1 /**
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17 package org.apache.openejb.util;
18
19 import java.io.IOException;
20
21 /**
22 * Provides Base64 encoding and decoding as defined by RFC 2045.
23 *
24 * <p>This class implements section <cite>6.8. Base64 Content-Transfer-Encoding</cite>
25 * from RFC 2045 <cite>Multipurpose Internet Mail Extensions (MIME) Part One:
26 * Format of Internet Message Bodies</cite> by Freed and Borenstein.</p>
27 *
28 * @see <a href="http://www.ietf.org/rfc/rfc2045.txt">RFC 2045</a>
29 * @author Apache Software Foundation
30 * @since 1.0-dev
31 * @version $Id: Base64.java 602704 2007-12-09 17:58:22Z jlaskowski $
32 */
33 public class Base64 {
34
35 /**
36 * Chunk size per RFC 2045 section 6.8.
37 *
38 * <p>The {@value} character limit does not count the trailing CRLF, but counts
39 * all other characters, including any equal signs.</p>
40 *
41 * @see <a href="http://www.ietf.org/rfc/rfc2045.txt">RFC 2045 section 6.8</a>
42 */
43 static final int CHUNK_SIZE = 76;
44
45 /**
46 * Chunk separator per RFC 2045 section 2.1.
47 *
48 * @see <a href="http://www.ietf.org/rfc/rfc2045.txt">RFC 2045 section 2.1</a>
49 */
50 static final byte[] CHUNK_SEPARATOR = "\r\n".getBytes();
51
52 /**
53 * The base length.
54 */
55 static final int BASELENGTH = 255;
56
57 /**
58 * Lookup length.
59 */
60 static final int LOOKUPLENGTH = 64;
61
62 /**
63 * Used to calculate the number of bits in a byte.
64 */
65 static final int EIGHTBIT = 8;
66
67 /**
68 * Used when encoding something which has fewer than 24 bits.
69 */
70 static final int SIXTEENBIT = 16;
71
72 /**
73 * Used to determine how many bits data contains.
74 */
75 static final int TWENTYFOURBITGROUP = 24;
76
77 /**
78 * Used to get the number of Quadruples.
79 */
80 static final int FOURBYTE = 4;
81
82 /**
83 * Used to test the sign of a byte.
84 */
85 static final int SIGN = -128;
86
87 /**
88 * Byte used to pad output.
89 */
90 static final byte PAD = (byte) '=';
91
92 /**
93 * Contains the Base64 values <code>0</code> through <code>63</code> accessed by using character encodings as
94 * indices.
95 * <p>
96 * For example, <code>base64Alphabet['+']</code> returns <code>62</code>.
97 * </p>
98 * <p>
99 * The value of undefined encodings is <code>-1</code>.
100 * </p>
101 */
102 private static byte[] base64Alphabet = new byte[BASELENGTH];
103
104 /**
105 * <p>
106 * Contains the Base64 encodings <code>A</code> through <code>Z</code>, followed by <code>a</code> through
107 * <code>z</code>, followed by <code>0</code> through <code>9</code>, followed by <code>+</code>, and
108 * <code>/</code>.
109 * </p>
110 * <p>
111 * This array is accessed by using character values as indices.
112 * </p>
113 * <p>
114 * For example, <code>lookUpBase64Alphabet[62] </code> returns <code>'+'</code>.
