Difference to Base64
From what I find online, there are different versions of the BinHex format. None of them is exactly the same as Base64. There are however major similarities. Taking for example the BinHex 4.0 specs, we see that the major binary content is encoded using an encoding scheme with base 64, thus encoing 3 octets to 4 characters. It uses a different alphabet, though:
BinHex: !"#$%&'()*+,-012345689@ABCDEFGHIJKLMNPQRSTUVXYZ[`abcdefhijklmpqr
Base64: ABCDEFGHIJKLMNOPQERTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/
So you'd either have to translate from one set of characters to the other, or do the decoding yourself.
Apart from the bulk of the binary content, there is some additional meta data included in the format. According to the spec, that includes delimiters between data and resource fork content, checksums, and some other bits of information.
Implementation
The following code will decode a BinHex input file and write its data fork to an output file. It should be easy enough to adjust this code to your needs.
import java.io.*;
import java.nio.charset.Charset;
public class BinHexDec {
private Charset charset;
private String filename;
private int dbegin, dend, rbegin, rend;
public BinHexDec(String... args) throws IOException {
InputStream in = System.in;
OutputStream out = System.out;
charset = Charset.defaultCharset();
if (args.length >= 1)
in = new FileInputStream(args[0]);
if (args.length >= 2)
out = new FileOutputStream(args[1]);
if (args.length >= 3)
charset = Charset.forName(args[2]);
Reader reader = new InputStreamReader(in, charset);
byte[] data;
data = unbase(reader);
reader.close();
if (data.length < 27)
throw new IOException("Too little data");
data = unrle(data);
parse(data);
out.write(data, dbegin, dend - dbegin);
}
private byte[] unbase(Reader in) throws IOException {
String digits = "!\"#$%&'()*+,-012345689@ABCDEFGHIJKLMNPQRSTUVXYZ" +
"[`abcdefhijklmpqr";
assert digits.length() == (1 << 6);
int[] value = new int[128];
for (int i = 0; i < value.length; ++i)
value[i] = -1;
for (int i = 0; i < digits.length(); ++i)
value[digits.charAt(i)] = i;
int state = 0;
int accum = 0;
int alen = 0;
ByteArrayOutputStream baos = new ByteArrayOutputStream();
while (true) {
int chr = in.read();
if (chr == -1)
throw new EOFException();
switch (state) {
case 0:
switch (chr) {
case '\n':
case '\r':
case '\t':
case ' ':
state = 1;
break;
}
break;
case 1:
switch (chr) {
case '\n':
case '\r':
case '\t':
case ' ':
break;
case ':':
state = 2;
break;
default:
state = 0;
}
break;
case 2:
int v = -1, c = chr & 0x7f;
if (c == chr) {
v = value[c];
if (v != -1) {
accum = (accum << 6) | v;
alen += 6;
if (alen > 8) {
alen -= 8;
baos.write((byte)(accum >>> alen));
}
break;
}
}
if (chr == ':') {
return baos.toByteArray();
}
break;
}
}
}
private static final byte RLE_MARK = (byte)0x90;
private byte[] unrle(byte[] in) throws IOException {
int len = in.length;
if (in[0] == 0x90)
throw new IOException("Incomplete RLE at beginning");
for (int i = 0; i < in.length; ++i) {
if (in[i] == RLE_MARK) {
++i;
if (i == in.length)
throw new IOException("Incomplete RLE at end");
int cnt = in[i] & 0xff;
if (cnt == 0)
len -= 1;
else
len += cnt - 3;
}
}
byte[] out = new byte[len];
for (int i = 0, o = 0; i < in.length; ++i) {
if (in[i] == RLE_MARK) {
++i;
int cnt = in[i] & 0xff;
if (cnt == 0)
out[o++] = RLE_MARK;
else {
byte b = out[o - 1];
for (int c = 1; c < cnt; ++c)
out[o++] = b;
}
}
else {
out[o++] = in[i];
}
}
return out;
}
private int get4(byte[] data, int offset) {
return ((data[offset ] & 0xff) << 24)
| ((data[offset + 1] & 0xff) << 16)
| ((data[offset + 2] & 0xff) << 8)
| ((data[offset + 3] & 0xff) );
}
private int get2(byte[] data, int offset) {
return ((data[offset] & 0xff) << 8)
| ((data[offset + 1] & 0xff));
}
private void crcCheck(byte[] data, int begin, int end) throws IOException {
int crc = 0;
for (int i = begin; i < end; ++i) {
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
crc = crc ^ (data[i] & 0xff);
}
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
crc = ((crc << 4)&0xffff) ^ (crc >> 12)*0x1021;
int expected = get2(data, end);
if (expected != crc)
throw new IOException("CRC mismatch");
}
private void parse(byte[] in) throws IOException {
int namelen = in[0] & 0xff;
int headlen = namelen + 20;
int dlen = get4(in, headlen - 8);
int rlen = get4(in, headlen - 4);
if (dlen < 0 || rlen < 0)
throw new IOException("Unsigned data lenhgths not supported");
dbegin = headlen + 2;
dend = dbegin + dlen;
rbegin = dend + 2;
rend = rbegin + rlen;
if (in.length != rend + 2 &&
(in.length != rend + 3 || in[rend + 2] != 0))
throw new IOException("Incorrect data size:" +
" expected " + (rend + 2) +
" but got " + in.length + " bytes");
crcCheck(in, 0, headlen);
crcCheck(in, dbegin, dend);
crcCheck(in, rbegin, rend);
filename = new String(in, 1, namelen, charset);
System.err.println(filename);
}
public static void main(String[] args) throws IOException {
new BinHexDec(args);
}
}
