1 /*
2 * Copyright (c) 1996, 2017, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 package java.util.zip;
27
28 /**
29 * This class provides support for general purpose compression using the
30 * popular ZLIB compression library. The ZLIB compression library was
31 * initially developed as part of the PNG graphics standard and is not
32 * protected by patents. It is fully described in the specifications at
33 * the <a href="package-summary.html#package.description">java.util.zip
34 * package description</a>.
35 *
36 * <p>The following code fragment demonstrates a trivial compression
37 * and decompression of a string using {@code Deflater} and
38 * {@code Inflater}.
39 *
40 * <blockquote><pre>
41 * try {
42 * // Encode a String into bytes
43 * String inputString = "blahblahblah";
44 * byte[] input = inputString.getBytes("UTF-8");
45 *
46 * // Compress the bytes
47 * byte[] output = new byte[100];
48 * Deflater compresser = new Deflater();
49 * compresser.setInput(input);
50 * compresser.finish();
51 * int compressedDataLength = compresser.deflate(output);
52 * compresser.end();
53 *
54 * // Decompress the bytes
55 * Inflater decompresser = new Inflater();
56 * decompresser.setInput(output, 0, compressedDataLength);
57 * byte[] result = new byte[100];
58 * int resultLength = decompresser.inflate(result);
59 * decompresser.end();
60 *
61 * // Decode the bytes into a String
62 * String outputString = new String(result, 0, resultLength, "UTF-8");
63 * } catch(java.io.UnsupportedEncodingException ex) {
64 * // handle
65 * } catch (java.util.zip.DataFormatException ex) {
66 * // handle
67 * }
68 * </pre></blockquote>
69 *
70 * @see Inflater
71 * @author David Connelly
72 * @since 1.1
73 */
74 public
75 class Deflater {
76
77 private final ZStreamRef zsRef;
78 private byte[] buf = new byte[0];
79 private int off, len;
80 private int level, strategy;
81 private boolean setParams;
82 private boolean finish, finished;
83 private long bytesRead;
84 private long bytesWritten;
85
86 /**
87 * Compression method for the deflate algorithm (the only one currently
88 * supported).
89 */
90 public static final int DEFLATED = 8;
91
92 /**
93 * Compression level for no compression.
94 */
95 public static final int NO_COMPRESSION = 0;
96
97 /**
98 * Compression level for fastest compression.
99 */
100 public static final int BEST_SPEED = 1;
101
102 /**
103 * Compression level for best compression.
104 */
105 public static final int BEST_COMPRESSION = 9;
106
107 /**
108 * Default compression level.
109 */
110 public static final int DEFAULT_COMPRESSION = -1;
111
112 /**
113 * Compression strategy best used for data consisting mostly of small
114 * values with a somewhat random distribution. Forces more Huffman coding
115 * and less string matching.
116 */
117 public static final int FILTERED = 1;
118
119 /**
120 * Compression strategy for Huffman coding only.
121 */
122 public static final int HUFFMAN_ONLY = 2;
123
124 /**
125 * Default compression strategy.
126 */
127 public static final int DEFAULT_STRATEGY = 0;
128
129 /**
130 * Compression flush mode used to achieve best compression result.
131 *
132 * @see Deflater#deflate(byte[], int, int, int)
133 * @since 1.7
134 */
135 public static final int NO_FLUSH = 0;
136
137 /**
138 * Compression flush mode used to flush out all pending output; may
139 * degrade compression for some compression algorithms.
140 *
141 * @see Deflater#deflate(byte[], int, int, int)
142 * @since 1.7
143 */
144 public static final int SYNC_FLUSH = 2;
145
146 /**
147 * Compression flush mode used to flush out all pending output and
148 * reset the deflater. Using this mode too often can seriously degrade
149 * compression.
150 *
151 * @see Deflater#deflate(byte[], int, int, int)
152 * @since 1.7
153 */
154 public static final int FULL_FLUSH = 3;
155
156 static {
157 ZipUtils.loadLibrary();
158 initIDs();
159 }
160
161 /**
162 * Creates a new compressor using the specified compression level.
163 * If 'nowrap' is true then the ZLIB header and checksum fields will
164 * not be used in order to support the compression format used in
165 * both GZIP and PKZIP.
166 * @param level the compression level (0-9)
167 * @param nowrap if true then use GZIP compatible compression
168 */
169 public Deflater(int level, boolean nowrap) {
170 this.level = level;
171 this.strategy = DEFAULT_STRATEGY;
172 this.zsRef = new ZStreamRef(this,
173 () -> init(level, DEFAULT_STRATEGY, nowrap),
174 Deflater::end);
175 }
176
177 /**
178 * Creates a new compressor using the specified compression level.
