1 /*
2 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
3 *
4 * This code is free software; you can redistribute it and/or modify it
5 * under the terms of the GNU General Public License version 2 only, as
6 * published by the Free Software Foundation. Oracle designates this
7 * particular file as subject to the "Classpath" exception as provided
8 * by Oracle in the LICENSE file that accompanied this code.
9 *
10 * This code is distributed in the hope that it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
13 * version 2 for more details (a copy is included in the LICENSE file that
14 * accompanied this code).
15 *
16 * You should have received a copy of the GNU General Public License version
17 * 2 along with this work; if not, write to the Free Software Foundation,
18 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
19 *
20 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
21 * or visit www.oracle.com if you need additional information or have any
22 * questions.
23 */
24
25 /* deflate.c -- compress data using the deflation algorithm
26 * Copyright (C) 1995-2017 Jean-loup Gailly and Mark Adler
27 * For conditions of distribution and use, see copyright notice in zlib.h
28 */
29
30 /*
31 * ALGORITHM
32 *
33 * The "deflation" process depends on being able to identify portions
34 * of the input text which are identical to earlier input (within a
35 * sliding window trailing behind the input currently being processed).
36 *
37 * The most straightforward technique turns out to be the fastest for
38 * most input files: try all possible matches and select the longest.
39 * The key feature of this algorithm is that insertions into the string
40 * dictionary are very simple and thus fast, and deletions are avoided
41 * completely. Insertions are performed at each input character, whereas
42 * string matches are performed only when the previous match ends. So it
43 * is preferable to spend more time in matches to allow very fast string
44 * insertions and avoid deletions. The matching algorithm for small
45 * strings is inspired from that of Rabin & Karp. A brute force approach
46 * is used to find longer strings when a small match has been found.
47 * A similar algorithm is used in comic (by Jan-Mark Wams) and freeze
48 * (by Leonid Broukhis).
49 * A previous version of this file used a more sophisticated algorithm
50 * (by Fiala and Greene) which is guaranteed to run in linear amortized
51 * time, but has a larger average cost, uses more memory and is patented.
52 * However the F&G algorithm may be faster for some highly redundant
53 * files if the parameter max_chain_length (described below) is too large.
54 *
55 * ACKNOWLEDGEMENTS
56 *
57 * The idea of lazy evaluation of matches is due to Jan-Mark Wams, and
58 * I found it in 'freeze' written by Leonid Broukhis.
59 * Thanks to many people for bug reports and testing.
60 *
61 * REFERENCES
62 *
63 * Deutsch, L.P.,"DEFLATE Compressed Data Format Specification".
64 * Available in http://tools.ietf.org/html/rfc1951
65 *
66 * A description of the Rabin and Karp algorithm is given in the book
67 * "Algorithms" by R. Sedgewick, Addison-Wesley, p252.
68 *
69 * Fiala,E.R., and Greene,D.H.
70 * Data Compression with Finite Windows, Comm.ACM, 32,4 (1989) 490-595
71 *
72 */
73
74 /* @(#) $Id$ */
75
76 #include "deflate.h"
77
78 const char deflate_copyright[] =
79 " deflate 1.2.11 Copyright 1995-2017 Jean-loup Gailly and Mark Adler ";
80 /*
81 If you use the zlib library in a product, an acknowledgment is welcome
82 in the documentation of your product. If for some reason you cannot
83 include such an acknowledgment, I would appreciate that you keep this
84 copyright string in the executable of your product.
85 */
86
87 /* ===========================================================================
88 * Function prototypes.
89 */
90 typedef enum {
91 need_more, /* block not completed, need more input or more output */
92 block_done, /* block flush performed */
93 finish_started, /* finish started, need only more output at next deflate */
94 finish_done /* finish done, accept no more input or output */
95 } block_state;
96
97 typedef block_state (*compress_func) OF((deflate_state *s, int flush));
98 /* Compression function. Returns the block state after the call. */
99
100 local int deflateStateCheck OF((z_streamp strm));
101 local void slide_hash OF((deflate_state *s));
102 local void fill_window OF((deflate_state *s));
103 local block_state deflate_stored OF((deflate_state *s, int flush));
104 local block_state deflate_fast OF((deflate_state *s, int flush));
105 #ifndef FASTEST
106 local block_state deflate_slow OF((deflate_state *s, int flush));
107 #endif
108 local block_state deflate_rle OF((deflate_state *s, int flush));
109 local block_state deflate_huff OF((deflate_state *s, int flush));
110 local void lm_init OF((deflate_state *s));
111 local void putShortMSB OF((deflate_state *s, uInt b));
112 local void flush_pending OF((z_streamp strm));
113 local unsigned read_buf OF((z_streamp strm, Bytef *buf, unsigned size));
114 #ifdef ASMV
115 # pragma message("Assembler code may have bugs -- use at your own risk")
116 void match_init OF((void)); /* asm code initialization */
117 uInt longest_match OF((deflate_state *s, IPos cur_match));
118 #else
119 local uInt longest_match OF((deflate_state *s, IPos cur_match));
120 #endif
121
122 #ifdef ZLIB_DEBUG
123 local void check_match OF((deflate_state *s, IPos start, IPos match,
124 int length));
125 #endif
126
127 /* ===========================================================================
128 * Local data
129 */
130
131 #define NIL 0
132 /* Tail of hash chains */
133
134 #ifndef TOO_FAR
135 # define TOO_FAR 4096
136 #endif
137 /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */
138
139 /* Values for max_lazy_match, good_match and max_chain_length, depending on
140 * the desired pack level (0..9). The values given below have been tuned to
141 * exclude worst case performance for pathological files. Better values may be
142 * found for specific files.
143 */
144 typedef struct config_s {
145 ush good_length; /* reduce lazy search above this match length */
146 ush max_lazy; /* do not perform lazy search above this match length */
147 ush nice_length; /* quit search above this match length */
148 ush max_chain;
149 compress_func func;
150 } config;
151
152 #ifdef FASTEST
153 local const config configuration_table[2] = {
154 /* good lazy nice chain */
155 /* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */
156 /* 1 */ {4, 4, 8, 4, deflate_fast}}; /* max speed, no lazy matches */
157 #else
158 local const config configuration_table[10] = {
159 /* good lazy nice chain */
160 /* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */
161 /* 1 */ {4, 4, 8, 4, deflate_fast}, /* max speed, no lazy matches */
162 /* 2 */ {4, 5, 16, 8, deflate_fast},
163 /* 3 */ {4, 6, 32, 32, deflate_fast},
164
165 /* 4 */ {4, 4, 16, 16, deflate_slow}, /* lazy matches */
166 /* 5 */ {8, 16, 32, 32, deflate_slow},
167 /* 6 */ {8, 16, 128, 128, deflate_slow},
168 /* 7 */ {8, 32, 128, 256, deflate_slow},
169 /* 8 */ {32, 128, 258, 1024, deflate_slow},
170 /* 9 */ {32, 258, 258, 4096, deflate_slow}}; /* max compression */
171 #endif
172
173 /* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4
174 * For deflate_fast() (levels <= 3) good is ignored and lazy has a different
175 * meaning.
176 */
177
178 /* rank Z_BLOCK between Z_NO_FLUSH and Z_PARTIAL_FLUSH */
179 #define RANK(f) (((f) * 2) - ((f) > 4 ? 9 : 0))
180
181 /* ===========================================================================
182 * Update a hash value with the given input byte
183 * IN assertion: all calls to UPDATE_HASH are made with consecutive input
184 * characters, so that a running hash key can be computed from the previous
185 * key instead of complete recalculation each time.
186 */
187 #define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask)
188
189
190 /* ===========================================================================
191 * Insert string str in the dictionary and set match_head to the previous head
192 * of the hash chain (the most recent string with same hash key). Return
193 * the previous length of the hash chain.
194 * If this file is compiled with -DFASTEST, the compression level is forced
195 * to 1, and no hash chains are maintained.
196 * IN assertion: all calls to INSERT_STRING are made with consecutive input
197 * characters and the first MIN_MATCH bytes of str are valid (except for
198 * the last MIN_MATCH-1 bytes of the input file).
199 */
200 #ifdef FASTEST
201 #define INSERT_STRING(s, str, match_head) \
202 (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \
203 match_head = s->head[s->ins_h], \
204 s->head[s->ins_h] = (Pos)(str))
205 #else
206 #define INSERT_STRING(s, str, match_head) \
207 (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \
208 match_head = s->prev[(str) & s->w_mask] = s->head[s->ins_h], \
209 s->head[s->ins_h] = (Pos)(str))
210 #endif
211
212 /* ===========================================================================
213 * Initialize the hash table (avoiding 64K overflow for 16 bit systems).
214 * prev[] will be initialized on the fly.
215 */
216 #define CLEAR_HASH(s) \
217 s->head[s->hash_size-1] = NIL; \
218 zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head));
219
220 /* ===========================================================================
221 * Slide the hash table when sliding the window down (could be avoided with 32
222 * bit values at the expense of memory usage). We slide even when level == 0 to
223 * keep the hash table consistent if we switch back to level > 0 later.
224 */
225 local void slide_hash(s)
226 deflate_state *s;
227 {
228 unsigned n, m;
229 Posf *p;
230 uInt wsize = s->w_size;
231
232 n = s->hash_size;
233 p = &s->head[n];
234 do {
235 m = *--p;
236 *p = (Pos)(m >= wsize ? m - wsize : NIL);
237 } while (--n);
238 n = wsize;
239 #ifndef FASTEST
240 p = &s->prev[n];
241 do {
242 m = *--p;
243 *p = (Pos)(m >= wsize ? m - wsize : NIL);
244 /* If n is not on any hash chain, prev[n] is garbage but
245 * its value will never be used.
