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 * https://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
18 package org.apache.commons.codec.binary;
19
20 import java.util.Arrays;
21
22 import org.apache.commons.codec.CodecPolicy;
23
24 /**
25 * Provides Base32 encoding and decoding as defined by <a href="https://www.ietf.org/rfc/rfc4648.txt">RFC 4648</a>.
26 *
27 * <p>
28 * The class can be parameterized in the following manner with various constructors:
29 * </p>
30 * <ul>
31 * <li>Whether to use the "base32hex" variant instead of the default "base32"</li>
32 * <li>Line length: Default 76. Line length that aren't multiples of 8 will still essentially end up being multiples of 8 in the encoded data.
33 * <li>Line separator: Default is CRLF ("\r\n")</li>
34 * </ul>
35 * <p>
36 * This class operates directly on byte streams, and not character streams.
37 * </p>
38 * <p>
39 * This class is thread-safe.
40 * </p>
41 * <p>
42 * To configure a new instance, use a {@link Builder}. For example:
43 * </p>
44 * <pre>
45 * Base32 base32 = Base32.builder()
46 * .setDecodingPolicy(DecodingPolicy.LENIENT) // default is lenient
47 * .setLineLength(0) // default is none
48 * .setLineSeparator('\r', '\n') // default is CR LF
49 * .setPadding('=') // default is '='
50 * .setEncodeTable(customEncodeTable) // default is RFC 4648 Section 6, Table 3: The Base 32 Alphabet
51 * .get()
52 * </pre>
53 *
54 * @see Base32InputStream
55 * @see Base32OutputStream
56 * @see <a href="https://www.ietf.org/rfc/rfc4648.txt">RFC 4648</a>
57 * @since 1.5
58 */
59 public class Base32 extends BaseNCodec {
60
61 /**
62 * Builds {@link Base32} instances.
63 *
64 * <p>
65 * To configure a new instance, use a {@link Builder}. For example:
66 * </p>
67 *
68 * <pre>
69 * Base32 base32 = Base32.builder()
70 * .setDecodingPolicy(DecodingPolicy.LENIENT) // default is lenient
71 * .setLineLength(0) // default is none
72 * .setLineSeparator('\r', '\n') // default is CR LF
73 * .setPadding('=') // default is '='
74 * .setEncodeTable(customEncodeTable) // default is RFC 4648 Section 6, Table 3: The Base 32 Alphabet
75 * .get()
76 * </pre>
77 *
78 * @since 1.17.0
79 */
80 public static class Builder extends AbstractBuilder<Base32, Builder> {
81
82 /**
83 * Constructs a new instance using <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
84 * Alphabet</a>.
85 */
86 public Builder() {
87 super(ENCODE_TABLE);
88 setDecodeTableRaw(DECODE_TABLE);
89 setEncodeTableRaw(ENCODE_TABLE);
90 setEncodedBlockSize(BYTES_PER_ENCODED_BLOCK);
91 setUnencodedBlockSize(BYTES_PER_UNENCODED_BLOCK);
92 }
93
94 @Override
95 public Base32 get() {
96 return new Base32(this);
97 }
98
99 /**
100 * Sets the encode table and derives the matching decode table.
101 * <p>
102 * The RFC 4648 Base32 and Base32 Hex tables keep their case-insensitive decoders in lenient mode. Strict decoding requires the encoding alphabet.
103 * </p>
104 *
105 * @param encodeTable The encode table with exactly 32 unique entries, null resets to the default.
106 * @return {@code this} instance.
107 * @throws IllegalArgumentException Thrown if the encode table does not contain exactly 32 unique entries.
108 */
109 @Override
110 public Builder setEncodeTable(final byte... encodeTable) {
111 super.setDecodeTableRaw(toDecodeTable(encodeTable));
112 return super.setEncodeTable(encodeTable);
113 }
114
115 /**
116 * Sets the encode and decode tables to use Base32 hexadecimal if {@code true}, otherwise use the Base32 alphabet.
117 * <p>
118 * This overrides a value previously set with {@link #setEncodeTable(byte...)}.
119 * </p>
120 *
121 * @param useHex use Base32 hexadecimal if {@code true}, otherwise use the Base32 alphabet.
122 * @return {@code this} instance.
123 * @since 1.18.0
124 */
125 public Builder setHexDecodeTable(final boolean useHex) {
126 return setEncodeTable(encodeTable(useHex));
127 }
128
129 /**
130 * Sets the encode table to use Base32 hexadecimal if {@code true}, otherwise use the Base32 alphabet.
