1/* deflate.c -- compress data using the deflation algorithm 2 * Copyright (C) 1995-2024 Jean-loup Gailly and Mark Adler 3 * For conditions of distribution and use, see copyright notice in zlib.h 4 */ 5 6/* 7 * ALGORITHM 8 * 9 * The "deflation" process depends on being able to identify portions 10 * of the input text which are identical to earlier input (within a 11 * sliding window trailing behind the input currently being processed). 12 * 13 * The most straightforward technique turns out to be the fastest for 14 * most input files: try all possible matches and select the longest. 15 * The key feature of this algorithm is that insertions into the string 16 * dictionary are very simple and thus fast, and deletions are avoided 17 * completely. Insertions are performed at each input character, whereas 18 * string matches are performed only when the previous match ends. So it 19 * is preferable to spend more time in matches to allow very fast string 20 * insertions and avoid deletions. The matching algorithm for small 21 * strings is inspired from that of Rabin & Karp. A brute force approach 22 * is used to find longer strings when a small match has been found. 23 * A similar algorithm is used in comic (by Jan-Mark Wams) and freeze 24 * (by Leonid Broukhis). 25 * A previous version of this file used a more sophisticated algorithm 26 * (by Fiala and Greene) which is guaranteed to run in linear amortized 27 * time, but has a larger average cost, uses more memory and is patented. 28 * However the F&G algorithm may be faster for some highly redundant 29 * files if the parameter max_chain_length (described below) is too large. 30 * 31 * ACKNOWLEDGEMENTS 32 * 33 * The idea of lazy evaluation of matches is due to Jan-Mark Wams, and 34 * I found it in 'freeze' written by Leonid Broukhis. 35 * Thanks to many people for bug reports and testing. 36 * 37 * REFERENCES 38 * 39 * Deutsch, L.P.,"DEFLATE Compressed Data Format Specification". 40 * Available in http://tools.ietf.org/html/rfc1951 41 * 42 * A description of the Rabin and Karp algorithm is given in the book 43 * "Algorithms" by R. Sedgewick, Addison-Wesley, p252. 44 * 45 * Fiala,E.R., and Greene,D.H. 46 * Data Compression with Finite Windows, Comm.ACM, 32,4 (1989) 490-595 47 * 48 */ 49 50/* @(#) $Id$ */ 51 52#include "deflate.h" 53 54const char deflate_copyright[] = 55 " deflate 1.3.1 Copyright 1995-2024 Jean-loup Gailly and Mark Adler "; 56/* 57 If you use the zlib library in a product, an acknowledgment is welcome 58 in the documentation of your product. If for some reason you cannot 59 include such an acknowledgment, I would appreciate that you keep this 60 copyright string in the executable of your product. 61 */ 62 63typedef enum { 64 need_more, /* block not completed, need more input or more output */ 65 block_done, /* block flush performed */ 66 finish_started, /* finish started, need only more output at next deflate */ 67 finish_done /* finish done, accept no more input or output */ 68} block_state; 69 70typedef block_state (*compress_func)(deflate_state *s, int flush); 71/* Compression function. Returns the block state after the call. */ 72 73local block_state deflate_stored(deflate_state *s, int flush); 74local block_state deflate_fast(deflate_state *s, int flush); 75#ifndef FASTEST 76local block_state deflate_slow(deflate_state *s, int flush); 77#endif 78local block_state deflate_rle(deflate_state *s, int flush); 79local block_state deflate_huff(deflate_state *s, int flush); 80 81/* =========================================================================== 82 * Local data 83 */ 84 85#define NIL 0 86/* Tail of hash chains */ 87 88#ifndef TOO_FAR 89# define TOO_FAR 4096 90#endif 91/* Matches of length 3 are discarded if their distance exceeds TOO_FAR */ 92 93/* Values for max_lazy_match, good_match and max_chain_length, depending on 94 * the desired pack level (0..9). The values given below have been tuned to 95 * exclude worst case performance for pathological files. Better values may be 96 * found for specific files. 97 */ 98typedef struct config_s { 99 ush good_length; /* reduce lazy search above this match length */ 100 ush max_lazy; /* do not perform lazy search above this match length */ 101 ush nice_length; /* quit search above this match length */ 102 ush max_chain; 103 compress_func func; 104} config; 105 106#ifdef FASTEST 107local const config configuration_table[2] = { 108/* good lazy nice chain */ 109/* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */ 110/* 1 */ {4, 4, 8, 4, deflate_fast}}; /* max speed, no lazy matches */ 111#else 112local const config configuration_table[10] = { 113/* good lazy nice chain */ 114/* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */ 115/* 1 */ {4, 4, 8, 4, deflate_fast}, /* max speed, no lazy matches */ 116/* 2 */ {4, 5, 16, 8, deflate_fast}, 117/* 3 */ {4, 6, 32, 32, deflate_fast}, 118 119/* 4 */ {4, 4, 16, 16, deflate_slow}, /* lazy matches */ 120/* 5 */ {8, 16, 32, 32, deflate_slow}, 121/* 6 */ {8, 16, 128, 128, deflate_slow}, 122/* 7 */ {8, 32, 128, 256, deflate_slow}, 123/* 8 */ {32, 128, 258, 1024, deflate_slow}, 124/* 9 */ {32, 258, 258, 4096, deflate_slow}}; /* max compression */ 125#endif 126 127/* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4 128 * For deflate_fast() (levels <= 3) good is ignored and lazy has a different 129 * meaning. 130 */ 131 132/* rank Z_BLOCK between Z_NO_FLUSH and Z_PARTIAL_FLUSH */ 133#define RANK(f) (((f) * 2) - ((f) > 4 ? 9 : 0)) 134 135/* =========================================================================== 136 * Update a hash value with the given input byte 137 * IN assertion: all calls to UPDATE_HASH are made with consecutive input 138 * characters, so that a running hash key can be computed from the previous 139 * key instead of complete recalculation each time. 140 */ 141#define UPDATE_HASH(s,h,c) (h = (((h) << s->hash_shift) ^ (c)) & s->hash_mask) 142 143 144/* =========================================================================== 145 * Insert string str in the dictionary and set match_head to the previous head 146 * of the hash chain (the most recent string with same hash key). Return 147 * the previous length of the hash chain. 148 * If this file is compiled with -DFASTEST, the compression level is forced 149 * to 1, and no hash chains are maintained. 150 * IN assertion: all calls to INSERT_STRING are made with consecutive input 151 * characters and the first MIN_MATCH bytes of str are valid (except for 152 * the last MIN_MATCH-1 bytes of the input file). 153 */ 154#ifdef FASTEST 155#define INSERT_STRING(s, str, match_head) \ 156 (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ 157 match_head = s->head[s->ins_h], \ 158 s->head[s->ins_h] = (Pos)(str)) 159#else 160#define INSERT_STRING(s, str, match_head) \ 161 (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ 162 match_head = s->prev[(str) & s->w_mask] = s->head[s->ins_h], \ 163 s->head[s->ins_h] = (Pos)(str)) 164#endif 165 166/* =========================================================================== 167 * Initialize the hash table (avoiding 64K overflow for 16 bit systems). 168 * prev[] will be initialized on the fly. 169 */ 170#define CLEAR_HASH(s) \ 171 do { \ 172 s->head[s->hash_size - 1] = NIL; \ 173 zmemzero((Bytef *)s->head, \ 174 (unsigned)(s->hash_size - 1)*sizeof(*s->head)); \ 175 } while (0) 176 177/* =========================================================================== 178 * Slide the hash table when sliding the window down (could be avoided with 32 179 * bit values at the expense of memory usage). We slide even when level == 0 to 180 * keep the hash table consistent if we switch back to level > 0 later. 