246 */
247 } while (--n);
248 #endif
249 }
250
251 /* ========================================================================= */
252 int ZEXPORT deflateInit_(strm, level, version, stream_size)
253 z_streamp strm;
254 int level;
255 const char *version;
256 int stream_size;
257 {
258 return deflateInit2_(strm, level, Z_DEFLATED, MAX_WBITS, DEF_MEM_LEVEL,
259 Z_DEFAULT_STRATEGY, version, stream_size);
260 /* To do: ignore strm->next_in if we use it as window */
261 }
262
263 /* ========================================================================= */
264 int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
265 version, stream_size)
266 z_streamp strm;
267 int level;
268 int method;
269 int windowBits;
270 int memLevel;
271 int strategy;
272 const char *version;
273 int stream_size;
274 {
275 deflate_state *s;
276 int wrap = 1;
277 static const char my_version[] = ZLIB_VERSION;
278
279 ushf *overlay;
280 /* We overlay pending_buf and d_buf+l_buf. This works since the average
281 * output size for (length,distance) codes is <= 24 bits.
282 */
283
284 if (version == Z_NULL || version[0] != my_version[0] ||
285 stream_size != sizeof(z_stream)) {
286 return Z_VERSION_ERROR;
287 }
288 if (strm == Z_NULL) return Z_STREAM_ERROR;
289
290 strm->msg = Z_NULL;
291 if (strm->zalloc == (alloc_func)0) {
292 #ifdef Z_SOLO
293 return Z_STREAM_ERROR;
294 #else
295 strm->zalloc = zcalloc;
296 strm->opaque = (voidpf)0;
297 #endif
298 }
299 if (strm->zfree == (free_func)0)
300 #ifdef Z_SOLO
301 return Z_STREAM_ERROR;
302 #else
303 strm->zfree = zcfree;
304 #endif
305
306 #ifdef FASTEST
307 if (level != 0) level = 1;
308 #else
309 if (level == Z_DEFAULT_COMPRESSION) level = 6;
310 #endif
311
312 if (windowBits < 0) { /* suppress zlib wrapper */
313 wrap = 0;
314 windowBits = -windowBits;
315 }
316 #ifdef GZIP
317 else if (windowBits > 15) {
318 wrap = 2; /* write gzip wrapper instead */
319 windowBits -= 16;
320 }
321 #endif
322 if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED ||
323 windowBits < 8 || windowBits > 15 || level < 0 || level > 9 ||
324 strategy < 0 || strategy > Z_FIXED || (windowBits == 8 && wrap != 1)) {
325 return Z_STREAM_ERROR;
326 }
327 if (windowBits == 8) windowBits = 9; /* until 256-byte window bug fixed */
328 s = (deflate_state *) ZALLOC(strm, 1, sizeof(deflate_state));
329 if (s == Z_NULL) return Z_MEM_ERROR;
330 strm->state = (struct internal_state FAR *)s;
331 s->strm = strm;
332 s->status = INIT_STATE; /* to pass state test in deflateReset() */
333
334 s->wrap = wrap;
335 s->gzhead = Z_NULL;
336 s->w_bits = (uInt)windowBits;
337 s->w_size = 1 << s->w_bits;
338 s->w_mask = s->w_size - 1;
339
340 s->hash_bits = (uInt)memLevel + 7;
341 s->hash_size = 1 << s->hash_bits;
342 s->hash_mask = s->hash_size - 1;
343 s->hash_shift = ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH);
344
345 s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte));
346 s->prev = (Posf *) ZALLOC(strm, s->w_size, sizeof(Pos));
347 s->head = (Posf *) ZALLOC(strm, s->hash_size, sizeof(Pos));
348
349 s->high_water = 0; /* nothing written to s->window yet */
350
351 s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */
352
353 overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2);
354 s->pending_buf = (uchf *) overlay;
355 s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L);
356
357 if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL ||
358 s->pending_buf == Z_NULL) {
359 s->status = FINISH_STATE;
360 strm->msg = ERR_MSG(Z_MEM_ERROR);
361 deflateEnd (strm);
362 return Z_MEM_ERROR;
363 }
364 s->d_buf = overlay + s->lit_bufsize/sizeof(ush);
365 s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize;
366
367 s->level = level;
368 s->strategy = strategy;
369 s->method = (Byte)method;
370
371 return deflateReset(strm);
372 }
373
374 /* =========================================================================
375 * Check for a valid deflate stream state. Return 0 if ok, 1 if not.
376 */
377 local int deflateStateCheck (strm)
378 z_streamp strm;
379 {
380 deflate_state *s;
381 if (strm == Z_NULL ||
382 strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0)
383 return 1;
384 s = strm->state;
385 if (s == Z_NULL || s->strm != strm || (s->status != INIT_STATE &&
386 #ifdef GZIP
387 s->status != GZIP_STATE &&
388 #endif
389 s->status != EXTRA_STATE &&
390 s->status != NAME_STATE &&
391 s->status != COMMENT_STATE &&
392 s->status != HCRC_STATE &&
393 s->status != BUSY_STATE &&
394 s->status != FINISH_STATE))
395 return 1;
396 return 0;
397 }
398
399 /* ========================================================================= */
400 int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength)
401 z_streamp strm;
402 const Bytef *dictionary;
403 uInt dictLength;
404 {
405 deflate_state *s;
406 uInt str, n;
407 int wrap;
408 unsigned avail;
409 z_const unsigned char *next;
410
411 if (deflateStateCheck(strm) || dictionary == Z_NULL)
412 return Z_STREAM_ERROR;
413 s = strm->state;
414 wrap = s->wrap;
415 if (wrap == 2 || (wrap == 1 && s->status != INIT_STATE) || s->lookahead)
416 return Z_STREAM_ERROR;
417
418 /* when using zlib wrappers, compute Adler-32 for provided dictionary */
419 if (wrap == 1)
420 strm->adler = adler32(strm->adler, dictionary, dictLength);
421 s->wrap = 0; /* avoid computing Adler-32 in read_buf */
422
423 /* if dictionary would fill window, just replace the history */
424 if (dictLength >= s->w_size) {
425 if (wrap == 0) { /* already empty otherwise */
426 CLEAR_HASH(s);
427 s->strstart = 0;
428 s->block_start = 0L;
429 s->insert = 0;
430 }
431 dictionary += dictLength - s->w_size; /* use the tail */
432 dictLength = s->w_size;
433 }
434
435 /* insert dictionary into window and hash */
436 avail = strm->avail_in;
437 next = strm->next_in;
438 strm->avail_in = dictLength;
439 strm->next_in = (z_const Bytef *)dictionary;
440 fill_window(s);
441 while (s->lookahead >= MIN_MATCH) {
442 str = s->strstart;
443 n = s->lookahead - (MIN_MATCH-1);
444 do {
445 UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]);
446 #ifndef FASTEST
447 s->prev[str & s->w_mask] = s->head[s->ins_h];
448 #endif
449 s->head[s->ins_h] = (Pos)str;
450 str++;
451 } while (--n);
452 s->strstart = str;
453 s->lookahead = MIN_MATCH-1;
454 fill_window(s);
455 }
456 s->strstart += s->lookahead;
457 s->block_start = (long)s->strstart;
458 s->insert = s->lookahead;
459 s->lookahead = 0;
460 s->match_length = s->prev_length = MIN_MATCH-1;
461 s->match_available = 0;
462 strm->next_in = next;
463 strm->avail_in = avail;
464 s->wrap = wrap;
465 return Z_OK;
466 }
467
468 /* ========================================================================= */
469 int ZEXPORT deflateGetDictionary (strm, dictionary, dictLength)
470 z_streamp strm;
471 Bytef *dictionary;
472 uInt *dictLength;
473 {
474 deflate_state *s;
475 uInt len;
476
477 if (deflateStateCheck(strm))
478 return Z_STREAM_ERROR;
479 s = strm->state;
480 len = s->strstart + s->lookahead;
481 if (len > s->w_size)
482 len = s->w_size;
483 if (dictionary != Z_NULL && len)
484 zmemcpy(dictionary, s->window + s->strstart + s->lookahead - len, len);
485 if (dictLength != Z_NULL)
486 *dictLength = len;
487 return Z_OK;
488 }
489
490 /* ========================================================================= */
491 int ZEXPORT deflateResetKeep (strm)
492 z_streamp strm;
493 {
494 deflate_state *s;
495
496 if (deflateStateCheck(strm)) {
497 return Z_STREAM_ERROR;
498 }
499
500 strm->total_in = strm->total_out = 0;
501 strm->msg = Z_NULL; /* use zfree if we ever allocate msg dynamically */
502 strm->data_type = Z_UNKNOWN;
503
504 s = (deflate_state *)strm->state;
505 s->pending = 0;
506 s->pending_out = s->pending_buf;
507
508 s->high_water = 0; /* reset to its inital value 0 */
509
510 if (s->wrap < 0) {
511 s->wrap = -s->wrap; /* was made negative by deflate(..., Z_FINISH); */
512 }
513 s->status =
514 #ifdef GZIP
515 s->wrap == 2 ? GZIP_STATE :
516 #endif
517 s->wrap ? INIT_STATE : BUSY_STATE;
518 strm->adler =
519 #ifdef GZIP
520 s->wrap == 2 ? crc32(0L, Z_NULL, 0) :
521 #endif
522 adler32(0L, Z_NULL, 0);
523 s->last_flush = Z_NO_FLUSH;
524
525 _tr_init(s);
526
527 return Z_OK;
528 }
529
530 /* ========================================================================= */
531 int ZEXPORT deflateReset (strm)
532 z_streamp strm;
533 {
534 int ret;
535
536 ret = deflateResetKeep(strm);
537 if (ret == Z_OK)
538 lm_init(strm->state);
539 return ret;
540 }
541
542 /* ========================================================================= */
543 int ZEXPORT deflateSetHeader (strm, head)
544 z_streamp strm;
545 gz_headerp head;
546 {
547 if (deflateStateCheck(strm) || strm->state->wrap != 2)
548 return Z_STREAM_ERROR;
549 strm->state->gzhead = head;
550 return Z_OK;
551 }
552
553 /* ========================================================================= */
554 int ZEXPORT deflatePending (strm, pending, bits)
555 unsigned *pending;
556 int *bits;
557 z_streamp strm;
558 {
559 if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
560 if (pending != Z_NULL)
561 *pending = strm->state->pending;
562 if (bits != Z_NULL)
563 *bits = strm->state->bi_valid;
564 return Z_OK;
565 }
566
567 /* ========================================================================= */
568 int ZEXPORT deflatePrime (strm, bits, value)
569 z_streamp strm;
570 int bits;
571 int value;
572 {
573 deflate_state *s;
574 int put;
575
576 if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
577 s = strm->state;
578 if ((Bytef *)(s->d_buf) < s->pending_out + ((Buf_size + 7) >> 3))
579 return Z_BUF_ERROR;
580 do {
581 put = Buf_size - s->bi_valid;
582 if (put > bits)
583 put = bits;
584 s->bi_buf |= (ush)((value & ((1 << put) - 1)) << s->bi_valid);
585 s->bi_valid += put;
586 _tr_flush_bits(s);
587 value >>= put;
588 bits -= put;
589 } while (bits);
590 return Z_OK;
591 }
592
593 /* ========================================================================= */
594 int ZEXPORT deflateParams(strm, level, strategy)
595 z_streamp strm;
596 int level;
597 int strategy;
598 {
599 deflate_state *s;
600 compress_func func;
601
602 if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
603 s = strm->state;
604
605 #ifdef FASTEST
606 if (level != 0) level = 1;
607 #else
608 if (level == Z_DEFAULT_COMPRESSION) level = 6;
609 #endif
610 if (level < 0 || level > 9 || strategy < 0 || strategy > Z_FIXED) {
611 return Z_STREAM_ERROR;
612 }
613 func = configuration_table[s->level].func;
614
615 if ((strategy != s->strategy || func != configuration_table[level].func) &&
616 s->high_water) {
617 /* Flush the last buffer: */
618 int err = deflate(strm, Z_BLOCK);
619 if (err == Z_STREAM_ERROR)
620 return err;
621 if (strm->avail_out == 0)
622 return Z_BUF_ERROR;
623 }
624 if (s->level != level) {
625 if (s->level == 0 && s->matches != 0) {
626 if (s->matches == 1)
627 slide_hash(s);
628 else
629 CLEAR_HASH(s);
630 s->matches = 0;
631 }
632 s->level = level;
633 s->max_lazy_match = configuration_table[level].max_lazy;
634 s->good_match = configuration_table[level].good_length;
635 s->nice_match = configuration_table[level].nice_length;
636 s->max_chain_length = configuration_table[level].max_chain;
637 }
638 s->strategy = strategy;
639 return Z_OK;
640 }
641
642 /* ========================================================================= */
643 int ZEXPORT deflateTune(strm, good_length, max_lazy, nice_length, max_chain)
644 z_streamp strm;
645 int good_length;
646 int max_lazy;
647 int nice_length;
648 int max_chain;
649 {
650 deflate_state *s;
651
652 if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
653 s = strm->state;
654 s->good_match = (uInt)good_length;
655 s->max_lazy_match = (uInt)max_lazy;
656 s->nice_match = nice_length;
657 s->max_chain_length = (uInt)max_chain;
658 return Z_OK;
659 }
660
661 /* =========================================================================
662 * For the default windowBits of 15 and memLevel of 8, this function returns
663 * a close to exact, as well as small, upper bound on the compressed size.