115 * </p>
116 */
117 private static byte[] lookUpBase64Alphabet = new byte[LOOKUPLENGTH];
118
119 // Populating the lookup and character arrays
120 static {
121 for (int i = 0; i < BASELENGTH; i++) {
122 base64Alphabet[i] = (byte) -1;
123 }
124 for (int i = 'Z'; i >= 'A'; i--) {
125 base64Alphabet[i] = (byte) (i - 'A');
126 }
127 for (int i = 'z'; i >= 'a'; i--) {
128 base64Alphabet[i] = (byte) (i - 'a' + 26);
129 }
130 for (int i = '9'; i >= '0'; i--) {
131 base64Alphabet[i] = (byte) (i - '0' + 52);
132 }
133
134 base64Alphabet['+'] = 62;
135 base64Alphabet['/'] = 63;
136
137 for (int i = 0; i <= 25; i++) {
138 lookUpBase64Alphabet[i] = (byte) ('A' + i);
139 }
140
141 for (int i = 26, j = 0; i <= 51; i++, j++) {
142 lookUpBase64Alphabet[i] = (byte) ('a' + j);
143 }
144
145 for (int i = 52, j = 0; i <= 61; i++, j++) {
146 lookUpBase64Alphabet[i] = (byte) ('0' + j);
147 }
148
149 lookUpBase64Alphabet[62] = (byte) '+';
150 lookUpBase64Alphabet[63] = (byte) '/';
151 }
152
153 /**
154 * Returns whether or not the <code>octect</code> is in the base 64 alphabet.
155 *
156 * @param octect The value to test
157 * @return <code>true</code> if the value is defined in the the base 64 alphabet, <code>false</code> otherwise.
158 */
159 private static boolean isBase64(byte octect) {
160 if (octect == PAD) {
161 return true;
162 } else if (octect < 0 || base64Alphabet[octect] == -1) {
163 return false;
164 } else {
165 return true;
166 }
167 }
168
169 /**
170 * Tests a given byte array to see if it contains
171 * only valid characters within the Base64 alphabet.
172 *
173 * @param arrayOctect byte array to test
174 * @return <code>true</code> if all bytes are valid characters in the Base64
175 * alphabet or if the byte array is empty; false, otherwise
176 */
177 public static boolean isArrayByteBase64(byte[] arrayOctect) {
178
179 arrayOctect = discardWhitespace(arrayOctect);
180
181 int length = arrayOctect.length;
182 if (length == 0) {
183 // shouldn't a 0 length array be valid base64 data?
184 // return false;
185 return true;
186 }
187 for (int i = 0; i < length; i++) {
188 if (!isBase64(arrayOctect[i])) {
189 return false;
190 }
191 }
192 return true;
193 }
194
195 /**
196 * Encodes binary data using the base64 algorithm but
197 * does not chunk the output.
198 *
199 * @param binaryData binary data to encode
200 * @return Base64 characters
201 */
202 public static byte[] encodeBase64(byte[] binaryData) {
203 return encodeBase64(binaryData, false);
204 }
205
206 /**
207 * Encodes binary data using the base64 algorithm and chunks
208 * the encoded output into 76 character blocks
209 *
210 * @param binaryData binary data to encode
211 * @return Base64 characters chunked in 76 character blocks
212 */
213 public static byte[] encodeBase64Chunked(byte[] binaryData) {
214 return encodeBase64(binaryData, true);
215 }
216
217
218 /**
219 * Decodes an Object using the base64 algorithm. This method
220 * is provided in order to satisfy the requirements of the
221 * Decoder interface, and will throw a DecoderException if the
222 * supplied object is not of type byte[].
223 *
224 * @param pObject Object to decode
225 * @return An object (of type byte[]) containing the
226 * binary data which corresponds to the byte[] supplied.
227 * @throws IOException if the parameter supplied is not
228 * of type byte[]
229 */
230 public Object decode(Object pObject) throws IOException {
231 if (!(pObject instanceof byte[])) {
232 throw new IOException("Parameter supplied to Base64 decode is not a byte[]");
233 }
234 return decode((byte[]) pObject);
235 }
236
237 /**
238 * Decodes a byte[] containing containing
239 * characters in the Base64 alphabet.
240 *
241 * @param pArray A byte array containing Base64 character data
242 * @return a byte array containing binary data
243 */
244 public byte[] decode(byte[] pArray) {
245 return decodeBase64(pArray);
246 }
247
248 /**
249 * Encodes binary data using the base64 algorithm, optionally
250 * chunking the output into 76 character blocks.