181 */ 182#if defined(__has_feature) 183# if __has_feature(memory_sanitizer) 184 __attribute__((no_sanitize("memory"))) 185# endif 186#endif 187local void slide_hash(deflate_state *s) { 188 unsigned n, m; 189 Posf *p; 190 uInt wsize = s->w_size; 191 192 n = s->hash_size; 193 p = &s->head[n]; 194 do { 195 m = *--p; 196 *p = (Pos)(m >= wsize ? m - wsize : NIL); 197 } while (--n); 198 n = wsize; 199#ifndef FASTEST 200 p = &s->prev[n]; 201 do { 202 m = *--p; 203 *p = (Pos)(m >= wsize ? m - wsize : NIL); 204 /* If n is not on any hash chain, prev[n] is garbage but 205 * its value will never be used. 206 */ 207 } while (--n); 208#endif 209} 210 211/* =========================================================================== 212 * Read a new buffer from the current input stream, update the adler32 213 * and total number of bytes read. All deflate() input goes through 214 * this function so some applications may wish to modify it to avoid 215 * allocating a large strm->next_in buffer and copying from it. 216 * (See also flush_pending()). 217 */ 218local unsigned read_buf(z_streamp strm, Bytef *buf, unsigned size) { 219 unsigned len = strm->avail_in; 220 221 if (len > size) len = size; 222 if (len == 0) return 0; 223 224 strm->avail_in -= len; 225 226 zmemcpy(buf, strm->next_in, len); 227 if (strm->state->wrap == 1) { 228 strm->adler = adler32(strm->adler, buf, len); 229 } 230#ifdef GZIP 231 else if (strm->state->wrap == 2) { 232 strm->adler = crc32(strm->adler, buf, len); 233 } 234#endif 235 strm->next_in += len; 236 strm->total_in += len; 237 238 return len; 239} 240 241/* =========================================================================== 242 * Fill the window when the lookahead becomes insufficient. 243 * Updates strstart and lookahead. 244 * 245 * IN assertion: lookahead < MIN_LOOKAHEAD 246 * OUT assertions: strstart <= window_size-MIN_LOOKAHEAD 247 * At least one byte has been read, or avail_in == 0; reads are 248 * performed for at least two bytes (required for the zip translate_eol 249 * option -- not supported here). 250 */ 251local void fill_window(deflate_state *s) { 252 unsigned n; 253 unsigned more; /* Amount of free space at the end of the window. */ 254 uInt wsize = s->w_size; 255 256 Assert(s->lookahead < MIN_LOOKAHEAD, "already enough lookahead"); 257 258 do { 259 more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart); 260 261 /* Deal with !@#$% 64K limit: */ 262 if (sizeof(int) <= 2) { 263 if (more == 0 && s->strstart == 0 && s->lookahead == 0) { 264 more = wsize; 265 266 } else if (more == (unsigned)(-1)) { 267 /* Very unlikely, but possible on 16 bit machine if 268 * strstart == 0 && lookahead == 1 (input done a byte at time) 269 */ 270 more--; 271 } 272 } 273 274 /* If the window is almost full and there is insufficient lookahead, 275 * move the upper half to the lower one to make room in the upper half. 276 */ 277 if (s->strstart >= wsize + MAX_DIST(s)) { 278 279 zmemcpy(s->window, s->window + wsize, (unsigned)wsize - more); 280 s->match_start -= wsize; 281 s->strstart -= wsize; /* we now have strstart >= MAX_DIST */ 282 s->block_start -= (long) wsize; 283 if (s->insert > s->strstart) 284 s->insert = s->strstart; 285 slide_hash(s); 286 more += wsize; 287 } 288 if (s->strm->avail_in == 0) break; 289 290 /* If there was no sliding: 291 * strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 && 292 * more == window_size - lookahead - strstart 293 * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1) 294 * => more >= window_size - 2*WSIZE + 2 295 * In the BIG_MEM or MMAP case (not yet supported), 296 * window_size == input_size + MIN_LOOKAHEAD && 297 * strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD. 298 * Otherwise, window_size == 2*WSIZE so more >= 2. 299 * If there was sliding, more >= WSIZE. So in all cases, more >= 2. 300 */ 301 Assert(more >= 2, "more < 2"); 302 303 n = read_buf(s->strm, s->window + s->strstart + s->lookahead, more); 304 s->lookahead += n; 305 306 /* Initialize the hash value now that we have some input: */ 307 if (s->lookahead + s->insert >= MIN_MATCH) { 308 uInt str = s->strstart - s->insert; 309 s->ins_h = s->window[str]; 310 UPDATE_HASH(s, s->ins_h, s->window[str + 1]); 311#if MIN_MATCH != 3 312 Call UPDATE_HASH() MIN_MATCH-3 more times 313#endif 314 while (s->insert) { 315 UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]); 316#ifndef FASTEST 317 s->prev[str & s->w_mask] = s->head[s->ins_h]; 318#endif 319 s->head[s->ins_h] = (Pos)str; 320 str++; 321 s->insert--; 322 if (s->lookahead + s->insert < MIN_MATCH) 323 break; 324 } 325 } 326 /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage, 327 * but this is not important since only literal bytes will be emitted. 328 */ 329 330 } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0); 331 332 /* If the WIN_INIT bytes after the end of the current data have never been 333 * written, then zero those bytes in order to avoid memory check reports of 334 * the use of uninitialized (or uninitialised as Julian writes) bytes by 335 * the longest match routines. Update the high water mark for the next 336 * time through here. WIN_INIT is set to MAX_MATCH since the longest match 337 * routines allow scanning to strstart + MAX_MATCH, ignoring lookahead. 338 */ 339 if (s->high_water < s->window_size) { 340 ulg curr = s->strstart + (ulg)(s->lookahead); 341 ulg init; 342 343 if (s->high_water < curr) { 344 /* Previous high water mark below current data -- zero WIN_INIT 345 * bytes or up to end of window, whichever is less. 346 */ 347 init = s->window_size - curr; 348 if (init > WIN_INIT) 349 init = WIN_INIT; 350 zmemzero(s->window + curr, (unsigned)init); 351 s->high_water = curr + init; 352 } 353 else if (s->high_water < (ulg)curr + WIN_INIT) { 354 /* High water mark at or above current data, but below current data 355 * plus WIN_INIT -- zero out to current data plus WIN_INIT, or up 356 * to end of window, whichever is less. 357 */ 358 init = (ulg)curr + WIN_INIT - s->high_water; 359 if (init > s->window_size - s->high_water) 360 init = s->window_size - s->high_water; 361 zmemzero(s->window + s->high_water, (unsigned)init); 362 s->high_water += init; 363 } 364 } 365 366 Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD, 367 "not enough room for search"); 368} 369 370/* ========================================================================= */ 371int ZEXPORT deflateInit_(z_streamp strm, int level, const char *version, 372 int stream_size) { 373 return deflateInit2_(strm, level, Z_DEFLATED, MAX_WBITS, DEF_MEM_LEVEL, 374 Z_DEFAULT_STRATEGY, version, stream_size); 375 /* To do: ignore strm->next_in if we use it as window */ 376} 377 378/* ========================================================================= */ 379int ZEXPORT deflateInit2_(z_streamp strm, int level, int method, 380 int windowBits, int memLevel, int strategy, 381 const char *version, int stream_size) { 382 deflate_state *s; 383 int wrap = 1; 384 static const char my_version[] = ZLIB_VERSION; 385 386 if (version == Z_NULL || version[0] != my_version[0] || 387 stream_size != sizeof(z_stream)) { 388 return Z_VERSION_ERROR; 389 } 390 if (strm == Z_NULL) return Z_STREAM_ERROR; 391 392 strm->msg = Z_NULL; 393 if (strm->zalloc == (alloc_func)0) { 394#if defined(Z_SOLO) && !defined(_KERNEL) 395 return Z_STREAM_ERROR; 396#else 397 strm->zalloc = zcalloc; 398 strm->opaque = (voidpf)0; 399#endif 400 } 401 if (strm->zfree == (free_func)0) 402#if defined(Z_SOLO) && !defined(_KERNEL) 403 return Z_STREAM_ERROR; 404#else 405 strm->zfree = zcfree; 406#endif 407 408#ifdef FASTEST 409 if (level != 