179 * Compressed data will be generated in ZLIB format.
180 * @param level the compression level (0-9)
181 */
182 public Deflater(int level) {
183 this(level, false);
184 }
185
186 /**
187 * Creates a new compressor with the default compression level.
188 * Compressed data will be generated in ZLIB format.
189 */
190 public Deflater() {
191 this(DEFAULT_COMPRESSION, false);
192 }
193
194 /**
195 * Sets input data for compression. This should be called whenever
196 * needsInput() returns true indicating that more input data is required.
197 * @param b the input data bytes
198 * @param off the start offset of the data
199 * @param len the length of the data
200 * @see Deflater#needsInput
201 */
202 public void setInput(byte[] b, int off, int len) {
203 if (b== null) {
204 throw new NullPointerException();
205 }
206 if (off < 0 || len < 0 || off > b.length - len) {
207 throw new ArrayIndexOutOfBoundsException();
208 }
209 synchronized (zsRef) {
210 this.buf = b;
211 this.off = off;
212 this.len = len;
213 }
214 }
215
216 /**
217 * Sets input data for compression. This should be called whenever
218 * needsInput() returns true indicating that more input data is required.
219 * @param b the input data bytes
220 * @see Deflater#needsInput
221 */
222 public void setInput(byte[] b) {
223 setInput(b, 0, b.length);
224 }
225
226 /**
227 * Sets preset dictionary for compression. A preset dictionary is used
228 * when the history buffer can be predetermined. When the data is later
229 * uncompressed with Inflater.inflate(), Inflater.getAdler() can be called
230 * in order to get the Adler-32 value of the dictionary required for
231 * decompression.
232 * @param b the dictionary data bytes
233 * @param off the start offset of the data
234 * @param len the length of the data
235 * @see Inflater#inflate
236 * @see Inflater#getAdler
237 */
238 public void setDictionary(byte[] b, int off, int len) {
239 if (b == null) {
240 throw new NullPointerException();
241 }
242 if (off < 0 || len < 0 || off > b.length - len) {
243 throw new ArrayIndexOutOfBoundsException();
244 }
245 synchronized (zsRef) {
246 ensureOpen();
247 setDictionary(zsRef.address(), b, off, len);
248 }
249 }
250
251 /**
252 * Sets preset dictionary for compression. A preset dictionary is used
253 * when the history buffer can be predetermined. When the data is later
254 * uncompressed with Inflater.inflate(), Inflater.getAdler() can be called
255 * in order to get the Adler-32 value of the dictionary required for
256 * decompression.
257 * @param b the dictionary data bytes
258 * @see Inflater#inflate
259 * @see Inflater#getAdler
260 */
261 public void setDictionary(byte[] b) {
262 setDictionary(b, 0, b.length);
263 }
264
265 /**
266 * Sets the compression strategy to the specified value.
267 *
268 * <p> If the compression strategy is changed, the next invocation
269 * of {@code deflate} will compress the input available so far with
270 * the old strategy (and may be flushed); the new strategy will take
271 * effect only after that invocation.
272 *
273 * @param strategy the new compression strategy
274 * @exception IllegalArgumentException if the compression strategy is
275 * invalid
276 */
277 public void setStrategy(int strategy) {
278 switch (strategy) {
279 case DEFAULT_STRATEGY:
280 case FILTERED:
281 case HUFFMAN_ONLY:
282 break;
283 default:
284 throw new IllegalArgumentException();
285 }
286 synchronized (zsRef) {
287 if (this.strategy != strategy) {
288 this.strategy = strategy;
289 setParams = true;
290 }
291 }
292 }
293
294 /**
295 * Sets the compression level to the specified value.
296 *
297 * <p> If the compression level is changed, the next invocation
298 * of {@code deflate} will compress the input available so far
299 * with the old level (and may be flushed); the new level will
300 * take effect only after that invocation.
301 *
302 * @param level the new compression level (0-9)
303 * @exception IllegalArgumentException if the compression level is invalid
304 */
305 public void setLevel(int level) {
306 if ((level < 0 || level > 9) && level != DEFAULT_COMPRESSION) {
307 throw new IllegalArgumentException("invalid compression level");
308 }
309 synchronized (zsRef) {
310 if (this.level != level) {
311 this.level = level;
312 setParams = true;
313 }
314 }
315 }
316
317 /**
318 * Returns true if the input data buffer is empty and setInput()
319 * should be called in order to provide more input.
