1 /**
2 *
3 * Licensed to the Apache Software Foundation (ASF) under one
4 * or more contributor license agreements. See the NOTICE file
5 * distributed with this work for additional information
6 * regarding copyright ownership. The ASF licenses this file
7 * to you under the Apache License, Version 2.0 (the
8 * "License"); you may not use this file except in compliance
9 * with the License. You may obtain a copy of the License at
10 *
11 * http://www.apache.org/licenses/LICENSE-2.0
12 *
13 * Unless required by applicable law or agreed to in writing, software
14 * distributed under the License is distributed on an "AS IS" BASIS,
15 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
16 * See the License for the specific language governing permissions and
17 * limitations under the License.
18 */
19
20 package org.apache.hadoop.hbase.util;
21
22 import org.apache.commons.logging.Log;
23 import org.apache.commons.logging.LogFactory;
24 import org.apache.hadoop.classification.InterfaceAudience;
25 import org.apache.hadoop.classification.InterfaceStability;
26
27 import java.io.BufferedInputStream;
28 import java.io.BufferedOutputStream;
29 import java.io.ByteArrayInputStream;
30 import java.io.ByteArrayOutputStream;
31 import java.io.File;
32 import java.io.FileInputStream;
33 import java.io.FileOutputStream;
34 import java.io.FilterInputStream;
35 import java.io.FilterOutputStream;
36 import java.io.IOException;
37 import java.io.InputStream;
38 import java.io.ObjectInputStream;
39 import java.io.ObjectOutputStream;
40 import java.io.OutputStream;
41 import java.io.Serializable;
42 import java.io.UnsupportedEncodingException;
43 import java.util.zip.GZIPInputStream;
44 import java.util.zip.GZIPOutputStream;
45
46 /**
47 * Encodes and decodes to and from Base64 notation.
48 *
49 * <p>
50 * Homepage: <a href="http://iharder.net/base64">http://iharder.net/base64</a>.
51 * </p>
52 *
53 * <p>
54 * Change Log:
55 * </p>
56 * <ul>
57 * <li>v2.2.1 - Fixed bug using URL_SAFE and ORDERED encodings. Fixed bug
58 * when using very small files (~< 40 bytes).</li>
59 * <li>v2.2 - Added some helper methods for encoding/decoding directly from
60 * one file to the next. Also added a main() method to support command
61 * line encoding/decoding from one file to the next. Also added these
62 * Base64 dialects:
63 * <ol>
64 * <li>The default is RFC3548 format.</li>
65 * <li>Using Base64.URLSAFE generates URL and file name friendly format as
66 * described in Section 4 of RFC3548.
67 * http://www.faqs.org/rfcs/rfc3548.html</li>
68 * <li>Using Base64.ORDERED generates URL and file name friendly format
69 * that preserves lexical ordering as described in
70 * http://www.faqs.org/qa/rfcc-1940.html</li>
71 * </ol>
72 * <p>
73 * Special thanks to Jim Kellerman at <a href="http://www.powerset.com/">
74 * http://www.powerset.com/</a> for contributing the new Base64 dialects.
75 * </li>
76 *
77 * <li>v2.1 - Cleaned up javadoc comments and unused variables and methods.
78 * Added some convenience methods for reading and writing to and from files.
79 * </li>
80 * <li>v2.0.2 - Now specifies UTF-8 encoding in places where the code fails on
81 * systems with other encodings (like EBCDIC).</li>
82 * <li>v2.0.1 - Fixed an error when decoding a single byte, that is, when the
83 * encoded data was a single byte.</li>
84 * <li>v2.0 - I got rid of methods that used booleans to set options. Now
85 * everything is more consolidated and cleaner. The code now detects when
86 * data that's being decoded is gzip-compressed and will decompress it
87 * automatically. Generally things are cleaner. You'll probably have to
88 * change some method calls that you were making to support the new options
89 * format (<tt>int</tt>s that you "OR" together).</li>
90 * <li>v1.5.1 - Fixed bug when decompressing and decoding to a byte[] using
91 * <tt>decode( String s, boolean gzipCompressed )</tt>. Added the ability to
92 * "suspend" encoding in the Output Stream so you can turn on and off the
93 * encoding if you need to embed base64 data in an otherwise "normal" stream
94 * (like an XML file).</li>
95 * <li>v1.5 - Output stream pases on flush() command but doesn't do anything
96 * itself. This helps when using GZIP streams. Added the ability to
97 * GZip-compress objects before encoding them.</li>
98 * <li>v1.4 - Added helper methods to read/write files.</li>
99 * <li>v1.3.6 - Fixed OutputStream.flush() so that 'position' is reset.</li>
100 * <li>v1.3.5 - Added flag to turn on and off line breaks. Fixed bug in input
101 * stream where last buffer being read, if not completely full, was not
102 * returned.</li>
103 * <li>v1.3.4 - Fixed when "improperly padded stream" error was thrown at the
104 * wrong time.</li>
105 * <li>v1.3.3 - Fixed I/O streams which were totally messed up.</li>
106 * </ul>
107 *
108 * <p>
109 * I am placing this code in the Public Domain. Do with it as you will. This
110 * software comes with no guarantees or warranties but with plenty of
111 * well-wishing instead!
112 * <p>
113 * Please visit <a href="http://iharder.net/base64">http://iharder.net/base64</a>
114 * periodically to check for updates or to contribute improvements.
115 * <p>
116 * author: Robert Harder, rob@iharder.net
117 * <br>
118 * version: 2.2.1
119 */
120 @InterfaceAudience.Public
121 @InterfaceStability.Stable
122 public class Base64 {
123
124 /* ******** P U B L I C F I E L D S ******** */
125
126 /** No options specified. Value is zero. */
127 public final static int NO_OPTIONS = 0;
128
129 /** Specify encoding. */
130 public final static int ENCODE = 1;
131
132 /** Specify decoding. */
133 public final static int DECODE = 0;
134
135 /** Specify that data should be gzip-compressed. */
136 public final static int GZIP = 2;
137
138 /** Don't break lines when encoding (violates strict Base64 specification) */
139 public final static int DONT_BREAK_LINES = 8;
140
141 /**
142 * Encode using Base64-like encoding that is URL and Filename safe as
143 * described in Section 4 of RFC3548:
144 * <a href="http://www.faqs.org/rfcs/rfc3548.html">
145 * http://www.faqs.org/rfcs/rfc3548.html</a>.
146 * It is important to note that data encoded this way is <em>not</em>
147 * officially valid Base64, or at the very least should not be called Base64
148 * without also specifying that is was encoded using the URL and
149 * Filename safe dialect.
150 */
151 public final static int URL_SAFE = 16;
152
153 /**
154 * Encode using the special "ordered" dialect of Base64 described here:
155 * <a href="http://www.faqs.org/qa/rfcc-1940.html">
156 * http://www.faqs.org/qa/rfcc-1940.html</a>.
157 */
158 public final static int ORDERED = 32;
159
160 /* ******** P R I V A T E F I E L D S ******** */
161
162 private static final Log LOG = LogFactory.getLog(Base64.class);
163
164 /** Maximum line length (76) of Base64 output. */
165 private final static int MAX_LINE_LENGTH = 76;
166
167 /** The equals sign (=) as a byte. */
168 private final static byte EQUALS_SIGN = (byte) '=';
169
170 /** The new line character (\n) as a byte. */
171 private final static byte NEW_LINE = (byte) '\n';
172
173 /** Preferred encoding. */
174 private final static String PREFERRED_ENCODING = "UTF-8";
175
176 private final static byte WHITE_SPACE_ENC = -5; // Indicates white space
177 private final static byte EQUALS_SIGN_ENC = -1; // Indicates equals sign
178
179 /* ******** S T A N D A R D B A S E 6 4 A L P H A B E T ******** */
180
181 /** The 64 valid Base64 values. */
182
183 /*
184 * Host platform may be something funny like EBCDIC, so we hardcode these
185 * values.