664 * They are coded as constants here for a reason--if the #define's are
665 * changed, then this function needs to be changed as well. The return
666 * value for 15 and 8 only works for those exact settings.
667 *
668 * For any setting other than those defaults for windowBits and memLevel,
669 * the value returned is a conservative worst case for the maximum expansion
670 * resulting from using fixed blocks instead of stored blocks, which deflate
671 * can emit on compressed data for some combinations of the parameters.
672 *
673 * This function could be more sophisticated to provide closer upper bounds for
674 * every combination of windowBits and memLevel. But even the conservative
675 * upper bound of about 14% expansion does not seem onerous for output buffer
676 * allocation.
677 */
678 uLong ZEXPORT deflateBound(strm, sourceLen)
679 z_streamp strm;
680 uLong sourceLen;
681 {
682 deflate_state *s;
683 uLong complen, wraplen;
684
685 /* conservative upper bound for compressed data */
686 complen = sourceLen +
687 ((sourceLen + 7) >> 3) + ((sourceLen + 63) >> 6) + 5;
688
689 /* if can't get parameters, return conservative bound plus zlib wrapper */
690 if (deflateStateCheck(strm))
691 return complen + 6;
692
693 /* compute wrapper length */
694 s = strm->state;
695 switch (s->wrap) {
696 case 0: /* raw deflate */
697 wraplen = 0;
698 break;
699 case 1: /* zlib wrapper */
700 wraplen = 6 + (s->strstart ? 4 : 0);
701 break;
702 #ifdef GZIP
703 case 2: /* gzip wrapper */
704 wraplen = 18;
705 if (s->gzhead != Z_NULL) { /* user-supplied gzip header */
706 Bytef *str;
707 if (s->gzhead->extra != Z_NULL)
708 wraplen += 2 + s->gzhead->extra_len;
709 str = s->gzhead->name;
710 if (str != Z_NULL)
711 do {
712 wraplen++;
713 } while (*str++);
714 str = s->gzhead->comment;
715 if (str != Z_NULL)
716 do {
717 wraplen++;
718 } while (*str++);
719 if (s->gzhead->hcrc)
720 wraplen += 2;
721 }
722 break;
723 #endif
724 default: /* for compiler happiness */
725 wraplen = 6;
726 }
727
728 /* if not default parameters, return conservative bound */
729 if (s->w_bits != 15 || s->hash_bits != 8 + 7)
730 return complen + wraplen;
731
732 /* default settings: return tight bound for that case */
733 return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
734 (sourceLen >> 25) + 13 - 6 + wraplen;
735 }
736
737 /* =========================================================================
738 * Put a short in the pending buffer. The 16-bit value is put in MSB order.
739 * IN assertion: the stream state is correct and there is enough room in
740 * pending_buf.
741 */
742 local void putShortMSB (s, b)
743 deflate_state *s;
744 uInt b;
745 {
746 put_byte(s, (Byte)(b >> 8));
747 put_byte(s, (Byte)(b & 0xff));
748 }
749
750 /* =========================================================================
751 * Flush as much pending output as possible. All deflate() output, except for
752 * some deflate_stored() output, goes through this function so some
753 * applications may wish to modify it to avoid allocating a large
754 * strm->next_out buffer and copying into it. (See also read_buf()).
755 */
756 local void flush_pending(strm)
757 z_streamp strm;
758 {
759 unsigned len;
760 deflate_state *s = strm->state;
761
762 _tr_flush_bits(s);
763 len = s->pending;
764 if (len > strm->avail_out) len = strm->avail_out;
765 if (len == 0) return;
766
767 zmemcpy(strm->next_out, s->pending_out, len);
768 strm->next_out += len;
769 s->pending_out += len;
770 strm->total_out += len;
771 strm->avail_out -= len;
772 s->pending -= len;
773 if (s->pending == 0) {
774 s->pending_out = s->pending_buf;
775 }
776 }
777
778 /* ===========================================================================
779 * Update the header CRC with the bytes s->pending_buf[beg..s->pending - 1].
780 */
781 #define HCRC_UPDATE(beg) \
782 do { \
783 if (s->gzhead->hcrc && s->pending > (beg)) \
784 strm->adler = crc32(strm->adler, s->pending_buf + (beg), \
785 s->pending - (beg)); \
786 } while (0)
787
788 /* ========================================================================= */
789 int ZEXPORT deflate (strm, flush)
790 z_streamp strm;
791 int flush;
792 {
793 int old_flush; /* value of flush param for previous deflate call */
794 deflate_state *s;
795
796 if (deflateStateCheck(strm) || flush > Z_BLOCK || flush < 0) {
797 return Z_STREAM_ERROR;
798 }
799 s = strm->state;
800
801 if (strm->next_out == Z_NULL ||
802 (strm->avail_in != 0 && strm->next_in == Z_NULL) ||
803 (s->status == FINISH_STATE && flush != Z_FINISH)) {
804 ERR_RETURN(strm, Z_STREAM_ERROR);
805 }
806 if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR);
807
808 old_flush = s->last_flush;
809 s->last_flush = flush;
810
811 /* Flush as much pending output as possible */
812 if (s->pending != 0) {
813 flush_pending(strm);
814 if (strm->avail_out == 0) {
815 /* Since avail_out is 0, deflate will be called again with
816 * more output space, but possibly with both pending and
817 * avail_in equal to zero. There won't be anything to do,
818 * but this is not an error situation so make sure we
819 * return OK instead of BUF_ERROR at next call of deflate:
820 */
821 s->last_flush = -1;
822 return Z_OK;
823 }
824
825 /* Make sure there is something to do and avoid duplicate consecutive
826 * flushes. For repeated and useless calls with Z_FINISH, we keep
827 * returning Z_STREAM_END instead of Z_BUF_ERROR.