131 * <p>
132 * This overrides a value previously set with {@link #setEncodeTable(byte...)}.
133 * </p>
134 *
135 * @param useHex
136 * <ul>
137 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding
138 * with Extended Hex Alphabet</a></li>
139 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
140 * Alphabet</a></li>
141 * </ul>
142 * @return {@code this} instance.
143 * @since 1.18.0
144 */
145 public Builder setHexEncodeTable(final boolean useHex) {
146 return setEncodeTable(encodeTable(useHex));
147 }
148 }
149
150 /**
151 * BASE32 characters are 5 bits in length. They are formed by taking a block of five octets to form a 40-bit string, which is converted into eight BASE32
152 * characters.
153 */
154 private static final int BITS_PER_ENCODED_BYTE = 5;
155
156 private static final int BYTES_PER_ENCODED_BLOCK = 8;
157 private static final int BYTES_PER_UNENCODED_BLOCK = 5;
158 private static final int DECODING_TABLE_LENGTH = 256;
159 private static final int ENCODING_TABLE_LENGTH = 1 << BITS_PER_ENCODED_BYTE;
160
161 /**
162 * This array is a lookup table that translates Unicode characters drawn from the "Base32 Alphabet" (as specified in Table 3 of RFC 4648) into their 5-bit
163 * positive integer equivalents. Characters that are not in the Base32 alphabet but fall within the bounds of the array are translated to -1.
164 */
165 // @formatter:off
166 private static final byte[] DECODE_TABLE = {
167 // 0 1 2 3 4 5 6 7 8 9 A B C D E F
168 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 00-0f
169 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 10-1f
170 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 20-2f
171 -1, -1, 26, 27, 28, 29, 30, 31, -1, -1, -1, -1, -1, -1, -1, -1, // 30-3f 2-7
172 -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, // 40-4f A-O
173 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, // 50-5a P-Z
174 -1, -1, -1, -1, -1, // 5b-5f
175 -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, // 60-6f a-o
176 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, // 70-7a p-z
177 };
178 // @formatter:on
179
180 /**
181 * This array is a lookup table that translates 5-bit positive integer index values into their "Base32 Alphabet" equivalents as specified in
182 * <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32 Alphabet</a>.
183 *
184 * @see <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32 Alphabet</a>
185 */
186 // @formatter:off
187 private static final byte[] ENCODE_TABLE = {
188 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',
189 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z',
190 '2', '3', '4', '5', '6', '7',
191 };
192 // @formatter:on
193
194 /**
195 * This array is a lookup table that translates Unicode characters drawn from the "Base32 Hex Alphabet" (as specified in Table 4 of RFC 4648) into their
196 * 5-bit positive integer equivalents. Characters that are not in the Base32 Hex alphabet but fall within the bounds of the array are translated to -1.
197 */
198 // @formatter:off
199 private static final byte[] HEX_DECODE_TABLE = {
200 // 0 1 2 3 4 5 6 7 8 9 A B C D E F
201 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 00-0f
202 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 10-1f
203 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 20-2f
204 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, -1, -1, -1, -1, -1, // 30-3f 0-9
205 -1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, // 40-4f A-O
206 25, 26, 27, 28, 29, 30, 31, // 50-56 P-V
207 -1, -1, -1, -1, -1, -1, -1, -1, -1, // 57-5f
208 -1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, // 60-6f a-o
209 25, 26, 27, 28, 29, 30, 31 // 70-76 p-v
210 };
211 // @formatter:on
212
213 /**
214 * This array is a lookup table that translates 5-bit positive integer index values into their "Base 32 Encoding with Extended Hex Alphabet" equivalents as
215 * specified in <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with Extended Hex
216 * Alphabet</a>.
217 *
218 * @see <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with Extended Hex Alphabet</a>
219 */
220 // @formatter:off
221 private static final byte[] HEX_ENCODE_TABLE = {
222 '0', '1', '2', '3', '4', '5', '6', '7', '8', '9',
223 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',
224 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V',
225 };
226 // @formatter:on
227
228 /** Mask used to extract 5 bits, used when encoding Base32 bytes */
229 private static final int MASK_5_BITS = 0x1f;
230
231 /** Mask used to extract 4 bits, used when decoding final trailing character. */
232 private static final long MASK_4_BITS = 0x0fL;
233
234 /** Mask used to extract 3 bits, used when decoding final trailing character. */
235 private static final long MASK_3_BITS = 0x07L;
236
237 /** Mask used to extract 2 bits, used when decoding final trailing character. */
238 private static final long MASK_2_BITS = 0x03L;
239
240 /** Mask used to extract 1 bits, used when decoding final trailing character. */
241 private static final long MASK_1_BITS = 0x01L;
242
243 // The static final fields above are used for the original static byte[] methods on Base32.