186 */
187 private final static byte[] _STANDARD_ALPHABET = { (byte) 'A', (byte) 'B',
188 (byte) 'C', (byte) 'D', (byte) 'E', (byte) 'F', (byte) 'G', (byte) 'H',
189 (byte) 'I', (byte) 'J', (byte) 'K', (byte) 'L', (byte) 'M', (byte) 'N',
190 (byte) 'O', (byte) 'P', (byte) 'Q', (byte) 'R', (byte) 'S', (byte) 'T',
191 (byte) 'U', (byte) 'V', (byte) 'W', (byte) 'X', (byte) 'Y', (byte) 'Z',
192 (byte) 'a', (byte) 'b', (byte) 'c', (byte) 'd', (byte) 'e', (byte) 'f',
193 (byte) 'g', (byte) 'h', (byte) 'i', (byte) 'j', (byte) 'k', (byte) 'l',
194 (byte) 'm', (byte) 'n', (byte) 'o', (byte) 'p', (byte) 'q', (byte) 'r',
195 (byte) 's', (byte) 't', (byte) 'u', (byte) 'v', (byte) 'w', (byte) 'x',
196 (byte) 'y', (byte) 'z', (byte) '0', (byte) '1', (byte) '2', (byte) '3',
197 (byte) '4', (byte) '5', (byte) '6', (byte) '7', (byte) '8', (byte) '9',
198 (byte) '+', (byte) '/'
199 };
200
201 /**
202 * Translates a Base64 value to either its 6-bit reconstruction value or a
203 * negative number indicating some other meaning.
204 */
205 private final static byte[] _STANDARD_DECODABET = {
206 -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 0 - 8
207 -5, -5, // Whitespace: Tab, Newline
208 -9, -9, // Decimal 11 - 12
209 -5, // Whitespace: Return
210 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 14 - 26
211 -9, -9, -9, -9, -9, // Decimal 27 - 31
212 -5, // Whitespace: Space
213 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 33 - 42
214 62, // Plus sign at decimal 43
215 -9, -9, -9, // Decimal 44 - 46
216 63, // Slash at decimal 47
217 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, // Numbers zero - nine
218 -9, -9, -9, // Decimal 58 - 60
219 -1, // Equals sign at decimal 61
220 -9, -9, -9, // Decimal 62 - 64
221 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, // Letters 'A' - 'N'
222 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, // Letters 'O' - 'Z'
223 -9, -9, -9, -9, -9, -9, // Decimal 91 - 96
224 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, // Letters 'a' - 'm'
225 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, // Letters 'n' -'z'
226 -9, -9, -9, -9 // Decimal 123 - 126
227 };
228
229 /* ******** U R L S A F E B A S E 6 4 A L P H A B E T ******** */
230
231 /**
232 * Used in the URL and Filename safe dialect described in Section 4 of RFC3548
233 * <a href="http://www.faqs.org/rfcs/rfc3548.html">
234 * http://www.faqs.org/rfcs/rfc3548.html</a>.
235 * Notice that the last two bytes become "hyphen" and "underscore" instead of
236 * "plus" and "slash."
237 */
238 private final static byte[] _URL_SAFE_ALPHABET = { (byte) 'A', (byte) 'B',
239 (byte) 'C', (byte) 'D', (byte) 'E', (byte) 'F', (byte) 'G', (byte) 'H',
240 (byte) 'I', (byte) 'J', (byte) 'K', (byte) 'L', (byte) 'M', (byte) 'N',
241 (byte) 'O', (byte) 'P', (byte) 'Q', (byte) 'R', (byte) 'S', (byte) 'T',
242 (byte) 'U', (byte) 'V', (byte) 'W', (byte) 'X', (byte) 'Y', (byte) 'Z',
243 (byte) 'a', (byte) 'b', (byte) 'c', (byte) 'd', (byte) 'e', (byte) 'f',
244 (byte) 'g', (byte) 'h', (byte) 'i', (byte) 'j', (byte) 'k', (byte) 'l',
245 (byte) 'm', (byte) 'n', (byte) 'o', (byte) 'p', (byte) 'q', (byte) 'r',
246 (byte) 's', (byte) 't', (byte) 'u', (byte) 'v', (byte) 'w', (byte) 'x',
247 (byte) 'y', (byte) 'z', (byte) '0', (byte) '1', (byte) '2', (byte) '3',
248 (byte) '4', (byte) '5', (byte) '6', (byte) '7', (byte) '8', (byte) '9',
249 (byte) '-', (byte) '_'
250 };
251
252 /**
253 * Used in decoding URL and Filename safe dialects of Base64.
254 */
255 private final static byte[] _URL_SAFE_DECODABET = {
256 -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 0 - 8
257 -5, -5, // Whitespace: Tab, Newline
258 -9, -9, // Decimal 11 - 12
259 -5, // Whitespace: Return
260 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 14 - 26
261 -9, -9, -9, -9, -9, // Decimal 27 - 31
262 -5, // Whitespace: Space
263 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 33 - 42
264 -9, // Plus sign at 43
265 -9, // Decimal 44
266 62, // Minus sign at 45
267 -9, // Decimal 46
268 -9, // Slash at 47
269 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, // Numbers 0 - 9
270 -9, -9, -9, // Decimal 58 - 60
271 -1, // Equals sign at 61
272 -9, -9, -9, // Decimal 62 - 64
273 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, // Letters 'A' - 'N'
274 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, // Letters 'O' - 'Z'
275 -9, -9, -9, -9, // Decimal 91 - 94
276 63, // Underscore at 95
277 -9, // Decimal 96
278 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, // Letters 'a' - 'm'
279 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, // Letters 'n' - 'z'
280 -9, -9, -9, -9 // Decimal 123 - 126
281 };
282
283 /* ******** O R D E R E D B A S E 6 4 A L P H A B E T ******** */
284
285 /**
286 * In addition to being URL and file name friendly, this encoding preserves
287 * the sort order of encoded values. Whatever is input, be it string or
288 * just an array of bytes, when you use this encoding, the encoded value sorts
289 * exactly the same as the input value. It is described in the RFC change
290 * request: <a href="http://www.faqs.org/qa/rfcc-1940.html">
291 * http://www.faqs.org/qa/rfcc-1940.html</a>.
292 *
293 * It replaces "plus" and "slash" with "hyphen" and "underscore" and
294 * rearranges the alphabet so that the characters are in their natural sort
295 * order.
296 */
297 private final static byte[] _ORDERED_ALPHABET = { (byte) '-', (byte) '0',
298 (byte) '1', (byte) '2', (byte) '3', (byte) '4', (byte) '5', (byte) '6',
299 (byte) '7', (byte) '8', (byte) '9', (byte) 'A', (byte) 'B', (byte) 'C',
300 (byte) 'D', (byte) 'E', (byte) 'F', (byte) 'G', (byte) 'H', (byte) 'I',
301 (byte) 'J', (byte) 'K', (byte) 'L', (byte) 'M', (byte) 'N', (byte) 'O',
302 (byte) 'P', (byte) 'Q', (byte) 'R', (byte) 'S', (byte) 'T', (byte) 'U',
303 (byte) 'V', (byte) 'W', (byte) 'X', (byte) 'Y', (byte) 'Z', (byte) '_',
304 (byte) 'a', (byte) 'b', (byte) 'c', (byte) 'd', (byte) 'e', (byte) 'f',
305 (byte) 'g', (byte) 'h', (byte) 'i', (byte) 'j', (byte) 'k', (byte) 'l',
306 (byte) 'm', (byte) 'n', (byte) 'o', (byte) 'p', (byte) 'q', (byte) 'r',
307 (byte) 's', (byte) 't', (byte) 'u', (byte) 'v', (byte) 'w', (byte) 'x',
308 (byte) 'y', (byte) 'z'
309 };
310
311 /**
312 * Used in decoding the "ordered" dialect of Base64.
