View Javadoc
1   package com.randomnoun.common;
2   
3   /* (c) 2013 randomnoun. All Rights Reserved. This work is licensed under a
4    * BSD Simplified License. (http://www.randomnoun.com/bsd-simplified.html)
5    */
6   
7   // this class was taken from the widen valet (aws route53) package 
8   
9   /**
10   * <p>Encodes and decodes to and from Base64 notation.</p>
11   * <p>Homepage: <a href="http://iharder.net/base64">http://iharder.net/base64</a>.</p>
12   * 
13   * <p>Example:</p>
14   * 
15   * <p><code>String encoded = Base64.encode( myByteArray );</code></p>
16   * 
17   * <p><code>byte[] myByteArray = Base64.decode( encoded );</code></p>
18   *
19   * <p>The <code>options</code> parameter, which appears in a few places, is used to pass 
20   * several pieces of information to the encoder. In the "higher level" methods such as 
21   * encodeBytes( bytes, options ) the options parameter can be used to indicate such 
22   * things as first gzipping the bytes before encoding them, not inserting linefeeds,
23   * and encoding using the URL-safe and Ordered dialects.</p>
24   *
25   * <p>Note, according to <a href="http://www.faqs.org/rfcs/rfc3548.html">RFC3548</a>,
26   * Section 2.1, implementations should not add line feeds unless explicitly told
27   * to do so. I've got Base64 set to this behavior now, although earlier versions
28   * broke lines by default.</p>
29   *
30   * <p>The constants defined in Base64 can be OR-ed together to combine options, so you 
31   * might make a call like this:</p>
32   *
33   * <code>String encoded = Base64.encodeBytes( mybytes, Base64.GZIP | Base64.DO_BREAK_LINES );</code>
34   * <p>to compress the data before encoding it and then making the output have newline characters.</p>
35   * <p>Also...</p>
36   * <code>String encoded = Base64.encodeBytes( crazyString.getBytes() );</code>
37   *
38   *
39   *
40   * <p>
41   * Change Log:
42   * </p>
43   * <ul>
44   *  <li>v2.3.7 - Fixed subtle bug when base 64 input stream contained the
45   *   value 01111111, which is an invalid base 64 character but should not
46   *   throw an ArrayIndexOutOfBoundsException either. Led to discovery of
47   *   mishandling (or potential for better handling) of other bad input
48   *   characters. You should now get an IOException if you try decoding
49   *   something that has bad characters in it.</li>
50   *  <li>v2.3.6 - Fixed bug when breaking lines and the final byte of the encoded
51   *   string ended in the last column; the buffer was not properly shrunk and
52   *   contained an extra (null) byte that made it into the string.</li>
53   *  <li>v2.3.5 - Fixed bug in {@link #encodeFromFile} where estimated buffer size
54   *   was wrong for files of size 31, 34, and 37 bytes.</li>
55   *  <li>v2.3.4 - Fixed bug when working with gzipped streams whereby flushing
56   *   the Base64.OutputStream closed the Base64 encoding (by padding with equals
57   *   signs) too soon. Also added an option to suppress the automatic decoding
58   *   of gzipped streams. Also added experimental support for specifying a
59   *   class loader when using the
60   *   {@link #decodeToObject(String, int, ClassLoader)}
61   *   method.</li>
62   *  <li>v2.3.3 - Changed default char encoding to US-ASCII which reduces the internal Java
63   *   footprint with its CharEncoders and so forth. Fixed some javadocs that were
64   *   inconsistent. Removed imports and specified things like java.io.IOException
65   *   explicitly inline.</li>
66   *  <li>v2.3.2 - Reduced memory footprint! Finally refined the "guessing" of how big the
67   *   final encoded data will be so that the code doesn't have to create two output
68   *   arrays: an oversized initial one and then a final, exact-sized one. Big win
69   *   when using the {@link #encodeBytesToBytes(byte[])} family of methods (and not
70   *   using the gzip options which uses a different mechanism with streams and stuff).</li>
71   *  <li>v2.3.1 - Added {@link #encodeBytesToBytes(byte[], int, int, int)} and some
72   *   similar helper methods to be more efficient with memory by not returning a
73   *   String but just a byte array.</li>
74   *  <li>v2.3 - <strong>This is not a drop-in replacement!</strong> This is two years of comments
75   *   and bug fixes queued up and finally executed. Thanks to everyone who sent
76   *   me stuff, and I'm sorry I wasn't able to distribute your fixes to everyone else.
77   *   Much bad coding was cleaned up including throwing exceptions where necessary 
78   *   instead of returning null values or something similar. Here are some changes
79   *   that may affect you:
80   *   <ul>
81   *    <li><em>Does not break lines, by default.</em> This is to keep in compliance with
82   *      <a href="http://www.faqs.org/rfcs/rfc3548.html">RFC3548</a>.</li>
83   *    <li><em>Throws exceptions instead of returning null values.</em> Because some operations
84   *      (especially those that may permit the GZIP option) use IO streams, there
85   *      is a possiblity of an java.io.IOException being thrown. After some discussion and
86   *      thought, I've changed the behavior of the methods to throw java.io.IOExceptions
87   *      rather than return null if ever there's an error. I think this is more
88   *      appropriate, though it will require some changes to your code. Sorry,
89   *      it should have been done this way to begin with.</li>
90   *    <li><em>Removed all references to System.out, System.err, and the like.</em>
91   *      Shame on me. All I can say is sorry they were ever there.</li>
92   *    <li><em>Throws NullPointerExceptions and IllegalArgumentExceptions</em> as needed
93   *      such as when passed arrays are null or offsets are invalid.</li>
94   *    <li>Cleaned up as much javadoc as I could to avoid any javadoc warnings.
95   *      This was especially annoying before for people who were thorough in their
96   *      own projects and then had gobs of javadoc warnings on this file.</li>
97   *   </ul>
98   *  <li>v2.2.1 - Fixed bug using URL_SAFE and ORDERED encodings. Fixed bug
99   *   when using very small files (~&lt; 40 bytes).</li>
100  *  <li>v2.2 - Added some helper methods for encoding/decoding directly from
101  *   one file to the next. Also added a main() method to support command line
102  *   encoding/decoding from one file to the next. Also added these Base64 dialects:
103  *   <ol>
104  *   <li>The default is RFC3548 format.</li>
105  *   <li>Calling Base64.setFormat(Base64.BASE64_FORMAT.URLSAFE_FORMAT) generates
106  *   URL and file name friendly format as described in Section 4 of RFC3548.
107  *   http://www.faqs.org/rfcs/rfc3548.html</li>
108  *   <li>Calling Base64.setFormat(Base64.BASE64_FORMAT.ORDERED_FORMAT) generates
109  *   URL and file name friendly format that preserves lexical ordering as described
110  *   in http://www.faqs.org/qa/rfcc-1940.html</li>
111  *   </ol>
112  *   Special thanks to Jim Kellerman at <a href="http://www.powerset.com/">http://www.powerset.com/</a>
113  *   for contributing the new Base64 dialects.
114  *  </li>
115  * 
116  *  <li>v2.1 - Cleaned up javadoc comments and unused variables and methods. Added
117  *   some convenience methods for reading and writing to and from files.</li>
118  *  <li>v2.0.2 - Now specifies UTF-8 encoding in places where the code fails on systems
119  *   with other encodings (like EBCDIC).</li>
120  *  <li>v2.0.1 - Fixed an error when decoding a single byte, that is, when the
121  *   encoded data was a single byte.</li>
122  *  <li>v2.0 - I got rid of methods that used booleans to set options. 
123  *   Now everything is more consolidated and cleaner. The code now detects
124  *   when data that's being decoded is gzip-compressed and will decompress it
125  *   automatically. Generally things are cleaner. You'll probably have to
126  *   change some method calls that you were making to support the new
127  *   options format (<code>int</code>s that you "OR" together).</li>
128  *  <li>v1.5.1 - Fixed bug when decompressing and decoding to a             
129  *   byte[] using <code>decode( String s, boolean gzipCompressed )</code>.      
130  *   Added the ability to "suspend" encoding in the Output Stream so        
131  *   you can turn on and off the encoding if you need to embed base64       
132  *   data in an otherwise "normal" stream (like an XML file).</li>  
133  *  <li>v1.5 - Output stream pases on flush() command but doesn't do anything itself.
134  *      This helps when using GZIP streams.
135  *      Added the ability to GZip-compress objects before encoding them.</li>
136  *  <li>v1.4 - Added helper methods to read/write files.</li>
137  *  <li>v1.3.6 - Fixed OutputStream.flush() so that 'position' is reset.</li>
138  *  <li>v1.3.5 - Added flag to turn on and off line breaks. Fixed bug in input stream
139  *      where last buffer being read, if not completely full, was not returned.</li>
140  *  <li>v1.3.4 - Fixed when "improperly padded stream" error was thrown at the wrong time.</li>
141  *  <li>v1.3.3 - Fixed I/O streams which were totally messed up.</li>
142  * </ul>
143  *
144  * <p>
145  * I am placing this code in the Public Domain. Do with it as you will.
146  * This software comes with no guarantees or warranties but with
147  * plenty of well-wishing instead!
148  * Please visit <a href="http://iharder.net/base64">http://iharder.net/base64</a>
149  * periodically to check for updates or to contribute improvements.
150  * </p>
151  *
152  * @author Robert Harder
153  * @author rob@iharder.net
154  * @version 2.3.7
155  */
156 public class Base64
157 {
158     
159     
160 
161     
162 /* ********  P U B L I C   F I E L D S  ******** */   
163     
164     
165     /** No options specified. Value is zero. */
166     public final static int NO_OPTIONS = 0;
167     
168     /** Specify encoding in first bit. Value is one. */
169     public final static int ENCODE = 1;
170     
171     
172     /** Specify decoding in first bit. Value is zero. */
173     public final static int DECODE = 0;
174     
175 
176     /** Specify that data should be gzip-compressed in second bit. Value is two. */
177     public final static int GZIP = 2;
178 
179     /** Specify that gzipped data should <em>not</em> be automatically gunzipped. */
180     public final static int DONT_GUNZIP = 4;
181     
182     
183     /** Do break lines when encoding. Value is 8. */
184     public final static int DO_BREAK_LINES = 8;
185 	
186     /** 
187      * Encode using Base64-like encoding that is URL- and Filename-safe as described
188      * in Section 4 of RFC3548: 
189      * <a href="http://www.faqs.org/rfcs/rfc3548.html">http://www.faqs.org/rfcs/rfc3548.html</a>.
190      * It is important to note that data encoded this way is <em>not</em> officially valid Base64, 
191      * or at the very least should not be called Base64 without also specifying that is
192      * was encoded using the URL- and Filename-safe dialect.
