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 *    https://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    /** code characters for values 0..63 */
036    private static final char[] ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/="
037        .toCharArray();
038
039    /** lookup table for converting base64 characters to value in range 0..63 */
040    private static final byte[] CODES = new byte[256];
041
042    static
043    {
044        for ( int ii = 0; ii < 256; ii++ )
045        {
046            CODES[ii] = -1;
047        }
048
049        for ( int ii = 'A'; ii <= 'Z'; ii++ )
050        {
051            CODES[ii] = ( byte ) ( ii - 'A' );
052        }
053
054        for ( int ii = 'a'; ii <= 'z'; ii++ )
055        {
056            CODES[ii] = ( byte ) ( 26 + ii - 'a' );
057        }
058
059        for ( int ii = '0'; ii <= '9'; ii++ )
060        {
061            CODES[ii] = ( byte ) ( 52 + ii - '0' );
062        }
063
064        CODES['+'] = 62;
065        CODES['/'] = 63;
066    }
067
068    /**
069     * Private constructor.
070     */
071    private Base64()
072    {
073    }
074
075
076    /**
077     * Encodes binary data to a Base64 encoded characters.
078     * 
079     * @param data
080     *            the array of bytes to encode
081     * @return base64-coded character array.
082     */
083    public static char[] encode( byte[] data )
084    {
085        char[] out = new char[( ( data.length + 2 ) / 3 ) * 4];
086
087        //
088        // 3 bytes encode to 4 chars. Output is always an even
089        // multiple of 4 characters.
090        //
091        for ( int i = 0, index = 0; i < data.length; i += 3, index += 4 )
092        {
093            boolean isQuadrupel = false;
094            boolean isTripel = false;
095
096            int val = 0xFF & data[i];
097            val <<= 8;
098            if ( ( i + 1 ) < data.length )
099            {
100                val |= ( 0xFF & data[i + 1] );
101                isTripel = true;
102            }
103
104            val <<= 8;
105            if ( ( i + 2 ) < data.length )
106            {
107                val |= ( 0xFF & data[i + 2] );
108                isQuadrupel = true;
109            }
110
111            out[index + 3] = ALPHABET[ isQuadrupel ? ( val & 0x3F ) : 64 ];
112            val >>= 6;
113            out[index + 2] = ALPHABET[ isTripel ? ( val & 0x3F ) : 64 ];
114            val >>= 6;
115            out[index + 1] = ALPHABET[val & 0x3F];
116            val >>= 6;
117            out[index + 0] = ALPHABET[val & 0x3F];
118        }
119        return out;
120    }
121
122
123    /**
124     * Decodes a BASE-64 encoded stream to recover the original data. White
125     * space before and after will be trimmed away, but no other manipulation of
126     * the input will be performed. As of version 1.2 this method will properly
127     * handle input containing junk characters (newlines and the like) rather
128     * than throwing an error. It does this by pre-parsing the input and
129     * generating from that a count of VALID input characters.
130     * 
131     * @param data
132     *            data to decode.
133     * @return the decoded binary data.
134     */
135    public static byte[] decode( char[] data )
136    {
137        // as our input could contain non-BASE64 data (newlines,
138        // whitespace of any sort, whatever) we must first adjust
139        // our count of USABLE data so that...
140        // (a) we don't misallocate the output array, and
141        // (b) think that we miscalculated our data length
142        // just because of extraneous throw-away junk
143
144        int tempLen = data.length;
145
146        for ( char c : data )
147        {
148            if ( ( c > 255 ) || CODES[c] < 0 )
149            {
150                // ignore non-valid chars and padding
151                --tempLen;
152            }
153        }
154        // calculate required length:
155        // -- 3 bytes for every 4 valid base64 chars
156        // -- plus 2 bytes if there are 3 extra base64 chars,
157        // or plus 1 byte if there are 2 extra.
158
159        int len = ( tempLen / 4 ) * 3;
160
161        if ( ( tempLen % 4 ) == 3 )
162        {
163            len += 2;
164        }
165
166        if ( ( tempLen % 4 ) == 2 )
167        {
168            len += 1;
169        }
170
171        byte[] out = new byte[len];
172
173        // # of excess bits stored in accum excess bits
174        int shift = 0;
175        int accum = 0;
176        int index = 0;
177
178        // we now go through the entire array (NOT using the 'tempLen' value)
179        for ( char c : data )
180        {
181            int value = ( c > 255 ) ? -1 : CODES[c];
182
183            // skip over non-code bits 
184            if ( value >= 0 )
185            {
186                // shift up by 6 each time thru
187                // loop, with new bits being put in
188                // at the bottom. whenever there
189                // are 8 or more shifted in, write them
190                // out (from the top, leaving any excess
191                // at the bottom for next iteration.
192                accum <<= 6;
193                shift += 6;
194                accum |= value;
195
196                if ( shift >= 8 )
197                {
198                    shift -= 8;
199                    out[index++] = ( byte ) ( ( accum >> shift ) & 0xff );
200                }
201            }
202            // we will also have skipped processing a padding null byte ('=') here;
203            // these are used ONLY for padding to an even length and do not legally
204            // occur as encoded data. for this reason we can ignore the fact
205            // that no index++ operation occurs in that special case: the out[] array
206            // is initialized to all-zero bytes to start with and that works to our
207            // advantage in this combination.
208        }
209
210        // if there is STILL something wrong we just have to throw up now!
211        if ( index != out.length )
212        {
213            throw new Error( I18n.err( I18n.ERR_17027_WRONG_DATA_LENGTH, index, out.length ) );
214        }
215
216        return out;
217    }
218}