decode Unicode - Java Internationalization

Java examples for Internationalization:Charset

  1. HOME
  2. Java
  3. Internationalization
  4. Charset

Description

decode Unicode

Demo Code


//package com.java2s;
import java.nio.ByteBuffer;
import java.nio.charset.Charset;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class Main {
    public static void main(String[] argv) throws Exception {
        String str = "java2s.com";
        System.out.println(decodeUnicode(str));
    }//  w w  w.ja  v  a2 s  .  c  o m

    public static String decodeUnicode(String str) {
        Charset set = Charset.forName("UTF-16");
        Pattern p = Pattern.compile("\\\\u([0-9a-fA-F]{4})");
        Matcher m = p.matcher(str);
        int start = 0;
        int start2;
        StringBuffer sb = new StringBuffer();
        while (m.find(start)) {
            start2 = m.start();
            if (start2 > start) {
                String seg = str.substring(start, start2);
                sb.append(seg);
            }
            String code = m.group(1);
            int i = Integer.valueOf(code, 16);
            byte[] bb = new byte[4];
            bb[0] = (byte) ((i >> 8) & 0xFF);
            bb[1] = (byte) (i & 0xFF);
            ByteBuffer b = ByteBuffer.wrap(bb);
            sb.append(String.valueOf(set.decode(b)).trim());
            start = m.end();
        }
        start2 = str.length();
        if (start2 > start) {
            String seg = str.substring(start, start2);
            sb.append(seg);
        }
        return sb.toString();
    }
}

Related Tutorials