244 // The private member fields below are used with the new streaming approach, which requires
245 // some state be preserved between calls of encode() and decode().
246
247 /**
248 * Creates a new Builder.
249 *
250 * <p>
251 * To configure a new instance, use a {@link Builder}. For example:
252 * </p>
253 *
254 * <pre>
255 * Base32 base32 = Base32.builder()
256 * .setDecodingPolicy(DecodingPolicy.LENIENT) // default is lenient
257 * .setLineLength(0) // default is none
258 * .setLineSeparator('\r', '\n') // default is CR LF
259 * .setPadding('=') // default is '='
260 * .setEncodeTable(customEncodeTable) // default is RFC 4648 Section 6, Table 3: The Base 32 Alphabet
261 * .get()
262 * </pre>
263 *
264 * @return A new Builder.
265 * @since 1.17.0
266 */
267 public static Builder builder() {
268 return new Builder();
269 }
270
271 /**
272 * Calculates a decode table for a given encode table.
273 *
274 * @param encodeTable that is used to determine decode lookup table.
275 * @return A new decode table.
276 * @throws IllegalArgumentException Thrown if the encode table does not contain exactly 32 unique entries.
277 */
278 private static byte[] calculateDecodeTable(final byte[] encodeTable) {
279 if (encodeTable.length != ENCODING_TABLE_LENGTH) {
280 throw new IllegalArgumentException("encodeTable must have exactly 32 entries.");
281 }
282 final byte[] decodeTable = new byte[DECODING_TABLE_LENGTH];
283 Arrays.fill(decodeTable, (byte) -1);
284 for (int i = 0; i < encodeTable.length; i++) {
285 final int encodedByte = encodeTable[i] & 0xff;
286 if (decodeTable[encodedByte] != -1) {
287 throw new IllegalArgumentException("encodeTable must not contain duplicate entries.");
288 }
289 decodeTable[encodedByte] = (byte) i;
290 }
291 return decodeTable;
292 }
293
294 private static byte[] decodeTable(final boolean useHex) {
295 return useHex ? HEX_DECODE_TABLE : DECODE_TABLE;
296 }
297
298 /**
299 * Gets the encoding table that matches {@code useHex}.
300 *
301 * @param useHex
302 * <ul>
303 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
304 * Extended Hex Alphabet</a></li>
305 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
306 * Alphabet</a></li>
307 * </ul>
308 * @return The encoding table that matches {@code useHex}.
309 */
310 private static byte[] encodeTable(final boolean useHex) {
311 return useHex ? HEX_ENCODE_TABLE : ENCODE_TABLE;
312 }
313
314 /**
315 * Gets the decode table that matches the given encode table.
316 *
317 * @param encodeTable that is used to determine decode lookup table.
318 * @return The matching decode table.
319 */
320 private static byte[] toDecodeTable(final byte[] encodeTable) {
321 final byte[] table = encodeTable != null ? encodeTable : ENCODE_TABLE;
322 if (Arrays.equals(table, ENCODE_TABLE)) {
323 return DECODE_TABLE;
324 }
325 if (Arrays.equals(table, HEX_ENCODE_TABLE)) {
326 return HEX_DECODE_TABLE;
327 }
328 return calculateDecodeTable(table);
329 }
330
331 /**
332 * Convenience variable to help us determine when our buffer is going to run out of room and needs resizing. {@code encodeSize = {@link
333 * #BYTES_PER_ENCODED_BLOCK} + lineSeparator.length;}
334 */
335 private final int encodeSize;
336
337 /**
338 * Line separator for encoding and strict decoding. Only used if lineLength > 0.
339 */
340 private final byte[] lineSeparator;
341
342 /**
343 * Constructs a Base32 codec used for decoding and encoding.
344 * <p>
345 * When encoding the line length is 0 (no chunking).
346 * </p>
347 */
348 public Base32() {
349 this(false);
350 }
351
352 /**
353 * Constructs a Base32 codec used for decoding and encoding.