251 *
252 * @param binaryData Array containing binary data to encode.
253 * @param isChunked if <code>true</code> this encoder will chunk
254 * the base64 output into 76 character blocks
255 * @return Base64-encoded data.
256 */
257 public static byte[] encodeBase64(byte[] binaryData, boolean isChunked) {
258 int lengthDataBits = binaryData.length * EIGHTBIT;
259 int fewerThan24bits = lengthDataBits % TWENTYFOURBITGROUP;
260 int numberTriplets = lengthDataBits / TWENTYFOURBITGROUP;
261 byte encodedData[] = null;
262 int encodedDataLength = 0;
263 int nbrChunks = 0;
264
265 if (fewerThan24bits != 0) {
266 //data not divisible by 24 bit
267 encodedDataLength = (numberTriplets + 1) * 4;
268 } else {
269 // 16 or 8 bit
270 encodedDataLength = numberTriplets * 4;
271 }
272
273 // If the output is to be "chunked" into 76 character sections,
274 // for compliance with RFC 2045 MIME, then it is important to
275 // allow for extra length to account for the separator(s)
276 if (isChunked) {
277
278 nbrChunks =
279 (CHUNK_SEPARATOR.length == 0 ? 0 : (int) Math.ceil((float) encodedDataLength / CHUNK_SIZE));
280 encodedDataLength += nbrChunks * CHUNK_SEPARATOR.length;
281 }
282
283 encodedData = new byte[encodedDataLength];
284
285 byte k = 0, l = 0, b1 = 0, b2 = 0, b3 = 0;
286
287 int encodedIndex = 0;
288 int dataIndex = 0;
289 int i = 0;
290 int nextSeparatorIndex = CHUNK_SIZE;
291 int chunksSoFar = 0;
292
293 //log.debug("number of triplets = " + numberTriplets);
294 for (i = 0; i < numberTriplets; i++) {
295 dataIndex = i * 3;
296 b1 = binaryData[dataIndex];
297 b2 = binaryData[dataIndex + 1];
298 b3 = binaryData[dataIndex + 2];
299
300 //log.debug("b1= " + b1 +", b2= " + b2 + ", b3= " + b3);
301
302 l = (byte) (b2 & 0x0f);
303 k = (byte) (b1 & 0x03);
304
305 byte val1 =
306 ((b1 & SIGN) == 0) ? (byte) (b1 >> 2) : (byte) ((b1) >> 2 ^ 0xc0);
307 byte val2 =
308 ((b2 & SIGN) == 0) ? (byte) (b2 >> 4) : (byte) ((b2) >> 4 ^ 0xf0);
309 byte val3 =
310 ((b3 & SIGN) == 0) ? (byte) (b3 >> 6) : (byte) ((b3) >> 6 ^ 0xfc);
311
312 encodedData[encodedIndex] = lookUpBase64Alphabet[val1];
313 //log.debug( "val2 = " + val2 );
314 //log.debug( "k4 = " + (k<<4) );
315 //log.debug( "vak = " + (val2 | (k<<4)) );
316 encodedData[encodedIndex + 1] =
317 lookUpBase64Alphabet[val2 | (k << 4)];
318 encodedData[encodedIndex + 2] =
319 lookUpBase64Alphabet[(l << 2) | val3];
320 encodedData[encodedIndex + 3] = lookUpBase64Alphabet[b3 & 0x3f];
321
322 encodedIndex += 4;
323
324 // If we are chunking, let's put a chunk separator down.