313 */
314 private final static byte[] _ORDERED_DECODABET = {
315 -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 0 - 8
316 -5, -5, // Whitespace: Tab, Newline
317 -9, -9, // Decimal 11 - 12
318 -5, // Whitespace: Return
319 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 14 - 26
320 -9, -9, -9, -9, -9, // Decimal 27 - 31
321 -5, // Whitespace: Space
322 -9, -9, -9, -9, -9, -9, -9, -9, -9, -9, // Decimal 33 - 42
323 -9, // Plus sign at 43
324 -9, // Decimal 44
325 0, // Minus sign at 45
326 -9, // Decimal 46
327 -9, // Slash at decimal 47
328 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, // Numbers 0 - 9
329 -9, -9, -9, // Decimal 58 - 60
330 -1, // Equals sign at 61
331 -9, -9, -9, // Decimal 62 - 64
332 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, // Letters 'A' - 'M'
333 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, // Letters 'N' - 'Z'
334 -9, -9, -9, -9, // Decimal 91 - 94
335 37, // Underscore at 95
336 -9, // Decimal 96
337 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, // Letters 'a' - 'm'
338 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, // Letters 'n' - 'z'
339 -9, -9, -9, -9 // Decimal 123 - 126
340 };
341
342 /* ******** D E T E R M I N E W H I C H A L H A B E T ******** */
343
344 /**
345 * Returns one of the _SOMETHING_ALPHABET byte arrays depending on the options
346 * specified. It's possible, though silly, to specify ORDERED and URLSAFE in
347 * which case one of them will be picked, though there is no guarantee as to
348 * which one will be picked.
349 *
350 * @param options URL_SAFE or ORDERED
351 * @return alphabet array to use
352 */
353 protected static byte[] getAlphabet(int options) {
354 if ((options & URL_SAFE) == URL_SAFE) {
355 return _URL_SAFE_ALPHABET;
356
357 } else if ((options & ORDERED) == ORDERED) {
358 return _ORDERED_ALPHABET;
359
360 } else {
361 return _STANDARD_ALPHABET;
362 }
363 } // end getAlphabet
364
365 /**
366 * Returns one of the _SOMETHING_DECODABET byte arrays depending on the
367 * options specified. It's possible, though silly, to specify ORDERED and
368 * URL_SAFE in which case one of them will be picked, though there is no
369 * guarantee as to which one will be picked.
370 * @param options URL_SAFE or ORDERED
371 * @return alphabet array to use
372 */
373 protected static byte[] getDecodabet(int options) {
374 if ((options & URL_SAFE) == URL_SAFE) {
375 return _URL_SAFE_DECODABET;
376
377 } else if ((options & ORDERED) == ORDERED) {
378 return _ORDERED_DECODABET;
379
380 } else {
381 return _STANDARD_DECODABET;
382 }
383 } // end getDecodabet
384
385 /** Defeats instantiation. */
386 private Base64() {}
387
388 /**
389 * Main program. Used for testing.
390 *
391 * Encodes or decodes two files from the command line
392 *
393 * @param args command arguments
394 */
395 public static void main(String[] args) {
396 if (args.length < 3) {
397 usage("Not enough arguments.");
398
399 } else {
400 String flag = args[0];
401 String infile = args[1];
402 String outfile = args[2];
403 if (flag.equals("-e")) { // encode
404 encodeFileToFile(infile, outfile);
405
406 } else if (flag.equals("-d")) { // decode
407 decodeFileToFile(infile, outfile);
408
409 } else {
410 usage("Unknown flag: " + flag);
411 }
412 }
413 } // end main
414
415 /**
416 * Prints command line usage.
417 *
418 * @param msg A message to include with usage info.
419 */
420 private static void usage(String msg) {
421 System.err.println(msg);
422 System.err.println("Usage: java Base64 -e|-d inputfile outputfile");
423 } // end usage
424
425 /* ******** E N C O D I N G M E T H O D S ******** */
426
427 /**
428 * Encodes up to the first three bytes of array <var>threeBytes</var> and
429 * returns a four-byte array in Base64 notation. The actual number of
430 * significant bytes in your array is given by <var>numSigBytes</var>. The
431 * array <var>threeBytes</var> needs only be as big as <var>numSigBytes</var>.
432 * Code can reuse a byte array by passing a four-byte array as <var>b4</var>.
433 *
434 * @param b4 A reusable byte array to reduce array instantiation
435 * @param threeBytes the array to convert
436 * @param numSigBytes the number of significant bytes in your array
437 * @param options options for get alphabet
438 * @return four byte array in Base64 notation.
439 * @since 1.5.1
440 */
441 protected static byte[] encode3to4(byte[] b4, byte[] threeBytes,
442 int numSigBytes, int options) {
443 encode3to4(threeBytes, 0, numSigBytes, b4, 0, options);
444 return b4;
445 } // end encode3to4
446
447 /**
448 * Encodes up to three bytes of the array <var>source</var> and writes the
449 * resulting four Base64 bytes to <var>destination</var>. The source and
450 * destination arrays can be manipulated anywhere along their length by
451 * specifying <var>srcOffset</var> and <var>destOffset</var>. This method
452 * does not check to make sure your arrays are large enough to accomodate
453 * <var>srcOffset</var> + 3 for the <var>source</var> array or
454 * <var>destOffset</var> + 4 for the <var>destination</var> array. The
455 * actual number of significant bytes in your array is given by
456 * <var>numSigBytes</var>.
457 * <p>
458 * This is the lowest level of the encoding methods with all possible
459 * parameters.
460 *
461 * @param source the array to convert
462 * @param srcOffset the index where conversion begins
463 * @param numSigBytes the number of significant bytes in your array
464 * @param destination the array to hold the conversion
465 * @param destOffset the index where output will be put
466 * @param options options for get alphabet
467 * @return the <var>destination</var> array
468 * @since 1.3
469 */
470 protected static byte[] encode3to4(byte[] source, int srcOffset,
471 int numSigBytes, byte[] destination, int destOffset, int options) {
472 byte[] ALPHABET = getAlphabet(options);
473
474 // 1 2 3
475 // 01234567890123456789012345678901 Bit position
476 // --------000000001111111122222222 Array position from threeBytes
477 // --------| || || || | Six bit groups to index ALPHABET
478 // >>18 >>12 >> 6 >> 0 Right shift necessary
479 // 0x3f 0x3f 0x3f Additional AND
480
481 // Create buffer with zero-padding if there are only one or two
482 // significant bytes passed in the array.
483 // We have to shift left 24 in order to flush out the 1's that appear
484 // when Java treats a value as negative that is cast from a byte to an int.
485 int inBuff =
486 (numSigBytes > 0 ? ((source[srcOffset] << 24) >>> 8) : 0)
487 | (numSigBytes > 1 ? ((source[srcOffset + 1] << 24) >>> 16) : 0)
488 | (numSigBytes > 2 ? ((source[srcOffset + 2] << 24) >>> 24) : 0);
489
490 switch (numSigBytes) {
491 case 3:
492 destination[destOffset] = ALPHABET[(inBuff >>> 18)];
493 destination[destOffset + 1] = ALPHABET[(inBuff >>> 12) & 0x3f];
494 destination[destOffset + 2] = ALPHABET[(inBuff >>> 6) & 0x3f];
495 destination[destOffset + 3] = ALPHABET[(inBuff) & 0x3f];
496 return destination;
497
498 case 2:
499 destination[destOffset] = ALPHABET[(inBuff >>> 18)];
500 destination[destOffset + 1] = ALPHABET[(inBuff >>> 12) & 0x3f];
501 destination[destOffset + 2] = ALPHABET[(inBuff >>> 6) & 0x3f];
502 destination[destOffset + 3] = EQUALS_SIGN;
503 return destination;
504
505 case 1:
506 destination[destOffset] = ALPHABET[(inBuff >>> 18)];
507 destination[destOffset + 1] = ALPHABET[(inBuff >>> 12) & 0x3f];
508 destination[destOffset + 2] = EQUALS_SIGN;
509 destination[destOffset + 3] = EQUALS_SIGN;
510 return destination;
511
512 default:
513 return destination;
514 } // end switch
515 } // end encode3to4
516
517 /**
518 * Serializes an object and returns the Base64-encoded version of that
519 * serialized object. If the object cannot be serialized or there is another
520 * error, the method will return <tt>null</tt>. The object is not
521 * GZip-compressed before being encoded.