193      */
194      public final static int URL_SAFE = 16;
195 
196 
197      /**
198       * Encode using the special "ordered" dialect of Base64 described here:
199       * <a href="http://www.faqs.org/qa/rfcc-1940.html">http://www.faqs.org/qa/rfcc-1940.html</a>.
200       */
201      public final static int ORDERED = 32;
202     
203     
204 /* ********  P R I V A T E   F I E L D S  ******** */  
205     
206     
207     /** Maximum line length (76) of Base64 output. */
208     private final static int MAX_LINE_LENGTH = 76;
209     
210     
211     /** The equals sign (=) as a byte. */
212     private final static byte EQUALS_SIGN = (byte)'=';
213     
214     
215     /** The new line character (\n) as a byte. */
216     private final static byte NEW_LINE = (byte)'\n';
217     
218     
219     /** Preferred encoding. */
220     private final static String PREFERRED_ENCODING = "US-ASCII";
221     
222 	
223     private final static byte WHITE_SPACE_ENC = -5; // Indicates white space in encoding
224     private final static byte EQUALS_SIGN_ENC = -1; // Indicates equals sign in encoding
225 	
226 	
227 /* ********  S T A N D A R D   B A S E 6 4   A L P H A B E T  ******** */	
228     
229     /** The 64 valid Base64 values. */
230     /* Host platform me be something funny like EBCDIC, so we hardcode these values. */
231     private final static byte[] _STANDARD_ALPHABET = {
232         (byte)'A', (byte)'B', (byte)'C', (byte)'D', (byte)'E', (byte)'F', (byte)'G',
233         (byte)'H', (byte)'I', (byte)'J', (byte)'K', (byte)'L', (byte)'M', (byte)'N',
234         (byte)'O', (byte)'P', (byte)'Q', (byte)'R', (byte)'S', (byte)'T', (byte)'U', 
235         (byte)'V', (byte)'W', (byte)'X', (byte)'Y', (byte)'Z',
236         (byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f', (byte)'g',
237         (byte)'h', (byte)'i', (byte)'j', (byte)'k', (byte)'l', (byte)'m', (byte)'n',
238         (byte)'o', (byte)'p', (byte)'q', (byte)'r', (byte)'s', (byte)'t', (byte)'u', 
239         (byte)'v', (byte)'w', (byte)'x', (byte)'y', (byte)'z',
240         (byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4', (byte)'5', 
241         (byte)'6', (byte)'7', (byte)'8', (byte)'9', (byte)'+', (byte)'/'
242     };
243 	
244     
245     /** 
246      * Translates a Base64 value to either its 6-bit reconstruction value
247      * or a negative number indicating some other meaning.
248      **/
249     private final static byte[] _STANDARD_DECODABET = {
250         -9,-9,-9,-9,-9,-9,-9,-9,-9,                 // Decimal  0 -  8
251         -5,-5,                                      // Whitespace: Tab and Linefeed
252         -9,-9,                                      // Decimal 11 - 12
253         -5,                                         // Whitespace: Carriage Return
254         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 14 - 26
255         -9,-9,-9,-9,-9,                             // Decimal 27 - 31
256         -5,                                         // Whitespace: Space
257         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,              // Decimal 33 - 42
258         62,                                         // Plus sign at decimal 43
259         -9,-9,-9,                                   // Decimal 44 - 46
260         63,                                         // Slash at decimal 47
261         52,53,54,55,56,57,58,59,60,61,              // Numbers zero through nine
262         -9,-9,-9,                                   // Decimal 58 - 60
263         -1,                                         // Equals sign at decimal 61
264         -9,-9,-9,                                      // Decimal 62 - 64
265         0,1,2,3,4,5,6,7,8,9,10,11,12,13,            // Letters 'A' through 'N'
266         14,15,16,17,18,19,20,21,22,23,24,25,        // Letters 'O' through 'Z'
267         -9,-9,-9,-9,-9,-9,                          // Decimal 91 - 96
268         26,27,28,29,30,31,32,33,34,35,36,37,38,     // Letters 'a' through 'm'
269         39,40,41,42,43,44,45,46,47,48,49,50,51,     // Letters 'n' through 'z'
270         -9,-9,-9,-9,-9                              // Decimal 123 - 127
271         ,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,       // Decimal 128 - 139
272         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 140 - 152
273         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 153 - 165
274         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 166 - 178
275         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 179 - 191
276         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 192 - 204
277         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 205 - 217
278         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 218 - 230
279         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 231 - 243
280         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9         // Decimal 244 - 255 
281     };
282 	
283 	
284 /* ********  U R L   S A F E   B A S E 6 4   A L P H A B E T  ******** */
285 	
286     /**
287      * Used in the URL- and Filename-safe dialect described in Section 4 of RFC3548: 
288      * <a href="http://www.faqs.org/rfcs/rfc3548.html">http://www.faqs.org/rfcs/rfc3548.html</a>.
289      * Notice that the last two bytes become "hyphen" and "underscore" instead of "plus" and "slash."
290      */
291     private final static byte[] _URL_SAFE_ALPHABET = {
292       (byte)'A', (byte)'B', (byte)'C', (byte)'D', (byte)'E', (byte)'F', (byte)'G',
293       (byte)'H', (byte)'I', (byte)'J', (byte)'K', (byte)'L', (byte)'M', (byte)'N',
294       (byte)'O', (byte)'P', (byte)'Q', (byte)'R', (byte)'S', (byte)'T', (byte)'U', 
295       (byte)'V', (byte)'W', (byte)'X', (byte)'Y', (byte)'Z',
296       (byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f', (byte)'g',
297       (byte)'h', (byte)'i', (byte)'j', (byte)'k', (byte)'l', (byte)'m', (byte)'n',
298       (byte)'o', (byte)'p', (byte)'q', (byte)'r', (byte)'s', (byte)'t', (byte)'u', 
299       (byte)'v', (byte)'w', (byte)'x', (byte)'y', (byte)'z',
300       (byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4', (byte)'5', 
301       (byte)'6', (byte)'7', (byte)'8', (byte)'9', (byte)'-', (byte)'_'
302     };
303 	
304     /**
305      * Used in decoding URL- and Filename-safe dialects of Base64.
306      */
307     private final static byte[] _URL_SAFE_DECODABET = {
308       -9,-9,-9,-9,-9,-9,-9,-9,-9,                 // Decimal  0 -  8
309       -5,-5,                                      // Whitespace: Tab and Linefeed
310       -9,-9,                                      // Decimal 11 - 12
311       -5,                                         // Whitespace: Carriage Return
312       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 14 - 26
313       -9,-9,-9,-9,-9,                             // Decimal 27 - 31
314       -5,                                         // Whitespace: Space
315       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,              // Decimal 33 - 42
316       -9,                                         // Plus sign at decimal 43
317       -9,                                         // Decimal 44
318       62,                                         // Minus sign at decimal 45
319       -9,                                         // Decimal 46
320       -9,                                         // Slash at decimal 47
321       52,53,54,55,56,57,58,59,60,61,              // Numbers zero through nine
322       -9,-9,-9,                                   // Decimal 58 - 60
323       -1,                                         // Equals sign at decimal 61
324       -9,-9,-9,                                   // Decimal 62 - 64
325       0,1,2,3,4,5,6,7,8,9,10,11,12,13,            // Letters 'A' through 'N'
326       14,15,16,17,18,19,20,21,22,23,24,25,        // Letters 'O' through 'Z'
327       -9,-9,-9,-9,                                // Decimal 91 - 94
328       63,                                         // Underscore at decimal 95
329       -9,                                         // Decimal 96
330       26,27,28,29,30,31,32,33,34,35,36,37,38,     // Letters 'a' through 'm'
331       39,40,41,42,43,44,45,46,47,48,49,50,51,     // Letters 'n' through 'z'
332       -9,-9,-9,-9,-9                              // Decimal 123 - 127
333       ,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 128 - 139
334       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 140 - 152
335       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 153 - 165
336       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 166 - 178
337       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 179 - 191
338       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 192 - 204
339       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 205 - 217
340       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 218 - 230
341       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 231 - 243
342       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9         // Decimal 244 - 255 
343     };
344 
345 
346 
347 /* ********  O R D E R E D   B A S E 6 4   A L P H A B E T  ******** */
348 
349     /**
350      * I don't get the point of this technique, but someone requested it,
351      * and it is described here:
352      * <a href="http://www.faqs.org/qa/rfcc-1940.html">http://www.faqs.org/qa/rfcc-1940.html</a>.
353      */
354     private final static byte[] _ORDERED_ALPHABET = {
355       (byte)'-',
356       (byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4',
357       (byte)'5', (byte)'6', (byte)'7', (byte)'8', (byte)'9',
358       (byte)'A', (byte)'B', (byte)'C', (byte)'D', (byte)'E', (byte)'F', (byte)'G',
359       (byte)'H', (byte)'I', (byte)'J', (byte)'K', (byte)'L', (byte)'M', (byte)'N',
360       (byte)'O', (byte)'P', (byte)'Q', (byte)'R', (byte)'S', (byte)'T', (byte)'U',
361       (byte)'V', (byte)'W', (byte)'X', (byte)'Y', (byte)'Z',
362       (byte)'_',
363       (byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f', (byte)'g',
364       (byte)'h', (byte)'i', (byte)'j', (byte)'k', (byte)'l', (byte)'m', (byte)'n',
365       (byte)'o', (byte)'p', (byte)'q', (byte)'r', (byte)'s', (byte)'t', (byte)'u',
366       (byte)'v', (byte)'w', (byte)'x', (byte)'y', (byte)'z'
367     };
368 	
369     /**
370      * Used in decoding the "ordered" dialect of Base64.