0) level = 1; 410#else 411 if (level == Z_DEFAULT_COMPRESSION) level = 6; 412#endif 413 414 if (windowBits < 0) { /* suppress zlib wrapper */ 415 wrap = 0; 416 if (windowBits < -15) 417 return Z_STREAM_ERROR; 418 windowBits = -windowBits; 419 } 420#ifdef GZIP 421 else if (windowBits > 15) { 422 wrap = 2; /* write gzip wrapper instead */ 423 windowBits -= 16; 424 } 425#endif 426 if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED || 427 windowBits < 8 || windowBits > 15 || level < 0 || level > 9 || 428 strategy < 0 || strategy > Z_FIXED || (windowBits == 8 && wrap != 1)) { 429 return Z_STREAM_ERROR; 430 } 431 if (windowBits == 8) windowBits = 9; /* until 256-byte window bug fixed */ 432 s = (deflate_state *) ZALLOC(strm, 1, sizeof(deflate_state)); 433 if (s == Z_NULL) return Z_MEM_ERROR; 434 strm->state = (struct internal_state FAR *)s; 435 s->strm = strm; 436 s->status = INIT_STATE; /* to pass state test in deflateReset() */ 437 438 s->wrap = wrap; 439 s->gzhead = Z_NULL; 440 s->w_bits = (uInt)windowBits; 441 s->w_size = 1 << s->w_bits; 442 s->w_mask = s->w_size - 1; 443 444 s->hash_bits = (uInt)memLevel + 7; 445 s->hash_size = 1 << s->hash_bits; 446 s->hash_mask = s->hash_size - 1; 447 s->hash_shift = ((s->hash_bits + MIN_MATCH-1) / MIN_MATCH); 448 449 s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte)); 450 s->prev = (Posf *) ZALLOC(strm, s->w_size, sizeof(Pos)); 451 s->head = (Posf *) ZALLOC(strm, s->hash_size, sizeof(Pos)); 452 453 s->high_water = 0; /* nothing written to s->window yet */ 454 455 s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */ 456 457 /* We overlay pending_buf and sym_buf. This works since the average size 458 * for length/distance pairs over any compressed block is assured to be 31 459 * bits or less. 460 * 461 * Analysis: The longest fixed codes are a length code of 8 bits plus 5 462 * extra bits, for lengths 131 to 257. The longest fixed distance codes are 463 * 5 bits plus 13 extra bits, for distances 16385 to 32768. The longest 464 * possible fixed-codes length/distance pair is then 31 bits total. 465 * 466 * sym_buf starts one-fourth of the way into pending_buf. So there are 467 * three bytes in sym_buf for every four bytes in pending_buf. Each symbol 468 * in sym_buf is three bytes -- two for the distance and one for the 469 * literal/length. As each symbol is consumed, the pointer to the next 470 * sym_buf value to read moves forward three bytes. From that symbol, up to 471 * 31 bits are written to pending_buf. The closest the written pending_buf 472 * bits gets to the next sym_buf symbol to read is just before the last 473 * code is written. At that time, 31*(n - 2) bits have been written, just 474 * after 24*(n - 2) bits have been consumed from sym_buf. sym_buf starts at 475 * 8*n bits into pending_buf. (Note that the symbol buffer fills when n - 1 476 * symbols are written.) The closest the writing gets to what is unread is 477 * then n + 14 bits. Here n is lit_bufsize, which is 16384 by default, and 478 * can range from 128 to 32768. 479 * 480 * Therefore, at a minimum, there are 142 bits of space between what is 481 * written and what is read in the overlain buffers, so the symbols cannot 482 * be overwritten by the compressed data. That space is actually 139 bits, 483 * due to the three-bit fixed-code block header. 484 * 485 * That covers the case where either Z_FIXED is specified, forcing fixed 486 * codes, or when the use of fixed codes is chosen, because that choice 487 * results in a smaller compressed block than dynamic codes. That latter 488 * condition then assures that the above analysis also covers all dynamic 489 * blocks. A dynamic-code block will only be chosen to be emitted if it has 490 * fewer bits than a fixed-code block would for the same set of symbols. 491 * Therefore its average symbol length is assured to be less than 31. So 492 * the compressed data for a dynamic block also cannot overwrite the 493 * symbols from which it is being constructed. 494 */ 495 496 s->pending_buf = (uchf *) ZALLOC(strm, s->lit_bufsize, LIT_BUFS); 497 s->pending_buf_size = (ulg)s->lit_bufsize * 4; 498 499 if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL || 500 s->pending_buf == Z_NULL) { 501 s->status = FINISH_STATE; 502 strm->msg = ERR_MSG(Z_MEM_ERROR); 503 deflateEnd (strm); 504 return Z_MEM_ERROR; 505 } 506#ifdef LIT_MEM 507 s->d_buf = (ushf *)(s->pending_buf + (s->lit_bufsize << 1)); 508 s->l_buf = s->pending_buf + (s->lit_bufsize << 2); 509 s->sym_end = s->lit_bufsize - 1; 510#else 511 s->sym_buf = s->pending_buf + s->lit_bufsize; 512 s->sym_end = (s->lit_bufsize - 1) * 3; 513#endif 514 /* We avoid equality with lit_bufsize*3 because of wraparound at 64K 515 * on 16 bit machines and because stored blocks are restricted to 516 * 64K-1 bytes. 517 */ 518 519 s->level = level; 520 s->strategy = strategy; 521 s->method = (Byte)method; 522 523 return deflateReset(strm); 524} 525 526/* ========================================================================= 527 * Check for a valid deflate stream state. Return 0 if ok, 1 if not. 528 */ 529local int deflateStateCheck(z_streamp strm) { 530 deflate_state *s; 531 if (strm == Z_NULL || 532 strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0) 533 return 1; 534 s = strm->state; 535 if (s == Z_NULL || s->strm != strm || (s->status != INIT_STATE && 536#ifdef GZIP 537 s->status != GZIP_STATE && 538#endif 539 s->status != EXTRA_STATE && 540 s->status != NAME_STATE && 541 s->status != COMMENT_STATE && 542 s->status != HCRC_STATE && 543 s->status != BUSY_STATE && 544 s->status != FINISH_STATE)) 545 return 1; 546 return 0; 547} 548 549/* ========================================================================= */ 550int ZEXPORT deflateSetDictionary(z_streamp strm, const Bytef *dictionary, 551 uInt dictLength) { 552 deflate_state *s; 553 uInt str, n; 554 int wrap; 555 unsigned avail; 556 z_const unsigned char *next; 557 558 if (deflateStateCheck(strm) || dictionary == Z_NULL) 559 return Z_STREAM_ERROR; 560 s = strm->state; 561 wrap = s->wrap; 562 if (wrap == 2 || (wrap == 1 && s->status != INIT_STATE) || s->lookahead) 563 return Z_STREAM_ERROR; 564 565 /* when using zlib wrappers, compute Adler-32 for provided dictionary */ 566 if (wrap == 1) 567 strm->adler = adler32(strm->adler, dictionary, dictLength); 568 s->wrap = 0; /* avoid computing Adler-32 in read_buf */ 569 570 /* if dictionary would fill window, just replace the history */ 571 if (dictLength >= s->w_size) { 572 if (wrap == 0) { /* already empty otherwise */ 573 CLEAR_HASH(s); 574 s->strstart = 0; 575 s->block_start = 0L; 576 s->insert = 0; 577 } 578 dictionary += dictLength - s->w_size; /* use the tail */ 579 dictLength = s->w_size; 580 } 581 582 /* insert dictionary into window and hash */ 583 avail = strm->avail_in; 584 next = strm->next_in; 585 strm->avail_in = dictLength; 586 strm->next_in = (z_const Bytef *)dictionary; 587 fill_window(s); 588 while (s->lookahead >= MIN_MATCH) { 589 str = s->strstart; 590 n = s->lookahead - (MIN_MATCH-1); 591 do { 592 UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]); 593#ifndef FASTEST 594 s->prev[str & s->w_mask] = s->head[s->ins_h]; 595#endif 596 s->head[s->ins_h] = (Pos)str; 597 str++; 598 } while (--n); 599 s->strstart = str; 600 s->lookahead = MIN_MATCH-1; 601 fill_window(s); 602 } 603 s->strstart += s->lookahead; 604 s->block_start = (long)s->strstart; 605 s->insert = s->lookahead; 606 s->lookahead = 0; 607 s->match_length = s->prev_length = MIN_MATCH-1; 608 s->match_available = 0; 609 strm->next_in = next; 610 strm->avail_in = avail; 611 s->wrap = wrap; 612 return Z_OK; 613} 614 