828 */
829 } else if (strm->avail_in == 0 && RANK(flush) <= RANK(old_flush) &&
830 flush != Z_FINISH) {
831 ERR_RETURN(strm, Z_BUF_ERROR);
832 }
833
834 /* User must not provide more input after the first FINISH: */
835 if (s->status == FINISH_STATE && strm->avail_in != 0) {
836 ERR_RETURN(strm, Z_BUF_ERROR);
837 }
838
839 /* Write the header */
840 if (s->status == INIT_STATE) {
841 /* zlib header */
842 uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8;
843 uInt level_flags;
844
845 if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2)
846 level_flags = 0;
847 else if (s->level < 6)
848 level_flags = 1;
849 else if (s->level == 6)
850 level_flags = 2;
851 else
852 level_flags = 3;
853 header |= (level_flags << 6);
854 if (s->strstart != 0) header |= PRESET_DICT;
855 header += 31 - (header % 31);
856
857 putShortMSB(s, header);
858
859 /* Save the adler32 of the preset dictionary: */
860 if (s->strstart != 0) {
861 putShortMSB(s, (uInt)(strm->adler >> 16));
862 putShortMSB(s, (uInt)(strm->adler & 0xffff));
863 }
864 strm->adler = adler32(0L, Z_NULL, 0);
865 s->status = BUSY_STATE;
866
867 /* Compression must start with an empty pending buffer */
868 flush_pending(strm);
869 if (s->pending != 0) {
870 s->last_flush = -1;
871 return Z_OK;
872 }
873 }
874 #ifdef GZIP
875 if (s->status == GZIP_STATE) {
876 /* gzip header */
877 strm->adler = crc32(0L, Z_NULL, 0);
878 put_byte(s, 31);
879 put_byte(s, 139);
880 put_byte(s, 8);
881 if (s->gzhead == Z_NULL) {
882 put_byte(s, 0);
883 put_byte(s, 0);
884 put_byte(s, 0);
885 put_byte(s, 0);
886 put_byte(s, 0);
887 put_byte(s, s->level == 9 ? 2 :
888 (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
889 4 : 0));
890 put_byte(s, OS_CODE);
891 s->status = BUSY_STATE;
892
893 /* Compression must start with an empty pending buffer */
894 flush_pending(strm);
895 if (s->pending != 0) {
896 s->last_flush = -1;
897 return Z_OK;
898 }
899 }
900 else {
901 put_byte(s, (s->gzhead->text ? 1 : 0) +
902 (s->gzhead->hcrc ? 2 : 0) +
903 (s->gzhead->extra == Z_NULL ? 0 : 4) +
904 (s->gzhead->name == Z_NULL ? 0 : 8) +
905 (s->gzhead->comment == Z_NULL ? 0 : 16)
906 );
907 put_byte(s, (Byte)(s->gzhead->time & 0xff));
908 put_byte(s, (Byte)((s->gzhead->time >> 8) & 0xff));
909 put_byte(s, (Byte)((s->gzhead->time >> 16) & 0xff));
910 put_byte(s, (Byte)((s->gzhead->time >> 24) & 0xff));
911 put_byte(s, s->level == 9 ? 2 :
912 (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
913 4 : 0));
914 put_byte(s, s->gzhead->os & 0xff);
915 if (s->gzhead->extra != Z_NULL) {
916 put_byte(s, s->gzhead->extra_len & 0xff);
917 put_byte(s, (s->gzhead->extra_len >> 8) & 0xff);
918 }
919 if (s->gzhead->hcrc)
920 strm->adler = crc32(strm->adler, s->pending_buf,
921 s->pending);
922 s->gzindex = 0;
923 s->status = EXTRA_STATE;
924 }
925 }
926 if (s->status == EXTRA_STATE) {
927 if (s->gzhead->extra != Z_NULL) {
928 ulg beg = s->pending; /* start of bytes to update crc */
929 uInt left = (s->gzhead->extra_len & 0xffff) - s->gzindex;
930 while (s->pending + left > s->pending_buf_size) {
931 uInt copy = s->pending_buf_size - s->pending;
932 zmemcpy(s->pending_buf + s->pending,
933 s->gzhead->extra + s->gzindex, copy);
934 s->pending = s->pending_buf_size;
935 HCRC_UPDATE(beg);
936 s->gzindex += copy;
937 flush_pending(strm);
938 if (s->pending != 0) {
939 s->last_flush = -1;
940 return Z_OK;
941 }
942 beg = 0;
943 left -= copy;
944 }
945 zmemcpy(s->pending_buf + s->pending,
946 s->gzhead->extra + s->gzindex, left);
947 s->pending += left;
948 HCRC_UPDATE(beg);
949 s->gzindex = 0;
950 }
951 s->status = NAME_STATE;
952 }
953 if (s->status == NAME_STATE) {
954 if (s->gzhead->name != Z_NULL) {
955 ulg beg = s->pending; /* start of bytes to update crc */
956 int val;
957 do {
958 if (s->pending == s->pending_buf_size) {
959 HCRC_UPDATE(beg);
960 flush_pending(strm);
961 if (s->pending != 0) {
962 s->last_flush = -1;
963 return Z_OK;
964 }
965 beg = 0;
966 }
967 val = s->gzhead->name[s->gzindex++];
968 put_byte(s, val);
969 } while (val != 0);
970 HCRC_UPDATE(beg);
971 s->gzindex = 0;
972 }
973 s->status = COMMENT_STATE;
974 }
975 if (s->status == COMMENT_STATE) {
976 if (s->gzhead->comment != Z_NULL) {
977 ulg beg = s->pending; /* start of bytes to update crc */
978 int val;
979 do {
980 if (s->pending == s->pending_buf_size) {
981 HCRC_UPDATE(beg);
982 flush_pending(strm);
983 if (s->pending != 0) {
984 s->last_flush = -1;
985 return Z_OK;
986 }
987 beg = 0;
988 }
989 val = s->gzhead->comment[s->gzindex++];
990 put_byte(s, val);
991 } while (val != 0);
992 HCRC_UPDATE(beg);
993 }
994 s->status = HCRC_STATE;
995 }
996 if (s->status == HCRC_STATE) {
997 if (s->gzhead->hcrc) {
998 if (s->pending + 2 > s->pending_buf_size) {
999 flush_pending(strm);
1000 if (s->pending != 0) {
1001 s->last_flush = -1;
1002 return Z_OK;
1003 }
1004 }
1005 put_byte(s, (Byte)(strm->adler & 0xff));
1006 put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
1007 strm->adler = crc32(0L, Z_NULL, 0);
1008 }
1009 s->status = BUSY_STATE;
1010
1011 /* Compression must start with an empty pending buffer */
1012 flush_pending(strm);
1013 if (s->pending != 0) {
1014 s->last_flush = -1;
1015 return Z_OK;
1016 }
1017 }
1018 #endif
1019
1020 /* Start a new block or continue the current one.
1021 */
1022 if (strm->avail_in != 0 || s->lookahead != 0 ||
1023 (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) {
1024 block_state bstate;
1025
1026 bstate = s->level == 0 ? deflate_stored(s, flush) :
1027 s->strategy == Z_HUFFMAN_ONLY ? deflate_huff(s, flush) :
1028 s->strategy == Z_RLE ? deflate_rle(s, flush) :
1029 (*(configuration_table[s->level].func))(s, flush);
1030
1031 if (bstate == finish_started || bstate == finish_done) {
1032 s->status = FINISH_STATE;
1033 }
1034 if (bstate == need_more || bstate == finish_started) {
1035 if (strm->avail_out == 0) {
1036 s->last_flush = -1; /* avoid BUF_ERROR next call, see above */
1037 }
1038 return Z_OK;
1039 /* If flush != Z_NO_FLUSH && avail_out == 0, the next call
1040 * of deflate should use the same flush parameter to make sure
1041 * that the flush is complete. So we don't have to output an
1042 * empty block here, this will be done at next call. This also
1043 * ensures that for a very small output buffer, we emit at most
1044 * one empty block.
1045 */
1046 }
1047 if (bstate == block_done) {
1048 if (flush == Z_PARTIAL_FLUSH) {
1049 _tr_align(s);
1050 } else if (flush != Z_BLOCK) { /* FULL_FLUSH or SYNC_FLUSH */
1051 _tr_stored_block(s, (char*)0, 0L, 0);
1052 /* For a full flush, this empty block will be recognized
1053 * as a special marker by inflate_sync().
1054 */
1055 if (flush == Z_FULL_FLUSH) {
1056 CLEAR_HASH(s); /* forget history */
1057 if (s->lookahead == 0) {
1058 s->strstart = 0;
1059 s->block_start = 0L;
1060 s->insert = 0;
1061 }
1062 }
1063 }
1064 flush_pending(strm);
1065 if (strm->avail_out == 0) {
1066 s->last_flush = -1; /* avoid BUF_ERROR at next call, see above */
1067 return Z_OK;
1068 }
1069 }
1070 }
1071
1072 if (flush != Z_FINISH) return Z_OK;
1073 if (s->wrap <= 0) return Z_STREAM_END;
1074
1075 /* Write the trailer */
1076 #ifdef GZIP
1077 if (s->wrap == 2) {
1078 put_byte(s, (Byte)(strm->adler & 0xff));
1079 put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
1080 put_byte(s, (Byte)((strm->adler >> 16) & 0xff));
1081 put_byte(s, (Byte)((strm->adler >> 24) & 0xff));
1082 put_byte(s, (Byte)(strm->total_in & 0xff));
1083 put_byte(s, (Byte)((strm->total_in >> 8) & 0xff));
1084 put_byte(s, (Byte)((strm->total_in >> 16) & 0xff));
1085 put_byte(s, (Byte)((strm->total_in >> 24) & 0xff));
1086 }
1087 else
1088 #endif
1089 {
1090 putShortMSB(s, (uInt)(strm->adler >> 16));
1091 putShortMSB(s, (uInt)(strm->adler & 0xffff));
1092 }
1093 flush_pending(strm);
1094 /* If avail_out is zero, the application will call deflate again
1095 * to flush the rest.
1096 */
1097 if (s->wrap > 0) s->wrap = -s->wrap; /* write the trailer only once! */
1098 return s->pending != 0 ? Z_OK : Z_STREAM_END;
1099 }
1100
1101 /* ========================================================================= */
1102 int ZEXPORT deflateEnd (strm)
1103 z_streamp strm;
1104 {
1105 int status;
1106
1107 if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
1108
1109 status = strm->state->status;
1110
1111 /* Deallocate in reverse order of allocations: */
1112 TRY_FREE(strm, strm->state->pending_buf);
1113 TRY_FREE(strm, strm->state->head);
1114 TRY_FREE(strm, strm->state->prev);
1115 TRY_FREE(strm, strm->state->window);
1116
1117 ZFREE(strm, strm->state);
1118 strm->state = Z_NULL;
1119
1120 return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK;
1121 }
1122
1123 /* =========================================================================
1124 * Copy the source state to the destination state.
1125 * To simplify the source, this is not supported for 16-bit MSDOS (which
1126 * doesn't have enough memory anyway to duplicate compression states).