371      */
372     private final static byte[] _ORDERED_DECODABET = {
373       -9,-9,-9,-9,-9,-9,-9,-9,-9,                 // Decimal  0 -  8
374       -5,-5,                                      // Whitespace: Tab and Linefeed
375       -9,-9,                                      // Decimal 11 - 12
376       -5,                                         // Whitespace: Carriage Return
377       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 14 - 26
378       -9,-9,-9,-9,-9,                             // Decimal 27 - 31
379       -5,                                         // Whitespace: Space
380       -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,              // Decimal 33 - 42
381       -9,                                         // Plus sign at decimal 43
382       -9,                                         // Decimal 44
383       0,                                          // Minus sign at decimal 45
384       -9,                                         // Decimal 46
385       -9,                                         // Slash at decimal 47
386       1,2,3,4,5,6,7,8,9,10,                       // Numbers zero through nine
387       -9,-9,-9,                                   // Decimal 58 - 60
388       -1,                                         // Equals sign at decimal 61
389       -9,-9,-9,                                   // Decimal 62 - 64
390       11,12,13,14,15,16,17,18,19,20,21,22,23,     // Letters 'A' through 'M'
391       24,25,26,27,28,29,30,31,32,33,34,35,36,     // Letters 'N' through 'Z'
392       -9,-9,-9,-9,                                // Decimal 91 - 94
393       37,                                         // Underscore at decimal 95
394       -9,                                         // Decimal 96
395       38,39,40,41,42,43,44,45,46,47,48,49,50,     // Letters 'a' through 'm'
396       51,52,53,54,55,56,57,58,59,60,61,62,63,     // Letters 'n' through 'z'
397       -9,-9,-9,-9,-9                                 // Decimal 123 - 127
398        ,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 128 - 139
399         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 140 - 152
400         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 153 - 165
401         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 166 - 178
402         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 179 - 191
403         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 192 - 204
404         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 205 - 217
405         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 218 - 230
406         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,     // Decimal 231 - 243
407         -9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9,-9         // Decimal 244 - 255 
408     };
409 
410 	
411 /* ********  D E T E R M I N E   W H I C H   A L H A B E T  ******** */
412 
413 
414     /**
415      * Returns one of the _SOMETHING_ALPHABET byte arrays depending on
416      * the options specified.
417      * It's possible, though silly, to specify ORDERED <b>and</b> URLSAFE
418      * in which case one of them will be picked, though there is
419      * no guarantee as to which one will be picked.
420      */
421     private final static byte[] getAlphabet( int options ) {
422         if ((options & URL_SAFE) == URL_SAFE) {
423             return _URL_SAFE_ALPHABET;
424         } else if ((options & ORDERED) == ORDERED) {
425             return _ORDERED_ALPHABET;
426         } else {
427             return _STANDARD_ALPHABET;
428         }
429     }	// end getAlphabet
430 
431 
432     /**
433      * Returns one of the _SOMETHING_DECODABET byte arrays depending on
434      * the options specified.
435      * It's possible, though silly, to specify ORDERED and URL_SAFE
436      * in which case one of them will be picked, though there is
437      * no guarantee as to which one will be picked.
438      */
439     private final static byte[] getDecodabet( int options ) {
440         if( (options & URL_SAFE) == URL_SAFE) {
441             return _URL_SAFE_DECODABET;
442         } else if ((options & ORDERED) == ORDERED) {
443             return _ORDERED_DECODABET;
444         } else {
445             return _STANDARD_DECODABET;
446         }
447     }	// end getAlphabet
448 
449 
450     
451     /** Defeats instantiation. */
452     private Base64(){}
453     
454 
455     
456     
457 /* ********  E N C O D I N G   M E T H O D S  ******** */    
458     
459     
460     /**
461      * Encodes up to the first three bytes of array <var>threeBytes</var>
462      * and returns a four-byte array in Base64 notation.
463      * The actual number of significant bytes in your array is
464      * given by <var>numSigBytes</var>.
465      * The array <var>threeBytes</var> needs only be as big as
466      * <var>numSigBytes</var>.
467      * Code can reuse a byte array by passing a four-byte array as <var>b4</var>.
468      *
469      * @param b4 A reusable byte array to reduce array instantiation
470      * @param threeBytes the array to convert
471      * @param numSigBytes the number of significant bytes in your array
472      * @return four byte array in Base64 notation.
473      * @since 1.5.1
474      */
475     private static byte[] encode3to4( byte[] b4, byte[] threeBytes, int numSigBytes, int options ) {
476         encode3to4( threeBytes, 0, numSigBytes, b4, 0, options );
477         return b4;
478     }   // end encode3to4
479 
480     
481     /**
482      * <p>Encodes up to three bytes of the array <var>source</var>
483      * and writes the resulting four Base64 bytes to <var>destination</var>.
484      * The source and destination arrays can be manipulated
485      * anywhere along their length by specifying 
486      * <var>srcOffset</var> and <var>destOffset</var>.
487      * This method does not check to make sure your arrays
488      * are large enough to accomodate <var>srcOffset</var> + 3 for
489      * the <var>source</var> array or <var>destOffset</var> + 4 for
490      * the <var>destination</var> array.
491      * The actual number of significant bytes in your array is
492      * given by <var>numSigBytes</var>.</p>
493 	 * <p>This is the lowest level of the encoding methods with
494 	 * all possible parameters.</p>
495      *
496      * @param source the array to convert
497      * @param srcOffset the index where conversion begins
498      * @param numSigBytes the number of significant bytes in your array
499      * @param destination the array to hold the conversion
500      * @param destOffset the index where output will be put
501      * @return the <var>destination</var> array
502      * @since 1.3
503      */
504     private static byte[] encode3to4( 
505     byte[] source, int srcOffset, int numSigBytes,
506     byte[] destination, int destOffset, int options ) {
507         
508 	byte[] ALPHABET = getAlphabet( options ); 
509 	
510         //           1         2         3  
511         // 01234567890123456789012345678901 Bit position
512         // --------000000001111111122222222 Array position from threeBytes
513         // --------|    ||    ||    ||    | Six bit groups to index ALPHABET
514         //          >>18  >>12  >> 6  >> 0  Right shift necessary
515         //                0x3f  0x3f  0x3f  Additional AND
516         
517         // Create buffer with zero-padding if there are only one or two
518         // significant bytes passed in the array.
519         // We have to shift left 24 in order to flush out the 1's that appear
520         // when Java treats a value as negative that is cast from a byte to an int.
521         int inBuff =   ( numSigBytes > 0 ? ((source[ srcOffset     ] << 24) >>>  8) : 0 )
522                      | ( numSigBytes > 1 ? ((source[ srcOffset + 1 ] << 24) >>> 16) : 0 )
523                      | ( numSigBytes > 2 ? ((source[ srcOffset + 2 ] << 24) >>> 24) : 0 );
524 
525         switch( numSigBytes )
526         {
527             case 3:
528                 destination[ destOffset     ] = ALPHABET[ (inBuff >>> 18)        ];
529                 destination[ destOffset + 1 ] = ALPHABET[ (inBuff >>> 12) & 0x3f ];
530                 destination[ destOffset + 2 ] = ALPHABET[ (inBuff >>>  6) & 0x3f ];
531                 destination[ destOffset + 3 ] = ALPHABET[ (inBuff       ) & 0x3f ];
532                 return destination;
533                 
534             case 2:
535                 destination[ destOffset     ] = ALPHABET[ (inBuff >>> 18)        ];
536                 destination[ destOffset + 1 ] = ALPHABET[ (inBuff >>> 12) & 0x3f ];
537                 destination[ destOffset + 2 ] = ALPHABET[ (inBuff >>>  6) & 0x3f ];
538                 destination[ destOffset + 3 ] = EQUALS_SIGN;
539                 return destination;
540                 
541             case 1:
542                 destination[ destOffset     ] = ALPHABET[ (inBuff >>> 18)        ];
543                 destination[ destOffset + 1 ] = ALPHABET[ (inBuff >>> 12) & 0x3f ];
544                 destination[ destOffset + 2 ] = EQUALS_SIGN;
545                 destination[ destOffset + 3 ] = EQUALS_SIGN;
546                 return destination;
547                 
548             default:
549                 return destination;
550         }   // end switch
551     }   // end encode3to4
552 
553 
554 
555     /**
556      * Performs Base64 encoding on the <code>raw</code> ByteBuffer,
557      * writing it to the <code>encoded</code> ByteBuffer.
558      * This is an experimental feature. Currently it does not
559      * pass along any options (such as {@link #DO_BREAK_LINES}
560      * or {@link #GZIP}.
561      *
562      * @param raw input buffer
563      * @param encoded output buffer
564      * @since 2.3
565      */
566     public static void encode( java.nio.ByteBuffer raw, java.nio.ByteBuffer encoded ){
567         byte[] raw3 = new byte[3];
568         byte[] enc4 = new byte[4];
569 
570         while( raw.hasRemaining() ){
571             int rem = Math.min(3,raw.remaining());
572             raw.get(raw3,0,rem);
573             Base64.encode3to4(enc4, raw3, rem, Base64.NO_OPTIONS );
574             encoded.put(enc4);
575         }   // end input remaining
576     }
577 
578 
579     /**
580      * Performs Base64 encoding on the <code>raw</code> ByteBuffer,
581      * writing it to the <code>encoded</code> CharBuffer.
582      * This is an experimental feature. Currently it does not
583      * pass along any options (such as {@link #DO_BREAK_LINES}
584      * or {@link #GZIP}.
585      *
586      * @param raw input buffer
587      * @param encoded output buffer
588      * @since 2.3
589      */
590     public static void encode( java.nio.ByteBuffer raw, java.nio.CharBuffer encoded ){
591         byte[] raw3 = new byte[3];
592         byte[] enc4 = new byte[4];
593 
594         while( raw.hasRemaining() ){
595             int rem = Math.min(3,raw.remaining());
596             raw.get(raw3,0,rem);
597             Base64.encode3to4(enc4, raw3, rem, Base64.NO_OPTIONS );
598             for( int i = 0; i < 4; i++ ){
599                 encoded.put( (char)(enc4[i] & 0xFF) );
600             }
601         }   // end input remaining
602     }
603 
604 
605     
606     
607     /**
608      * Serializes an object and returns the Base64-encoded
609      * version of that serialized object.  
610      *  
611      * <p>As of v 2.3, if the object
612      * cannot be serialized or there is another error,
613      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
614      * In earlier versions, it just returned a null value, but
615      * in retrospect that's a pretty poor way to handle it.</p>
616      * 
617      * The object is not GZip-compressed before being encoded.
618      *
619      * @param serializableObject The object to encode
620      * @return The Base64-encoded object
621      * @throws java.io.IOException if there is an error
622      * @throws NullPointerException if serializedObject is null
623      * @since 1.4
624      */
625     public static String encodeObject( java.io.Serializable serializableObject )
626     throws java.io.IOException {
627         return encodeObject( serializableObject, NO_OPTIONS );
628     }   // end encodeObject
629     
630 
631 
632     /**
633      * Serializes an object and returns the Base64-encoded
634      * version of that serialized object.
635      *  
636      * <p>As of v 2.3, if the object
637      * cannot be serialized or there is another error,
638      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
639      * In earlier versions, it just returned a null value, but
640      * in retrospect that's a pretty poor way to handle it.</p>
641      * 
642      * The object is not GZip-compressed before being encoded.
643      * <p>
644      * Example options:<pre>
645      *   GZIP: gzip-compresses object before encoding it.