615/* ========================================================================= */ 616int ZEXPORT deflateGetDictionary(z_streamp strm, Bytef *dictionary, 617 uInt *dictLength) { 618 deflate_state *s; 619 uInt len; 620 621 if (deflateStateCheck(strm)) 622 return Z_STREAM_ERROR; 623 s = strm->state; 624 len = s->strstart + s->lookahead; 625 if (len > s->w_size) 626 len = s->w_size; 627 if (dictionary != Z_NULL && len) 628 zmemcpy(dictionary, s->window + s->strstart + s->lookahead - len, len); 629 if (dictLength != Z_NULL) 630 *dictLength = len; 631 return Z_OK; 632} 633 634/* ========================================================================= */ 635int ZEXPORT deflateResetKeep(z_streamp strm) { 636 deflate_state *s; 637 638 if (deflateStateCheck(strm)) { 639 return Z_STREAM_ERROR; 640 } 641 642 strm->total_in = strm->total_out = 0; 643 strm->msg = Z_NULL; /* use zfree if we ever allocate msg dynamically */ 644 strm->data_type = Z_UNKNOWN; 645 646 s = (deflate_state *)strm->state; 647 s->pending = 0; 648 s->pending_out = s->pending_buf; 649 650 if (s->wrap < 0) { 651 s->wrap = -s->wrap; /* was made negative by deflate(..., Z_FINISH); */ 652 } 653 s->status = 654#ifdef GZIP 655 s->wrap == 2 ? GZIP_STATE : 656#endif 657 INIT_STATE; 658 strm->adler = 659#ifdef GZIP 660 s->wrap == 2 ? crc32(0L, Z_NULL, 0) : 661#endif 662 adler32(0L, Z_NULL, 0); 663 s->last_flush = -2; 664 665 _tr_init(s); 666 667 return Z_OK; 668} 669 670/* =========================================================================== 671 * Initialize the "longest match" routines for a new zlib stream 672 */ 673local void lm_init(deflate_state *s) { 674 s->window_size = (ulg)2L*s->w_size; 675 676 CLEAR_HASH(s); 677 678 /* Set the default configuration parameters: 679 */ 680 s->max_lazy_match = configuration_table[s->level].max_lazy; 681 s->good_match = configuration_table[s->level].good_length; 682 s->nice_match = configuration_table[s->level].nice_length; 683 s->max_chain_length = configuration_table[s->level].max_chain; 684 685 s->strstart = 0; 686 s->block_start = 0L; 687 s->lookahead = 0; 688 s->insert = 0; 689 s->match_length = s->prev_length = MIN_MATCH-1; 690 s->match_available = 0; 691 s->ins_h = 0; 692} 693 694/* ========================================================================= */ 695int ZEXPORT deflateReset(z_streamp strm) { 696 int ret; 697 698 ret = deflateResetKeep(strm); 699 if (ret == Z_OK) 700 lm_init(strm->state); 701 return ret; 702} 703 704/* ========================================================================= */ 705int ZEXPORT deflateSetHeader(z_streamp strm, gz_headerp head) { 706 if (deflateStateCheck(strm) || strm->state->wrap != 2) 707 return Z_STREAM_ERROR; 708 strm->state->gzhead = head; 709 return Z_OK; 710} 711 712/* ========================================================================= */ 713int ZEXPORT deflatePending(z_streamp strm, unsigned *pending, int *bits) { 714 if (deflateStateCheck(strm)) return Z_STREAM_ERROR; 715 if (pending != Z_NULL) 716 *pending = strm->state->pending; 717 if (bits != Z_NULL) 718 *bits = strm->state->bi_valid; 719 return Z_OK; 720} 721 722/* ========================================================================= */ 723int ZEXPORT deflatePrime(z_streamp strm, int bits, int value) { 724 deflate_state *s; 725 int put; 726 727 if (deflateStateCheck(strm)) return Z_STREAM_ERROR; 728 s = strm->state; 729#ifdef LIT_MEM 730 if (bits < 0 || bits > 16 || 731 (uchf *)s->d_buf < s->pending_out + ((Buf_size + 7) >> 3)) 732 return Z_BUF_ERROR; 733#else 734 if (bits < 0 || bits > 16 || 735 s->sym_buf < s->pending_out + ((Buf_size + 7) >> 3)) 736 return Z_BUF_ERROR; 737#endif 738 do { 739 put = Buf_size - s->bi_valid; 740 if (put > bits) 741 put = bits; 742 s->bi_buf |= (ush)((value & ((1 << put) - 1)) << s->bi_valid); 743 s->bi_valid += put; 744 _tr_flush_bits(s); 745 value >>= put; 746 bits -= put; 747 } while (bits); 748 return Z_OK; 749} 750 751/* ========================================================================= */ 752int ZEXPORT deflateParams(z_streamp strm, int level, int strategy) { 753 deflate_state *s; 754 compress_func func; 755 756 if (deflateStateCheck(strm)) return Z_STREAM_ERROR; 757 s = strm->state; 758 759#ifdef FASTEST 760 if (level != 0) level = 1; 761#else 762 if (level == Z_DEFAULT_COMPRESSION) level = 6; 763#endif 764 if (level < 0 || level > 9 || strategy < 0 || strategy > Z_FIXED) { 765 return Z_STREAM_ERROR; 766 } 767 func = configuration_table[s->level].func; 768 769 if ((strategy != s->strategy || func != configuration_table[level].func) && 770 s->last_flush != -2) { 771 /* Flush the last buffer: */ 772 int err = deflate(strm, Z_BLOCK); 773 if (err == Z_STREAM_ERROR) 774 return err; 775 if (strm->avail_in || (s->strstart - s->block_start) + s->lookahead) 776 return Z_BUF_ERROR; 777 } 778 if (s->level != level) { 779 if (s->level == 0 && s->matches != 0) { 780 if (s->matches == 1) 781 slide_hash(s); 782 else 783 CLEAR_HASH(s); 784 s->matches = 0; 785 } 786 s->level = level; 787 s->max_lazy_match = configuration_table[level].max_lazy; 788 s->good_match = configuration_table[level].good_length; 789 s->nice_match = configuration_table[level].nice_length; 790 s->max_chain_length = configuration_table[level].max_chain; 791 } 792 s->strategy = strategy; 793 return Z_OK; 794} 795 796/* ========================================================================= */ 797int ZEXPORT deflateTune(z_streamp strm, int good_length, int max_lazy, 798 int nice_length, int max_chain) { 799 deflate_state *s; 800 801 if (deflateStateCheck(strm)) return Z_STREAM_ERROR; 802 s = strm->state; 803 s->good_match = (uInt)good_length; 804 s->max_lazy_match = (uInt)max_lazy; 805 s->nice_match = nice_length; 806 s->max_chain_length = (uInt)max_chain; 807 return Z_OK; 808} 809 810/* ========================================================================= 811 * For the default windowBits of 15 and memLevel of 8, this function returns a 812 * close to exact, as well as small, upper bound on the compressed size. This 813 * is an expansion of ~0.03%, plus a small constant. 814 * 815 * For any setting other than those defaults for windowBits and memLevel, one 816 * of two worst case bounds is returned. This is at most an expansion of ~4% or 817 * ~13%, plus a small constant. 818 * 819 * Both the 0.03% and 4% derive from the overhead of stored blocks. The first 820 * one is for stored blocks of 16383 bytes (memLevel == 8), whereas the second 821 * is for stored blocks of 127 bytes (the worst case memLevel == 1). The 822 * expansion results from five bytes of header for each stored block. 823 * 824 * The larger expansion of 13% results from a window size less than or equal to 825 * the symbols buffer size (windowBits <= memLevel + 7). In that case some of 826 * the data being compressed may have slid out of the sliding window, impeding 827 * a stored block from being emitted. Then the only choice is a fixed or 828 * dynamic block, where a fixed block limits the maximum expansion to 9 bits 829 * per 8-bit byte, plus 10 bits for every block. The smallest block size for 830 * which this can occur is 255 (memLevel == 2). 831 * 832 * Shifts are used to approximate divisions, for speed. 