Base2¹⁵

binary-to-text encoding scheme that represents binary data in an unicode string format. Each unicode character represents 15 bits of binary data.

Example

Encoding	Data	chararacters
Plain text	Base2e15 is awesome!	20
Base215	嗺둽嬖蟝巍媖疌켉溁닽壪	11
Base64	QmFzZTJlMTUgaXMgYXdlc29tZSE=	27+1

Mapping table

Every character represents 15 bits of data, except the last character, which is 7 or 15 bits.

Binary	Unicode	Unicode Range Name
15 bits mapping
0x0000 ~ 0x1935	U+3480 ~ U+4DB5	CJK Unified Ideographs Extension A
0x1936 ~ 0x545B	U+4E00 ~ U+8925	CJK Unified Ideographs
0x545C ~ 0x7FFF	U+AC00 ~ U+D7A3	Hangul Syllables
7 bits mapping
0x00 ~ 0x7F	U+3400 ~ U+347F	CJK Unified Ideographs Extension A

Usage

A simple usage example in dart:

import 'dart:convert';
import 'package:base2e15/base2e15.dart';

main() {
  String msg = 'Base2e15 is awesome!';
  String encoded = Base2e15.encode(UTF8.encode(msg));
  String decoded = UTF8.decode(Base2e15.decode(encoded));
}

example in c:

#include <stdio.h>
#include "base2e15.h"
int main(int argc, char *argv[])
{
    char c[] = "Base2e15 is awesome!";
    int len = strlen(c);

    int encodeLen = base2e15_encode_length(len) + 1; \\ one more char for the \0
    wchar_t* encoded = malloc(encodeLen * sizeof(wchar_t));
    base2e15_encode(c, len, encoded, encodeLen);
    
    // int buffLen = base2e15_decode_length(encoded, -1);
    int buffLen = 256; 
    char* decoded = malloc(buffLen * sizeof(char));
    int decodeLen = base2e15_decode(encoded, -1, decoded, buffLen);
    // if buffLen is not big enough, decodeLen will be -1
}

Compare

Compare	Base215	Base64
bits per character	15	6
bits per char width	7.5 (15/2)	6 (6/1)
bits per UTF8 byte	5 (15/3)	6 (6/1)
bits per UTF16 byte	7.5 (15/2)	3 (6/2)

Why not Base2¹⁶?

The unicode range CJK Unified Ideographs Extension B contains 42711 characters (U+20000 ~ U+2A6D6), together with the characters used by Base2¹⁵, there are more than 65536 usable characters to encode 16 bits in each character.

However, font support for CJK Unified Ideographs Extension B is missing in most mobile devices and using this code range will also reduce the bits capacity in UTF8 and UTF16 encoding, since those characters require one more byte in UTF8 and 2 more bytes in UTF16.