646      *   DO_BREAK_LINES: break lines at 76 characters
647      * </pre>
648      * <p>
649      * Example: <code>encodeObject( myObj, Base64.GZIP )</code> or
650      * <p>
651      * Example: <code>encodeObject( myObj, Base64.GZIP | Base64.DO_BREAK_LINES )</code>
652      *
653      * @param serializableObject The object to encode
654      * @param options Specified options
655      * @return The Base64-encoded object
656      * @see Base64#GZIP
657      * @see Base64#DO_BREAK_LINES
658      * @throws java.io.IOException if there is an error
659      * @since 2.0
660      */
661     public static String encodeObject( java.io.Serializable serializableObject, int options )
662     throws java.io.IOException {
663 
664         if( serializableObject == null ){
665             throw new NullPointerException( "Cannot serialize a null object." );
666         }   // end if: null
667         
668         // Streams
669         java.io.ByteArrayOutputStream  baos  = null; 
670         java.io.OutputStream           b64os = null;
671         java.util.zip.GZIPOutputStream gzos  = null;
672         java.io.ObjectOutputStream     oos   = null;
673         
674         
675         try {
676             // ObjectOutputStream -> (GZIP) -> Base64 -> ByteArrayOutputStream
677             baos  = new java.io.ByteArrayOutputStream();
678             b64os = new OutputStream( baos, ENCODE | options );
679             if( (options & GZIP) != 0 ){
680                 // Gzip
681                 gzos = new java.util.zip.GZIPOutputStream(b64os);
682                 oos = new java.io.ObjectOutputStream( gzos );
683             } else {
684                 // Not gzipped
685                 oos = new java.io.ObjectOutputStream( b64os );
686             }
687             oos.writeObject( serializableObject );
688         }   // end try
689         catch( java.io.IOException e ) {
690             // Catch it and then throw it immediately so that
691             // the finally{} block is called for cleanup.
692             throw e;
693         }   // end catch
694         finally {
695             try{ oos.close();   } catch( Exception e ){}
696             try{ gzos.close();  } catch( Exception e ){}
697             try{ b64os.close(); } catch( Exception e ){}
698             try{ baos.close();  } catch( Exception e ){}
699         }   // end finally
700         
701         // Return value according to relevant encoding.
702         try {
703             return new String( baos.toByteArray(), PREFERRED_ENCODING );
704         }   // end try
705         catch (java.io.UnsupportedEncodingException uue){
706             // Fall back to some Java default
707             return new String( baos.toByteArray() );
708         }   // end catch
709         
710     }   // end encode
711     
712     
713 
714     /**
715      * Encodes a byte array into Base64 notation.
716      * Does not GZip-compress data.
717      *  
718      * @param source The data to convert
719      * @return The data in Base64-encoded form
720      * @throws NullPointerException if source array is null
721      * @since 1.4
722      */
723     public static String encodeBytes( byte[] source ) {
724         // Since we're not going to have the GZIP encoding turned on,
725         // we're not going to have an java.io.IOException thrown, so
726         // we should not force the user to have to catch it.
727         String encoded = null;
728         try {
729             encoded = encodeBytes(source, 0, source.length, NO_OPTIONS);
730         } catch (java.io.IOException ex) {
731             assert false : ex.getMessage();
732         }   // end catch
733         assert encoded != null;
734         return encoded;
735     }   // end encodeBytes
736     
737 
738 
739     /**
740      * Encodes a byte array into Base64 notation.
741      * <p>
742      * Example options:<pre>
743      *   GZIP: gzip-compresses object before encoding it.
744      *   DO_BREAK_LINES: break lines at 76 characters
745      *     <i>Note: Technically, this makes your encoding non-compliant.</i>
746      * </pre>
747      * <p>
748      * Example: <code>encodeBytes( myData, Base64.GZIP )</code> or
749      * <p>
750      * Example: <code>encodeBytes( myData, Base64.GZIP | Base64.DO_BREAK_LINES )</code>
751      *
752      *  
753      * <p>As of v 2.3, if there is an error with the GZIP stream,
754      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
755      * In earlier versions, it just returned a null value, but
756      * in retrospect that's a pretty poor way to handle it.</p>
757      * 
758      *
759      * @param source The data to convert
760      * @param options Specified options
761      * @return The Base64-encoded data as a String
762      * @see Base64#GZIP
763      * @see Base64#DO_BREAK_LINES
764      * @throws java.io.IOException if there is an error
765      * @throws NullPointerException if source array is null
766      * @since 2.0
767      */
768     public static String encodeBytes( byte[] source, int options ) throws java.io.IOException {
769         return encodeBytes( source, 0, source.length, options );
770     }   // end encodeBytes
771     
772     
773     /**
774      * Encodes a byte array into Base64 notation.
775      * Does not GZip-compress data.
776      *  
777      * <p>As of v 2.3, if there is an error,
778      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
779      * In earlier versions, it just returned a null value, but
780      * in retrospect that's a pretty poor way to handle it.</p>
781      * 
782      *
783      * @param source The data to convert
784      * @param off Offset in array where conversion should begin
785      * @param len Length of data to convert
786      * @return The Base64-encoded data as a String
787      * @throws NullPointerException if source array is null
788      * @throws IllegalArgumentException if source array, offset, or length are invalid
789      * @since 1.4
790      */
791     public static String encodeBytes( byte[] source, int off, int len ) {
792         // Since we're not going to have the GZIP encoding turned on,
793         // we're not going to have an java.io.IOException thrown, so
794         // we should not force the user to have to catch it.
795         String encoded = null;
796         try {
797             encoded = encodeBytes( source, off, len, NO_OPTIONS );
798         } catch (java.io.IOException ex) {
799             assert false : ex.getMessage();
800         }   // end catch
801         assert encoded != null;
802         return encoded;
803     }   // end encodeBytes
804     
805     
806 
807     /**
808      * Encodes a byte array into Base64 notation.
809      * <p>
810      * Example options:<pre>
811      *   GZIP: gzip-compresses object before encoding it.
812      *   DO_BREAK_LINES: break lines at 76 characters
813      *     <i>Note: Technically, this makes your encoding non-compliant.</i>
814      * </pre>
815      * <p>
816      * Example: <code>encodeBytes( myData, Base64.GZIP )</code> or
817      * <p>
818      * Example: <code>encodeBytes( myData, Base64.GZIP | Base64.DO_BREAK_LINES )</code>
819      *
820      *  
821      * <p>As of v 2.3, if there is an error with the GZIP stream,
822      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
823      * In earlier versions, it just returned a null value, but
824      * in retrospect that's a pretty poor way to handle it.</p>
825      * 
826      *
827      * @param source The data to convert
828      * @param off Offset in array where conversion should begin
829      * @param len Length of data to convert
830      * @param options Specified options
831      * @return The Base64-encoded data as a String
832      * @see Base64#GZIP
833      * @see Base64#DO_BREAK_LINES
834      * @throws java.io.IOException if there is an error
835      * @throws NullPointerException if source array is null
836      * @throws IllegalArgumentException if source array, offset, or length are invalid
837      * @since 2.0
838      */
839     public static String encodeBytes( byte[] source, int off, int len, int options ) throws java.io.IOException {
840         byte[] encoded = encodeBytesToBytes( source, off, len, options );
841 
842         // Return value according to relevant encoding.
843         try {
844             return new String( encoded, PREFERRED_ENCODING );
845         }   // end try
846         catch (java.io.UnsupportedEncodingException uue) {
847             return new String( encoded );
848         }   // end catch
849         
850     }   // end encodeBytes
851 
852 
853 
854 
855     /**
856      * Similar to {@link #encodeBytes(byte[])} but returns
857      * a byte array instead of instantiating a String. This is more efficient
858      * if you're working with I/O streams and have large data sets to encode.
859      *
860      *
861      * @param source The data to convert
862      * @return The Base64-encoded data as a byte[] (of ASCII characters)
863      * @throws NullPointerException if source array is null
864      * @since 2.3.1
865      */
866     public static byte[] encodeBytesToBytes( byte[] source ) {
867         byte[] encoded = null;
868         try {
869             encoded = encodeBytesToBytes( source, 0, source.length, Base64.NO_OPTIONS );
870         } catch( java.io.IOException ex ) {
871             assert false : "IOExceptions only come from GZipping, which is turned off: " + ex.getMessage();
872         }
873         return encoded;
874     }
875 
876 
877     /**
878      * Similar to {@link #encodeBytes(byte[], int, int, int)} but returns
879      * a byte array instead of instantiating a String. This is more efficient
880      * if you're working with I/O streams and have large data sets to encode.
881      *
882      *
883      * @param source The data to convert
884      * @param off Offset in array where conversion should begin
885      * @param len Length of data to convert
886      * @param options Specified options
887      * @return The Base64-encoded data as a String
888      * @see Base64#GZIP
889      * @see Base64#DO_BREAK_LINES
890      * @throws java.io.IOException if there is an error
891      * @throws NullPointerException if source array is null
892      * @throws IllegalArgumentException if source array, offset, or length are invalid
893      * @since 2.3.1
894      */
895     public static byte[] encodeBytesToBytes( byte[] source, int off, int len, int options ) throws java.io.IOException {
896 
897         if( source == null ){
898             throw new NullPointerException( "Cannot serialize a null array." );
899         }   // end if: null
900 
901         if( off < 0 ){
902             throw new IllegalArgumentException( "Cannot have negative offset: " + off );
903         }   // end if: off < 0
904 
905         if( len < 0 ){
906             throw new IllegalArgumentException( "Cannot have length offset: " + len );
907         }   // end if: len < 0
908 
909         if( off + len > source.length  ){
910             throw new IllegalArgumentException(
911             String.format( "Cannot have offset of %d and length of %d with array of length %d", off,len,source.length));
912         }   // end if: off < 0
913 
914 
915 
916         // Compress?
917         if( (options & GZIP) != 0 ) {
918             java.io.ByteArrayOutputStream  baos  = null;
919             java.util.zip.GZIPOutputStream gzos  = null;
920             OutputStream            b64os = null;
921 
922             try {
923                 // GZip -> Base64 -> ByteArray
924                 baos = new java.io.ByteArrayOutputStream();
925                 b64os = new OutputStream( baos, ENCODE | options );
926                 gzos  = new java.util.zip.GZIPOutputStream( b64os );
927 
928                 gzos.write( source, off, len );
929                 gzos.close();
930             }   // end try
931             catch( java.io.IOException e ) {
932                 // Catch it and then throw it immediately so that
933                 // the finally{} block is called for cleanup.
934                 throw e;
935             }   // end catch
936             finally {
937                 try{ gzos.close();  } catch( Exception e ){}
938                 try{ b64os.close(); } catch( Exception e ){}
939                 try{ baos.close();  } catch( Exception e ){}
940             }   // end finally
941 
942             return baos.toByteArray();
943         }   // end if: compress
944 
945         // Else, don't compress. Better not to use streams at all then.