522 *
523 * @param serializableObject The object to encode
524 * @return The Base64-encoded object
525 * @since 1.4
526 */
527 public static String encodeObject(Serializable serializableObject) {
528 return encodeObject(serializableObject, NO_OPTIONS);
529 } // end encodeObject
530
531 /**
532 * Serializes an object and returns the Base64-encoded version of that
533 * serialized object. If the object cannot be serialized or there is another
534 * error, the method will return <tt>null</tt>.
535 * <p>
536 * Valid options:
537 * <ul>
538 * <li>GZIP: gzip-compresses object before encoding it.</li>
539 * <li>DONT_BREAK_LINES: don't break lines at 76 characters. <i>Note:
540 * Technically, this makes your encoding non-compliant.</i></li>
541 * </ul>
542 * <p>
543 * Example: <code>encodeObject( myObj, Base64.GZIP )</code> or
544 * <p>
545 * Example:
546 * <code>encodeObject( myObj, Base64.GZIP | Base64.DONT_BREAK_LINES )</code>
547 *
548 * @param serializableObject The object to encode
549 * @param options Specified options
550 * @see Base64#GZIP
551 * @see Base64#DONT_BREAK_LINES
552 * @return The Base64-encoded object
553 * @since 2.0
554 */
555 @SuppressWarnings({"ConstantConditions"})
556 public static String encodeObject(Serializable serializableObject,
557 int options) {
558
559 ByteArrayOutputStream baos = new ByteArrayOutputStream();
560 OutputStream b64os = null;
561 ObjectOutputStream oos = null;
562 try {
563 // ObjectOutputStream -> (GZIP) -> Base64 -> ByteArrayOutputStream
564 b64os = new Base64OutputStream(baos, ENCODE | options);
565
566 oos = ((options & GZIP) == GZIP) ?
567 new ObjectOutputStream(new GZIPOutputStream(b64os)) :
568 new ObjectOutputStream(b64os);
569
570 oos.writeObject(serializableObject);
571 return new String(baos.toByteArray(), PREFERRED_ENCODING);
572
573 } catch (UnsupportedEncodingException uue) {
574 return new String(baos.toByteArray());
575
576 } catch (IOException e) {
577 LOG.error("error encoding object", e);
578 return null;
579
580 } finally {
581 if (oos != null) {
582 try {
583 oos.close();
584 } catch (Exception e) {
585 LOG.error("error closing ObjectOutputStream", e);
586 }
587 }
588 if (b64os != null) {
589 try {
590 b64os.close();
591 } catch (Exception e) {
592 LOG.error("error closing Base64OutputStream", e);
593 }
594 }
595 try {
596 baos.close();
597 } catch (Exception e) {
598 LOG.error("error closing ByteArrayOutputStream", e);
599 }
600 } // end finally
601 } // end encode
602
603 /**
604 * Encodes a byte array into Base64 notation. Does not GZip-compress data.
605 *
606 * @param source The data to convert
607 * @return encoded byte array
608 * @since 1.4
609 */
610 public static String encodeBytes(byte[] source) {
611 return encodeBytes(source, 0, source.length, NO_OPTIONS);
612 } // end encodeBytes
613
614 /**
615 * Encodes a byte array into Base64 notation.
616 * <p>
617 * Valid options:
618 * <ul>
619 * <li>GZIP: gzip-compresses object before encoding it.</li>
620 * <li>DONT_BREAK_LINES: don't break lines at 76 characters. <i>Note:
621 * Technically, this makes your encoding non-compliant.</i></li>
622 * </ul>
623 *
624 * <p>
625 * Example: <code>encodeBytes( myData, Base64.GZIP )</code> or
626 * <p>
627 * Example:
628 * <code>encodeBytes( myData, Base64.GZIP | Base64.DONT_BREAK_LINES )</code>
629 *
630 * @param source The data to convert
631 * @param options Specified options
632 * @see Base64#GZIP
633 * @see Base64#DONT_BREAK_LINES
634 * @see Base64#URL_SAFE
635 * @see Base64#ORDERED
636 * @return encoded byte array
637 * @since 2.0
638 */
639 public static String encodeBytes(byte[] source, int options) {
640 return encodeBytes(source, 0, source.length, options);
641 } // end encodeBytes
642
643 /**
644 * Encodes a byte array into Base64 notation. Does not GZip-compress data.
645 *
646 * @param source The data to convert
647 * @param off Offset in array where conversion should begin
648 * @param len Length of data to convert
649 * @return encoded byte array
650 * @since 1.4
651 */
652 public static String encodeBytes(byte[] source, int off, int len) {
653 return encodeBytes(source, off, len, NO_OPTIONS);
654 } // end encodeBytes
655
656 /**
657 * Encodes a byte array into Base64 notation.
658 * <p>
659 * Valid options:
660 * <ul>
661 * <li>GZIP: gzip-compresses object before encoding it.</li>
662 * <li>DONT_BREAK_LINES: don't break lines at 76 characters. <i>Note:
663 * Technically, this makes your encoding non-compliant.</i></li>
664 * </ul>
665 *
666 * <p>
667 * Example: <code>encodeBytes( myData, Base64.GZIP )</code> or
668 * <p>
669 * Example:
670 * <code>encodeBytes( myData, Base64.GZIP | Base64.DONT_BREAK_LINES )</code>
671 *
672 * @param source The data to convert
673 * @param off Offset in array where conversion should begin
674 * @param len Length of data to convert
675 * @param options Specified options
676 * @see Base64#GZIP
677 * @see Base64#DONT_BREAK_LINES
678 * @see Base64#URL_SAFE
679 * @see Base64#ORDERED
680 * @return encoded byte array
681 * @since 2.0
682 */
683 public static String encodeBytes(byte[] source, int off, int len, int options) {
684 if ((options & GZIP) == GZIP) { // Compress?
685 // GZip -> Base64 -> ByteArray
686 ByteArrayOutputStream baos = new ByteArrayOutputStream();
687 GZIPOutputStream gzos = null;
688
689 try {
690 gzos =
691 new GZIPOutputStream(new Base64OutputStream(baos, ENCODE | options));
692
693 gzos.write(source, off, len);
694 gzos.close();
695 gzos = null;
696 return new String(baos.toByteArray(), PREFERRED_ENCODING);
697
698 } catch (UnsupportedEncodingException uue) {
699 return new String(baos.toByteArray());
700
701 } catch (IOException e) {
702 LOG.error("error encoding byte array", e);
703 return null;
704
705 } finally {
706 if (gzos != null) {
707 try {
708 gzos.close();
709 } catch (Exception e) {
710 LOG.error("error closing GZIPOutputStream", e);
711 }
712 }
713 try {
714 baos.close();
715 } catch (Exception e) {
716 LOG.error("error closing ByteArrayOutputStream", e);
717 }
718 } // end finally
719
720 } // end Compress
721
722 // Don't compress. Better not to use streams at all then.