320 * @return true if the input data buffer is empty and setInput()
321 * should be called in order to provide more input
322 */
323 public boolean needsInput() {
324 synchronized (zsRef) {
325 return len <= 0;
326 }
327 }
328
329 /**
330 * When called, indicates that compression should end with the current
331 * contents of the input buffer.
332 */
333 public void finish() {
334 synchronized (zsRef) {
335 finish = true;
336 }
337 }
338
339 /**
340 * Returns true if the end of the compressed data output stream has
341 * been reached.
342 * @return true if the end of the compressed data output stream has
343 * been reached
344 */
345 public boolean finished() {
346 synchronized (zsRef) {
347 return finished;
348 }
349 }
350
351 /**
352 * Compresses the input data and fills specified buffer with compressed
353 * data. Returns actual number of bytes of compressed data. A return value
354 * of 0 indicates that {@link #needsInput() needsInput} should be called
355 * in order to determine if more input data is required.
356 *
357 * <p>This method uses {@link #NO_FLUSH} as its compression flush mode.
358 * An invocation of this method of the form {@code deflater.deflate(b, off, len)}
359 * yields the same result as the invocation of
360 * {@code deflater.deflate(b, off, len, Deflater.NO_FLUSH)}.
361 *
362 * @param b the buffer for the compressed data
363 * @param off the start offset of the data
364 * @param len the maximum number of bytes of compressed data
365 * @return the actual number of bytes of compressed data written to the
366 * output buffer
367 */
368 public int deflate(byte[] b, int off, int len) {
369 return deflate(b, off, len, NO_FLUSH);
370 }
371
372 /**
373 * Compresses the input data and fills specified buffer with compressed
374 * data. Returns actual number of bytes of compressed data. A return value
375 * of 0 indicates that {@link #needsInput() needsInput} should be called
376 * in order to determine if more input data is required.
377 *
378 * <p>This method uses {@link #NO_FLUSH} as its compression flush mode.
379 * An invocation of this method of the form {@code deflater.deflate(b)}
380 * yields the same result as the invocation of
381 * {@code deflater.deflate(b, 0, b.length, Deflater.NO_FLUSH)}.
382 *
383 * @param b the buffer for the compressed data
384 * @return the actual number of bytes of compressed data written to the
385 * output buffer
386 */
387 public int deflate(byte[] b) {
388 return deflate(b, 0, b.length, NO_FLUSH);
389 }
390
391 /**
392 * Compresses the input data and fills the specified buffer with compressed
393 * data. Returns actual number of bytes of data compressed.
394 *
395 * <p>Compression flush mode is one of the following three modes:
396 *
397 * <ul>
398 * <li>{@link #NO_FLUSH}: allows the deflater to decide how much data
399 * to accumulate, before producing output, in order to achieve the best
400 * compression (should be used in normal use scenario). A return value
401 * of 0 in this flush mode indicates that {@link #needsInput()} should
402 * be called in order to determine if more input data is required.
403 *
404 * <li>{@link #SYNC_FLUSH}: all pending output in the deflater is flushed,
405 * to the specified output buffer, so that an inflater that works on
406 * compressed data can get all input data available so far (In particular
407 * the {@link #needsInput()} returns {@code true} after this invocation
408 * if enough output space is provided). Flushing with {@link #SYNC_FLUSH}
409 * may degrade compression for some compression algorithms and so it
410 * should be used only when necessary.
411 *
412 * <li>{@link #FULL_FLUSH}: all pending output is flushed out as with
413 * {@link #SYNC_FLUSH}. The compression state is reset so that the inflater
414 * that works on the compressed output data can restart from this point
415 * if previous compressed data has been damaged or if random access is
416 * desired. Using {@link #FULL_FLUSH} too often can seriously degrade
417 * compression.
418 * </ul>
419 *
420 * <p>In the case of {@link #FULL_FLUSH} or {@link #SYNC_FLUSH}, if
421 * the return value is {@code len}, the space available in output
422 * buffer {@code b}, this method should be invoked again with the same
423 * {@code flush} parameter and more output space. Make sure that
424 * {@code len} is greater than 6 to avoid flush marker (5 bytes) being
425 * repeatedly output to the output buffer every time this method is
426 * invoked.