946         else {
947             boolean breakLines = (options & DO_BREAK_LINES) != 0;
948 
949             //int    len43   = len * 4 / 3;
950             //byte[] outBuff = new byte[   ( len43 )                      // Main 4:3
951             //                           + ( (len % 3) > 0 ? 4 : 0 )      // Account for padding
952             //                           + (breakLines ? ( len43 / MAX_LINE_LENGTH ) : 0) ]; // New lines
953             // Try to determine more precisely how big the array needs to be.
954             // If we get it right, we don't have to do an array copy, and
955             // we save a bunch of memory.
956             int encLen = ( len / 3 ) * 4 + ( len % 3 > 0 ? 4 : 0 ); // Bytes needed for actual encoding
957             if( breakLines ){
958                 encLen += encLen / MAX_LINE_LENGTH; // Plus extra newline characters
959             }
960             byte[] outBuff = new byte[ encLen ];
961 
962 
963             int d = 0;
964             int e = 0;
965             int len2 = len - 2;
966             int lineLength = 0;
967             for( ; d < len2; d+=3, e+=4 ) {
968                 encode3to4( source, d+off, 3, outBuff, e, options );
969 
970                 lineLength += 4;
971                 if( breakLines && lineLength >= MAX_LINE_LENGTH )
972                 {
973                     outBuff[e+4] = NEW_LINE;
974                     e++;
975                     lineLength = 0;
976                 }   // end if: end of line
977             }   // en dfor: each piece of array
978 
979             if( d < len ) {
980                 encode3to4( source, d+off, len - d, outBuff, e, options );
981                 e += 4;
982             }   // end if: some padding needed
983 
984 
985             // Only resize array if we didn't guess it right.
986             if( e <= outBuff.length - 1 ){
987                 // If breaking lines and the last byte falls right at
988                 // the line length (76 bytes per line), there will be
989                 // one extra byte, and the array will need to be resized.
990                 // Not too bad of an estimate on array size, I'd say.
991                 byte[] finalOut = new byte[e];
992                 System.arraycopy(outBuff,0, finalOut,0,e);
993                 //System.err.println("Having to resize array from " + outBuff.length + " to " + e );
994                 return finalOut;
995             } else {
996                 //System.err.println("No need to resize array.");
997                 return outBuff;
998             }
999         
1000         }   // end else: don't compress
1001 
1002     }   // end encodeBytesToBytes
1003     
1004 
1005     
1006     
1007     
1008 /* ********  D E C O D I N G   M E T H O D S  ******** */
1009     
1010     
1011     /**
1012      * Decodes four bytes from array <var>source</var>
1013      * and writes the resulting bytes (up to three of them)
1014      * to <var>destination</var>.
1015      * The source and destination arrays can be manipulated
1016      * anywhere along their length by specifying 
1017      * <var>srcOffset</var> and <var>destOffset</var>.
1018      * This method does not check to make sure your arrays
1019      * are large enough to accomodate <var>srcOffset</var> + 4 for
1020      * the <var>source</var> array or <var>destOffset</var> + 3 for
1021      * the <var>destination</var> array.
1022      * This method returns the actual number of bytes that 
1023      * were converted from the Base64 encoding.
1024 	 * <p>This is the lowest level of the decoding methods with
1025 	 * all possible parameters.</p>
1026      * 
1027      *
1028      * @param source the array to convert
1029      * @param srcOffset the index where conversion begins
1030      * @param destination the array to hold the conversion
1031      * @param destOffset the index where output will be put
1032 	 * @param options alphabet type is pulled from this (standard, url-safe, ordered)
1033      * @return the number of decoded bytes converted
1034      * @throws NullPointerException if source or destination arrays are null
1035      * @throws IllegalArgumentException if srcOffset or destOffset are invalid
1036      *         or there is not enough room in the array.
1037      * @since 1.3
1038      */
1039     private static int decode4to3( 
1040     byte[] source, int srcOffset, 
1041     byte[] destination, int destOffset, int options ) {
1042         
1043         // Lots of error checking and exception throwing
1044         if( source == null ){
1045             throw new NullPointerException( "Source array was null." );
1046         }   // end if
1047         if( destination == null ){
1048             throw new NullPointerException( "Destination array was null." );
1049         }   // end if
1050         if( srcOffset < 0 || srcOffset + 3 >= source.length ){
1051             throw new IllegalArgumentException( String.format(
1052             "Source array with length %d cannot have offset of %d and still process four bytes.", source.length, srcOffset ) );
1053         }   // end if
1054         if( destOffset < 0 || destOffset +2 >= destination.length ){
1055             throw new IllegalArgumentException( String.format(
1056             "Destination array with length %d cannot have offset of %d and still store three bytes.", destination.length, destOffset ) );
1057         }   // end if
1058         
1059         
1060         byte[] DECODABET = getDecodabet( options ); 
1061 	
1062         // Example: Dk==
1063         if( source[ srcOffset + 2] == EQUALS_SIGN ) {
1064             // Two ways to do the same thing. Don't know which way I like best.
1065           //int outBuff =   ( ( DECODABET[ source[ srcOffset    ] ] << 24 ) >>>  6 )
1066           //              | ( ( DECODABET[ source[ srcOffset + 1] ] << 24 ) >>> 12 );
1067             int outBuff =   ( ( DECODABET[ source[ srcOffset    ] ] & 0xFF ) << 18 )
1068                           | ( ( DECODABET[ source[ srcOffset + 1] ] & 0xFF ) << 12 );
1069             
1070             destination[ destOffset ] = (byte)( outBuff >>> 16 );
1071             return 1;
1072         }
1073         
1074         // Example: DkL=
1075         else if( source[ srcOffset + 3 ] == EQUALS_SIGN ) {
1076             // Two ways to do the same thing. Don't know which way I like best.
1077           //int outBuff =   ( ( DECODABET[ source[ srcOffset     ] ] << 24 ) >>>  6 )
1078           //              | ( ( DECODABET[ source[ srcOffset + 1 ] ] << 24 ) >>> 12 )
1079           //              | ( ( DECODABET[ source[ srcOffset + 2 ] ] << 24 ) >>> 18 );
1080             int outBuff =   ( ( DECODABET[ source[ srcOffset     ] ] & 0xFF ) << 18 )
1081                           | ( ( DECODABET[ source[ srcOffset + 1 ] ] & 0xFF ) << 12 )
1082                           | ( ( DECODABET[ source[ srcOffset + 2 ] ] & 0xFF ) <<  6 );
1083             
1084             destination[ destOffset     ] = (byte)( outBuff >>> 16 );
1085             destination[ destOffset + 1 ] = (byte)( outBuff >>>  8 );
1086             return 2;
1087         }
1088         
1089         // Example: DkLE
1090         else {
1091             // Two ways to do the same thing. Don't know which way I like best.
1092           //int outBuff =   ( ( DECODABET[ source[ srcOffset     ] ] << 24 ) >>>  6 )
1093           //              | ( ( DECODABET[ source[ srcOffset + 1 ] ] << 24 ) >>> 12 )
1094           //              | ( ( DECODABET[ source[ srcOffset + 2 ] ] << 24 ) >>> 18 )
1095           //              | ( ( DECODABET[ source[ srcOffset + 3 ] ] << 24 ) >>> 24 );
1096             int outBuff =   ( ( DECODABET[ source[ srcOffset     ] ] & 0xFF ) << 18 )
1097                           | ( ( DECODABET[ source[ srcOffset + 1 ] ] & 0xFF ) << 12 )
1098                           | ( ( DECODABET[ source[ srcOffset + 2 ] ] & 0xFF ) <<  6)
1099                           | ( ( DECODABET[ source[ srcOffset + 3 ] ] & 0xFF )      );
1100 
1101             
1102             destination[ destOffset     ] = (byte)( outBuff >> 16 );
1103             destination[ destOffset + 1 ] = (byte)( outBuff >>  8 );
1104             destination[ destOffset + 2 ] = (byte)( outBuff       );
1105 
1106             return 3;
1107         }
1108     }   // end decodeToBytes
1109     
1110 
1111 
1112 
1113 
1114     /**
1115      * Low-level access to decoding ASCII characters in
1116      * the form of a byte array. <strong>Ignores GUNZIP option, if
1117      * it's set.</strong> This is not generally a recommended method,
1118      * although it is used internally as part of the decoding process.
1119      * Special case: if len = 0, an empty array is returned. Still,
1120      * if you need more speed and reduced memory footprint (and aren't
1121      * gzipping), consider this method.
1122      *
1123      * @param source The Base64 encoded data
1124      * @return decoded data
1125      * @since 2.3.1
1126      */
1127     public static byte[] decode( byte[] source )
1128     throws java.io.IOException {
1129         byte[] decoded = null;
1130 //        try {
1131             decoded = decode( source, 0, source.length, Base64.NO_OPTIONS );
1132 //        } catch( java.io.IOException ex ) {
1133 //            assert false : "IOExceptions only come from GZipping, which is turned off: " + ex.getMessage();
1134 //        }
1135         return decoded;
1136     }
1137 
1138     
1139     
1140     /**
1141      * Low-level access to decoding ASCII characters in
1142      * the form of a byte array. <strong>Ignores GUNZIP option, if
1143      * it's set.</strong> This is not generally a recommended method,
1144      * although it is used internally as part of the decoding process.
1145      * Special case: if len = 0, an empty array is returned. Still,
1146      * if you need more speed and reduced memory footprint (and aren't
1147      * gzipping), consider this method.
1148      *
1149      * @param source The Base64 encoded data
1150      * @param off    The offset of where to begin decoding
1151      * @param len    The length of characters to decode
1152      * @param options Can specify options such as alphabet type to use
1153      * @return decoded data
1154      * @throws java.io.IOException If bogus characters exist in source data
1155      * @since 1.3
1156      */
1157     public static byte[] decode( byte[] source, int off, int len, int options )
1158     throws java.io.IOException {
1159         
1160         // Lots of error checking and exception throwing
1161         if( source == null ){
1162             throw new NullPointerException( "Cannot decode null source array." );
1163         }   // end if
1164         if( off < 0 || off + len > source.length ){
1165             throw new IllegalArgumentException( String.format(
1166             "Source array with length %d cannot have offset of %d and process %d bytes.", source.length, off, len ) );
1167         }   // end if
1168         
1169         if( len == 0 ){
1170             return new byte[0];
1171         }else if( len < 4 ){
1172             throw new IllegalArgumentException( 
1173             "Base64-encoded string must have at least four characters, but length specified was " + len );
1174         }   // end if
1175         
1176         byte[] DECODABET = getDecodabet( options );
1177 	
1178         int    len34   = len * 3 / 4;       // Estimate on array size
1179         byte[] outBuff = new byte[ len34 ]; // Upper limit on size of output
1180         int    outBuffPosn = 0;             // Keep track of where we're writing
1181         
1182         byte[] b4        = new byte[4];     // Four byte buffer from source, eliminating white space
1183         int    b4Posn    = 0;               // Keep track of four byte input buffer
1184         int    i         = 0;               // Source array counter
1185         byte   sbiDecode = 0;               // Special value from DECODABET
1186         
1187         for( i = off; i < off+len; i++ ) {  // Loop through source
1188             
1189             sbiDecode = DECODABET[ source[i]&0xFF ];
1190             
1191             // White space, Equals sign, or legit Base64 character
1192             // Note the values such as -5 and -9 in the
1193             // DECODABETs at the top of the file.