723
724 boolean breakLines = ((options & DONT_BREAK_LINES) == 0);
725
726 int len43 = len * 4 / 3;
727 byte[] outBuff =
728 new byte[(len43) // Main 4:3
729 + ((len % 3) > 0 ? 4 : 0) // padding
730 + (breakLines ? (len43 / MAX_LINE_LENGTH) : 0)]; // New lines
731 int d = 0;
732 int e = 0;
733 int len2 = len - 2;
734 int lineLength = 0;
735 for (; d < len2; d += 3, e += 4) {
736 encode3to4(source, d + off, 3, outBuff, e, options);
737
738 lineLength += 4;
739 if (breakLines && lineLength == MAX_LINE_LENGTH) {
740 outBuff[e + 4] = NEW_LINE;
741 e++;
742 lineLength = 0;
743 } // end if: end of line
744 } // end for: each piece of array
745
746 if (d < len) {
747 encode3to4(source, d + off, len - d, outBuff, e, options);
748 e += 4;
749 } // end if: some padding needed
750
751 // Return value according to relevant encoding.
752 try {
753 return new String(outBuff, 0, e, PREFERRED_ENCODING);
754
755 } catch (UnsupportedEncodingException uue) {
756 return new String(outBuff, 0, e);
757 }
758 } // end encodeBytes
759
760 /* ******** D E C O D I N G M E T H O D S ******** */
761
762 /**
763 * Decodes four bytes from array <var>source</var> and writes the resulting
764 * bytes (up to three of them) to <var>destination</var>. The source and
765 * destination arrays can be manipulated anywhere along their length by
766 * specifying <var>srcOffset</var> and <var>destOffset</var>. This method
767 * does not check to make sure your arrays are large enough to accomodate
768 * <var>srcOffset</var> + 4 for the <var>source</var> array or
769 * <var>destOffset</var> + 3 for the <var>destination</var> array. This
770 * method returns the actual number of bytes that were converted from the
771 * Base64 encoding.
772 * <p>
773 * This is the lowest level of the decoding methods with all possible
774 * parameters.
775 * </p>
776 *
777 * @param source the array to convert
778 * @param srcOffset the index where conversion begins
779 * @param destination the array to hold the conversion
780 * @param destOffset the index where output will be put
781 * @param options options for getDecoabet
782 * @see Base64#URL_SAFE
783 * @see Base64#ORDERED
784 * @return the number of decoded bytes converted
785 * @since 1.3
786 */
787 @SuppressWarnings({"ConstantConditions"})
788 protected static int decode4to3(byte[] source, int srcOffset,
789 byte[] destination, int destOffset, int options) {
790 byte[] DECODABET = getDecodabet(options);
791
792 if (source[srcOffset + 2] == EQUALS_SIGN) { // Example: Dk==
793 // Two ways to do the same thing. Don't know which way I like best.
794 // int outBuff = ( ( DECODABET[ source[ srcOffset ] ] << 24 ) >>> 6 )
795 // | ( ( DECODABET[ source[ srcOffset + 1] ] << 24 ) >>> 12 );
796 int outBuff =
797 ((DECODABET[source[srcOffset]] & 0xFF) << 18)
798 | ((DECODABET[source[srcOffset + 1]] & 0xFF) << 12);
799
800 destination[destOffset] = (byte) (outBuff >>> 16);
801 return 1;
802
803 } else if (source[srcOffset + 3] == EQUALS_SIGN) { // Example: DkL=
804 // Two ways to do the same thing. Don't know which way I like best.
805 // int outBuff = ( ( DECODABET[ source[ srcOffset ] ] << 24 ) >>> 6 )
806 // | ( ( DECODABET[ source[ srcOffset + 1 ] ] << 24 ) >>> 12 )
807 // | ( ( DECODABET[ source[ srcOffset + 2 ] ] << 24 ) >>> 18 );
808 int outBuff =
809 ((DECODABET[source[srcOffset]] & 0xFF) << 18)
810 | ((DECODABET[source[srcOffset + 1]] & 0xFF) << 12)
811 | ((DECODABET[source[srcOffset + 2]] & 0xFF) << 6);
812
813 destination[destOffset] = (byte) (outBuff >>> 16);
814 destination[destOffset + 1] = (byte) (outBuff >>> 8);
815 return 2;
816
817 } else { // Example: DkLE
818 try {
819 // Two ways to do the same thing. Don't know which way I like best.
820 // int outBuff = ( ( DECODABET[ source[ srcOffset ] ] << 24 ) >>> 6 )
821 // | ( ( DECODABET[ source[ srcOffset + 1 ] ] << 24 ) >>> 12 )
822 // | ( ( DECODABET[ source[ srcOffset + 2 ] ] << 24 ) >>> 18 )
823 // | ( ( DECODABET[ source[ srcOffset + 3 ] ] << 24 ) >>> 24 );
824 int outBuff =
825 ((DECODABET[source[srcOffset]] & 0xFF) << 18)
826 | ((DECODABET[source[srcOffset + 1]] & 0xFF) << 12)
827 | ((DECODABET[source[srcOffset + 2]] & 0xFF) << 6)
828 | ((DECODABET[source[srcOffset + 3]] & 0xFF));
829
830 destination[destOffset] = (byte) (outBuff >> 16);
831 destination[destOffset + 1] = (byte) (outBuff >> 8);
832 destination[destOffset + 2] = (byte) (outBuff);
833
834 return 3;
835
836 } catch (Exception e) {
837 LOG.error("error decoding bytes at " + source[srcOffset] + ": " +
838 (DECODABET[source[srcOffset]]) + ", " + source[srcOffset + 1] +
839 ": " + (DECODABET[source[srcOffset + 1]]) + ", " +
840 source[srcOffset + 2] + ": " + (DECODABET[source[srcOffset + 2]]) +
841 ", " + source[srcOffset + 3] + ": " +
842 (DECODABET[source[srcOffset + 3]]), e);
843 return -1;
844 } // end catch
845 }
846 } // end decodeToBytes
847
848 /**
849 * Very low-level access to decoding ASCII characters in the form of a byte
850 * array. Does not support automatically gunzipping or any other "fancy"
851 * features.
852 *
853 * @param source The Base64 encoded data
854 * @param off The offset of where to begin decoding
855 * @param len The length of characters to decode
856 * @param options options for getDecodabet
857 * @see Base64#URL_SAFE
858 * @see Base64#ORDERED
859 * @return decoded data
860 * @since 1.3
861 */
862 public static byte[] decode(byte[] source, int off, int len, int options) {
863 byte[] DECODABET = getDecodabet(options);
864
865 int len34 = len * 3 / 4;
866 byte[] outBuff = new byte[len34]; // Upper limit on size of output
867 int outBuffPosn = 0;
868
869 byte[] b4 = new byte[4];
870 int b4Posn = 0;
871 int i;
872 byte sbiCrop;
873 byte sbiDecode;
874 for (i = off; i < off + len; i++) {
875 sbiCrop = (byte) (source[i] & 0x7f); // Only the low seven bits
876 sbiDecode = DECODABET[sbiCrop];
877
878 if (sbiDecode >= WHITE_SPACE_ENC) { // Whitespace, Equals or better
879 if (sbiDecode >= EQUALS_SIGN_ENC) { // Equals or better
880 b4[b4Posn++] = sbiCrop;
881 if (b4Posn > 3) {
882 outBuffPosn += decode4to3(b4, 0, outBuff, outBuffPosn, options);
883 b4Posn = 0;
884
885 // If that was the equals sign, break out of 'for' loop
886 if (sbiCrop == EQUALS_SIGN)
887 break;
888 } // end if: quartet built
889 } // end if: equals sign or better
890 } else {
891 LOG.error("Bad Base64 input character at " + i + ": " + source[i] +
892 "(decimal)");
893 return null;
894 } // end else:
895 } // each input character
896
897 byte[] out = new byte[outBuffPosn];
898 System.arraycopy(outBuff, 0, out, 0, outBuffPosn);
899 return out;
900 } // end decode
901
902 /**
903 * Decodes data from Base64 notation, automatically detecting gzip-compressed
904 * data and decompressing it.
905 *
906 * @param s the string to decode
907 * @return the decoded data
908 * @since 1.4
909 */
910 public static byte[] decode(String s) {
911 return decode(s, NO_OPTIONS);
912 }
913
914 /**
915 * Decodes data from Base64 notation, automatically detecting gzip-compressed
916 * data and decompressing it.