833 */ 834uLong ZEXPORT deflateBound(z_streamp strm, uLong sourceLen) { 835 deflate_state *s; 836 uLong fixedlen, storelen, wraplen; 837 838 /* upper bound for fixed blocks with 9-bit literals and length 255 839 (memLevel == 2, which is the lowest that may not use stored blocks) -- 840 ~13% overhead plus a small constant */ 841 fixedlen = sourceLen + (sourceLen >> 3) + (sourceLen >> 8) + 842 (sourceLen >> 9) + 4; 843 844 /* upper bound for stored blocks with length 127 (memLevel == 1) -- 845 ~4% overhead plus a small constant */ 846 storelen = sourceLen + (sourceLen >> 5) + (sourceLen >> 7) + 847 (sourceLen >> 11) + 7; 848 849 /* if can't get parameters, return larger bound plus a zlib wrapper */ 850 if (deflateStateCheck(strm)) 851 return (fixedlen > storelen ? fixedlen : storelen) + 6; 852 853 /* compute wrapper length */ 854 s = strm->state; 855 switch (s->wrap) { 856 case 0: /* raw deflate */ 857 wraplen = 0; 858 break; 859 case 1: /* zlib wrapper */ 860 wraplen = 6 + (s->strstart ? 4 : 0); 861 break; 862#ifdef GZIP 863 case 2: /* gzip wrapper */ 864 wraplen = 18; 865 if (s->gzhead != Z_NULL) { /* user-supplied gzip header */ 866 Bytef *str; 867 if (s->gzhead->extra != Z_NULL) 868 wraplen += 2 + s->gzhead->extra_len; 869 str = s->gzhead->name; 870 if (str != Z_NULL) 871 do { 872 wraplen++; 873 } while (*str++); 874 str = s->gzhead->comment; 875 if (str != Z_NULL) 876 do { 877 wraplen++; 878 } while (*str++); 879 if (s->gzhead->hcrc) 880 wraplen += 2; 881 } 882 break; 883#endif 884 default: /* for compiler happiness */ 885 wraplen = 6; 886 } 887 888 /* if not default parameters, return one of the conservative bounds */ 889 if (s->w_bits != 15 || s->hash_bits != 8 + 7) 890 return (s->w_bits <= s->hash_bits && s->level ? fixedlen : storelen) + 891 wraplen; 892 893 /* default settings: return tight bound for that case -- ~0.03% overhead 894 plus a small constant */ 895 return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) + 896 (sourceLen >> 25) + 13 - 6 + wraplen; 897} 898 899/* ========================================================================= 900 * Put a short in the pending buffer. The 16-bit value is put in MSB order. 901 * IN assertion: the stream state is correct and there is enough room in 902 * pending_buf. 903 */ 904local void putShortMSB(deflate_state *s, uInt b) { 905 put_byte(s, (Byte)(b >> 8)); 906 put_byte(s, (Byte)(b & 0xff)); 907} 908 909/* ========================================================================= 910 * Flush as much pending output as possible. All deflate() output, except for 911 * some deflate_stored() output, goes through this function so some 912 * applications may wish to modify it to avoid allocating a large 913 * strm->next_out buffer and copying into it. (See also read_buf()). 914 */ 915local void flush_pending(z_streamp strm) { 916 unsigned len; 917 deflate_state *s = strm->state; 918 919 _tr_flush_bits(s); 920 len = s->pending; 921 if (len > strm->avail_out) len = strm->avail_out; 922 if (len == 0) return; 923 924 zmemcpy(strm->next_out, s->pending_out, len); 925 strm->next_out += len; 926 s->pending_out += len; 927 strm->total_out += len; 928 strm->avail_out -= len; 929 s->pending -= len; 930 if (s->pending == 0) { 931 s->pending_out = s->pending_buf; 932 } 933} 934 935/* =========================================================================== 936 * Update the header CRC with the bytes s->pending_buf[beg..s->pending - 1]. 937 */ 938#define HCRC_UPDATE(beg) \ 939 do { \ 940 if (s->gzhead->hcrc && s->pending > (beg)) \ 941 strm->adler = crc32(strm->adler, s->pending_buf + (beg), \ 942 s->pending - (beg)); \ 943 } while (0) 944 945/* ========================================================================= */ 946int ZEXPORT deflate(z_streamp strm, int flush) { 947 int old_flush; /* value of flush param for previous deflate call */ 948 deflate_state *s; 949 950 if (deflateStateCheck(strm) || flush > Z_BLOCK || flush < 0) { 951 return Z_STREAM_ERROR; 952 } 953 s = strm->state; 954 955 if (strm->next_out == Z_NULL || 956 (strm->avail_in != 0 && strm->next_in == Z_NULL) || 957 (s->status == FINISH_STATE && flush != Z_FINISH)) { 958 ERR_RETURN(strm, Z_STREAM_ERROR); 959 } 960 if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR); 961 962 old_flush = s->last_flush; 963 s->last_flush = flush; 964 965 /* Flush as much pending output as possible */ 966 if (s->pending != 0) { 967 flush_pending(strm); 968 if (strm->avail_out == 0) { 969 /* Since avail_out is 0, deflate will be called again with 970 * more output space, but possibly with both pending and 971 * avail_in equal to zero. There won't be anything to do, 972 * but this is not an error situation so make sure we 973 * return OK instead of BUF_ERROR at next call of deflate: 974 */ 975 s->last_flush = -1; 976 return Z_OK; 977 } 978 979 /* Make sure there is something to do and avoid duplicate consecutive 980 * flushes. For repeated and useless calls with Z_FINISH, we keep 981 * returning Z_STREAM_END instead of Z_BUF_ERROR. 982 */ 983 } else if (strm->avail_in == 0 && RANK(flush) <= RANK(old_flush) && 984 flush != Z_FINISH) { 985 ERR_RETURN(strm, Z_BUF_ERROR); 986 } 987 988 /* User must not provide more input after the first FINISH: */ 989 if (s->status == FINISH_STATE && strm->avail_in != 0) { 990 ERR_RETURN(strm, Z_BUF_ERROR); 991 } 992 993 /* Write the header */ 994 if (s->status == INIT_STATE && s->wrap == 0) 995 s->status = BUSY_STATE; 996 if (s->status == INIT_STATE) { 997 /* zlib header */ 998 uInt header = (Z_DEFLATED + ((s->w_bits - 8) << 4)) << 8; 999 uInt level_flags; 1000 1001 if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2) 1002 level_flags = 0; 1003 else if (s->level < 6) 1004 level_flags = 1; 1005 else if (s->level == 6) 1006 level_flags = 2; 1007 else 1008 level_flags = 3; 1009 header |= (level_flags << 6); 1010 if (s->strstart != 0) header |= PRESET_DICT; 1011 header += 31 - (header % 31); 1012 1013 putShortMSB(s, header); 1014 1015 /* Save the adler32 of the preset dictionary: */ 1016 if (s->strstart != 0) { 1017 putShortMSB(s, (uInt)(strm->adler >> 16)); 1018 putShortMSB(s, (uInt)(strm->adler & 0xffff)); 1019 } 1020 strm->adler = adler32(0L, Z_NULL, 0); 1021 s->status = BUSY_STATE; 1022 1023 /* Compression must start with an empty pending buffer */ 1024 flush_pending(strm); 1025 if (s->pending != 0) { 1026 s->last_flush = -1; 1027 return Z_OK; 1028 } 1029 } 1030#ifdef GZIP 1031 if (s->status == GZIP_STATE) { 1032 /* gzip header */ 1033 strm->adler = crc32(0L, Z_NULL, 0); 1034 put_byte(s, 31); 1035 put_byte(s, 139); 1036 put_byte(s, 8); 1037 if (s->gzhead == Z_NULL) { 1038 put_byte(s, 0); 1039 put_byte(s, 0); 1040 put_byte(s, 0); 1041 put_byte(s, 0); 1042 put_byte(s, 0); 1043 put_byte(s, s->level == 9 ? 2 : 1044 (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ? 1045 4 : 0)); 1046 put_byte(s, OS_CODE); 1047 s->status = BUSY_STATE; 1048 1049 /* Compression must start with an empty pending buffer */ 1050 flush_pending(strm); 1051 if (s->pending != 0) { 1052 s->last_flush = -1; 1053 return Z_OK; 1054 } 1055 } 1056 else { 1057 put_byte(s, (s->gzhead->text ? 1 : 0) + 1058 (s->gzhead->hcrc ? 2 : 0) + 1059 (s->gzhead->extra == Z_NULL ? 0 : 4) + 1060 (s->gzhead->name == Z_NULL ? 0 : 8) + 1061 (s->gzhead->comment == Z_NULL ? 0 : 16) 1062 ); 1063 put_byte(s, (Byte)(s->gzhead->time & 0xff)); 1064 put_byte(s, (Byte)((s->gzhead->time >> 8) & 0xff)); 1065 put_byte(s, (Byte)((s->gzhead->time >> 16) & 0xff)); 1066 put_byte(s, (Byte)((s->gzhead->time >> 24) & 0xff)); 1067 put_byte(s, s->level == 9 ? 2 : 1068 (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ? 