354 * <p>
355 * When encoding the line length is 0 (no chunking).
356 * </p>
357 *
358 * @param useHex
359 * <ul>
360 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
361 * Extended Hex Alphabet</a></li>
362 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
363 * Alphabet</a></li>
364 * </ul>
365 * @deprecated Use {@link #builder()} and {@link Builder}.
366 */
367 @Deprecated
368 public Base32(final boolean useHex) {
369 this(0, null, useHex, PAD_DEFAULT);
370 }
371
372 /**
373 * Constructs a Base32 codec used for decoding and encoding.
374 * <p>
375 * When encoding the line length is 0 (no chunking).
376 * </p>
377 *
378 * @param useHex
379 * <ul>
380 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
381 * Extended Hex Alphabet</a></li>
382 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
383 * Alphabet</a></li>
384 * </ul>
385 * @param padding byte used as padding byte.
386 * @deprecated Use {@link #builder()} and {@link Builder}.
387 */
388 @Deprecated
389 public Base32(final boolean useHex, final byte padding) {
390 this(0, null, useHex, padding);
391 }
392
393 private Base32(final Builder builder) {
394 super(builder);
395 if (builder.getLineLength() > 0) {
396 final byte[] lineSeparator = builder.getLineSeparator();
397 // Must be done after initializing the tables
398 if (containsAlphabetOrPad(lineSeparator)) {
399 final String sep = StringUtils.newStringUtf8(lineSeparator);
400 throw new IllegalArgumentException("lineSeparator must not contain Base32 characters: [" + sep + "]");
401 }
402 this.encodeSize = BYTES_PER_ENCODED_BLOCK + lineSeparator.length;
403 this.lineSeparator = lineSeparator;
404 } else {
405 this.encodeSize = BYTES_PER_ENCODED_BLOCK;
406 this.lineSeparator = null;
407 }
408 if (isInAlphabet(builder.getPadding()) || Character.isWhitespace(builder.getPadding())) {
409 throw new IllegalArgumentException("pad must not be in alphabet or whitespace");
410 }
411 }
412
413 /**
414 * Constructs a Base32 codec used for decoding and encoding.
415 * <p>
416 * When encoding the line length is 0 (no chunking).
417 * </p>
418 *
419 * @param pad byte used as padding byte.
420 * @deprecated Use {@link #builder()} and {@link Builder}.
421 */
422 @Deprecated
423 public Base32(final byte pad) {
424 this(false, pad);
425 }
426
427 /**
428 * Constructs a Base32 codec used for decoding and encoding.
429 * <p>
430 * When encoding the line length is given in the constructor, the line separator is CRLF.
431 * </p>
432 *
433 * @param lineLength Each line of encoded data will be at most of the given length (rounded down to the nearest multiple of 8). If lineLength <= 0, then
434 * the output will not be divided into lines (chunks). Ignored when decoding leniently.
435 * @deprecated Use {@link #builder()} and {@link Builder}.
436 */
437 @Deprecated
438 public Base32(final int lineLength) {
439 this(lineLength, CHUNK_SEPARATOR);
440 }
441
442 /**
443 * Constructs a Base32 codec used for decoding and encoding.
444 * <p>
445 * When encoding the line length and line separator are given in the constructor.
446 * </p>
447 * <p>
448 * Line lengths that aren't multiples of 8 will still essentially end up being multiples of 8 in the encoded data.
449 * </p>
450 *
451 * @param lineLength Each line of encoded data will be at most of the given length (rounded down to the nearest multiple of 8). If lineLength <= 0,
452 * then the output will not be divided into lines (chunks). Ignored when decoding leniently.
453 * @param lineSeparator Each line of encoded data will end with this sequence of bytes.
454 * @throws IllegalArgumentException Thrown when the {@code lineSeparator} contains Base32 characters.
455 * @deprecated Use {@link #builder()} and {@link Builder}.
456 */
457 @Deprecated
458 public Base32(final int lineLength, final byte[] lineSeparator) {
459 this(lineLength, lineSeparator, false, PAD_DEFAULT);
460 }
461
462 /**
463 * Constructs a Base32 / Base32 Hex codec used for decoding and encoding.
464 * <p>
465 * When encoding the line length and line separator are given in the constructor.
466 * </p>
467 * <p>
468 * Line lengths that aren't multiples of 8 will still essentially end up being multiples of 8 in the encoded data.