1194             if( sbiDecode >= WHITE_SPACE_ENC )  {
1195                 if( sbiDecode >= EQUALS_SIGN_ENC ) {
1196                     b4[ b4Posn++ ] = source[i];         // Save non-whitespace
1197                     if( b4Posn > 3 ) {                  // Time to decode?
1198                         outBuffPosn += decode4to3( b4, 0, outBuff, outBuffPosn, options );
1199                         b4Posn = 0;
1200                         
1201                         // If that was the equals sign, break out of 'for' loop
1202                         if( source[i] == EQUALS_SIGN ) {
1203                             break;
1204                         }   // end if: equals sign
1205                     }   // end if: quartet built
1206                 }   // end if: equals sign or better
1207             }   // end if: white space, equals sign or better
1208             else {
1209                 // There's a bad input character in the Base64 stream.
1210                 throw new java.io.IOException( String.format(
1211                 "Bad Base64 input character decimal %d in array position %d", ((int)source[i])&0xFF, i ) );
1212             }   // end else: 
1213         }   // each input character
1214                                    
1215         byte[] out = new byte[ outBuffPosn ];
1216         System.arraycopy( outBuff, 0, out, 0, outBuffPosn ); 
1217         return out;
1218     }   // end decode
1219     
1220     
1221 	
1222 	
1223     /**
1224      * Decodes data from Base64 notation, automatically
1225      * detecting gzip-compressed data and decompressing it.
1226      *
1227      * @param s the string to decode
1228      * @return the decoded data
1229      * @throws java.io.IOException If there is a problem
1230      * @since 1.4
1231      */
1232     public static byte[] decode( String s ) throws java.io.IOException {
1233         return decode( s, NO_OPTIONS );
1234     }
1235 
1236     
1237     
1238     /**
1239      * Decodes data from Base64 notation, automatically
1240      * detecting gzip-compressed data and decompressing it.
1241      *
1242      * @param s the string to decode
1243      * @param options encode options such as URL_SAFE
1244      * @return the decoded data
1245      * @throws java.io.IOException if there is an error
1246      * @throws NullPointerException if <code>s</code> is null
1247      * @since 1.4
1248      */
1249     public static byte[] decode( String s, int options ) throws java.io.IOException {
1250         
1251         if( s == null ){
1252             throw new NullPointerException( "Input string was null." );
1253         }   // end if
1254         
1255         byte[] bytes;
1256         try {
1257             bytes = s.getBytes( PREFERRED_ENCODING );
1258         }   // end try
1259         catch( java.io.UnsupportedEncodingException uee ) {
1260             bytes = s.getBytes();
1261         }   // end catch
1262 		//</change>
1263         
1264         // Decode
1265         bytes = decode( bytes, 0, bytes.length, options );
1266         
1267         // Check to see if it's gzip-compressed
1268         // GZIP Magic Two-Byte Number: 0x8b1f (35615)
1269         boolean dontGunzip = (options & DONT_GUNZIP) != 0;
1270         if( (bytes != null) && (bytes.length >= 4) && (!dontGunzip) ) {
1271             
1272             int head = ((int)bytes[0] & 0xff) | ((bytes[1] << 8) & 0xff00);
1273             if( java.util.zip.GZIPInputStream.GZIP_MAGIC == head )  {
1274                 java.io.ByteArrayInputStream  bais = null;
1275                 java.util.zip.GZIPInputStream gzis = null;
1276                 java.io.ByteArrayOutputStream baos = null;
1277                 byte[] buffer = new byte[2048];
1278                 int    length = 0;
1279 
1280                 try {
1281                     baos = new java.io.ByteArrayOutputStream();
1282                     bais = new java.io.ByteArrayInputStream( bytes );
1283                     gzis = new java.util.zip.GZIPInputStream( bais );
1284 
1285                     while( ( length = gzis.read( buffer ) ) >= 0 ) {
1286                         baos.write(buffer,0,length);
1287                     }   // end while: reading input
1288 
1289                     // No error? Get new bytes.
1290                     bytes = baos.toByteArray();
1291 
1292                 }   // end try
1293                 catch( java.io.IOException e ) {
1294                     e.printStackTrace();
1295                     // Just return originally-decoded bytes
1296                 }   // end catch
1297                 finally {
1298                     try{ baos.close(); } catch( Exception e ){}
1299                     try{ gzis.close(); } catch( Exception e ){}
1300                     try{ bais.close(); } catch( Exception e ){}
1301                 }   // end finally
1302 
1303             }   // end if: gzipped
1304         }   // end if: bytes.length >= 2
1305         
1306         return bytes;
1307     }   // end decode
1308 
1309 
1310 
1311     /**
1312      * Attempts to decode Base64 data and deserialize a Java
1313      * Object within. Returns <code>null</code> if there was an error.
1314      *
1315      * @param encodedObject The Base64 data to decode
1316      * @return The decoded and deserialized object
1317      * @throws NullPointerException if encodedObject is null
1318      * @throws java.io.IOException if there is a general error
1319      * @throws ClassNotFoundException if the decoded object is of a
1320      *         class that cannot be found by the JVM
1321      * @since 1.5
1322      */
1323     public static Object decodeToObject( String encodedObject )
1324     throws java.io.IOException, ClassNotFoundException {
1325         return decodeToObject(encodedObject,NO_OPTIONS,null);
1326     }
1327     
1328 
1329     /**
1330      * Attempts to decode Base64 data and deserialize a Java
1331      * Object within. Returns <code>null</code> if there was an error.
1332      * If <code>loader</code> is not null, it will be the class loader
1333      * used when deserializing.
1334      *
1335      * @param encodedObject The Base64 data to decode
1336      * @param options Various parameters related to decoding
1337      * @param loader Optional class loader to use in deserializing classes.
1338      * @return The decoded and deserialized object
1339      * @throws NullPointerException if encodedObject is null
1340      * @throws java.io.IOException if there is a general error
1341      * @throws ClassNotFoundException if the decoded object is of a 
1342      *         class that cannot be found by the JVM
1343      * @since 2.3.4
1344      */
1345     public static Object decodeToObject( 
1346     String encodedObject, int options, final ClassLoader loader )
1347     throws java.io.IOException, ClassNotFoundException {
1348         
1349         // Decode and gunzip if necessary
1350         byte[] objBytes = decode( encodedObject, options );
1351         
1352         java.io.ByteArrayInputStream  bais = null;
1353         java.io.ObjectInputStream     ois  = null;
1354         Object obj = null;
1355         
1356         try {
1357             bais = new java.io.ByteArrayInputStream( objBytes );
1358 
1359             // If no custom class loader is provided, use Java's builtin OIS.
1360             if( loader == null ){
1361                 ois  = new java.io.ObjectInputStream( bais );
1362             }   // end if: no loader provided
1363 
1364             // Else make a customized object input stream that uses
1365             // the provided class loader.
1366             else {
1367                 ois = new java.io.ObjectInputStream(bais){
1368                     @Override
1369                     public Class<?> resolveClass(java.io.ObjectStreamClass streamClass)
1370                     throws java.io.IOException, ClassNotFoundException {
1371                         Class<?> c = Class.forName(streamClass.getName(), false, loader);
1372                         if( c == null ){
1373                             return super.resolveClass(streamClass);
1374                         } else {
1375                             return c;   // Class loader knows of this class.
1376                         }   // end else: not null
1377                     }   // end resolveClass
1378                 };  // end ois
1379             }   // end else: no custom class loader
1380         
1381             obj = ois.readObject();
1382         }   // end try
1383         catch( java.io.IOException e ) {
1384             throw e;    // Catch and throw in order to execute finally{}
1385         }   // end catch
1386         catch( ClassNotFoundException e ) {
1387             throw e;    // Catch and throw in order to execute finally{}
1388         }   // end catch
1389         finally {
1390             try{ bais.close(); } catch( Exception e ){}
1391             try{ ois.close();  } catch( Exception e ){}
1392         }   // end finally
1393         
1394         return obj;
1395     }   // end decodeObject
1396     
1397     
1398     
1399     /**
1400      * Convenience method for encoding data to a file.
1401      *
1402      * <p>As of v 2.3, if there is a error,
1403      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
1404      * In earlier versions, it just returned false, but
1405      * in retrospect that's a pretty poor way to handle it.</p>
1406      * 
1407      * @param dataToEncode byte array of data to encode in base64 form
1408      * @param filename Filename for saving encoded data
1409      * @throws java.io.IOException if there is an error
1410      * @throws NullPointerException if dataToEncode is null
1411      * @since 2.1
1412      */
1413     public static void encodeToFile( byte[] dataToEncode, String filename )
1414     throws java.io.IOException {
1415         
1416         if( dataToEncode == null ){
1417             throw new NullPointerException( "Data to encode was null." );
1418         }   // end iff
1419         
1420         OutputStream bos = null;
1421         try {
1422             bos = new OutputStream(
1423                   new java.io.FileOutputStream( filename ), Base64.ENCODE );
1424             bos.write( dataToEncode );
1425         }   // end try
1426         catch( java.io.IOException e ) {
1427             throw e; // Catch and throw to execute finally{} block
1428         }   // end catch: java.io.IOException
1429         finally {
1430             try{ bos.close(); } catch( Exception e ){}
1431         }   // end finally
1432         
1433     }   // end encodeToFile
1434     
1435     
1436     /**
1437      * Convenience method for decoding data to a file.