1127 */
1128 int ZEXPORT deflateCopy (dest, source)
1129 z_streamp dest;
1130 z_streamp source;
1131 {
1132 #ifdef MAXSEG_64K
1133 return Z_STREAM_ERROR;
1134 #else
1135 deflate_state *ds;
1136 deflate_state *ss;
1137 ushf *overlay;
1138
1139
1140 if (deflateStateCheck(source) || dest == Z_NULL) {
1141 return Z_STREAM_ERROR;
1142 }
1143
1144 ss = source->state;
1145
1146 zmemcpy((voidpf)dest, (voidpf)source, sizeof(z_stream));
1147
1148 ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state));
1149 if (ds == Z_NULL) return Z_MEM_ERROR;
1150 dest->state = (struct internal_state FAR *) ds;
1151 zmemcpy((voidpf)ds, (voidpf)ss, sizeof(deflate_state));
1152 ds->strm = dest;
1153
1154 ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte));
1155 ds->prev = (Posf *) ZALLOC(dest, ds->w_size, sizeof(Pos));
1156 ds->head = (Posf *) ZALLOC(dest, ds->hash_size, sizeof(Pos));
1157 overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2);
1158 ds->pending_buf = (uchf *) overlay;
1159
1160 if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL ||
1161 ds->pending_buf == Z_NULL) {
1162 deflateEnd (dest);
1163 return Z_MEM_ERROR;
1164 }
1165 /* following zmemcpy do not work for 16-bit MSDOS */
1166 zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte));
1167 zmemcpy((voidpf)ds->prev, (voidpf)ss->prev, ds->w_size * sizeof(Pos));
1168 zmemcpy((voidpf)ds->head, (voidpf)ss->head, ds->hash_size * sizeof(Pos));
1169 zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size);
1170
1171 ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf);
1172 ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush);
1173 ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize;
1174
1175 ds->l_desc.dyn_tree = ds->dyn_ltree;
1176 ds->d_desc.dyn_tree = ds->dyn_dtree;
1177 ds->bl_desc.dyn_tree = ds->bl_tree;
1178
1179 return Z_OK;
1180 #endif /* MAXSEG_64K */
1181 }
1182
1183 /* ===========================================================================
1184 * Read a new buffer from the current input stream, update the adler32
1185 * and total number of bytes read. All deflate() input goes through
1186 * this function so some applications may wish to modify it to avoid
1187 * allocating a large strm->next_in buffer and copying from it.
1188 * (See also flush_pending()).
1189 */
1190 local unsigned read_buf(strm, buf, size)
1191 z_streamp strm;
1192 Bytef *buf;
1193 unsigned size;
1194 {
1195 unsigned len = strm->avail_in;
1196
1197 if (len > size) len = size;
1198 if (len == 0) return 0;
1199
1200 strm->avail_in -= len;
1201
1202 zmemcpy(buf, strm->next_in, len);
1203 if (strm->state->wrap == 1) {
1204 strm->adler = adler32(strm->adler, buf, len);
1205 }
1206 #ifdef GZIP
1207 else if (strm->state->wrap == 2) {
1208 strm->adler = crc32(strm->adler, buf, len);
1209 }
1210 #endif
1211 strm->next_in += len;
1212 strm->total_in += len;
1213
1214 return len;
1215 }
1216
1217 /* ===========================================================================
1218 * Initialize the "longest match" routines for a new zlib stream
1219 */
1220 local void lm_init (s)
1221 deflate_state *s;
1222 {
1223 s->window_size = (ulg)2L*s->w_size;
1224
1225 CLEAR_HASH(s);
1226
1227 /* Set the default configuration parameters:
1228 */
1229 s->max_lazy_match = configuration_table[s->level].max_lazy;
1230 s->good_match = configuration_table[s->level].good_length;
1231 s->nice_match = configuration_table[s->level].nice_length;
1232 s->max_chain_length = configuration_table[s->level].max_chain;
1233
1234 s->strstart = 0;
1235 s->block_start = 0L;
1236 s->lookahead = 0;
1237 s->insert = 0;
1238 s->match_length = s->prev_length = MIN_MATCH-1;
1239 s->match_available = 0;
1240 s->ins_h = 0;
1241 #ifndef FASTEST
1242 #ifdef ASMV
1243 match_init(); /* initialize the asm code */
1244 #endif
1245 #endif
1246 }
1247
1248 #ifndef FASTEST
1249 /* ===========================================================================
1250 * Set match_start to the longest match starting at the given string and
1251 * return its length. Matches shorter or equal to prev_length are discarded,
1252 * in which case the result is equal to prev_length and match_start is
1253 * garbage.
1254 * IN assertions: cur_match is the head of the hash chain for the current
1255 * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
1256 * OUT assertion: the match length is not greater than s->lookahead.
1257 */
1258 #ifndef ASMV
1259 /* For 80x86 and 680x0, an optimized version will be provided in match.asm or
1260 * match.S. The code will be functionally equivalent.
1261 */
1262 local uInt longest_match(s, cur_match)
1263 deflate_state *s;
1264 IPos cur_match; /* current match */
1265 {
1266 unsigned chain_length = s->max_chain_length;/* max hash chain length */
1267 register Bytef *scan = s->window + s->strstart; /* current string */
1268 register Bytef *match; /* matched string */
1269 register int len; /* length of current match */
1270 int best_len = (int)s->prev_length; /* best match length so far */
1271 int nice_match = s->nice_match; /* stop if match long enough */
1272 IPos limit = s->strstart > (IPos)MAX_DIST(s) ?
1273 s->strstart - (IPos)MAX_DIST(s) : NIL;
1274 /* Stop when cur_match becomes <= limit. To simplify the code,
1275 * we prevent matches with the string of window index 0.
1276 */
1277 Posf *prev = s->prev;
1278 uInt wmask = s->w_mask;
1279
1280 #ifdef UNALIGNED_OK
1281 /* Compare two bytes at a time. Note: this is not always beneficial.
1282 * Try with and without -DUNALIGNED_OK to check.
1283 */
1284 register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1;
1285 register ush scan_start = *(ushf*)scan;
1286 register ush scan_end = *(ushf*)(scan+best_len-1);
1287 #else
1288 register Bytef *strend = s->window + s->strstart + MAX_MATCH;
1289 register Byte scan_end1 = scan[best_len-1];
1290 register Byte scan_end = scan[best_len];
1291 #endif
1292
1293 /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
1294 * It is easy to get rid of this optimization if necessary.
1295 */
1296 Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");
1297
1298 /* Do not waste too much time if we already have a good match: */
1299 if (s->prev_length >= s->good_match) {
1300 chain_length >>= 2;
1301 }
1302 /* Do not look for matches beyond the end of the input. This is necessary
1303 * to make deflate deterministic.
1304 */
1305 if ((uInt)nice_match > s->lookahead) nice_match = (int)s->lookahead;
1306
1307 Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
1308
1309 do {
1310 Assert(cur_match < s->strstart, "no future");
1311 match = s->window + cur_match;
1312
1313 /* Skip to next match if the match length cannot increase
1314 * or if the match length is less than 2. Note that the checks below
1315 * for insufficient lookahead only occur occasionally for performance
1316 * reasons. Therefore uninitialized memory will be accessed, and
1317 * conditional jumps will be made that depend on those values.
1318 * However the length of the match is limited to the lookahead, so
1319 * the output of deflate is not affected by the uninitialized values.
1320 */
1321 #if (defined(UNALIGNED_OK) && MAX_MATCH == 258)
1322 /* This code assumes sizeof(unsigned short) == 2. Do not use
1323 * UNALIGNED_OK if your compiler uses a different size.
1324 */
1325 if (*(ushf*)(match+best_len-1) != scan_end ||
1326 *(ushf*)match != scan_start) continue;
1327
1328 /* It is not necessary to compare scan[2] and match[2] since they are
1329 * always equal when the other bytes match, given that the hash keys
1330 * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at
1331 * strstart+3, +5, ... up to strstart+257. We check for insufficient
1332 * lookahead only every 4th comparison; the 128th check will be made
1333 * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is
1334 * necessary to put more guard bytes at the end of the window, or
1335 * to check more often for insufficient lookahead.
1336 */
1337 Assert(scan[2] == match[2], "scan[2]?");
1338 scan++, match++;
1339 do {
1340 } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
1341 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
1342 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
1343 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
1344 scan < strend);
1345 /* The funny "do {}" generates better code on most compilers */
1346
1347 /* Here, scan <= window+strstart+257 */
1348 Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
1349 if (*scan == *match) scan++;
1350
1351 len = (MAX_MATCH - 1) - (int)(strend-scan);
1352 scan = strend - (MAX_MATCH-1);
1353
1354 #else /* UNALIGNED_OK */
1355
1356 if (match[best_len] != scan_end ||
1357 match[best_len-1] != scan_end1 ||
1358 *match != *scan ||
1359 *++match != scan[1]) continue;
1360
1361 /* The check at best_len-1 can be removed because it will be made
1362 * again later. (This heuristic is not always a win.)
1363 * It is not necessary to compare scan[2] and match[2] since they
1364 * are always equal when the other bytes match, given that
1365 * the hash keys are equal and that HASH_BITS >= 8.
1366 */
1367 scan += 2, match++;
1368 Assert(*scan == *match, "match[2]?");
1369
1370 /* We check for insufficient lookahead only every 8th comparison;
1371 * the 256th check will be made at strstart+258.
1372 */
1373 do {
1374 } while (*++scan == *++match && *++scan == *++match &&
1375 *++scan == *++match && *++scan == *++match &&
1376 *++scan == *++match && *++scan == *++match &&
1377 *++scan == *++match && *++scan == *++match &&
1378 scan < strend);
1379
1380 Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
1381
1382 len = MAX_MATCH - (int)(strend - scan);
1383 scan = strend - MAX_MATCH;
1384
1385 #endif /* UNALIGNED_OK */
1386
1387 if (len > best_len) {
1388 s->match_start = cur_match;
1389 best_len = len;
1390 if (len >= nice_match) break;
1391 #ifdef UNALIGNED_OK
1392 scan_end = *(ushf*)(scan+best_len-1);
1393 #else
1394 scan_end1 = scan[best_len-1];
1395 scan_end = scan[best_len];
1396 #endif
1397 }
1398 } while ((cur_match = prev[cur_match & wmask]) > limit
1399 && --chain_length != 0);
1400
1401 if ((uInt)best_len <= s->lookahead) return (uInt)best_len;
1402 return s->lookahead;
1403 }
1404 #endif /* ASMV */
1405
1406 #else /* FASTEST */
1407
1408 /* ---------------------------------------------------------------------------
1409 * Optimized version for FASTEST only
1410 */
1411 local uInt longest_match(s, cur_match)
1412 deflate_state *s;
1413 IPos cur_match; /* current match */
1414 {
1415 register Bytef *scan = s->window + s->strstart; /* current string */
1416 register Bytef *match; /* matched string */
1417 register int len; /* length of current match */
1418 register Bytef *strend = s->window + s->strstart + MAX_MATCH;
1419
1420 /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
1421 * It is easy to get rid of this optimization if necessary.