469 * </p>
470 *
471 * @param lineLength Each line of encoded data will be at most of the given length (rounded down to the nearest multiple of 8). If lineLength <= 0,
472 * then the output will not be divided into lines (chunks). Ignored when decoding leniently.
473 * @param lineSeparator Each line of encoded data will end with this sequence of bytes.
474 * @param useHex
475 * <ul>
476 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
477 * Extended Hex Alphabet</a></li>
478 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
479 * Alphabet</a></li>
480 * </ul>
481 * @throws IllegalArgumentException Thrown when the {@code lineSeparator} contains Base32 characters. Or the lineLength > 0 and lineSeparator is null.
482 * @deprecated Use {@link #builder()} and {@link Builder}.
483 */
484 @Deprecated
485 public Base32(final int lineLength, final byte[] lineSeparator, final boolean useHex) {
486 this(lineLength, lineSeparator, useHex, PAD_DEFAULT);
487 }
488
489 /**
490 * Constructs a Base32 / Base32 Hex codec used for decoding and encoding.
491 * <p>
492 * When encoding the line length and line separator are given in the constructor.
493 * </p>
494 * <p>
495 * Line lengths that aren't multiples of 8 will still essentially end up being multiples of 8 in the encoded data.
496 * </p>
497 *
498 * @param lineLength Each line of encoded data will be at most of the given length (rounded down to the nearest multiple of 8). If lineLength <= 0,
499 * then the output will not be divided into lines (chunks). Ignored when decoding leniently.
500 * @param lineSeparator Each line of encoded data will end with this sequence of bytes.
501 * @param useHex
502 * <ul>
503 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
504 * Extended Hex Alphabet</a></li>
505 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
506 * Alphabet</a></li>
507 * </ul>
508 * @param padding padding byte.
509 * @throws IllegalArgumentException Thrown when the {@code lineSeparator} contains Base32 characters. Or the lineLength > 0 and lineSeparator is null.
510 * @deprecated Use {@link #builder()} and {@link Builder}.
511 */
512 @Deprecated
513 public Base32(final int lineLength, final byte[] lineSeparator, final boolean useHex, final byte padding) {
514 this(lineLength, lineSeparator, useHex, padding, DECODING_POLICY_DEFAULT);
515 }
516
517 /**
518 * Constructs a Base32 / Base32 Hex codec used for decoding and encoding.
519 * <p>
520 * When encoding the line length and line separator are given in the constructor.
521 * </p>
522 * <p>
523 * Line lengths that aren't multiples of 8 will still essentially end up being multiples of 8 in the encoded data.
524 * </p>
525 *
526 * @param lineLength Each line of encoded data will be at most of the given length (rounded down to the nearest multiple of 8). If lineLength <= 0,
527 * then the output will not be divided into lines (chunks). Ignored when decoding leniently.
528 * @param lineSeparator Each line of encoded data will end with this sequence of bytes.
529 * @param useHex
530 * <ul>
531 * <li>If true, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-7">RFC 4648 Section 7, Table 4: Base 32 Encoding with
532 * Extended Hex Alphabet</a></li>
533 * <li>If false, then use <a href="https://datatracker.ietf.org/doc/html/rfc4648#section-6">RFC 4648 Section 6, Table 3: The Base 32
534 * Alphabet</a></li>
535 * </ul>
536 * @param padding padding byte.
537 * @param decodingPolicy The decoding policy.
538 * @throws IllegalArgumentException Thrown when the {@code lineSeparator} contains Base32 characters. Or the lineLength > 0 and lineSeparator is null.
539 * @since 1.15
540 * @deprecated Use {@link #builder()} and {@link Builder}.
541 */
542 @Deprecated
543 public Base32(final int lineLength, final byte[] lineSeparator, final boolean useHex, final byte padding, final CodecPolicy decodingPolicy) {
544 // @formatter:off
545 this(builder()
546 .setLineLength(lineLength)
547 .setLineSeparator(lineSeparator != null ? lineSeparator : EMPTY_BYTE_ARRAY)
548 .setDecodeTable(decodeTable(useHex))
549 .setEncodeTableRaw(encodeTable(useHex))
550 .setPadding(padding)
551 .setDecodingPolicy(decodingPolicy));
552 // @formatter:on
553 }
554
555 /**
556 * <p>
557 * Decodes all of the provided data, starting at inPos, for inAvail bytes. Should be called at least twice: once with the data to decode, and once with
558 * inAvail set to "-1" to alert decoder that EOF has been reached. Strict decoding requires the "-1" call to validate the complete input.
