2. You set a unique suffix for each device

Well because all of the elements in a dbus name have certain restrictions, for example US-ASCII, how could you name a device with french or greek or chinese?

Goal is to have a list of all surrounding devices with their names. This is easy if all devices have a plain english name but what about other languages or even special characters and spaces?

Answer 1

In case the strings you are referring to are simple names I guess their length is relatively small and therefore performance is not a issue.

Using the code below you can encode-decode any utf8 string to and from English letters.

object LetterConverter {
    private def toTwoLetters(byte: Byte): String = {
      val num = byte + 128;
      //range now from 0 to 255
      val division = num / 26;
      val modulo = num % 26;
      val firstletter = ('a' + division).toChar.toString;
      val secondletter = ('a' + modulo).toChar.toString;
      firstletter + secondletter;
    }

    val byteRange = (-128 to 127).map(_.toByte);

    val letters = byteRange.map(n => toTwoLetters(n));

    private val byteToLetters = byteRange.zip(letters).toMap;
    require(byteToLetters.size == 256, "size really is: " + byteToLetters.size);

    private val lettersToByte = byteToLetters.map(_.swap);
    require(lettersToByte.size == 256, "size really is: " + lettersToByte.size);

    def bytesToLetters(bytes: Seq[Byte]): String = {
      bytes.map(byte => byteToLetters(byte)).mkString;
    }

    def lettersToBytes(letters: String): Seq[Byte] = {
      letters.sliding(2, 2).map(twoLetters => lettersToByte(twoLetters)).toSeq;
    }

    def encode(str: String, charset: Charset = Codec.UTF8): String = {
      bytesToLetters(str.getBytes(charset));
    }

    def decode(letters: String, charset: Charset = Codec.UTF8): String = {
      new String(lettersToBytes(letters).toArray, charset);
    }
}

Therefore let's say you have the prefix: com.pligor.service
Therefore having a non english name name let's say: Γιώργος
this would be translated to something like that: fmaiefoiafmaiofa
And you may just append that to get: com.pligor.service.fmaiefoiafmaiofa

After receiving the name on the client part you may extract the postfix fmaiefoiafmaiofa and by decoding it you get back the original greek name Γιώργος

Enjoy!