1069 4 : 0)); 1070 put_byte(s, s->gzhead->os & 0xff); 1071 if (s->gzhead->extra != Z_NULL) { 1072 put_byte(s, s->gzhead->extra_len & 0xff); 1073 put_byte(s, (s->gzhead->extra_len >> 8) & 0xff); 1074 } 1075 if (s->gzhead->hcrc) 1076 strm->adler = crc32(strm->adler, s->pending_buf, 1077 s->pending); 1078 s->gzindex = 0; 1079 s->status = EXTRA_STATE; 1080 } 1081 } 1082 if (s->status == EXTRA_STATE) { 1083 if (s->gzhead->extra != Z_NULL) { 1084 ulg beg = s->pending; /* start of bytes to update crc */ 1085 uInt left = (s->gzhead->extra_len & 0xffff) - s->gzindex; 1086 while (s->pending + left > s->pending_buf_size) { 1087 uInt copy = s->pending_buf_size - s->pending; 1088 zmemcpy(s->pending_buf + s->pending, 1089 s->gzhead->extra + s->gzindex, copy); 1090 s->pending = s->pending_buf_size; 1091 HCRC_UPDATE(beg); 1092 s->gzindex += copy; 1093 flush_pending(strm); 1094 if (s->pending != 0) { 1095 s->last_flush = -1; 1096 return Z_OK; 1097 } 1098 beg = 0; 1099 left -= copy; 1100 } 1101 zmemcpy(s->pending_buf + s->pending, 1102 s->gzhead->extra + s->gzindex, left); 1103 s->pending += left; 1104 HCRC_UPDATE(beg); 1105 s->gzindex = 0; 1106 } 1107 s->status = NAME_STATE; 1108 } 1109 if (s->status == NAME_STATE) { 1110 if (s->gzhead->name != Z_NULL) { 1111 ulg beg = s->pending; /* start of bytes to update crc */ 1112 int val; 1113 do { 1114 if (s->pending == s->pending_buf_size) { 1115 HCRC_UPDATE(beg); 1116 flush_pending(strm); 1117 if (s->pending != 0) { 1118 s->last_flush = -1; 1119 return Z_OK; 1120 } 1121 beg = 0; 1122 } 1123 val = s->gzhead->name[s->gzindex++]; 1124 put_byte(s, val); 1125 } while (val != 0); 1126 HCRC_UPDATE(beg); 1127 s->gzindex = 0; 1128 } 1129 s->status = COMMENT_STATE; 1130 } 1131 if (s->status == COMMENT_STATE) { 1132 if (s->gzhead->comment != Z_NULL) { 1133 ulg beg = s->pending; /* start of bytes to update crc */ 1134 int val; 1135 do { 1136 if (s->pending == s->pending_buf_size) { 1137 HCRC_UPDATE(beg); 1138 flush_pending(strm); 1139 if (s->pending != 0) { 1140 s->last_flush = -1; 1141 return Z_OK; 1142 } 1143 beg = 0; 1144 } 1145 val = s->gzhead->comment[s->gzindex++]; 1146 put_byte(s, val); 1147 } while (val != 0); 1148 HCRC_UPDATE(beg); 1149 } 1150 s->status = HCRC_STATE; 1151 } 1152 if (s->status == HCRC_STATE) { 1153 if (s->gzhead->hcrc) { 1154 if (s->pending + 2 > s->pending_buf_size) { 1155 flush_pending(strm); 1156 if (s->pending != 0) { 1157 s->last_flush = -1; 1158 return Z_OK; 1159 } 1160 } 1161 put_byte(s, (Byte)(strm->adler & 0xff)); 1162 put_byte(s, (Byte)((strm->adler >> 8) & 0xff)); 1163 strm->adler = crc32(0L, Z_NULL, 0); 1164 } 1165 s->status = BUSY_STATE; 1166 1167 /* Compression must start with an empty pending buffer */ 1168 flush_pending(strm); 1169 if (s->pending != 0) { 1170 s->last_flush = -1; 1171 return Z_OK; 1172 } 1173 } 1174#endif 1175 1176 /* Start a new block or continue the current one. 1177 */ 1178 if (strm->avail_in != 0 || s->lookahead != 0 || 1179 (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) { 1180 block_state bstate; 1181 1182 bstate = s->level == 0 ? deflate_stored(s, flush) : 1183 s->strategy == Z_HUFFMAN_ONLY ? deflate_huff(s, flush) : 1184 s->strategy == Z_RLE ? deflate_rle(s, flush) : 1185 (*(configuration_table[s->level].func))(s, flush); 1186 1187 if (bstate == finish_started || bstate == finish_done) { 1188 s->status = FINISH_STATE; 1189 } 1190 if (bstate == need_more || bstate == finish_started) { 1191 if (strm->avail_out == 0) { 1192 s->last_flush = -1; /* avoid BUF_ERROR next call, see above */ 1193 } 1194 return Z_OK; 1195 /* If flush != Z_NO_FLUSH && avail_out == 0, the next call 1196 * of deflate should use the same flush parameter to make sure 1197 * that the flush is complete. So we don't have to output an 1198 * empty block here, this will be done at next call. This also 1199 * ensures that for a very small output buffer, we emit at most 1200 * one empty block. 1201 */ 1202 } 1203 if (bstate == block_done) { 1204 if (flush == Z_PARTIAL_FLUSH) { 1205 _tr_align(s); 1206 } else if (flush != Z_BLOCK) { /* FULL_FLUSH or SYNC_FLUSH */ 1207 _tr_stored_block(s, (char*)0, 0L, 0); 1208 /* For a full flush, this empty block will be recognized 1209 * as a special marker by inflate_sync(). 1210 */ 1211 if (flush == Z_FULL_FLUSH) { 1212 CLEAR_HASH(s); /* forget history */ 1213 if (s->lookahead == 0) { 1214 s->strstart = 0; 1215 s->block_start = 0L; 1216 s->insert = 0; 1217 } 1218 } 1219 } 1220 flush_pending(strm); 1221 if (strm->avail_out == 0) { 1222 s->last_flush = -1; /* avoid BUF_ERROR at next call, see above */ 1223 return Z_OK; 1224 } 1225 } 1226 } 1227 1228 if (flush != Z_FINISH) return Z_OK; 1229 if (s->wrap <= 0) return Z_STREAM_END; 1230 1231 /* Write the trailer */ 1232#ifdef GZIP 1233 if (s->wrap == 2) { 1234 put_byte(s, (Byte)(strm->adler & 0xff)); 1235 put_byte(s, (Byte)((strm->adler >> 8) & 0xff)); 1236 put_byte(s, (Byte)((strm->adler >> 16) & 0xff)); 1237 put_byte(s, (Byte)((strm->adler >> 24) & 0xff)); 1238 put_byte(s, (Byte)(strm->total_in & 0xff)); 1239 put_byte(s, (Byte)((strm->total_in >> 8) & 0xff)); 1240 put_byte(s, (Byte)((strm->total_in >> 16) & 0xff)); 1241 put_byte(s, (Byte)((strm->total_in >> 24) & 0xff)); 1242 } 1243 else 1244#endif 1245 { 1246 putShortMSB(s, (uInt)(strm->adler >> 16)); 1247 putShortMSB(s, (uInt)(strm->adler & 0xffff)); 1248 } 1249 flush_pending(strm); 1250 /* If avail_out is zero, the application will call deflate again 1251 * to flush the rest. 1252 */ 1253 if (s->wrap > 0) s->wrap = -s->wrap; /* write the trailer only once! */ 1254 return s->pending != 0 ? Z_OK : Z_STREAM_END; 1255} 1256 1257/* ========================================================================= */ 1258int ZEXPORT deflateEnd(z_streamp strm) { 1259 int status; 1260 1261 if (deflateStateCheck(strm)) return Z_STREAM_ERROR; 1262 1263 status = strm->state->status; 1264 1265 /* Deallocate in reverse order of allocations: */ 1266 TRY_FREE(strm, strm->state->pending_buf); 1267 TRY_FREE(strm, strm->state->head); 1268 TRY_FREE(strm, strm->state->prev); 1269 TRY_FREE(strm, strm->state->window); 1270 1271 ZFREE(strm, strm->state); 1272 strm->state = Z_NULL; 1273 1274 return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK; 1275} 1276 1277/* ========================================================================= 1278 * Copy the source state to the destination state. 1279 * To simplify the source, this is not supported for 16-bit MSDOS (which 1280 * doesn't have enough memory anyway to duplicate compression states). 1281 */ 1282int ZEXPORT deflateCopy(z_streamp dest, z_streamp source) { 1283#ifdef MAXSEG_64K 1284 (void)dest; 1285 (void)source; 1286 return Z_STREAM_ERROR; 1287#else 1288 deflate_state *ds; 1289 deflate_state *ss; 1290 1291 1292 if (deflateStateCheck(source) || dest == Z_NULL) { 1293 return Z_STREAM_ERROR; 1294 } 1295 1296 ss = source->state; 1297 1298 zmemcpy((voidpf)dest, (voidpf)source, sizeof(z_stream)); 1299 1300 ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state)); 1301 if (ds == Z_NULL) return Z_MEM_ERROR; 1302 dest->state = (struct internal_state FAR *) ds; 1303 zmemcpy((voidpf)ds, (voidpf)ss, sizeof(deflate_state)); 1304 ds->strm = dest; 1305 1306 ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte)); 1307 ds->prev = (Posf *) ZALLOC(dest, ds->w_size, sizeof(Pos)); 1308 ds->head = (Posf *) ZALLOC(dest, ds->hash_size, sizeof(Pos)); 1309 ds->pending_buf = (uchf *) ZALLOC(dest, ds->lit_bufsize, LIT_BUFS); 1310 1311 if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL || 1312 ds->pending_buf == Z_NULL) { 1313 deflateEnd (dest); 1314 return Z_MEM_ERROR; 1315 } 1316 /* following zmemcpy do not work for 16-bit MSDOS */ 1317 zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte)); 1318 zmemcpy((voidpf)ds->prev, (voidpf)ss->prev, ds->w_size * sizeof(Pos)); 1319 zmemcpy((voidpf)ds->head, (voidpf)ss->head, ds->hash_size * sizeof(Pos)); 1320 zmemcpy(ds->pending_buf, ss->pending_buf, ds->lit_bufsize * LIT_BUFS); 1321 1322 ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf); 1323#ifdef LIT_MEM 1324 ds->d_buf = (ushf *)(ds->pending_buf + (ds->lit_bufsize << 1)); 1325 ds->l_buf = ds->pending_buf + (ds->lit_bufsize << 2); 1326#else 1327 ds->sym_buf = ds->pending_buf + ds->lit_bufsize; 1328#endif 1329 1330 ds->l_desc.dyn_tree = ds->dyn_ltree; 1331 ds->d_desc.dyn_tree = ds->dyn_dtree; 1332 ds->bl_desc.dyn_tree = ds->bl_tree; 1333 1334 return Z_OK; 1335#endif /* MAXSEG_64K */ 1336} 1337 1338#ifndef FASTEST 1339/* =========================================================================== 1340 * Set match_start to the longest match starting at the given string and 1341 * return its length. Matches shorter or equal to prev_length are discarded, 1342 * in which case the result is equal to prev_length and match_start is 1343 * garbage. 