325 if (isChunked) {
326 // this assumes that CHUNK_SIZE % 4 == 0
327 if (encodedIndex == nextSeparatorIndex) {
328 System.arraycopy(
329 CHUNK_SEPARATOR,
330 0,
331 encodedData,
332 encodedIndex,
333 CHUNK_SEPARATOR.length);
334 chunksSoFar++;
335 nextSeparatorIndex =
336 (CHUNK_SIZE * (chunksSoFar + 1)) +
337 (chunksSoFar * CHUNK_SEPARATOR.length);
338 encodedIndex += CHUNK_SEPARATOR.length;
339 }
340 }
341 }
342
343 // form integral number of 6-bit groups
344 dataIndex = i * 3;
345
346 if (fewerThan24bits == EIGHTBIT) {
347 b1 = binaryData[dataIndex];
348 k = (byte) (b1 & 0x03);
349 //log.debug("b1=" + b1);
350 //log.debug("b1<<2 = " + (b1>>2) );
351 byte val1 =
352 ((b1 & SIGN) == 0) ? (byte) (b1 >> 2) : (byte) ((b1) >> 2 ^ 0xc0);
353 encodedData[encodedIndex] = lookUpBase64Alphabet[val1];
354 encodedData[encodedIndex + 1] = lookUpBase64Alphabet[k << 4];
355 encodedData[encodedIndex + 2] = PAD;
356 encodedData[encodedIndex + 3] = PAD;
357 } else if (fewerThan24bits == SIXTEENBIT) {
358
359 b1 = binaryData[dataIndex];
360 b2 = binaryData[dataIndex + 1];
361 l = (byte) (b2 & 0x0f);
362 k = (byte) (b1 & 0x03);
363
364 byte val1 =
365 ((b1 & SIGN) == 0) ? (byte) (b1 >> 2) : (byte) ((b1) >> 2 ^ 0xc0);
366 byte val2 =
367 ((b2 & SIGN) == 0) ? (byte) (b2 >> 4) : (byte) ((b2) >> 4 ^ 0xf0);
368
369 encodedData[encodedIndex] = lookUpBase64Alphabet[val1];
370 encodedData[encodedIndex + 1] =
371 lookUpBase64Alphabet[val2 | (k << 4)];
372 encodedData[encodedIndex + 2] = lookUpBase64Alphabet[l << 2];
373 encodedData[encodedIndex + 3] = PAD;
374 }
375
376 if (isChunked) {
377 // we also add a separator to the end of the final chunk.
378 if (chunksSoFar < nbrChunks) {
379 System.arraycopy(
380 CHUNK_SEPARATOR,
381 0,
382 encodedData,
383 encodedDataLength - CHUNK_SEPARATOR.length,
384 CHUNK_SEPARATOR.length);
385 }
386 }
387
388 return encodedData;
389 }
390
391 /**
392 * Decodes Base64 data into octects
393 *
394 * @param base64Data Byte array containing Base64 data
395 * @return Array containing decoded data.
396 */
397 public static byte[] decodeBase64(byte[] base64Data) {
398 // RFC 2045 requires that we discard ALL non-Base64 characters
399 base64Data = discardNonBase64(base64Data);
400
401 // handle the edge case, so we don't have to worry about it later
402 if (base64Data.length == 0) {
403 return new byte[0];
404 }
405
406 int numberQuadruple = base64Data.length / FOURBYTE;
407 byte decodedData[] = null;
408 byte b1 = 0, b2 = 0, b3 = 0, b4 = 0, marker0 = 0, marker1 = 0;
409
410 // Throw away anything not in base64Data
411
412 int encodedIndex = 0;
413 int dataIndex = 0;
414 {
415 // this sizes the output array properly - rlw
416 int lastData = base64Data.length;
417 // ignore the '=' padding
418 while (base64Data[lastData - 1] == PAD) {
419 if (--lastData == 0) {
420 return new byte[0];
421 }
422 }
423 decodedData = new byte[lastData - numberQuadruple];
424 }
425
426 for (int i = 0; i < numberQuadruple; i++) {
427 dataIndex = i * 4;
428 marker0 = base64Data[dataIndex + 2];
429 marker1 = base64Data[dataIndex + 3];
430
431 b1 = base64Alphabet[base64Data[dataIndex]];
432 b2 = base64Alphabet[base64Data[dataIndex + 1]];
433
434 if (marker0 != PAD && marker1 != PAD) {
435 //No PAD e.g 3cQl
436 b3 = base64Alphabet[marker0];
437 b4 = base64Alphabet[marker1];
438
439 decodedData[encodedIndex] = (byte) (b1 << 2 | b2 >> 4);
440 decodedData[encodedIndex + 1] =
441 (byte) (((b2 & 0xf) << 4) | ((b3 >> 2) & 0xf));
442 decodedData[encodedIndex + 2] = (byte) (b3 << 6 | b4);
443 } else if (marker0 == PAD) {
444 //Two PAD e.g. 3c[Pad][Pad]
445 decodedData[encodedIndex] = (byte) (b1 << 2 | b2 >> 4);
446 } else if (marker1 == PAD) {
447 //One PAD e.g. 3cQ[Pad]
448 b3 = base64Alphabet[marker0];
449
450 decodedData[encodedIndex] = (byte) (b1 << 2 | b2 >> 4);
451 decodedData[encodedIndex + 1] =
452 (byte) (((b2 & 0xf) << 4) | ((b3 >> 2) & 0xf));
453 }
454 encodedIndex += 3;
455 }
456 return decodedData;
457 }
458
459 /**
460 * Discards any whitespace from a base-64 encoded block.