559 * </p>
560 * <p>
561 * Lenient decoding ignores non-alphabet characters and stops at the first padding byte. Strict decoding accepts only the canonical form produced by this
562 * instance's encoder, including its alphabet, padding, and line separators.
563 * </p>
564 * <p>
565 * Output is written to {@link BaseNCodec.Context#buffer Context#buffer} as 8-bit octets, using
566 * {@link BaseNCodec.Context#pos Context#pos} as the buffer position
567 * </p>
568 *
569 * @param input byte[] array of ASCII data to Base32 decode.
570 * @param inPos Position to start reading data from.
571 * @param inAvail Amount of bytes available from input for decoding.
572 * @param context The context to be used.
573 * @throws IllegalArgumentException Thrown when a problem is detected processing data.
574 */
575 @Override
576 void decode(final byte[] input, int inPos, final int inAvail, final Context context) {
577 // package protected for access from I/O streams
578 if (context.eof) {
579 return;
580 }
581 if (inAvail < 0) {
582 context.eof = true;
583 if (isStrictDecoding()) {
584 validateCanonicalEnd(true, context);
585 }
586 }
587 final int decodeSize = this.encodeSize - 1;
588 for (int i = 0; i < inAvail; i++) {
589 final int b = input[inPos++] & 0xff;
590 if (isStrictDecoding()) {
591 if (!validateCanonicalByte(b, lineSeparator, true, context)) {
592 continue;
593 }
594 } else if (b == (pad & 0xff)) {
595 // We're done.
596 context.eof = true;
597 break;
598 }
599 final byte[] buffer = ensureBufferSize(decodeSize, context);
600 if (b < this.decodeTable.length) {
601 final int result = this.decodeTable[b];
602 if (result >= 0) {
603 context.modulus = (context.modulus + 1) % BYTES_PER_ENCODED_BLOCK;
604 // collect decoded bytes
605 context.lbitWorkArea = (context.lbitWorkArea << BITS_PER_ENCODED_BYTE) + result;
606 if (context.modulus == 0) { // we can output the 5 bytes
607 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 32 & MASK_8BITS);
608 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 24 & MASK_8BITS);
609 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 16 & MASK_8BITS);
610 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 8 & MASK_8BITS);
611 buffer[context.pos++] = (byte) (context.lbitWorkArea & MASK_8BITS);
612 }
613 }
614 }
615 }
616 // Strict decoding waits for physical EOF to validate the complete input.
617 // Lenient decoding also treats the first padding byte as EOF.
618 if (context.eof && context.modulus > 0) { // if modulus == 0, nothing to do
619 final byte[] buffer = ensureBufferSize(decodeSize, context);
620 // We ignore partial bytes, i.e. only multiples of 8 count.
621 // Any combination not part of a valid encoding is either partially decoded
622 // or will raise an exception. Possible trailing characters are 2, 4, 5, 7.
623 // It is not possible to encode with 1, 3, 6 trailing characters.
624 // For backwards compatibility 3 & 6 chars are decoded anyway rather than discarded.
625 // See the encode(byte[]) method EOF section.