917 *
918 * @param s the string to decode
919 * @param options options for decode
920 * @see Base64#URL_SAFE
921 * @see Base64#ORDERED
922 * @return the decoded data
923 * @since 1.4
924 */
925 public static byte[] decode(String s, int options) {
926 byte[] bytes;
927 try {
928 bytes = s.getBytes(PREFERRED_ENCODING);
929
930 } catch (UnsupportedEncodingException uee) {
931 bytes = s.getBytes();
932 } // end catch
933
934 // Decode
935
936 bytes = decode(bytes, 0, bytes.length, options);
937
938 // Check to see if it's gzip-compressed
939 // GZIP Magic Two-Byte Number: 0x8b1f (35615)
940
941 if (bytes != null && bytes.length >= 4) {
942 int head = (bytes[0] & 0xff) | ((bytes[1] << 8) & 0xff00);
943 if (GZIPInputStream.GZIP_MAGIC == head) {
944 GZIPInputStream gzis = null;
945 ByteArrayOutputStream baos = new ByteArrayOutputStream();
946 try {
947 gzis = new GZIPInputStream(new ByteArrayInputStream(bytes));
948
949 byte[] buffer = new byte[2048];
950 for (int length; (length = gzis.read(buffer)) >= 0; ) {
951 baos.write(buffer, 0, length);
952 } // end while: reading input
953
954 // No error? Get new bytes.
955 bytes = baos.toByteArray();
956
957 } catch (IOException e) {
958 // Just return originally-decoded bytes
959
960 } finally {
961 try {
962 baos.close();
963 } catch (Exception e) {
964 LOG.error("error closing ByteArrayOutputStream", e);
965 }
966 if (gzis != null) {
967 try {
968 gzis.close();
969 } catch (Exception e) {
970 LOG.error("error closing GZIPInputStream", e);
971 }
972 }
973 } // end finally
974 } // end if: gzipped
975 } // end if: bytes.length >= 2
976
977 return bytes;
978 } // end decode
979
980 /**
981 * Attempts to decode Base64 data and deserialize a Java Object within.
982 * Returns <tt>null</tt> if there was an error.
983 *
984 * @param encodedObject The Base64 data to decode
985 * @return The decoded and deserialized object
986 * @since 1.5
987 */
988 public static Object decodeToObject(String encodedObject) {
989 // Decode and gunzip if necessary
990 byte[] objBytes = decode(encodedObject);
991
992 Object obj = null;
993 ObjectInputStream ois = null;
994 try {
995 ois = new ObjectInputStream(new ByteArrayInputStream(objBytes));
996 obj = ois.readObject();
997
998 } catch (IOException e) {
999 LOG.error("error decoding object", e);
1000
1001 } catch (ClassNotFoundException e) {
1002 LOG.error("error decoding object", e);
1003
1004 } finally {
1005 if (ois != null) {
1006 try {
1007 ois.close();
1008 } catch (Exception e) {
1009 LOG.error("error closing ObjectInputStream", e);
1010 }
1011 }
1012 } // end finally
1013
1014 return obj;
1015 } // end decodeObject
1016
1017 /**
1018 * Convenience method for encoding data to a file.
1019 *
1020 * @param dataToEncode byte array of data to encode in base64 form
1021 * @param filename Filename for saving encoded data
1022 * @return <tt>true</tt> if successful, <tt>false</tt> otherwise
1023 *
1024 * @since 2.1
1025 */
1026 public static boolean encodeToFile(byte[] dataToEncode, String filename) {
1027 boolean success = false;
1028 Base64OutputStream bos = null;
1029 try {
1030 bos = new Base64OutputStream(new FileOutputStream(filename), ENCODE);
1031 bos.write(dataToEncode);
1032 success = true;
1033
1034 } catch (IOException e) {
1035 LOG.error("error encoding file: " + filename, e);
1036 success = false;
1037
1038 } finally {
1039 if (bos != null) {
1040 try {
1041 bos.close();
1042 } catch (Exception e) {
1043 LOG.error("error closing Base64OutputStream", e);
1044 }
1045 }
1046 } // end finally
1047
1048 return success;
1049 } // end encodeToFile
1050
1051 /**
1052 * Convenience method for decoding data to a file.
1053 *
1054 * @param dataToDecode Base64-encoded data as a string
1055 * @param filename Filename for saving decoded data
1056 * @return <tt>true</tt> if successful, <tt>false</tt> otherwise
1057 *
1058 * @since 2.1
1059 */
1060 public static boolean decodeToFile(String dataToDecode, String filename) {
1061 boolean success = false;
1062 Base64OutputStream bos = null;
1063 try {
1064 bos = new Base64OutputStream(new FileOutputStream(filename), DECODE);
1065 bos.write(dataToDecode.getBytes(PREFERRED_ENCODING));
1066 success = true;
1067
1068 } catch (IOException e) {
1069 LOG.error("error decoding to file: " + filename, e);
1070 success = false;
1071
1072 } finally {
1073 if (bos != null) {
1074 try {
1075 bos.close();
1076 } catch (Exception e) {
1077 LOG.error("error closing Base64OutputStream", e);
1078 }
1079 }
1080 } // end finally
1081
1082 return success;
1083 } // end decodeToFile
1084
1085 /**
1086 * Convenience method for reading a base64-encoded file and decoding it.
1087 *
1088 * @param filename Filename for reading encoded data
1089 * @return decoded byte array or null if unsuccessful
1090 *
1091 * @since 2.1
1092 */
1093 public static byte[] decodeFromFile(String filename) {
1094 byte[] decodedData = null;
1095 Base64InputStream bis = null;
1096 try {
1097 File file = new File(filename);
1098 byte[] buffer;
1099
1100 // Check the size of file
1101 if (file.length() > Integer.MAX_VALUE) {
1102 LOG.fatal("File is too big for this convenience method (" +
1103 file.length() + " bytes).");
1104 return null;
1105 } // end if: file too big for int index
1106
1107 buffer = new byte[(int) file.length()];
1108
1109 // Open a stream
1110
1111 bis = new Base64InputStream(new BufferedInputStream(
1112 new FileInputStream(file)), DECODE);
1113
1114 // Read until done
1115
1116 int length = 0;
1117 for (int numBytes; (numBytes = bis.read(buffer, length, 4096)) >= 0; ) {
1118 length += numBytes;
1119 }
1120
1121 // Save in a variable to return
1122
1123 decodedData = new byte[length];
1124 System.arraycopy(buffer, 0, decodedData, 0, length);
1125
1126 } catch (IOException e) {
1127 LOG.error("Error decoding from file " + filename, e);
1128
1129 } finally {
1130 if (bis != null) {
1131 try {
1132 bis.close();
1133 } catch (Exception e) {
1134 LOG.error("error closing Base64InputStream", e);
1135 }
1136 }
1137 } // end finally
1138
1139 return decodedData;
1140 } // end decodeFromFile
1141
1142 /**
1143 * Convenience method for reading a binary file and base64-encoding it.
1144 *
1145 * @param filename Filename for reading binary data
1146 * @return base64-encoded string or null if unsuccessful
1147 *
1148 * @since 2.1
1149 */
1150 public static String encodeFromFile(String filename) {
1151 String encodedData = null;
1152 Base64InputStream bis = null;
1153 try {
1154 File file = new File(filename);
1155
1156 // Need max() for math on small files (v2.2.1)
1157
1158 byte[] buffer = new byte[Math.max((int) (file.length() * 1.4), 40)];
1159
1160 // Open a stream
1161
1162 bis = new Base64InputStream(new BufferedInputStream(
1163 new FileInputStream(file)), ENCODE);
1164
1165 // Read until done
1166 int length = 0;
1167 for (int numBytes; (numBytes = bis.read(buffer, length, 4096)) >= 0; ) {
1168 length += numBytes;
1169 }
1170
1171 // Save in a variable to return
1172
1173 encodedData = new String(buffer, 0, length, PREFERRED_ENCODING);
1174
1175 } catch (IOException e) {
1176 LOG.error("Error encoding from file " + filename, e);
1177
1178 } finally {
1179 if (bis != null) {
1180 try {
1181 bis.close();
1182 } catch (Exception e) {
1183 LOG.error("error closing Base64InputStream", e);
1184 }
1185 }
1186 } // end finally
1187
1188 return encodedData;
1189 } // end encodeFromFile
1190
1191 /**
1192 * Reads <tt>infile</tt> and encodes it to <tt>outfile</tt>.
