001/* 002 * Licensed to the Apache Software Foundation (ASF) under one 003 * or more contributor license agreements. See the NOTICE file 004 * distributed with this work for additional information 005 * regarding copyright ownership. The ASF licenses this file 006 * to you under the Apache License, Version 2.0 (the 007 * "License"); you may not use this file except in compliance 008 * with the License. You may obtain a copy of the License at 009 * 010 * http://www.apache.org/licenses/LICENSE-2.0 011 * 012 * Unless required by applicable law or agreed to in writing, 013 * software distributed under the License is distributed on an 014 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY 015 * KIND, either express or implied. See the License for the 016 * specific language governing permissions and limitations 017 * under the License. 018 * 019 */ 020 021package org.apache.directory.api.util; 022 023 024import org.apache.directory.api.i18n.I18n; 025 026 027/** 028 * Encoding and decoding of Base64 characters to and from raw bytes. 029 * 030 * @author <a href="mailto:dev@directory.apache.org">Apache Directory Project</a> 031 */ 032public final class Base64 033{ 034 035 /** 036 * Private constructor. 037 */ 038 private Base64() 039 { 040 } 041 042 043 /** 044 * Encodes binary data to a Base64 encoded characters. 045 * 046 * @param data 047 * the array of bytes to encode 048 * @return base64-coded character array. 049 */ 050 public static char[] encode( byte[] data ) 051 { 052 char[] out = new char[( ( data.length + 2 ) / 3 ) * 4]; 053 054 // 055 // 3 bytes encode to 4 chars. Output is always an even 056 // multiple of 4 characters. 057 // 058 for ( int ii = 0, index = 0; ii < data.length; ii += 3, index += 4 ) 059 { 060 boolean isQuadrupel = false; 061 boolean isTripel = false; 062 063 int val = ( 0xFF & data[ii] ); 064 val <<= 8; 065 if ( ( ii + 1 ) < data.length ) 066 { 067 val |= ( 0xFF & data[ii + 1] ); 068 isTripel = true; 069 } 070 071 val <<= 8; 072 if ( ( ii + 2 ) < data.length ) 073 { 074 val |= ( 0xFF & data[ii + 2] ); 075 isQuadrupel = true; 076 } 077 078 out[index + 3] = ALPHABET[( isQuadrupel ? ( val & 0x3F ) : 64 )]; 079 val >>= 6; 080 out[index + 2] = ALPHABET[( isTripel ? ( val & 0x3F ) : 64 )]; 081 val >>= 6; 082 out[index + 1] = ALPHABET[val & 0x3F]; 083 val >>= 6; 084 out[index + 0] = ALPHABET[val & 0x3F]; 085 } 086 return out; 087 } 088 089 090 /** 091 * Decodes a BASE-64 encoded stream to recover the original data. White 092 * space before and after will be trimmed away, but no other manipulation of 093 * the input will be performed. As of version 1.2 this method will properly 094 * handle input containing junk characters (newlines and the like) rather 095 * than throwing an error. It does this by pre-parsing the input and 096 * generating from that a count of VALID input characters. 097 * 098 * @param data 099 * data to decode. 100 * @return the decoded binary data. 101 */ 102 public static byte[] decode( char[] data ) 103 { 104 // as our input could contain non-BASE64 data (newlines, 105 // whitespace of any sort, whatever) we must first adjust 106 // our count of USABLE data so that... 107 // (a) we don't misallocate the output array, and 108 // (b) think that we miscalculated our data length 109 // just because of extraneous throw-away junk 110 111 int tempLen = data.length; 112 113 for ( char c : data ) 114 { 115 if ( ( c > 255 ) || CODES[c] < 0 ) 116 { 117 --tempLen; // ignore non-valid chars and padding 118 } 119 } 120 // calculate required length: 121 // -- 3 bytes for every 4 valid base64 chars 122 // -- plus 2 bytes if there are 3 extra base64 chars, 123 // or plus 1 byte if there are 2 extra. 124 125 int len = ( tempLen / 4 ) * 3; 126 127 if ( ( tempLen % 4 ) == 3 ) 128 { 129 len += 2; 130 } 131 132 if ( ( tempLen % 4 ) == 2 ) 133 { 134 len += 1; 135 } 136 137 byte[] out = new byte[len]; 138 139 int shift = 0; // # of excess bits stored in accum 140 int accum = 0; // excess bits 141 int index = 0; 142 143 // we now go through the entire array (NOT using the 'tempLen' value) 144 for ( char c : data ) 145 { 146 int value = ( c > 255 ) ? -1 : CODES[c]; 147 148 if ( value >= 0 ) // skip over non-code 149 { 150 accum <<= 6; // bits shift up by 6 each time thru 151 shift += 6; // loop, with new bits being put in 152 accum |= value; // at the bottom. whenever there 153 if ( shift >= 8 ) // are 8 or more shifted in, write them 154 { 155 shift -= 8; // out (from the top, leaving any excess 156 out[index++] = // at the bottom for next iteration. 157 ( byte ) ( ( accum >> shift ) & 0xff ); 158 } 159 } 160 // we will also have skipped processing a padding null byte ('=') 161 // here; 162 // these are used ONLY for padding to an even length and do not 163 // legally 164 // occur as encoded data. for this reason we can ignore the fact 165 // that 166 // no index++ operation occurs in that special case: the out[] array 167 // is 168 // initialized to all-zero bytes to start with and that works to our 169 // advantage in this combination. 170 } 171 172 // if there is STILL something wrong we just have to throw up now! 173 if ( index != out.length ) 174 { 175 throw new Error( I18n.err( I18n.ERR_04348, index, out.length ) ); 176 } 177 178 return out; 179 } 180 181 /** code characters for values 0..63 */ 182 private static final char[] ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=" 183 .toCharArray(); 184 185 /** lookup table for converting base64 characters to value in range 0..63 */ 186 private static final byte[] CODES = new byte[256]; 187 188 static 189 { 190 for ( int ii = 0; ii < 256; ii++ ) 191 { 192 CODES[ii] = -1; 193 } 194 195 for ( int ii = 'A'; ii <= 'Z'; ii++ ) 196 { 197 CODES[ii] = ( byte ) ( ii - 'A' ); 198 } 199 200 for ( int ii = 'a'; ii <= 'z'; ii++ ) 201 { 202 CODES[ii] = ( byte ) ( 26 + ii - 'a' ); 203 } 204 205 for ( int ii = '0'; ii <= '9'; ii++ ) 206 { 207 CODES[ii] = ( byte ) ( 52 + ii - '0' ); 208 } 209 210 CODES['+'] = 62; 211 CODES['/'] = 63; 212 } 213}