com.wl4g.infra.common.codec.Base58 Maven / Gradle / Ivy
/*
* Copyright 2017 ~ 2025 the original author or authors. James Wong
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.wl4g.infra.common.codec;
import static com.google.common.base.Charsets.UTF_8;
import static org.apache.commons.lang3.StringUtils.EMPTY;
import static org.apache.commons.lang3.StringUtils.isBlank;
import java.io.UnsupportedEncodingException;
/**
* Base58 codec utility. {@link Base58}
*
* @author James Wong <[email protected]>
* @version 2020年3月25日 v1.0.0
* @see
*/
public abstract class Base58 {
/**
* Encodes the given bytes in base58. No checksum is appended.
*
* @param input
* @return
*/
public static String encodeBase58(String input) {
if (isBlank(input)) {
return EMPTY;
}
return encodeBase58(input.getBytes(UTF_8));
}
/**
* Encodes the given bytes in base58. No checksum is appended.
*
* @param input
* @return
*/
public static String encodeBase58(byte[] input) {
if (input.length == 0) {
return EMPTY;
}
input = copyOfRange(input, 0, input.length);
// Count leading zeroes.
int zeroCount = 0;
while (zeroCount < input.length && input[zeroCount] == 0) {
++zeroCount;
}
// The actual encoding.
byte[] temp = new byte[input.length * 2];
int j = temp.length;
int startAt = zeroCount;
while (startAt < input.length) {
byte mod = divmod58(input, startAt);
if (input[startAt] == 0) {
++startAt;
}
temp[--j] = (byte) ALPHABET[mod];
}
// Strip extra '1' if there are some after decoding.
while (j < temp.length && temp[j] == ALPHABET[0]) {
++j;
}
// Add as many leading '1' as there were leading zeros.
while (--zeroCount >= 0) {
temp[--j] = (byte) ALPHABET[0];
}
byte[] output = copyOfRange(temp, j, temp.length);
try {
return new String(output, "US-ASCII");
} catch (UnsupportedEncodingException e) {
throw new IllegalStateException(e); // Cannot happen.
}
}
/**
* Decode to base58 string.
*
* @param input
* @return
* @throws IllegalArgumentException
*/
public static byte[] decodeBase58(String input) throws IllegalArgumentException {
if (isBlank(input)) {
return new byte[0];
}
byte[] input58 = new byte[input.length()];
// Transform the String to a base58 byte sequence
for (int i = 0; i < input.length(); ++i) {
char c = input.charAt(i);
int digit58 = -1;
if (c >= 0 && c < 128) {
digit58 = INDEXES[c];
}
if (digit58 < 0) {
throw new IllegalArgumentException("Illegal character " + c + " at " + i);
}
input58[i] = (byte) digit58;
}
// Count leading zeroes
int zeroCount = 0;
while (zeroCount < input58.length && input58[zeroCount] == 0) {
++zeroCount;
}
// The encoding
byte[] temp = new byte[input.length()];
int j = temp.length;
int startAt = zeroCount;
while (startAt < input58.length) {
byte mod = divmod256(input58, startAt);
if (input58[startAt] == 0) {
++startAt;
}
temp[--j] = mod;
}
// Do no add extra leading zeroes, move j to first non null byte.
while (j < temp.length && temp[j] == 0) {
++j;
}
return copyOfRange(temp, j - zeroCount, temp.length);
}
/**
* number -> number / 58, returns number % 58
*
* @param number
* @param startAt
* @return
*/
private static byte divmod58(byte[] number, int startAt) {
int remainder = 0;
for (int i = startAt; i < number.length; i++) {
int digit256 = (int) number[i] & 0xFF;
int temp = remainder * 256 + digit256;
number[i] = (byte) (temp / 58);
remainder = temp % 58;
}
return (byte) remainder;
}
/**
* number -> number / 256, returns number % 256
*
* @param number58
* @param startAt
* @return
*/
private static byte divmod256(byte[] number58, int startAt) {
int remainder = 0;
for (int i = startAt; i < number58.length; i++) {
int digit58 = (int) number58[i] & 0xFF;
int temp = remainder * 58 + digit58;
number58[i] = (byte) (temp / 256);
remainder = temp % 256;
}
return (byte) remainder;
}
/**
* copyOfRange
*
* @param source
* @param from
* @param to
* @return
*/
private static byte[] copyOfRange(byte[] source, int from, int to) {
byte[] range = new byte[to - from];
System.arraycopy(source, from, range, 0, range.length);
return range;
}
final private static char[] ALPHABET = "123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz".toCharArray();
final private static int[] INDEXES = new int[128];
static {
for (int i = 0; i < INDEXES.length; i++) {
INDEXES[i] = -1;
}
for (int i = 0; i < ALPHABET.length; i++) {
INDEXES[ALPHABET[i]] = i;
}
}
}