1422 */
1423 Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");
1424
1425 Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
1426
1427 Assert(cur_match < s->strstart, "no future");
1428
1429 match = s->window + cur_match;
1430
1431 /* Return failure if the match length is less than 2:
1432 */
1433 if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1;
1434
1435 /* The check at best_len-1 can be removed because it will be made
1436 * again later. (This heuristic is not always a win.)
1437 * It is not necessary to compare scan[2] and match[2] since they
1438 * are always equal when the other bytes match, given that
1439 * the hash keys are equal and that HASH_BITS >= 8.
1440 */
1441 scan += 2, match += 2;
1442 Assert(*scan == *match, "match[2]?");
1443
1444 /* We check for insufficient lookahead only every 8th comparison;
1445 * the 256th check will be made at strstart+258.
1446 */
1447 do {
1448 } while (*++scan == *++match && *++scan == *++match &&
1449 *++scan == *++match && *++scan == *++match &&
1450 *++scan == *++match && *++scan == *++match &&
1451 *++scan == *++match && *++scan == *++match &&
1452 scan < strend);
1453
1454 Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
1455
1456 len = MAX_MATCH - (int)(strend - scan);
1457
1458 if (len < MIN_MATCH) return MIN_MATCH - 1;
1459
1460 s->match_start = cur_match;
1461 return (uInt)len <= s->lookahead ? (uInt)len : s->lookahead;
1462 }
1463
1464 #endif /* FASTEST */
1465
1466 #ifdef ZLIB_DEBUG
1467
1468 #define EQUAL 0
1469 /* result of memcmp for equal strings */
1470
1471 /* ===========================================================================
1472 * Check that the match at match_start is indeed a match.
1473 */
1474 local void check_match(s, start, match, length)
1475 deflate_state *s;
1476 IPos start, match;
1477 int length;
1478 {
1479 /* check that the match is indeed a match */
1480 if (zmemcmp(s->window + match,
1481 s->window + start, length) != EQUAL) {
1482 fprintf(stderr, " start %u, match %u, length %d\n",
1483 start, match, length);
1484 do {
1485 fprintf(stderr, "%c%c", s->window[match++], s->window[start++]);
1486 } while (--length != 0);
1487 z_error("invalid match");
1488 }
1489 if (z_verbose > 1) {
1490 fprintf(stderr,"\\[%d,%d]", start-match, length);
1491 do { putc(s->window[start++], stderr); } while (--length != 0);
1492 }
1493 }
1494 #else
1495 # define check_match(s, start, match, length)
1496 #endif /* ZLIB_DEBUG */
1497
1498 /* ===========================================================================
1499 * Fill the window when the lookahead becomes insufficient.
1500 * Updates strstart and lookahead.
1501 *
1502 * IN assertion: lookahead < MIN_LOOKAHEAD
1503 * OUT assertions: strstart <= window_size-MIN_LOOKAHEAD
1504 * At least one byte has been read, or avail_in == 0; reads are
1505 * performed for at least two bytes (required for the zip translate_eol
1506 * option -- not supported here).
1507 */
1508 local void fill_window(s)
1509 deflate_state *s;
1510 {
1511 unsigned n;
1512 unsigned more; /* Amount of free space at the end of the window. */
1513 uInt wsize = s->w_size;
1514
1515 Assert(s->lookahead < MIN_LOOKAHEAD, "already enough lookahead");
1516
1517 do {
1518 more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart);
1519
1520 /* Deal with !@#$% 64K limit: */
1521 if (sizeof(int) <= 2) {
1522 if (more == 0 && s->strstart == 0 && s->lookahead == 0) {
1523 more = wsize;
1524
1525 } else if (more == (unsigned)(-1)) {
1526 /* Very unlikely, but possible on 16 bit machine if
1527 * strstart == 0 && lookahead == 1 (input done a byte at time)
1528 */
1529 more--;
1530 }
1531 }
1532
1533 /* If the window is almost full and there is insufficient lookahead,
1534 * move the upper half to the lower one to make room in the upper half.
1535 */
1536 if (s->strstart >= wsize+MAX_DIST(s)) {
1537
1538 zmemcpy(s->window, s->window+wsize, (unsigned)wsize - more);
1539 s->match_start -= wsize;
1540 s->strstart -= wsize; /* we now have strstart >= MAX_DIST */
1541 s->block_start -= (long) wsize;
1542 slide_hash(s);
1543 more += wsize;
1544 }
1545 if (s->strm->avail_in == 0) break;
1546
1547 /* If there was no sliding:
1548 * strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
1549 * more == window_size - lookahead - strstart
1550 * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1)
1551 * => more >= window_size - 2*WSIZE + 2
1552 * In the BIG_MEM or MMAP case (not yet supported),
1553 * window_size == input_size + MIN_LOOKAHEAD &&
1554 * strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD.
1555 * Otherwise, window_size == 2*WSIZE so more >= 2.
1556 * If there was sliding, more >= WSIZE. So in all cases, more >= 2.
1557 */
1558 Assert(more >= 2, "more < 2");
1559
1560 n = read_buf(s->strm, s->window + s->strstart + s->lookahead, more);
1561 s->lookahead += n;
1562
1563 /* Initialize the hash value now that we have some input: */
1564 if (s->lookahead + s->insert >= MIN_MATCH) {
1565 uInt str = s->strstart - s->insert;
1566 s->ins_h = s->window[str];
1567 UPDATE_HASH(s, s->ins_h, s->window[str + 1]);
1568 #if MIN_MATCH != 3
1569 Call UPDATE_HASH() MIN_MATCH-3 more times
1570 #endif
1571 while (s->insert) {
1572 UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]);
1573 #ifndef FASTEST
1574 s->prev[str & s->w_mask] = s->head[s->ins_h];
1575 #endif
1576 s->head[s->ins_h] = (Pos)str;
1577 str++;
1578 s->insert--;
1579 if (s->lookahead + s->insert < MIN_MATCH)
1580 break;
1581 }
1582 }
1583 /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage,
1584 * but this is not important since only literal bytes will be emitted.
1585 */
1586
1587 } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0);
1588
1589 /* If the WIN_INIT bytes after the end of the current data have never been
1590 * written, then zero those bytes in order to avoid memory check reports of
1591 * the use of uninitialized (or uninitialised as Julian writes) bytes by
1592 * the longest match routines. Update the high water mark for the next
1593 * time through here. WIN_INIT is set to MAX_MATCH since the longest match
1594 * routines allow scanning to strstart + MAX_MATCH, ignoring lookahead.
1595 */
1596 if (s->high_water < s->window_size) {
1597 ulg curr = s->strstart + (ulg)(s->lookahead);
1598 ulg init;
1599
1600 if (s->high_water < curr) {
1601 /* Previous high water mark below current data -- zero WIN_INIT
1602 * bytes or up to end of window, whichever is less.
1603 */
1604 init = s->window_size - curr;
1605 if (init > WIN_INIT)
1606 init = WIN_INIT;
1607 zmemzero(s->window + curr, (unsigned)init);
1608 s->high_water = curr + init;
1609 }
1610 else if (s->high_water < (ulg)curr + WIN_INIT) {
1611 /* High water mark at or above current data, but below current data
1612 * plus WIN_INIT -- zero out to current data plus WIN_INIT, or up
1613 * to end of window, whichever is less.
1614 */
1615 init = (ulg)curr + WIN_INIT - s->high_water;
1616 if (init > s->window_size - s->high_water)
1617 init = s->window_size - s->high_water;
1618 zmemzero(s->window + s->high_water, (unsigned)init);
1619 s->high_water += init;
1620 }
1621 }
1622
1623 Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
1624 "not enough room for search");
1625 }
1626
1627 /* ===========================================================================
1628 * Flush the current block, with given end-of-file flag.
1629 * IN assertion: strstart is set to the end of the current match.
1630 */
1631 #define FLUSH_BLOCK_ONLY(s, last) { \
1632 _tr_flush_block(s, (s->block_start >= 0L ? \
1633 (charf *)&s->window[(unsigned)s->block_start] : \
1634 (charf *)Z_NULL), \
1635 (ulg)((long)s->strstart - s->block_start), \
1636 (last)); \
1637 s->block_start = s->strstart; \
1638 flush_pending(s->strm); \
1639 Tracev((stderr,"[FLUSH]")); \
1640 }
1641
1642 /* Same but force premature exit if necessary. */
1643 #define FLUSH_BLOCK(s, last) { \
1644 FLUSH_BLOCK_ONLY(s, last); \
1645 if (s->strm->avail_out == 0) return (last) ? finish_started : need_more; \
1646 }
1647
1648 /* Maximum stored block length in deflate format (not including header). */
1649 #define MAX_STORED 65535
1650
1651 /* Minimum of a and b. */
1652 #define MIN(a, b) ((a) > (b) ? (b) : (a))
1653
1654 /* ===========================================================================
1655 * Copy without compression as much as possible from the input stream, return
1656 * the current block state.
1657 *
1658 * In case deflateParams() is used to later switch to a non-zero compression
1659 * level, s->matches (otherwise unused when storing) keeps track of the number
1660 * of hash table slides to perform. If s->matches is 1, then one hash table
1661 * slide will be done when switching. If s->matches is 2, the maximum value
1662 * allowed here, then the hash table will be cleared, since two or more slides
1663 * is the same as a clear.
1664 *
1665 * deflate_stored() is written to minimize the number of times an input byte is
1666 * copied. It is most efficient with large input and output buffers, which
1667 * maximizes the opportunites to have a single copy from next_in to next_out.
1668 */
1669 local block_state deflate_stored(s, flush)
1670 deflate_state *s;
1671 int flush;
1672 {
1673 /* Smallest worthy block size when not flushing or finishing. By default
1674 * this is 32K. This can be as small as 507 bytes for memLevel == 1. For
1675 * large input and output buffers, the stored block size will be larger.