1193 *
1194 * @param infile Input file
1195 * @param outfile Output file
1196 * @since 2.2
1197 */
1198 public static void encodeFileToFile(String infile, String outfile) {
1199 String encoded = encodeFromFile(infile);
1200 OutputStream out = null;
1201 try {
1202 out = new BufferedOutputStream(new FileOutputStream(outfile));
1203 out.write(encoded.getBytes("US-ASCII")); // Strict, 7-bit output.
1204
1205 } catch (IOException e) {
1206 LOG.error("error encoding from file " + infile + " to " + outfile, e);
1207
1208 } finally {
1209 if (out != null) {
1210 try {
1211 out.close();
1212 } catch (Exception e) {
1213 LOG.error("error closing " + outfile, e);
1214 }
1215 }
1216 } // end finally
1217 } // end encodeFileToFile
1218
1219 /**
1220 * Reads <tt>infile</tt> and decodes it to <tt>outfile</tt>.
1221 *
1222 * @param infile Input file
1223 * @param outfile Output file
1224 * @since 2.2
1225 */
1226 public static void decodeFileToFile(String infile, String outfile) {
1227 byte[] decoded = decodeFromFile(infile);
1228 OutputStream out = null;
1229 try {
1230 out = new BufferedOutputStream(new FileOutputStream(outfile));
1231 out.write(decoded);
1232
1233 } catch (IOException e) {
1234 LOG.error("error decoding from file " + infile + " to " + outfile, e);
1235
1236 } finally {
1237 if (out != null) {
1238 try {
1239 out.close();
1240 } catch (Exception e) {
1241 LOG.error("error closing " + outfile, e);
1242 }
1243 }
1244 } // end finally
1245 } // end decodeFileToFile
1246
1247 /* ******** I N N E R C L A S S I N P U T S T R E A M ******** */
1248
1249 /**
1250 * A {@link Base64.Base64InputStream} will read data from another
1251 * <tt>InputStream</tt>, given in the constructor, and
1252 * encode/decode to/from Base64 notation on the fly.
1253 *
1254 * @see Base64
1255 * @since 1.3
1256 */
1257 public static class Base64InputStream extends FilterInputStream {
1258 private boolean encode; // Encoding or decoding
1259 private int position; // Current position in the buffer
1260 private byte[] buffer; // Buffer holding converted data
1261 private int bufferLength; // Length of buffer (3 or 4)
1262 private int numSigBytes; // Meaningful bytes in the buffer
1263 private int lineLength;
1264 private boolean breakLines; // Break lines at < 80 characters
1265 private int options; // Record options
1266 private byte[] decodabet; // Local copy avoids method calls
1267
1268 /**
1269 * Constructs a {@link Base64InputStream} in DECODE mode.
1270 *
1271 * @param in the <tt>InputStream</tt> from which to read data.
1272 * @since 1.3
1273 */
1274 public Base64InputStream(InputStream in) {
1275 this(in, DECODE);
1276 } // end constructor
1277
1278 /**
1279 * Constructs a {@link Base64.Base64InputStream} in either ENCODE or DECODE mode.
1280 * <p>
1281 * Valid options:
1282 *
1283 * <pre>
1284 * ENCODE or DECODE: Encode or Decode as data is read.
1285 * DONT_BREAK_LINES: don't break lines at 76 characters
1286 * (only meaningful when encoding)
1287 * <i>Note: Technically, this makes your encoding non-compliant.</i>
1288 * </pre>
1289 *
1290 * <p>
1291 * Example: <code>new Base64.Base64InputStream( in, Base64.DECODE )</code>
1292 *
1293 *
1294 * @param in the <tt>InputStream</tt> from which to read data.
1295 * @param options Specified options
1296 * @see Base64#ENCODE
1297 * @see Base64#DECODE
1298 * @see Base64#DONT_BREAK_LINES
1299 * @since 2.0
1300 */
1301 public Base64InputStream(InputStream in, int options) {
1302 super(in);
1303 this.breakLines = (options & DONT_BREAK_LINES) != DONT_BREAK_LINES;
1304 this.encode = (options & ENCODE) == ENCODE;
1305 this.bufferLength = encode ? 4 : 3;
1306 this.buffer = new byte[bufferLength];
1307 this.position = -1;
1308 this.lineLength = 0;
1309 this.options = options; // Record for later, mostly to determine which
1310 // alphabet to use
1311 this.decodabet = getDecodabet(options);
1312 } // end constructor
1313
1314 /**
1315 * Reads enough of the input stream to convert to/from Base64 and returns
1316 * the next byte.
1317 *
1318 * @return next byte
1319 * @since 1.3
1320 */
1321 @Override
1322 public int read() throws IOException {
1323 // Do we need to get data?
1324 if (position < 0) {
1325 if (encode) {
1326 byte[] b3 = new byte[3];
1327 int numBinaryBytes = 0;
1328 for (int i = 0; i < 3; i++) {
1329 try {
1330 int b = in.read();
1331
1332 // If end of stream, b is -1.
1333 if (b >= 0) {
1334 b3[i] = (byte) b;
1335 numBinaryBytes++;
1336 } // end if: not end of stream
1337
1338 } catch (IOException e) {
1339 // Only a problem if we got no data at all.
1340 if (i == 0)
1341 throw e;
1342
1343 } // end catch
1344 } // end for: each needed input byte
1345
1346 if (numBinaryBytes > 0) {
1347 encode3to4(b3, 0, numBinaryBytes, buffer, 0, options);
1348 position = 0;
1349 numSigBytes = 4;
1350
1351 } else {
1352 return -1;
1353 } // end else
1354
1355 } else {
1356 byte[] b4 = new byte[4];
1357 int i;
1358 for (i = 0; i < 4; i++) {
1359 // Read four "meaningful" bytes:
1360 int b;
1361 do {
1362 b = in.read();
1363 } while (b >= 0 && decodabet[b & 0x7f] <= WHITE_SPACE_ENC);
1364
1365 if (b < 0) {
1366 break; // Reads a -1 if end of stream
1367 }
1368
1369 b4[i] = (byte) b;
1370 } // end for: each needed input byte
1371
1372 if (i == 4) {
1373 numSigBytes = decode4to3(b4, 0, buffer, 0, options);
1374 position = 0;
1375
1376 } else if (i == 0) {
1377 return -1;
1378
1379 } else {
1380 // Must have broken out from above.
1381 throw new IOException("Improperly padded Base64 input.");
1382 } // end
1383 } // end else: decode
1384 } // end else: get data
1385
1386 // Got data?
1387 if (position >= 0) {
1388 // End of relevant data?
1389 if ( /* !encode && */position >= numSigBytes) {
1390 return -1;
1391 }
1392
1393 if (encode && breakLines && lineLength >= MAX_LINE_LENGTH) {
1394 lineLength = 0;
1395 return '\n';
1396
1397 }
1398 lineLength++; // This isn't important when decoding
1399 // but throwing an extra "if" seems
1400 // just as wasteful.
1401
1402 int b = buffer[position++];
1403
1404 if (position >= bufferLength)
1405 position = -1;
1406
1407 return b & 0xFF; // This is how you "cast" a byte that's
1408 // intended to be unsigned.
1409
1410 }
1411
1412 // When JDK1.4 is more accepted, use an assertion here.
