使用java的Unicode base 64编码

Question

I am trying to encode and decode a UTF8 string to base64. In theory not a problem but when decoding and never seem to output the correct characters but the ?.

        String original = "خهعسيبنتا";
        B64encoder benco = new B64encoder();
        String enc = benco.encode(original);
        try
        {
            String dec = new String(benco.decode(enc.toCharArray()), "UTF-8");
            PrintStream out = new PrintStream(System.out, true, "UTF-8");
            out.println("Original: " + original);
            prtHx("ara", original.getBytes());
            out.println("Encoded: " + enc);
            prtHx("enc", enc.getBytes());
            out.println("Decoded: " + dec);
            prtHx("dec", dec.getBytes());
        } catch (UnsupportedEncodingException e)
        {
            e.printStackTrace();
        }

The output to the console is as follow:

Original: خهعسيبنتا
ara = 3F, 3F, 3F, 3F, 3F, 3F, 3F, 3F, 3F
Encoded: Pz8/Pz8/Pz8/
enc = 50, 7A, 38, 2F, 50, 7A, 38, 2F, 50, 7A, 38, 2F
Decoded: ?????????
dec = 3F, 3F, 3F, 3F, 3F, 3F, 3F, 3F, 3F

prtHx simply writes the hex value of the bytes to the output. Am I doing something obviously wrong here?

---

Andreas pointed to the correct solution by highlighting that the getBytes() method uses the platform default encoding (Cp1252) even though the source file itself is UTF-8. By using the getBytes("UTF-8") I was able to notice that the bytes encoded and decoded were actually different. further investigation shown that the encode method used getBytes(). Changing this did the trick nicely.

try
        {
            String enc = benco.encode(original);
            String dec = new String(benco.decode(enc.toCharArray()), "UTF-8");
            PrintStream out = new PrintStream(System.out, true, "UTF-8");
            out.println("Original: " + original);
            prtHx("ori", original.getBytes("UTF-8"));
            out.println("Encoded: " + enc);
            prtHx("enc", enc.getBytes("UTF-8"));
            out.println("Decoded: " + dec);
            prtHx("dec", dec.getBytes("UTF-8"));

        } catch (UnsupportedEncodingException e)
        {
            e.printStackTrace();
        }

System encoding Cp1252
Original: خهعسيبنتا
ori = D8, AE, D9, 87, D8, B9, D8, B3, D9, 8A, D8, A8, D9, 86, D8, AA, D8, A7
Encoded: 2K7Zh9i52LPZitio2YbYqtin
enc = 32, 4B, 37, 5A, 68, 39, 69, 35, 32, 4C, 50, 5A, 69, 74, 69, 6F, 32, 59, 62, 59, 71, 74, 69, 6E
Decoded: خهعسيبنتا
dec = D8, AE, D9, 87, D8, B9, D8, B3, D9, 8A, D8, A8, D9, 86, D8, AA, D8, A7

Thanks.

Answer 1

String#getBytes() encodes the characters using the platform's default charset. The actual encoding of the String literal "خهعسيبنتا" is "defined" in the java source file (you choose a character encoding when you create or save the file)

This could be the reason, why ara is encode to 0x3f bytes..

Give this a try:

out.println("Original: " + original);
prtHx("ara", original.getBytes("UTF-8"));
out.println("Encoded: " + enc);
prtHx("enc", enc.getBytes("UTF-8"));
out.println("Decoded: " + dec);
prtHx("dec", dec.getBytes("UTF-8"));