626 switch (context.modulus) {
627 // case 0 : // impossible, as excluded above
628 case 1: // 5 bits - either ignore entirely, or raise an exception
629 validateTrailingCharacters();
630 // falls-through
631 case 2: // 10 bits, drop 2 and output one byte
632 validateCharacter(MASK_2_BITS, context);
633 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 2 & MASK_8BITS);
634 break;
635 case 3: // 15 bits, drop 7 and output 1 byte, or raise an exception
636 validateTrailingCharacters();
637 // Not possible from a valid encoding but decode anyway
638 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 7 & MASK_8BITS);
639 break;
640 case 4: // 20 bits = 2*8 + 4
641 validateCharacter(MASK_4_BITS, context);
642 context.lbitWorkArea = context.lbitWorkArea >> 4; // drop 4 bits
643 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 8 & MASK_8BITS);
644 buffer[context.pos++] = (byte) (context.lbitWorkArea & MASK_8BITS);
645 break;
646 case 5: // 25 bits = 3*8 + 1
647 validateCharacter(MASK_1_BITS, context);
648 context.lbitWorkArea = context.lbitWorkArea >> 1;
649 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 16 & MASK_8BITS);
650 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 8 & MASK_8BITS);
651 buffer[context.pos++] = (byte) (context.lbitWorkArea & MASK_8BITS);
652 break;
653 case 6: // 30 bits = 3*8 + 6, or raise an exception
654 validateTrailingCharacters();
655 // Not possible from a valid encoding but decode anyway
656 context.lbitWorkArea = context.lbitWorkArea >> 6;
657 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 16 & MASK_8BITS);
658 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 8 & MASK_8BITS);
659 buffer[context.pos++] = (byte) (context.lbitWorkArea & MASK_8BITS);
660 break;
661 case 7: // 35 bits = 4*8 +3
662 validateCharacter(MASK_3_BITS, context);
663 context.lbitWorkArea = context.lbitWorkArea >> 3;
664 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 24 & MASK_8BITS);
665 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 16 & MASK_8BITS);
666 buffer[context.pos++] = (byte) (context.lbitWorkArea >> 8 & MASK_8BITS);
667 buffer[context.pos++] = (byte) (context.lbitWorkArea & MASK_8BITS);
668 break;
669 default:
670 // modulus can be 0-7, and we excluded 0,1 already
671 throw new IllegalStateException("Impossible modulus " + context.modulus);
672 }
673 }
674 }
675
676 /**
677 * <p>
678 * Encodes all of the provided data, starting at inPos, for inAvail bytes. Must be called at least twice: once with the data to encode, and once with
679 * inAvail set to "-1" to alert encoder that EOF has been reached, so flush last remaining bytes (if not multiple of 5).
680 * </p>
681 *
682 * @param input byte[] array of binary data to Base32 encode.
683 * @param inPos Position to start reading data from.
684 * @param inAvail Amount of bytes available from input for encoding.
685 * @param context The context to be used.
686 * @throws IllegalArgumentException Thrown when a problem is detected processing data.
687 */
688 @Override
689 void encode(final byte[] input, int inPos, final int inAvail, final Context context) {
690 // package protected for access from I/O streams
691 if (context.eof) {
692 return;
693 }
694 // inAvail < 0 is how we're informed of EOF in the underlying data we're
695 // encoding.
696 if (inAvail < 0) {
697 context.eof = true;
698 if (0 == context.modulus && lineLength == 0) {
699 return; // no leftovers to process and not using chunking
700 }
701 final byte[] buffer = ensureBufferSize(encodeSize, context);
702 final int savedPos = context.pos;
703 switch (context.modulus) { // % 5
704 case 0:
705 break;
706 case 1: // Only 1 octet; take top 5 bits then remainder
707 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 3) & MASK_5_BITS]; // 8-1*5 = 3
708 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea << 2) & MASK_5_BITS]; // 5-3=2
709 buffer[context.pos++] = pad;
710 buffer[context.pos++] = pad;
711 buffer[context.pos++] = pad;
712 buffer[context.pos++] = pad;
713 buffer[context.pos++] = pad;
714 buffer[context.pos++] = pad;
715 break;
716 case 2: // 2 octets = 16 bits to use
717 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 11) & MASK_5_BITS]; // 16-1*5 = 11
718 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 6) & MASK_5_BITS]; // 16-2*5 = 6
719 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 1) & MASK_5_BITS]; // 16-3*5 = 1
720 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea << 4) & MASK_5_BITS]; // 5-1 = 4
721 buffer[context.pos++] = pad;
722 buffer[context.pos++] = pad;
723 buffer[context.pos++] = pad;
724 buffer[context.pos++] = pad;
725 break;
726 case 3: // 3 octets = 24 bits to use
727 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 19) & MASK_5_BITS]; // 24-1*5 = 19
728 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 14) & MASK_5_BITS]; // 24-2*5 = 14
729 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 9) & MASK_5_BITS]; // 24-3*5 = 9