1344 * IN assertions: cur_match is the head of the hash chain for the current 1345 * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1 1346 * OUT assertion: the match length is not greater than s->lookahead. 1347 */ 1348local uInt longest_match(deflate_state *s, IPos cur_match) { 1349 unsigned chain_length = s->max_chain_length;/* max hash chain length */ 1350 register Bytef *scan = s->window + s->strstart; /* current string */ 1351 register Bytef *match; /* matched string */ 1352 register int len; /* length of current match */ 1353 int best_len = (int)s->prev_length; /* best match length so far */ 1354 int nice_match = s->nice_match; /* stop if match long enough */ 1355 IPos limit = s->strstart > (IPos)MAX_DIST(s) ? 1356 s->strstart - (IPos)MAX_DIST(s) : NIL; 1357 /* Stop when cur_match becomes <= limit. To simplify the code, 1358 * we prevent matches with the string of window index 0. 1359 */ 1360 Posf *prev = s->prev; 1361 uInt wmask = s->w_mask; 1362 1363#ifdef UNALIGNED_OK 1364 /* Compare two bytes at a time. Note: this is not always beneficial. 1365 * Try with and without -DUNALIGNED_OK to check. 1366 */ 1367 register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1; 1368 register ush scan_start = *(ushf*)scan; 1369 register ush scan_end = *(ushf*)(scan + best_len - 1); 1370#else 1371 register Bytef *strend = s->window + s->strstart + MAX_MATCH; 1372 register Byte scan_end1 = scan[best_len - 1]; 1373 register Byte scan_end = scan[best_len]; 1374#endif 1375 1376 /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. 1377 * It is easy to get rid of this optimization if necessary. 1378 */ 1379 Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever"); 1380 1381 /* Do not waste too much time if we already have a good match: */ 1382 if (s->prev_length >= s->good_match) { 1383 chain_length >>= 2; 1384 } 1385 /* Do not look for matches beyond the end of the input. This is necessary 1386 * to make deflate deterministic. 1387 */ 1388 if ((uInt)nice_match > s->lookahead) nice_match = (int)s->lookahead; 1389 1390 Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD, 1391 "need lookahead"); 1392 1393 do { 1394 Assert(cur_match < s->strstart, "no future"); 1395 match = s->window + cur_match; 1396 1397 /* Skip to next match if the match length cannot increase 1398 * or if the match length is less than 2. Note that the checks below 1399 * for insufficient lookahead only occur occasionally for performance 1400 * reasons. Therefore uninitialized memory will be accessed, and 1401 * conditional jumps will be made that depend on those values. 1402 * However the length of the match is limited to the lookahead, so 1403 * the output of deflate is not affected by the uninitialized values. 1404 */ 1405#if (defined(UNALIGNED_OK) && MAX_MATCH == 258) 1406 /* This code assumes sizeof(unsigned short) == 2. Do not use 1407 * UNALIGNED_OK if your compiler uses a different size. 1408 */ 1409 if (*(ushf*)(match + best_len - 1) != scan_end || 1410 *(ushf*)match != scan_start) continue; 1411 1412 /* It is not necessary to compare scan[2] and match[2] since they are 1413 * always equal when the other bytes match, given that the hash keys 1414 * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at 1415 * strstart + 3, + 5, up to strstart + 257. We check for insufficient 1416 * lookahead only every 4th comparison; the 128th check will be made 1417 * at strstart + 257. If MAX_MATCH-2 is not a multiple of 8, it is 1418 * necessary to put more guard bytes at the end of the window, or 1419 * to check more often for insufficient lookahead. 1420 */ 1421 Assert(scan[2] == match[2], "scan[2]?"); 1422 scan++, match++; 1423 do { 1424 } while (*(ushf*)(scan += 2) == *(ushf*)(match += 2) && 1425 *(ushf*)(scan += 2) == *(ushf*)(match += 2) && 1426 *(ushf*)(scan += 2) == *(ushf*)(match += 2) && 1427 *(ushf*)(scan += 2) == *(ushf*)(match += 2) && 1428 scan < strend); 1429 /* The funny "do {}" generates better code on most compilers */ 1430 1431 /* Here, scan <= window + strstart + 257 */ 1432 Assert(scan <= s->window + (unsigned)(s->window_size - 1), 1433 "wild scan"); 1434 if (*scan == *match) scan++; 1435 1436 len = (MAX_MATCH - 1) - (int)(strend - scan); 1437 scan = strend - (MAX_MATCH-1); 1438 1439#else /* UNALIGNED_OK */ 1440 1441 if (match[best_len] != scan_end || 1442 match[best_len - 1] != scan_end1 || 1443 *match != *scan || 1444 *++match != scan[1]) continue; 1445 1446 /* The check at best_len - 1 can be removed because it will be made 1447 * again later. (This heuristic is not always a win.) 1448 * It is not necessary to compare scan[2] and match[2] since they 1449 * are always equal when the other bytes match, given that 1450 * the hash keys are equal and that HASH_BITS >= 8. 1451 */ 1452 scan += 2, match++; 1453 Assert(*scan == *match, "match[2]?"); 1454 1455 /* We check for insufficient lookahead only every 8th comparison; 1456 * the 256th check will be made at strstart + 258. 1457 */ 1458 do { 1459 } while (*++scan == *++match && *++scan == *++match && 1460 *++scan == *++match && *++scan == *++match && 1461 *++scan == *++match && *++scan == *++match && 1462 *++scan == *++match && *++scan == *++match && 1463 scan < strend); 1464 1465 Assert(scan <= s->window + (unsigned)(s->window_size - 1), 1466 "wild scan"); 1467 1468 len = MAX_MATCH - (int)(strend - scan); 1469 scan = strend - MAX_MATCH; 1470 1471#endif /* UNALIGNED_OK */ 1472 1473 if (len > best_len) { 1474 s->match_start = cur_match; 1475 best_len = len; 1476 if (len >= nice_match) break; 1477#ifdef UNALIGNED_OK 1478 scan_end = *(ushf*)(scan + best_len - 1); 1479#else 1480 scan_end1 = scan[best_len - 1]; 1481 scan_end = scan[best_len]; 1482#endif 1483 } 1484 } while ((cur_match = prev[cur_match & wmask]) > limit 1485 && --chain_length != 0); 1486 1487 if ((uInt)best_len <= s->lookahead) return (uInt)best_len; 1488 return s->lookahead; 1489} 1490 1491#else /* FASTEST */ 1492 1493/* --------------------------------------------------------------------------- 1494 * Optimized version for FASTEST only 1495 */ 1496local uInt longest_match(deflate_state *s, IPos cur_match) { 1497 register Bytef *scan = s->window + s->strstart; /* current string */ 1498 register Bytef *match; /* matched string */ 1499 register int len; /* length of current match */ 1500 register Bytef *strend = s->window + s->strstart + MAX_MATCH; 1501 1502 /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. 1503 * It is easy to get rid of this optimization if necessary. 1504 */ 1505 Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever"); 1506 1507 Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD, 1508 "need lookahead"); 1509 1510 Assert(cur_match < s->strstart, "no future"); 1511 1512 match = s->window + cur_match; 1513 1514 /* Return failure if the match length is less than 2: 1515 */ 1516 if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1; 1517 1518 /* The check at best_len - 1 can be removed because it will be made 1519 * again later. (This heuristic is not always a win.) 1520 * It is not necessary to compare scan[2] and match[2] since they 1521 * are always equal when the other bytes match, given that 1522 * the hash keys are equal and that HASH_BITS >= 8. 