427 *
428 * @param b the buffer for the compressed data
429 * @param off the start offset of the data
430 * @param len the maximum number of bytes of compressed data
431 * @param flush the compression flush mode
432 * @return the actual number of bytes of compressed data written to
433 * the output buffer
434 *
435 * @throws IllegalArgumentException if the flush mode is invalid
436 * @since 1.7
437 */
438 public int deflate(byte[] b, int off, int len, int flush) {
439 if (b == null) {
440 throw new NullPointerException();
441 }
442 if (off < 0 || len < 0 || off > b.length - len) {
443 throw new ArrayIndexOutOfBoundsException();
444 }
445 synchronized (zsRef) {
446 ensureOpen();
447 if (flush == NO_FLUSH || flush == SYNC_FLUSH ||
448 flush == FULL_FLUSH) {
449 int thisLen = this.len;
450 int n = deflateBytes(zsRef.address(), b, off, len, flush);
451 bytesWritten += n;
452 bytesRead += (thisLen - this.len);
453 return n;
454 }
455 throw new IllegalArgumentException();
456 }
457 }
458
459 /**
460 * Returns the ADLER-32 value of the uncompressed data.
461 * @return the ADLER-32 value of the uncompressed data
462 */
463 public int getAdler() {
464 synchronized (zsRef) {
465 ensureOpen();
466 return getAdler(zsRef.address());
467 }
468 }
469
470 /**
471 * Returns the total number of uncompressed bytes input so far.
472 *
473 * <p>Since the number of bytes may be greater than
474 * Integer.MAX_VALUE, the {@link #getBytesRead()} method is now
475 * the preferred means of obtaining this information.</p>
476 *
477 * @return the total number of uncompressed bytes input so far
478 */
479 public int getTotalIn() {
480 return (int) getBytesRead();
481 }
482
483 /**
484 * Returns the total number of uncompressed bytes input so far.
485 *
486 * @return the total (non-negative) number of uncompressed bytes input so far
487 * @since 1.5
488 */
489 public long getBytesRead() {
490 synchronized (zsRef) {
491 ensureOpen();
492 return bytesRead;
493 }
494 }
495
496 /**
497 * Returns the total number of compressed bytes output so far.
498 *
499 * <p>Since the number of bytes may be greater than
500 * Integer.MAX_VALUE, the {@link #getBytesWritten()} method is now
501 * the preferred means of obtaining this information.</p>
502 *
503 * @return the total number of compressed bytes output so far
504 */
505 public int getTotalOut() {
506 return (int) getBytesWritten();
507 }
508
509 /**
510 * Returns the total number of compressed bytes output so far.
511 *
512 * @return the total (non-negative) number of compressed bytes output so far
513 * @since 1.5
514 */
515 public long getBytesWritten() {
516 synchronized (zsRef) {
517 ensureOpen();
518 return bytesWritten;
519 }
520 }
521
522 /**
523 * Resets deflater so that a new set of input data can be processed.
524 * Keeps current compression level and strategy settings.
525 */
526 public void reset() {
527 synchronized (zsRef) {
528 ensureOpen();
529 reset(zsRef.address());
530 finish = false;
531 finished = false;
532 off = len = 0;
533 bytesRead = bytesWritten = 0;
534 }
535 }
536
537 /**
538 * Closes the compressor and discards any unprocessed input.
539 *
540 * This method should be called when the compressor is no longer
541 * being used. Once this method is called, the behavior of the
542 * Deflater object is undefined.
543 */
544 public void end() {
545 synchronized (zsRef) {
546 zsRef.clean();
547 buf = null;
548 }
549 }
550
551 /**
552 * Closes the compressor when garbage is collected.
553 *
554 * @deprecated The {@code finalize} method has been deprecated and
555 * implemented as a no-op. Subclasses that override {@code finalize}
556 * in order to perform cleanup should be modified to use alternative
557 * cleanup mechanisms and to remove the overriding {@code finalize}
558 * method. The recommended cleanup for compressor is to explicitly
559 * call {@code end} method when it is no longer in use. If the
560 * {@code end} is not invoked explicitly the resource of the compressor
561 * will be released when the instance becomes phantom-reachable.
562 */
563 @Deprecated(since="9", forRemoval=true)
564 protected void finalize() {}
565
566 private void ensureOpen() {
567 assert Thread.holdsLock(zsRef);
568 if (zsRef.address() == 0)
569 throw new NullPointerException("Deflater has been closed");
570 }
571
572 private static native void initIDs();
573 private static native long init(int level, int strategy, boolean nowrap);
574 private static native void setDictionary(long addr, byte[] b, int off, int len);
575 private native int deflateBytes(long addr, byte[] b, int off, int len,
576 int flush);
577 private static native int getAdler(long addr);
578 private static native void reset(long addr);
579 private static native void end(long addr);
580 }