730 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 4) & MASK_5_BITS]; // 24-4*5 = 4
731 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea << 1) & MASK_5_BITS]; // 5-4 = 1
732 buffer[context.pos++] = pad;
733 buffer[context.pos++] = pad;
734 buffer[context.pos++] = pad;
735 break;
736 case 4: // 4 octets = 32 bits to use
737 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 27) & MASK_5_BITS]; // 32-1*5 = 27
738 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 22) & MASK_5_BITS]; // 32-2*5 = 22
739 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 17) & MASK_5_BITS]; // 32-3*5 = 17
740 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 12) & MASK_5_BITS]; // 32-4*5 = 12
741 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 7) & MASK_5_BITS]; // 32-5*5 = 7
742 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 2) & MASK_5_BITS]; // 32-6*5 = 2
743 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea << 3) & MASK_5_BITS]; // 5-2 = 3
744 buffer[context.pos++] = pad;
745 break;
746 default:
747 throw new IllegalStateException("Impossible modulus " + context.modulus);
748 }
749 context.currentLinePos += context.pos - savedPos; // keep track of current line position
750 // if currentPos == 0 we are at the start of a line, so don't add CRLF
751 if (lineLength > 0 && context.currentLinePos > 0) { // add chunk separator if required
752 System.arraycopy(lineSeparator, 0, buffer, context.pos, lineSeparator.length);
753 context.pos += lineSeparator.length;
754 }
755 } else {
756 for (int i = 0; i < inAvail; i++) {
757 final byte[] buffer = ensureBufferSize(encodeSize, context);
758 context.modulus = (context.modulus + 1) % BYTES_PER_UNENCODED_BLOCK;
759 int b = input[inPos++];
760 if (b < 0) {
761 b += 256;
762 }
763 context.lbitWorkArea = (context.lbitWorkArea << 8) + b; // BITS_PER_BYTE
764 if (0 == context.modulus) { // we have enough bytes to create our output
765 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 35) & MASK_5_BITS];
766 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 30) & MASK_5_BITS];
767 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 25) & MASK_5_BITS];
768 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 20) & MASK_5_BITS];
769 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 15) & MASK_5_BITS];
770 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 10) & MASK_5_BITS];
771 buffer[context.pos++] = encodeTable[(int) (context.lbitWorkArea >> 5) & MASK_5_BITS];
772 buffer[context.pos++] = encodeTable[(int) context.lbitWorkArea & MASK_5_BITS];
773 context.currentLinePos += BYTES_PER_ENCODED_BLOCK;
774 if (lineLength > 0 && lineLength <= context.currentLinePos) {
775 System.arraycopy(lineSeparator, 0, buffer, context.pos, lineSeparator.length);
776 context.pos += lineSeparator.length;
777 context.currentLinePos = 0;
778 }
779 }
780 }
781 }
782 }
783
784 /**
785 * Gets the line separator (for testing only).
786 *
787 * @return The line separator.
788 */
789 byte[] getLineSeparator() {
790 return lineSeparator;
791 }
792
793 /**
794 * Tests whether the {@code octet} is in the Base32 alphabet.
795 *
796 * @param octet The value to test.
797 * @return {@code true} if the value is defined in the Base32 alphabet {@code false} otherwise.
798 */
799 @Override
800 public boolean isInAlphabet(final byte octet) {
801 final int value = octet & 0xff;
802 return value < decodeTable.length && decodeTable[value] != -1;
803 }
804
805 /**
806 * Validates whether decoding the final trailing character is possible in the context of the set of possible Base32 values.
807 * <p>
808 * The character is valid if the lower bits within the provided mask are zero. This is used to test the final trailing base-32 digit is zero in the bits
809 * that will be discarded.
810 * </p>
811 *
812 * @param emptyBitsMask The mask of the lower bits that should be empty.
813 * @param context The context to be used.
814 * @throws IllegalArgumentException Thrown if the bits being checked contain any non-zero value.
815 */
816 private void validateCharacter(final long emptyBitsMask, final Context context) {
817 // Use the long bit work area
818 if (isStrictDecoding() && (context.lbitWorkArea & emptyBitsMask) != 0) {
819 throw new IllegalArgumentException("Strict decoding: Last encoded character (before the paddings if any) is a valid " +
820 "Base32 alphabet but not a possible encoding. Expected the discarded bits from the character to be zero.");
821 }
822 }
823
824 /**
825 * Validates whether decoding allows final trailing characters that cannot be created during encoding.
826 *
827 * @throws IllegalArgumentException Thrown if strict decoding is enabled.
828 */
829 private void validateTrailingCharacters() {
830 if (isStrictDecoding()) {
831 throw new IllegalArgumentException("Strict decoding: Last encoded character(s) (before the paddings if any) are valid " +
832 "Base32 alphabet but not a possible encoding. Decoding requires either 2, 4, 5, or 7 trailing 5-bit characters to create bytes.");
833 }
834 }
835 }