461 *
462 * @param data The base-64 encoded data to discard the whitespace
463 * from.
464 * @return The data, less whitespace (see RFC 2045).
465 */
466 static byte[] discardWhitespace(byte[] data) {
467 byte groomedData[] = new byte[data.length];
468 int bytesCopied = 0;
469
470 for (int i = 0; i < data.length; i++) {
471 switch (data[i]) {
472 case (byte) ' ' :
473 case (byte) '\n' :
474 case (byte) '\r' :
475 case (byte) '\t' :
476 break;
477 default:
478 groomedData[bytesCopied++] = data[i];
479 }
480 }
481
482 byte packedData[] = new byte[bytesCopied];
483
484 System.arraycopy(groomedData, 0, packedData, 0, bytesCopied);
485
486 return packedData;
487 }
488
489 /**
490 * Discards any characters outside of the base64 alphabet, per
491 * the requirements on page 25 of RFC 2045 - "Any characters
492 * outside of the base64 alphabet are to be ignored in base64
493 * encoded data."
494 *
495 * @param data The base-64 encoded data to groom
496 * @return The data, less non-base64 characters (see RFC 2045).
497 */
498 static byte[] discardNonBase64(byte[] data) {
499 byte groomedData[] = new byte[data.length];
500 int bytesCopied = 0;
501
502 for (int i = 0; i < data.length; i++) {
503 if (isBase64(data[i])) {
504 groomedData[bytesCopied++] = data[i];
505 }
506 }
507
508 byte packedData[] = new byte[bytesCopied];
509
510 System.arraycopy(groomedData, 0, packedData, 0, bytesCopied);
511
512 return packedData;
513 }
514
515
516 // Implementation of the Encoder Interface
517
518 /**
519 * Encodes an Object using the base64 algorithm. This method
520 * is provided in order to satisfy the requirements of the
521 * Encoder interface, and will throw an EncoderException if the
522 * supplied object is not of type byte[].
523 *
524 * @param pObject Object to encode
525 * @return An object (of type byte[]) containing the
526 * base64 encoded data which corresponds to the byte[] supplied.
527 * @throws IOException if the parameter supplied is not
528 * of type byte[]
529 */
530 public Object encode(Object pObject) throws IOException {
531 if (!(pObject instanceof byte[])) {
532 throw new IOException(
533 "Parameter supplied to Base64 encode is not a byte[]");
534 }
535 return encode((byte[]) pObject);
536 }
537
538 /**
539 * Encodes a byte[] containing binary data, into a byte[] containing
540 * characters in the Base64 alphabet.
541 *
542 * @param pArray a byte array containing binary data
543 * @return A byte array containing only Base64 character data
544 */
545 public byte[] encode(byte[] pArray) {
546 return encodeBase64(pArray, false);
547 }
548
549 }