1413 throw new IOException("Error in Base64 code reading stream.");
1414
1415 } // end read
1416
1417 /**
1418 * Calls {@link #read()} repeatedly until the end of stream is reached or
1419 * <var>len</var> bytes are read. Returns number of bytes read into array
1420 * or -1 if end of stream is encountered.
1421 *
1422 * @param dest array to hold values
1423 * @param off offset for array
1424 * @param len max number of bytes to read into array
1425 * @return bytes read into array or -1 if end of stream is encountered.
1426 * @since 1.3
1427 */
1428 @Override
1429 public int read(byte[] dest, int off, int len) throws IOException {
1430 int i;
1431 int b;
1432 for (i = 0; i < len; i++) {
1433 b = read();
1434 if (b >= 0) {
1435 dest[off + i] = (byte) b;
1436 } else if (i == 0) {
1437 return -1;
1438 } else {
1439 break; // Out of 'for' loop
1440 }
1441 } // end for: each byte read
1442 return i;
1443 } // end read
1444
1445 } // end inner class InputStream
1446
1447 /* ******** I N N E R C L A S S O U T P U T S T R E A M ******** */
1448
1449 /**
1450 * A {@link Base64.Base64OutputStream} will write data to another
1451 * <tt>OutputStream</tt>, given in the constructor, and
1452 * encode/decode to/from Base64 notation on the fly.
1453 *
1454 * @see Base64
1455 * @since 1.3
1456 */
1457 public static class Base64OutputStream extends FilterOutputStream {
1458 private boolean encode;
1459 private int position;
1460 private byte[] buffer;
1461 private int bufferLength;
1462 private int lineLength;
1463 private boolean breakLines;
1464 private byte[] b4; // Scratch used in a few places
1465 private boolean suspendEncoding;
1466 private int options; // Record for later
1467 private byte[] decodabet; // Local copy avoids method calls
1468
1469 /**
1470 * Constructs a {@link Base64OutputStream} in ENCODE mode.
1471 *
1472 * @param out the <tt>OutputStream</tt> to which data will be written.
1473 * @since 1.3
1474 */
1475 public Base64OutputStream(OutputStream out) {
1476 this(out, ENCODE);
1477 } // end constructor
1478
1479 /**
1480 * Constructs a {@link Base64OutputStream} in either ENCODE or DECODE mode.
1481 * <p>
1482 * Valid options:
1483 *
1484 * <ul>
1485 * <li>ENCODE or DECODE: Encode or Decode as data is read.</li>
1486 * <li>DONT_BREAK_LINES: don't break lines at 76 characters (only
1487 * meaningful when encoding) <i>Note: Technically, this makes your
1488 * encoding non-compliant.</i></li>
1489 * </ul>
1490 *
1491 * <p>
1492 * Example: <code>new Base64.Base64OutputStream( out, Base64.ENCODE )</code>
1493 *
1494 * @param out the <tt>OutputStream</tt> to which data will be written.
1495 * @param options Specified options.
1496 * @see Base64#ENCODE
1497 * @see Base64#DECODE
1498 * @see Base64#DONT_BREAK_LINES
1499 * @since 1.3
1500 */
1501 public Base64OutputStream(OutputStream out, int options) {
1502 super(out);
1503 this.breakLines = (options & DONT_BREAK_LINES) != DONT_BREAK_LINES;
1504 this.encode = (options & ENCODE) == ENCODE;
1505 this.bufferLength = encode ? 3 : 4;
1506 this.buffer = new byte[bufferLength];
1507 this.position = 0;
1508 this.lineLength = 0;
1509 this.suspendEncoding = false;
1510 this.b4 = new byte[4];
1511 this.options = options;
1512 this.decodabet = getDecodabet(options);
1513 } // end constructor
1514
1515 /**
1516 * Writes the byte to the output stream after converting to/from Base64
1517 * notation. When encoding, bytes are buffered three at a time before the
1518 * output stream actually gets a write() call. When decoding, bytes are
1519 * buffered four at a time.
1520 *
1521 * @param theByte the byte to write
1522 * @since 1.3
1523 */
1524 @Override
1525 public void write(int theByte) throws IOException {
1526 // Encoding suspended?
1527 if (suspendEncoding) {
1528 super.out.write(theByte);
1529 return;
1530 } // end if: supsended
1531
1532 // Encode?
1533 if (encode) {
1534 buffer[position++] = (byte) theByte;
1535 if (position >= bufferLength) { // Enough to encode.
1536 out.write(encode3to4(b4, buffer, bufferLength, options));
1537 lineLength += 4;
1538 if (breakLines && lineLength >= MAX_LINE_LENGTH) {
1539 out.write(NEW_LINE);
1540 lineLength = 0;
1541 } // end if: end of line
1542
1543 position = 0;
1544 } // end if: enough to output
1545
1546 } else {
1547 // Meaningful Base64 character?
1548 if (decodabet[theByte & 0x7f] > WHITE_SPACE_ENC) {
1549 buffer[position++] = (byte) theByte;
1550 if (position >= bufferLength) { // Enough to output.
1551 int len = decode4to3(buffer, 0, b4, 0, options);
1552 out.write(b4, 0, len);
1553 position = 0;
1554 } // end if: enough to output
1555
1556 } else if (decodabet[theByte & 0x7f] != WHITE_SPACE_ENC) {
1557 throw new IOException("Invalid character in Base64 data.");
1558 } // end else: not white space either
1559 } // end else: decoding
1560 } // end write
1561
1562 /**
1563 * Calls {@link #write(int)} repeatedly until <var>len</var> bytes are
1564 * written.
1565 *
1566 * @param theBytes array from which to read bytes
1567 * @param off offset for array
1568 * @param len max number of bytes to read into array
1569 * @since 1.3
1570 */
1571 @Override
1572 public void write(byte[] theBytes, int off, int len) throws IOException {
1573 // Encoding suspended?
1574 if (suspendEncoding) {
1575 super.out.write(theBytes, off, len);
1576 return;
1577 } // end if: supsended
1578
1579 for (int i = 0; i < len; i++) {
1580 write(theBytes[off + i]);
1581 } // end for: each byte written
1582
1583 } // end write
1584
1585 /**
1586 * Method added by PHIL. [Thanks, PHIL. -Rob] This pads the buffer without
1587 * closing the stream.
1588 *
1589 * @throws IOException e
1590 */
1591 public void flushBase64() throws IOException {
1592 if (position > 0) {
1593 if (encode) {
1594 out.write(encode3to4(b4, buffer, position, options));
1595 position = 0;
1596
1597 } else {
1598 throw new IOException("Base64 input not properly padded.");
1599 } // end else: decoding
1600 } // end if: buffer partially full
1601
1602 } // end flush
1603
1604 /**
1605 * Flushes and closes (I think, in the superclass) the stream.
1606 *
1607 * @since 1.3
1608 */
1609 @Override
1610 public void close() throws IOException {
1611 // 1. Ensure that pending characters are written
1612 flushBase64();
1613
1614 // 2. Actually close the stream
1615 // Base class both flushes and closes.
1616 super.close();
1617
1618 buffer = null;
1619 out = null;
1620 } // end close
1621
1622 /**
1623 * Suspends encoding of the stream. May be helpful if you need to embed a
1624 * piece of base640-encoded data in a stream.
1625 *
1626 * @throws IOException e
1627 * @since 1.5.1
1628 */
1629 public void suspendEncoding() throws IOException {
1630 flushBase64();
1631 this.suspendEncoding = true;
1632 } // end suspendEncoding
1633
1634 /**
1635 * Resumes encoding of the stream. May be helpful if you need to embed a
1636 * piece of base640-encoded data in a stream.
1637 *
1638 * @since 1.5.1
1639 */
1640 public void resumeEncoding() {
1641 this.suspendEncoding = false;
1642 } // end resumeEncoding
1643
1644 } // end inner class OutputStream
1645
1646 } // end class Base64