Effectively encodes/decodes binary data to the text suitable to use as urls and file names, unlike Base64 which is case-sensitive

NameCode32.java

package net.sergeych.utils;

import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;

/**
 * Binary to text encoder suitable to use encoded strings as any parts of urls
 * and parts of file names safely even on case-insensitive systems. Human
 * friendly, easy to retype from print or from voice, treats confusing
 * characters (like No filling characters, byte granularity, minimal overhead
 * for base 32 family. Ideal for keys, ids and so on.
 * 
 * <p>
 * Based on the Crockford's base32 alphabet "
 * <code>0123456789ABCDEFGHJKMNPQRSTVWXYZ</code>" and provides more aliases to
 * reduce errors when the encoded data are entered by a human:
 * <code>"lLiI1", "oO0",
 * "uUvV"</code> are all the same while decoding. <b>Hyphens are ignored<b>
 * while decoding and can be inserted to make it more readable.
 * 
 * <p>
 * See also <a href='http://www.crockford.com/wrmg/base32.html'>Douglas
 * Crockford base32</a> algorithm.
 * 
 * @author sergeych ([email protected])
 * @license MIT license.
 */
public class NameCode32 {

	static private final char[] abc;
	static private final HashMap<Character, Integer> inAbc;
	static private final String abcString;

	static {
		abcString = "0123456789ABCDEFGHJKMNPQRSTVWXYZ";
		abc = abcString.toCharArray();
		inAbc = new HashMap<>();
		for (int i = 0; i < 32; i++) {
			char c = abc[i];
			inAbc.put(c, i);
			char cl = Character.toLowerCase(c);
			if (c != cl)
				inAbc.put(cl, i);
		}
		for (char c : "iIlL".toCharArray())
			inAbc.put(c, 1);
		for (char c : "oOоО".toCharArray())
			inAbc.put(c, 0);
		for (char c : "уУ".toCharArray())
			inAbc.put(c, 30);
		for (char c : "uU".toCharArray())
			inAbc.put(c, 27);
	}

	/**
	 * Decode encoded string back into byte[].
	 * 
	 * @param src
	 *            encoded string
	 * @return byte[] result
	 */
	public static byte[] decode(String src) {
		List<Byte> res = decodeList(src);
		byte bres[] = new byte[res.size()];
		int cnt = 0;

		for (byte x : res)
			bres[cnt++] = x;

		return bres;
	}

	/**
	 * Decode encoded string back as a List<Byte>. Less overhead than
	 * {@link #decode(String)}.
	 * 
	 * @param src
	 *            encoded string
	 * @return List<Byte> result
	 */
	public static List<Byte> decodeList(String src) {
		try {
			ArrayList<Byte> res = new ArrayList<>();
			int acc = 0;
			int bits = 0;

			for (char c : src.toCharArray()) {
				if (c != '-') {
					acc |= (inAbc.get(c) << bits);
					bits += 5;
					if (bits >= 8) {
						res.add((byte) (acc & 0xFF));
						acc >>>= 8;
						bits -= 8;
					}
				}
			}

			if (bits >= 8)
				res.add((byte) (acc & 0xFF));
			return res;
		} catch (NullPointerException x) {
			throw new IllegalArgumentException("Invalid encoded string: " + src);
		}
	}

	/**
	 * Encode binary data into sergeych's base32 string.
	 * 
	 * @param data
	 *            source data. Can be empty but can not be null.
	 * @return encoded string
	 */
	static public String encode(byte[] data) {
		if (data.length == 0)
			return "";
		StringBuilder b = new StringBuilder();
		int acc = 0;
		int bits = 0;
		int count = 0;

		while (count < data.length) {
			int x = (data[count++] + 256) & 0xff;
			acc |= (x << bits);
			bits += 8;
			while (bits >= 5) {
				b.append(abc[acc & 0x1F]);
				bits -= 5;
				acc >>>= 5;
			}
		}
		if (bits > 0)
			b.append(abc[acc & 0x1f]);
		return b.toString();
	}

	/**
	 * Return the encoding alphabet (no aliases included).
	 * 
	 * @return alhabet as string, e.g. first character has weight of 0, second 1
	 *         and so on.
	 */
	public static String getAlphabet() {
		return abcString;
	}
}