1438      *
1439      * <p>As of v 2.3, if there is a error,
1440      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
1441      * In earlier versions, it just returned false, but
1442      * in retrospect that's a pretty poor way to handle it.</p>
1443      * 
1444      * @param dataToDecode Base64-encoded data as a string
1445      * @param filename Filename for saving decoded data
1446      * @throws java.io.IOException if there is an error
1447      * @since 2.1
1448      */
1449     public static void decodeToFile( String dataToDecode, String filename )
1450     throws java.io.IOException {
1451         
1452         OutputStream bos = null;
1453         try{
1454             bos = new OutputStream(
1455                       new java.io.FileOutputStream( filename ), Base64.DECODE );
1456             bos.write( dataToDecode.getBytes( PREFERRED_ENCODING ) );
1457         }   // end try
1458         catch( java.io.IOException e ) {
1459             throw e; // Catch and throw to execute finally{} block
1460         }   // end catch: java.io.IOException
1461         finally {
1462                 try{ bos.close(); } catch( Exception e ){}
1463         }   // end finally
1464         
1465     }   // end decodeToFile
1466     
1467     
1468     
1469     
1470     /**
1471      * Convenience method for reading a base64-encoded
1472      * file and decoding it.
1473      *
1474      * <p>As of v 2.3, if there is a error,
1475      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
1476      * In earlier versions, it just returned false, but
1477      * in retrospect that's a pretty poor way to handle it.</p>
1478      * 
1479      * @param filename Filename for reading encoded data
1480      * @return decoded byte array
1481      * @throws java.io.IOException if there is an error
1482      * @since 2.1
1483      */
1484     public static byte[] decodeFromFile( String filename )
1485     throws java.io.IOException {
1486         
1487         byte[] decodedData = null;
1488         InputStream bis = null;
1489         try
1490         {
1491             // Set up some useful variables
1492             java.io.File file = new java.io.File( filename );
1493             byte[] buffer = null;
1494             int length   = 0;
1495             int numBytes = 0;
1496             
1497             // Check for size of file
1498             if( file.length() > Integer.MAX_VALUE )
1499             {
1500                 throw new java.io.IOException( "File is too big for this convenience method (" + file.length() + " bytes)." );
1501             }   // end if: file too big for int index
1502             buffer = new byte[ (int)file.length() ];
1503             
1504             // Open a stream
1505             bis = new InputStream(
1506                       new java.io.BufferedInputStream( 
1507                       new java.io.FileInputStream( file ) ), Base64.DECODE );
1508             
1509             // Read until done
1510             while( ( numBytes = bis.read( buffer, length, 4096 ) ) >= 0 ) {
1511                 length += numBytes;
1512             }   // end while
1513             
1514             // Save in a variable to return
1515             decodedData = new byte[ length ];
1516             System.arraycopy( buffer, 0, decodedData, 0, length );
1517             
1518         }   // end try
1519         catch( java.io.IOException e ) {
1520             throw e; // Catch and release to execute finally{}
1521         }   // end catch: java.io.IOException
1522         finally {
1523             try{ bis.close(); } catch( Exception e) {}
1524         }   // end finally
1525         
1526         return decodedData;
1527     }   // end decodeFromFile
1528     
1529     
1530     
1531     /**
1532      * Convenience method for reading a binary file
1533      * and base64-encoding it.
1534      *
1535      * <p>As of v 2.3, if there is a error,
1536      * the method will throw an java.io.IOException. <b>This is new to v2.3!</b>
1537      * In earlier versions, it just returned false, but
1538      * in retrospect that's a pretty poor way to handle it.</p>
1539      * 
1540      * @param filename Filename for reading binary data
1541      * @return base64-encoded string
1542      * @throws java.io.IOException if there is an error
1543      * @since 2.1
1544      */
1545     public static String encodeFromFile( String filename )
1546     throws java.io.IOException {
1547         
1548         String encodedData = null;
1549         InputStream bis = null;
1550         try
1551         {
1552             // Set up some useful variables
1553             java.io.File file = new java.io.File( filename );
1554             byte[] buffer = new byte[ Math.max((int)(file.length() * 1.4+1),40) ]; // Need max() for math on small files (v2.2.1); Need +1 for a few corner cases (v2.3.5)
1555             int length   = 0;
1556             int numBytes = 0;
1557             
1558             // Open a stream
1559             bis = new InputStream(
1560                       new java.io.BufferedInputStream( 
1561                       new java.io.FileInputStream( file ) ), Base64.ENCODE );
1562             
1563             // Read until done
1564             while( ( numBytes = bis.read( buffer, length, 4096 ) ) >= 0 ) {
1565                 length += numBytes;
1566             }   // end while
1567             
1568             // Save in a variable to return
1569             encodedData = new String( buffer, 0, length, Base64.PREFERRED_ENCODING );
1570                 
1571         }   // end try
1572         catch( java.io.IOException e ) {
1573             throw e; // Catch and release to execute finally{}
1574         }   // end catch: java.io.IOException
1575         finally {
1576             try{ bis.close(); } catch( Exception e) {}
1577         }   // end finally
1578         
1579         return encodedData;
1580         }   // end encodeFromFile
1581     
1582     /**
1583      * Reads <code>infile</code> and encodes it to <code>outfile</code>.
1584      *
1585      * @param infile Input file
1586      * @param outfile Output file
1587      * @throws java.io.IOException if there is an error
1588      * @since 2.2
1589      */
1590     public static void encodeFileToFile( String infile, String outfile )
1591     throws java.io.IOException {
1592         
1593         String encoded = Base64.encodeFromFile( infile );
1594         java.io.OutputStream out = null;
1595         try{
1596             out = new java.io.BufferedOutputStream(
1597                   new java.io.FileOutputStream( outfile ) );
1598             out.write( encoded.getBytes("US-ASCII") ); // Strict, 7-bit output.
1599         }   // end try
1600         catch( java.io.IOException e ) {
1601             throw e; // Catch and release to execute finally{}
1602         }   // end catch
1603         finally {
1604             try { out.close(); }
1605             catch( Exception ex ){}
1606         }   // end finally    
1607     }   // end encodeFileToFile
1608 
1609 
1610     /**
1611      * Reads <code>infile</code> and decodes it to <code>outfile</code>.
1612      *
1613      * @param infile Input file
1614      * @param outfile Output file
1615      * @throws java.io.IOException if there is an error
1616      * @since 2.2
1617      */
1618     public static void decodeFileToFile( String infile, String outfile )
1619     throws java.io.IOException {
1620         
1621         byte[] decoded = Base64.decodeFromFile( infile );
1622         java.io.OutputStream out = null;
1623         try{
1624             out = new java.io.BufferedOutputStream(
1625                   new java.io.FileOutputStream( outfile ) );
1626             out.write( decoded );
1627         }   // end try
1628         catch( java.io.IOException e ) {
1629             throw e; // Catch and release to execute finally{}
1630         }   // end catch
1631         finally {
1632             try { out.close(); }
1633             catch( Exception ex ){}
1634         }   // end finally    
1635     }   // end decodeFileToFile
1636     
1637     
1638     /* ********  I N N E R   C L A S S   I N P U T S T R E A M  ******** */
1639     
1640     
1641     
1642     /**
1643      * A {@link Base64.InputStream} will read data from another
1644      * <code>java.io.InputStream</code>, given in the constructor,
1645      * and encode/decode to/from Base64 notation on the fly.
1646      *
1647      * @see Base64
1648      * @since 1.3
1649      */
1650     public static class InputStream extends java.io.FilterInputStream {
1651         
1652         private boolean encode;         // Encoding or decoding
1653         private int     position;       // Current position in the buffer
1654         private byte[]  buffer;         // Small buffer holding converted data
1655         private int     bufferLength;   // Length of buffer (3 or 4)
1656         private int     numSigBytes;    // Number of meaningful bytes in the buffer
1657         private int     lineLength;
1658         private boolean breakLines;     // Break lines at less than 80 characters
1659         private int     options;        // Record options used to create the stream.
1660         private byte[]  decodabet;      // Local copies to avoid extra method calls
1661         
1662         
1663         /**
1664          * Constructs a {@link Base64.InputStream} in DECODE mode.
1665          *
1666          * @param in the <code>java.io.InputStream</code> from which to read data.
1667          * @since 1.3
1668          */
1669         public InputStream( java.io.InputStream in ) {
1670             this( in, DECODE );
1671         }   // end constructor
1672         
1673         
1674         /**
1675          * Constructs a {@link Base64.InputStream} in
1676          * either ENCODE or DECODE mode.
1677          * <p>
1678          * Valid options:<pre>
1679          *   ENCODE or DECODE: Encode or Decode as data is read.
1680          *   DO_BREAK_LINES: break lines at 76 characters
1681          *     (only meaningful when encoding)
1682          * </pre>
1683          * <p>
1684          * Example: <code>new Base64.InputStream( in, Base64.DECODE )</code>
1685          *
1686          *
1687          * @param in the <code>java.io.InputStream</code> from which to read data.
1688          * @param options Specified options
1689          * @see Base64#ENCODE
1690          * @see Base64#DECODE
1691          * @see Base64#DO_BREAK_LINES
1692          * @since 2.0
1693          */
1694         public InputStream( java.io.InputStream in, int options ) {
1695             
1696             super( in );
1697             this.options      = options; // Record for later
1698             this.breakLines   = (options & DO_BREAK_LINES) > 0;
1699             this.encode       = (options & ENCODE) > 0;
1700             this.bufferLength = encode ? 4 : 3;
1701             this.buffer       = new byte[ bufferLength ];
1702             this.position     = -1;
1703             this.lineLength   = 0;
1704             this.decodabet    = getDecodabet(options);
1705         }   // end constructor
1706         
1707         /**
1708          * Reads enough of the input stream to convert
1709          * to/from Base64 and returns the next byte.
1710          *
1711          * @return next byte
1712          * @since 1.3
1713          */
1714         @Override
1715         public int read() throws java.io.IOException  {
1716             
1717             // Do we need to get data?
1718             if( position < 0 ) {
1719                 if( encode ) {
1720                     byte[] b3 = new byte[3];
1721                     int numBinaryBytes = 0;
1722                     for( int i = 0; i < 3; i++ ) {
1723                         int b = in.read();
1724 
1725                         // If end of stream, b is -1.
1726                         if( b >= 0 ) {
1727                             b3[i] = (byte)b;
1728                             numBinaryBytes++;
1729                         } else {
1730                             break; // out of for loop
1731                         }   // end else: end of stream
1732                             
1733                     }   // end for: each needed input byte
1734                     
1735                     if( numBinaryBytes > 0 ) {
1736                         encode3to4( b3, 0, numBinaryBytes, buffer, 0, options );
1737                         position = 0;
1738                         numSigBytes = 4;
1739                     }   // end if: got data
1740                     else {
1741                         return -1;  // Must be end of stream
1742                     }   // end else
1743                 }   // end if: encoding
1744                 
1745                 // Else decoding
1746                 else {
1747                     byte[] b4 = new byte[4];
1748                     int i = 0;
1749                     for( i = 0; i < 4; i++ ) {
1750                         // Read four "meaningful" bytes:
1751                         int b = 0;
1752                         do{ b = in.read(); }
1753                         while( b >= 0 && decodabet[ b & 0x7f ] <= WHITE_SPACE_ENC );
1754                         
1755                         if( b < 0 ) {
1756                             break; // Reads a -1 if end of stream
1757                         }   // end if: end of stream
1758                         
1759                         b4[i] = (byte)b;
1760                     }   // end for: each needed input byte
1761                     
1762                     if( i == 4 ) {
1763                         numSigBytes = decode4to3( b4, 0, buffer, 0, options );
1764                         position = 0;
1765                     }   // end if: got four characters
1766                     else if( i == 0 ){
1767                         return -1;
1768                     }   // end else if: also padded correctly
1769                     else {
1770                         // Must have broken out from above.