1676 */
1677 unsigned min_block = MIN(s->pending_buf_size - 5, s->w_size);
1678
1679 /* Copy as many min_block or larger stored blocks directly to next_out as
1680 * possible. If flushing, copy the remaining available input to next_out as
1681 * stored blocks, if there is enough space.
1682 */
1683 unsigned len, left, have, last = 0;
1684 unsigned used = s->strm->avail_in;
1685 do {
1686 /* Set len to the maximum size block that we can copy directly with the
1687 * available input data and output space. Set left to how much of that
1688 * would be copied from what's left in the window.
1689 */
1690 len = MAX_STORED; /* maximum deflate stored block length */
1691 have = (s->bi_valid + 42) >> 3; /* number of header bytes */
1692 if (s->strm->avail_out < have) /* need room for header */
1693 break;
1694 /* maximum stored block length that will fit in avail_out: */
1695 have = s->strm->avail_out - have;
1696 left = s->strstart - s->block_start; /* bytes left in window */
1697 if (len > (ulg)left + s->strm->avail_in)
1698 len = left + s->strm->avail_in; /* limit len to the input */
1699 if (len > have)
1700 len = have; /* limit len to the output */
1701
1702 /* If the stored block would be less than min_block in length, or if
1703 * unable to copy all of the available input when flushing, then try
1704 * copying to the window and the pending buffer instead. Also don't
1705 * write an empty block when flushing -- deflate() does that.
1706 */
1707 if (len < min_block && ((len == 0 && flush != Z_FINISH) ||
1708 flush == Z_NO_FLUSH ||
1709 len != left + s->strm->avail_in))
1710 break;
1711
1712 /* Make a dummy stored block in pending to get the header bytes,
1713 * including any pending bits. This also updates the debugging counts.
1714 */
1715 last = flush == Z_FINISH && len == left + s->strm->avail_in ? 1 : 0;
1716 _tr_stored_block(s, (char *)0, 0L, last);
1717
1718 /* Replace the lengths in the dummy stored block with len. */
1719 s->pending_buf[s->pending - 4] = len;
1720 s->pending_buf[s->pending - 3] = len >> 8;
1721 s->pending_buf[s->pending - 2] = ~len;
1722 s->pending_buf[s->pending - 1] = ~len >> 8;
1723
1724 /* Write the stored block header bytes. */
1725 flush_pending(s->strm);
1726
1727 #ifdef ZLIB_DEBUG
1728 /* Update debugging counts for the data about to be copied. */
1729 s->compressed_len += len << 3;
1730 s->bits_sent += len << 3;
1731 #endif
1732
1733 /* Copy uncompressed bytes from the window to next_out. */
1734 if (left) {
1735 if (left > len)
1736 left = len;
1737 zmemcpy(s->strm->next_out, s->window + s->block_start, left);
1738 s->strm->next_out += left;
1739 s->strm->avail_out -= left;
1740 s->strm->total_out += left;
1741 s->block_start += left;
1742 len -= left;
1743 }
1744
1745 /* Copy uncompressed bytes directly from next_in to next_out, updating
1746 * the check value.
1747 */
1748 if (len) {
1749 read_buf(s->strm, s->strm->next_out, len);
1750 s->strm->next_out += len;
1751 s->strm->avail_out -= len;
1752 s->strm->total_out += len;
1753 }
1754 } while (last == 0);
1755
1756 /* Update the sliding window with the last s->w_size bytes of the copied
1757 * data, or append all of the copied data to the existing window if less
1758 * than s->w_size bytes were copied. Also update the number of bytes to
1759 * insert in the hash tables, in the event that deflateParams() switches to
1760 * a non-zero compression level.
1761 */
1762 used -= s->strm->avail_in; /* number of input bytes directly copied */
1763 if (used) {
1764 /* If any input was used, then no unused input remains in the window,
1765 * therefore s->block_start == s->strstart.
1766 */
1767 if (used >= s->w_size) { /* supplant the previous history */
1768 s->matches = 2; /* clear hash */
1769 zmemcpy(s->window, s->strm->next_in - s->w_size, s->w_size);
1770 s->strstart = s->w_size;
1771 }
1772 else {
1773 if (s->window_size - s->strstart <= used) {
1774 /* Slide the window down. */
1775 s->strstart -= s->w_size;
1776 zmemcpy(s->window, s->window + s->w_size, s->strstart);
1777 if (s->matches < 2)
1778 s->matches++; /* add a pending slide_hash() */
1779 }
1780 zmemcpy(s->window + s->strstart, s->strm->next_in - used, used);
1781 s->strstart += used;
1782 }
1783 s->block_start = s->strstart;
1784 s->insert += MIN(used, s->w_size - s->insert);
1785 }
1786 if (s->high_water < s->strstart)
1787 s->high_water = s->strstart;
1788
1789 /* If the last block was written to next_out, then done. */
1790 if (last)
1791 return finish_done;
1792
1793 /* If flushing and all input has been consumed, then done. */
1794 if (flush != Z_NO_FLUSH && flush != Z_FINISH &&
1795 s->strm->avail_in == 0 && (long)s->strstart == s->block_start)
1796 return block_done;
1797
1798 /* Fill the window with any remaining input. */
1799 have = s->window_size - s->strstart - 1;
1800 if (s->strm->avail_in > have && s->block_start >= (long)s->w_size) {
1801 /* Slide the window down. */
1802 s->block_start -= s->w_size;
1803 s->strstart -= s->w_size;
1804 zmemcpy(s->window, s->window + s->w_size, s->strstart);
1805 if (s->matches < 2)
1806 s->matches++; /* add a pending slide_hash() */
1807 have += s->w_size; /* more space now */
1808 }
1809 if (have > s->strm->avail_in)
1810 have = s->strm->avail_in;
1811 if (have) {
1812 read_buf(s->strm, s->window + s->strstart, have);
1813 s->strstart += have;
1814 }
1815 if (s->high_water < s->strstart)
1816 s->high_water = s->strstart;
1817
1818 /* There was not enough avail_out to write a complete worthy or flushed
1819 * stored block to next_out. Write a stored block to pending instead, if we
1820 * have enough input for a worthy block, or if flushing and there is enough
1821 * room for the remaining input as a stored block in the pending buffer.
1822 */
1823 have = (s->bi_valid + 42) >> 3; /* number of header bytes */
1824 /* maximum stored block length that will fit in pending: */
1825 have = MIN(s->pending_buf_size - have, MAX_STORED);
1826 min_block = MIN(have, s->w_size);
1827 left = s->strstart - s->block_start;
1828 if (left >= min_block ||
1829 ((left || flush == Z_FINISH) && flush != Z_NO_FLUSH &&
1830 s->strm->avail_in == 0 && left <= have)) {
1831 len = MIN(left, have);
1832 last = flush == Z_FINISH && s->strm->avail_in == 0 &&
1833 len == left ? 1 : 0;
1834 _tr_stored_block(s, (charf *)s->window + s->block_start, len, last);
1835 s->block_start += len;
1836 flush_pending(s->strm);
1837 }
1838
1839 /* We've done all we can with the available input and output. */
1840 return last ? finish_started : need_more;
1841 }
1842
1843 /* ===========================================================================
1844 * Compress as much as possible from the input stream, return the current
1845 * block state.
1846 * This function does not perform lazy evaluation of matches and inserts
1847 * new strings in the dictionary only for unmatched strings or for short
1848 * matches. It is used only for the fast compression options.
1849 */
1850 local block_state deflate_fast(s, flush)
1851 deflate_state *s;
1852 int flush;
1853 {
1854 IPos hash_head; /* head of the hash chain */
1855 int bflush; /* set if current block must be flushed */
1856
1857 for (;;) {
1858 /* Make sure that we always have enough lookahead, except
1859 * at the end of the input file. We need MAX_MATCH bytes
1860 * for the next match, plus MIN_MATCH bytes to insert the
1861 * string following the next match.
1862 */
1863 if (s->lookahead < MIN_LOOKAHEAD) {
1864 fill_window(s);
1865 if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) {
1866 return need_more;
1867 }
1868 if (s->lookahead == 0) break; /* flush the current block */
1869 }
1870
1871 /* Insert the string window[strstart .. strstart+2] in the
1872 * dictionary, and set hash_head to the head of the hash chain:
1873 */
1874 hash_head = NIL;
1875 if (s->lookahead >= MIN_MATCH) {
1876 INSERT_STRING(s, s->strstart, hash_head);
1877 }
1878
1879 /* Find the longest match, discarding those <= prev_length.
1880 * At this point we have always match_length < MIN_MATCH
1881 */
1882 if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) {
1883 /* To simplify the code, we prevent matches with the string
1884 * of window index 0 (in particular we have to avoid a match
1885 * of the string with itself at the start of the input file).
1886 */
1887 s->match_length = longest_match (s, hash_head);
1888 /* longest_match() sets match_start */
1889 }
1890 if (s->match_length >= MIN_MATCH) {
1891 check_match(s, s->strstart, s->match_start, s->match_length);
1892
1893 _tr_tally_dist(s, s->strstart - s->match_start,
1894 s->match_length - MIN_MATCH, bflush);
1895
1896 s->lookahead -= s->match_length;
1897
1898 /* Insert new strings in the hash table only if the match length
1899 * is not too large. This saves time but degrades compression.
1900 */
1901 #ifndef FASTEST
1902 if (s->match_length <= s->max_insert_length &&
1903 s->lookahead >= MIN_MATCH) {
1904 s->match_length--; /* string at strstart already in table */
1905 do {
1906 s->strstart++;
1907 INSERT_STRING(s, s->strstart, hash_head);
1908 /* strstart never exceeds WSIZE-MAX_MATCH, so there are
1909 * always MIN_MATCH bytes ahead.
1910 */
1911 } while (--s->match_length != 0);
1912 s->strstart++;
1913 } else
1914 #endif
1915 {
1916 s->strstart += s->match_length;
1917 s->match_length = 0;
1918 s->ins_h = s->window[s->strstart];
1919 UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
1920 #if MIN_MATCH != 3
1921 Call UPDATE_HASH() MIN_MATCH-3 more times
1922 #endif
1923 /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not
1924 * matter since it will be recomputed at next deflate call.
1925 */
1926 }
1927 } else {
1928 /* No match, output a literal byte */
1929 Tracevv((stderr,"%c", s->window[s->strstart]));
1930 _tr_tally_lit (s, s->window[s->strstart], bflush);
1931 s->lookahead--;
1932 s->strstart++;
1933 }
1934 if (bflush) FLUSH_BLOCK(s, 0);
1935 }
1936 s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1;
1937 if (flush == Z_FINISH) {
1938 FLUSH_BLOCK(s, 1);
1939 return finish_done;
1940 }
1941 if (s->last_lit)
1942 FLUSH_BLOCK(s, 0);
1943 return block_done;
1944 }
1945
1946 #ifndef FASTEST
1947 /* ===========================================================================
1948 * Same as above, but achieves better compression. We use a lazy
1949 * evaluation for matches: a match is finally adopted only if there is
1950 * no better match at the next window position.
1951 */
1952 local block_state deflate_slow(s, flush)
1953 deflate_state *s;
1954 int flush;
1955 {
1956 IPos hash_head; /* head of hash chain */
1957 int bflush; /* set if current block must be flushed */
1958
1959 /* Process the input block. */
1960 for (;;) {
1961 /* Make sure that we always have enough lookahead, except
1962 * at the end of the input file. We need MAX_MATCH bytes
1963 * for the next match, plus MIN_MATCH bytes to insert the
1964 * string following the next match.
1965 */
1966 if (s->lookahead < MIN_LOOKAHEAD) {
1967 fill_window(s);
1968 if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) {
1969 return need_more;
1970 }
1971 if (s->lookahead == 0) break; /* flush the current block */
1972 }
1973
1974 /* Insert the string window[strstart .. strstart+2] in the
1975 * dictionary, and set hash_head to the head of the hash chain:
1976 */
1977 hash_head = NIL;
1978 if (s->lookahead >= MIN_MATCH) {
1979 INSERT_STRING(s, s->strstart, hash_head);
1980 }
1981
1982 /* Find the longest match, discarding those <= prev_length.
1983 */
1984 s->prev_length = s->match_length, s->prev_match = s->match_start;
1985 s->match_length = MIN_MATCH-1;
1986
1987 if (hash_head != NIL && s->prev_length < s->max_lazy_match &&
1988 s->strstart - hash_head <= MAX_DIST(s)) {
1989 /* To simplify the code, we prevent matches with the string
1990 * of window index 0 (in particular we have to avoid a match
1991 * of the string with itself at the start of the input file).
1992 */
1993 s->match_length = longest_match (s, hash_head);
1994 /* longest_match() sets match_start */
1995
1996 if (s->match_length <= 5 && (s->strategy == Z_FILTERED
1997 #if TOO_FAR <= 32767
1998 || (s->match_length == MIN_MATCH &&
1999 s->strstart - s->match_start > TOO_FAR)
2000 #endif
2001 )) {
2002
2003 /* If prev_match is also MIN_MATCH, match_start is garbage
2004 * but we will ignore the current match anyway.
2005 */
2006 s->match_length = MIN_MATCH-1;
2007 }
2008 }
2009 /* If there was a match at the previous step and the current
2010 * match is not better, output the previous match:
2011 */
2012 if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) {
2013 uInt max_insert = s->strstart + s->lookahead - MIN_MATCH;
2014 /* Do not insert strings in hash table beyond this. */
2015
2016 check_match(s, s->strstart-1, s->prev_match, s->prev_length);
2017
2018 _tr_tally_dist(s, s->strstart -1 - s->prev_match,
2019 s->prev_length - MIN_MATCH, bflush);
2020
2021 /* Insert in hash table all strings up to the end of the match.
2022 * strstart-1 and strstart are already inserted. If there is not
2023 * enough lookahead, the last two strings are not inserted in
2024 * the hash table.
2025 */
2026 s->lookahead -= s->prev_length-1;
2027 s->prev_length -= 2;
2028 do {
2029 if (++s->strstart <= max_insert) {
2030 INSERT_STRING(s, s->strstart, hash_head);
2031 }
2032 } while (--s->prev_length != 0);
2033 s->match_available = 0;
2034 s->match_length = MIN_MATCH-1;
2035 s->strstart++;
2036
2037 if (bflush) FLUSH_BLOCK(s, 0);
2038
2039 } else if (s->match_available) {
2040 /* If there was no match at the previous position, output a
2041 * single literal. If there was a match but the current match
2042 * is longer, truncate the previous match to a single literal.
2043 */
2044 Tracevv((stderr,"%c", s->window[s->strstart-1]));
2045 _tr_tally_lit(s, s->window[s->strstart-1], bflush);
2046 if (bflush) {
2047 FLUSH_BLOCK_ONLY(s, 0);
2048 }
2049 s->strstart++;
2050 s->lookahead--;
2051 if (s->strm->avail_out == 0) return need_more;
2052 } else {
2053 /* There is no previous match to compare with, wait for
2054 * the next step to decide.
2055 */
2056 s->match_available = 1;
2057 s->strstart++;
2058 s->lookahead--;
2059 }
2060 }
2061 Assert (flush != Z_NO_FLUSH, "no flush?");
2062 if (s->match_available) {
2063 Tracevv((stderr,"%c", s->window[s->strstart-1]));
2064 _tr_tally_lit(s, s->window[s->strstart-1], bflush);
2065 s->match_available = 0;
2066 }
2067 s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1;
2068 if (flush == Z_FINISH) {
2069 FLUSH_BLOCK(s, 1);
2070 return finish_done;
2071 }
2072 if (s->last_lit)
2073 FLUSH_BLOCK(s, 0);
2074 return block_done;
2075 }
2076 #endif /* FASTEST */
2077
2078 /* ===========================================================================
2079 * For Z_RLE, simply look for runs of bytes, generate matches only of distance
2080 * one. Do not maintain a hash table. (It will be regenerated if this run of
2081 * deflate switches away from Z_RLE.)
2082 */
2083 local block_state deflate_rle(s, flush)
2084 deflate_state *s;
2085 int flush;
2086 {
2087 int bflush; /* set if current block must be flushed */
2088 uInt prev; /* byte at distance one to match */
2089 Bytef *scan, *strend; /* scan goes up to strend for length of run */
2090
2091 for (;;) {
2092 /* Make sure that we always have enough lookahead, except
2093 * at the end of the input file. We need MAX_MATCH bytes
2094 * for the longest run, plus one for the unrolled loop.
2095 */
2096 if (s->lookahead <= MAX_MATCH) {
2097 fill_window(s);
2098 if (s->lookahead <= MAX_MATCH && flush == Z_NO_FLUSH) {
2099 return need_more;
2100 }
2101 if (s->lookahead == 0) break; /* flush the current block */
2102 }
2103
2104 /* See how many times the previous byte repeats */
2105 s->match_length = 0;
2106 if (s->lookahead >= MIN_MATCH && s->strstart > 0) {
2107 scan = s->window + s->strstart - 1;
2108 prev = *scan;
2109 if (prev == *++scan && prev == *++scan && prev == *++scan) {
2110 strend = s->window + s->strstart + MAX_MATCH;
2111 do {
2112 } while (prev == *++scan && prev == *++scan &&
2113 prev == *++scan && prev == *++scan &&
2114 prev == *++scan && prev == *++scan &&
2115 prev == *++scan && prev == *++scan &&
2116 scan < strend);
2117 s->match_length = MAX_MATCH - (uInt)(strend - scan);
2118 if (s->match_length > s->lookahead)
2119 s->match_length = s->lookahead;
2120 }
2121 Assert(scan <= s->window+(uInt)(s->window_size-1), "wild scan");
2122 }
2123
2124 /* Emit match if have run of MIN_MATCH or longer, else emit literal */
2125 if (s->match_length >= MIN_MATCH) {
2126 check_match(s, s->strstart, s->strstart - 1, s->match_length);
2127
2128 _tr_tally_dist(s, 1, s->match_length - MIN_MATCH, bflush);
2129
2130 s->lookahead -= s->match_length;
2131 s->strstart += s->match_length;
2132 s->match_length = 0;
2133 } else {
2134 /* No match, output a literal byte */
2135 Tracevv((stderr,"%c", s->window[s->strstart]));
2136 _tr_tally_lit (s, s->window[s->strstart], bflush);
2137 s->lookahead--;
2138 s->strstart++;
2139 }
2140 if (bflush) FLUSH_BLOCK(s, 0);
2141 }
2142 s->insert = 0;
2143 if (flush == Z_FINISH) {
2144 FLUSH_BLOCK(s, 1);
2145 return finish_done;
2146 }
2147 if (s->last_lit)
2148 FLUSH_BLOCK(s, 0);
2149 return block_done;
2150 }
2151
2152 /* ===========================================================================
2153 * For Z_HUFFMAN_ONLY, do not look for matches. Do not maintain a hash table.
2154 * (It will be regenerated if this run of deflate switches away from Huffman.)
2155 */
2156 local block_state deflate_huff(s, flush)
2157 deflate_state *s;
2158 int flush;
2159 {
2160 int bflush; /* set if current block must be flushed */
2161
2162 for (;;) {
2163 /* Make sure that we have a literal to write. */
2164 if (s->lookahead == 0) {
2165 fill_window(s);
2166 if (s->lookahead == 0) {
2167 if (flush == Z_NO_FLUSH)
2168 return need_more;
2169 break; /* flush the current block */
2170 }
2171 }
2172
2173 /* Output a literal byte */
2174 s->match_length = 0;
2175 Tracevv((stderr,"%c", s->window[s->strstart]));
2176 _tr_tally_lit (s, s->window[s->strstart], bflush);
2177 s->lookahead--;
2178 s->strstart++;
2179 if (bflush) FLUSH_BLOCK(s, 0);
2180 }
2181 s->insert = 0;
2182 if (flush == Z_FINISH) {
2183 FLUSH_BLOCK(s, 1);
2184 return finish_done;
2185 }
2186 if (s->last_lit)
2187 FLUSH_BLOCK(s, 0);
2188 return block_done;
2189 }