1771                         throw new java.io.IOException( "Improperly padded Base64 input." );
1772                     }   // end 
1773                     
1774                 }   // end else: decode
1775             }   // end else: get data
1776             
1777             // Got data?
1778             if( position >= 0 ) {
1779                 // End of relevant data?
1780                 if( /*!encode &&*/ position >= numSigBytes ){
1781                     return -1;
1782                 }   // end if: got data
1783                 
1784                 if( encode && breakLines && lineLength >= MAX_LINE_LENGTH ) {
1785                     lineLength = 0;
1786                     return '\n';
1787                 }   // end if
1788                 else {
1789                     lineLength++;   // This isn't important when decoding
1790                                     // but throwing an extra "if" seems
1791                                     // just as wasteful.
1792                     
1793                     int b = buffer[ position++ ];
1794 
1795                     if( position >= bufferLength ) {
1796                         position = -1;
1797                     }   // end if: end
1798 
1799                     return b & 0xFF; // This is how you "cast" a byte that's
1800                                      // intended to be unsigned.
1801                 }   // end else
1802             }   // end if: position >= 0
1803             
1804             // Else error
1805             else {
1806                 throw new java.io.IOException( "Error in Base64 code reading stream." );
1807             }   // end else
1808         }   // end read
1809         
1810         
1811         /**
1812          * Calls {@link #read()} repeatedly until the end of stream
1813          * is reached or <var>len</var> bytes are read.
1814          * Returns number of bytes read into array or -1 if
1815          * end of stream is encountered.
1816          *
1817          * @param dest array to hold values
1818          * @param off offset for array
1819          * @param len max number of bytes to read into array
1820          * @return bytes read into array or -1 if end of stream is encountered.
1821          * @since 1.3
1822          */
1823         @Override
1824         public int read( byte[] dest, int off, int len ) 
1825         throws java.io.IOException {
1826             int i;
1827             int b;
1828             for( i = 0; i < len; i++ ) {
1829                 b = read();
1830                 
1831                 if( b >= 0 ) {
1832                     dest[off + i] = (byte) b;
1833                 }
1834                 else if( i == 0 ) {
1835                     return -1;
1836                 }
1837                 else {
1838                     break; // Out of 'for' loop
1839                 } // Out of 'for' loop
1840             }   // end for: each byte read
1841             return i;
1842         }   // end read
1843         
1844     }   // end inner class InputStream
1845     
1846     
1847     
1848     
1849     
1850     
1851     /* ********  I N N E R   C L A S S   O U T P U T S T R E A M  ******** */
1852     
1853     
1854     
1855     /**
1856      * A {@link Base64.OutputStream} will write data to another
1857      * <code>java.io.OutputStream</code>, given in the constructor,
1858      * and encode/decode to/from Base64 notation on the fly.
1859      *
1860      * @see Base64
1861      * @since 1.3
1862      */
1863     public static class OutputStream extends java.io.FilterOutputStream {
1864         
1865         private boolean encode;
1866         private int     position;
1867         private byte[]  buffer;
1868         private int     bufferLength;
1869         private int     lineLength;
1870         private boolean breakLines;
1871         private byte[]  b4;         // Scratch used in a few places
1872         private boolean suspendEncoding;
1873         private int     options;    // Record for later
1874         private byte[]  decodabet;  // Local copies to avoid extra method calls
1875         
1876         /**
1877          * Constructs a {@link Base64.OutputStream} in ENCODE mode.
1878          *
1879          * @param out the <code>java.io.OutputStream</code> to which data will be written.
1880          * @since 1.3
1881          */
1882         public OutputStream( java.io.OutputStream out ) {
1883             this( out, ENCODE );
1884         }   // end constructor
1885         
1886         
1887         /**
1888          * Constructs a {@link Base64.OutputStream} in
1889          * either ENCODE or DECODE mode.
1890          * <p>
1891          * Valid options:<pre>
1892          *   ENCODE or DECODE: Encode or Decode as data is read.
1893          *   DO_BREAK_LINES: don't break lines at 76 characters
1894          *     (only meaningful when encoding)
1895          * </pre>
1896          * <p>
1897          * Example: <code>new Base64.OutputStream( out, Base64.ENCODE )</code>
1898          *
1899          * @param out the <code>java.io.OutputStream</code> to which data will be written.
1900          * @param options Specified options.
1901          * @see Base64#ENCODE
1902          * @see Base64#DECODE
1903          * @see Base64#DO_BREAK_LINES
1904          * @since 1.3
1905          */
1906         public OutputStream( java.io.OutputStream out, int options ) {
1907             super( out );
1908             this.breakLines   = (options & DO_BREAK_LINES) != 0;
1909             this.encode       = (options & ENCODE) != 0;
1910             this.bufferLength = encode ? 3 : 4;
1911             this.buffer       = new byte[ bufferLength ];
1912             this.position     = 0;
1913             this.lineLength   = 0;
1914             this.suspendEncoding = false;
1915             this.b4           = new byte[4];
1916             this.options      = options;
1917             this.decodabet    = getDecodabet(options);
1918         }   // end constructor
1919         
1920         
1921         /**
1922          * Writes the byte to the output stream after
1923          * converting to/from Base64 notation.
1924          * When encoding, bytes are buffered three
1925          * at a time before the output stream actually
1926          * gets a write() call.
1927          * When decoding, bytes are buffered four
1928          * at a time.
1929          *
1930          * @param theByte the byte to write
1931          * @since 1.3
1932          */
1933         @Override
1934         public void write(int theByte) 
1935         throws java.io.IOException {
1936             // Encoding suspended?
1937             if( suspendEncoding ) {
1938                 this.out.write( theByte );
1939                 return;
1940             }   // end if: supsended
1941             
1942             // Encode?
1943             if( encode ) {
1944                 buffer[ position++ ] = (byte)theByte;
1945                 if( position >= bufferLength ) { // Enough to encode.
1946                 
1947                     this.out.write( encode3to4( b4, buffer, bufferLength, options ) );
1948 
1949                     lineLength += 4;
1950                     if( breakLines && lineLength >= MAX_LINE_LENGTH ) {
1951                         this.out.write( NEW_LINE );
1952                         lineLength = 0;
1953                     }   // end if: end of line
1954 
1955                     position = 0;
1956                 }   // end if: enough to output
1957             }   // end if: encoding
1958 
1959             // Else, Decoding
1960             else {
1961                 // Meaningful Base64 character?
1962                 if( decodabet[ theByte & 0x7f ] > WHITE_SPACE_ENC ) {
1963                     buffer[ position++ ] = (byte)theByte;
1964                     if( position >= bufferLength ) { // Enough to output.
1965                     
1966                         int len = Base64.decode4to3( buffer, 0, b4, 0, options );
1967                         out.write( b4, 0, len );
1968                         position = 0;
1969                     }   // end if: enough to output
1970                 }   // end if: meaningful base64 character
1971                 else if( decodabet[ theByte & 0x7f ] != WHITE_SPACE_ENC ) {
1972                     throw new java.io.IOException( "Invalid character in Base64 data." );
1973                 }   // end else: not white space either
1974             }   // end else: decoding
1975         }   // end write
1976         
1977         
1978         
1979         /**
1980          * Calls {@link #write(int)} repeatedly until <var>len</var> 
1981          * bytes are written.
1982          *
1983          * @param theBytes array from which to read bytes
1984          * @param off offset for array
1985          * @param len max number of bytes to read into array
1986          * @since 1.3
1987          */
1988         @Override
1989         public void write( byte[] theBytes, int off, int len ) 
1990         throws java.io.IOException {
1991             // Encoding suspended?
1992             if( suspendEncoding ) {
1993                 this.out.write( theBytes, off, len );
1994                 return;
1995             }   // end if: supsended
1996             
1997             for( int i = 0; i < len; i++ ) {
1998                 write( theBytes[ off + i ] );
1999             }   // end for: each byte written
2000             
2001         }   // end write
2002         
2003         
2004         
2005         /**
2006          * Method added by PHIL. [Thanks, PHIL. -Rob]
2007          * This pads the buffer without closing the stream.
2008          * @throws java.io.IOException  if there's an error.
2009          */
2010         public void flushBase64() throws java.io.IOException  {
2011             if( position > 0 ) {
2012                 if( encode ) {
2013                     out.write( encode3to4( b4, buffer, position, options ) );
2014                     position = 0;
2015                 }   // end if: encoding
2016                 else {
2017                     throw new java.io.IOException( "Base64 input not properly padded." );
2018                 }   // end else: decoding
2019             }   // end if: buffer partially full
2020 
2021         }   // end flush
2022 
2023         
2024         /** 
2025          * Flushes and closes (I think, in the superclass) the stream. 
2026          *
2027          * @since 1.3
2028          */
2029         @Override
2030         public void close() throws java.io.IOException {
2031             // 1. Ensure that pending characters are written
2032             flushBase64();
2033 
2034             // 2. Actually close the stream
2035             // Base class both flushes and closes.
2036             super.close();
2037             
2038             buffer = null;
2039             out    = null;
2040         }   // end close
2041         
2042         
2043         
2044         /**
2045          * Suspends encoding of the stream.
2046          * May be helpful if you need to embed a piece of
2047          * base64-encoded data in a stream.
2048          *
2049          * @throws java.io.IOException  if there's an error flushing
2050          * @since 1.5.1
2051          */
2052         public void suspendEncoding() throws java.io.IOException  {
2053             flushBase64();
2054             this.suspendEncoding = true;
2055         }   // end suspendEncoding
2056         
2057         
2058         /**
2059          * Resumes encoding of the stream.
2060          * May be helpful if you need to embed a piece of
2061          * base64-encoded data in a stream.
2062          *
2063          * @since 1.5.1
2064          */
2065         public void resumeEncoding() {
2066             this.suspendEncoding = false;
2067         }   // end resumeEncoding
2068         
2069         
2070         
2071     }   // end inner class OutputStream
2072     
2073     
2074 }   // end class Base64