1523 */ 1524 scan += 2, match += 2; 1525 Assert(*scan == *match, "match[2]?"); 1526 1527 /* We check for insufficient lookahead only every 8th comparison; 1528 * the 256th check will be made at strstart + 258. 1529 */ 1530 do { 1531 } while (*++scan == *++match && *++scan == *++match && 1532 *++scan == *++match && *++scan == *++match && 1533 *++scan == *++match && *++scan == *++match && 1534 *++scan == *++match && *++scan == *++match && 1535 scan < strend); 1536 1537 Assert(scan <= s->window + (unsigned)(s->window_size - 1), "wild scan"); 1538 1539 len = MAX_MATCH - (int)(strend - scan); 1540 1541 if (len < MIN_MATCH) return MIN_MATCH - 1; 1542 1543 s->match_start = cur_match; 1544 return (uInt)len <= s->lookahead ? (uInt)len : s->lookahead; 1545} 1546 1547#endif /* FASTEST */ 1548 1549#ifdef ZLIB_DEBUG 1550 1551#define EQUAL 0 1552/* result of memcmp for equal strings */ 1553 1554/* =========================================================================== 1555 * Check that the match at match_start is indeed a match. 1556 */ 1557local void check_match(deflate_state *s, IPos start, IPos match, int length) { 1558 /* check that the match is indeed a match */ 1559 Bytef *back = s->window + (int)match, *here = s->window + start; 1560 IPos len = length; 1561 if (match == (IPos)-1) { 1562 /* match starts one byte before the current window -- just compare the 1563 subsequent length-1 bytes */ 1564 back++; 1565 here++; 1566 len--; 1567 } 1568 if (zmemcmp(back, here, len) != EQUAL) { 1569 fprintf(stderr, " start %u, match %d, length %d\n", 1570 start, (int)match, length); 1571 do { 1572 fprintf(stderr, "(%02x %02x)", *back++, *here++); 1573 } while (--len != 0); 1574 z_error("invalid match"); 1575 } 1576 if (z_verbose > 1) { 1577 fprintf(stderr,"\\[%d,%d]", start - match, length); 1578 do { putc(s->window[start++], stderr); } while (--length != 0); 1579 } 1580} 1581#else 1582# define check_match(s, start, match, length) 1583#endif /* ZLIB_DEBUG */ 1584 1585/* =========================================================================== 1586 * Flush the current block, with given end-of-file flag. 1587 * IN assertion: strstart is set to the end of the current match. 1588 */ 1589#define FLUSH_BLOCK_ONLY(s, last) { \ 1590 _tr_flush_block(s, (s->block_start >= 0L ? \ 1591 (charf *)&s->window[(unsigned)s->block_start] : \ 1592 (charf *)Z_NULL), \ 1593 (ulg)((long)s->strstart - s->block_start), \ 1594 (last)); \ 1595 s->block_start = s->strstart; \ 1596 flush_pending(s->strm); \ 1597 Tracev((stderr,"[FLUSH]")); \ 1598} 1599 1600/* Same but force premature exit if necessary. */ 1601#define FLUSH_BLOCK(s, last) { \ 1602 FLUSH_BLOCK_ONLY(s, last); \ 1603 if (s->strm->avail_out == 0) return (last) ? finish_started : need_more; \ 1604} 1605 1606/* Maximum stored block length in deflate format (not including header). */ 1607#define MAX_STORED 65535 1608 1609#if !defined(MIN) 1610/* Minimum of a and b. */ 1611#define MIN(a, b) ((a) > (b) ? (b) : (a)) 1612#endif 1613 1614/* =========================================================================== 1615 * Copy without compression as much as possible from the input stream, return 1616 * the current block state. 1617 * 1618 * In case deflateParams() is used to later switch to a non-zero compression 1619 * level, s->matches (otherwise unused when storing) keeps track of the number 1620 * of hash table slides to perform. If s->matches is 1, then one hash table 1621 * slide will be done when switching. If s->matches is 2, the maximum value 1622 * allowed here, then the hash table will be cleared, since two or more slides 1623 * is the same as a clear. 1624 * 1625 * deflate_stored() is written to minimize the number of times an input byte is 1626 * copied. It is most efficient with large input and output buffers, which 1627 * maximizes the opportunities to have a single copy from next_in to next_out. 1628 */ 1629local block_state deflate_stored(deflate_state *s, int flush) { 1630 /* Smallest worthy block size when not flushing or finishing. By default 1631 * this is 32K. This can be as small as 507 bytes for memLevel == 1. For 1632 * large input and output buffers, the stored block size will be larger. 1633 */ 1634 unsigned min_block = MIN(s->pending_buf_size - 5, s->w_size); 1635 1636 /* Copy as many min_block or larger stored blocks directly to next_out as 1637 * possible. If flushing, copy the remaining available input to next_out as 1638 * stored blocks, if there is enough space. 1639 */ 1640 unsigned len, left, have, last = 0; 1641 unsigned used = s->strm->avail_in; 1642 do { 1643 /* Set len to the maximum size block that we can copy directly with the 1644 * available input data and output space. Set left to how much of that 1645 * would be copied from what's left in the window. 1646 */ 1647 len = MAX_STORED; /* maximum deflate stored block length */ 1648 have = (s->bi_valid + 42) >> 3; /* number of header bytes */ 1649 if (s->strm->avail_out < have) /* need room for header */ 1650 break; 1651 /* maximum stored block length that will fit in avail_out: */ 1652 have = s->strm->avail_out - have; 1653 left = s->strstart - s->block_start; /* bytes left in window */ 1654 if (len > (ulg)left + s->strm->avail_in) 1655 len = left + s->strm->avail_in; /* limit len to the input */ 1656 if (len > have) 1657 len = have; /* limit len to the output */ 1658 1659 /* If the stored block would be less than min_block in length, or if 1660 * unable to copy all of the available input when flushing, then try 1661 * copying to the window and the pending buffer instead. Also don't 1662 * write an empty block when flushing -- deflate() does that. 1663 */ 1664 if (len < min_block && ((len == 0 && flush != Z_FINISH) || 1665 flush == Z_NO_FLUSH || 1666 len != left + s->strm->avail_in)) 1667 break; 1668 1669 /* Make a dummy stored block in pending to get the header bytes, 1670 * including any pending bits. This also updates the debugging counts. 1671 */ 1672 last = flush == Z_FINISH && len == left + s->strm->avail_in ? 1 : 0; 1673 _tr_stored_block(s, (char *)0, 0L, last); 1674 1675 /* Replace the lengths in the dummy stored block with len. */ 1676 s->pending_buf[s->pending - 4] = len; 1677 s->pending_buf[s->pending - 3] = len >> 8; 1678 s->pending_buf[s->pending - 2] = ~len; 1679 s->pending_buf[s->pending - 1] = ~len >> 8; 1680 1681 /* Write the stored block header bytes. */ 1682 flush_pending(s->strm); 1683 1684#ifdef ZLIB_DEBUG 1685 /* Update debugging counts for the data about to be copied. */ 1686 s->compressed_len += len << 3; 1687 s->bits_sent += len << 3; 1688#endif 1689 1690 /* Copy uncompressed bytes from the window to next_out. */ 1691 if (left) { 1692 if (left > len) 1693 left = len; 1694 zmemcpy(s->strm->next_out, s->window + s->block_start, left); 1695 s->strm->next_out += left; 1696 s->strm->avail_out -= left; 1697 s->strm->total_out += left; 1698 s->block_start += left; 1699 len -= left; 1700 } 1701 1702 /* Copy uncompressed bytes directly from next_in to next_out, updating 1703 * the check value. 1704 */ 1705 if (len) { 1706 read_buf(s->strm, s->strm->next_out, len); 1707 s->strm->next_out += len; 1708 s->strm->avail_out -= len; 1709 s->strm->total_out += len; 1710 } 1711 } while (last == 0); 1712 1713 /* Update the sliding window with the last s->w_size bytes of the copied 1714 * data, or append all of the copied data to the existing window if less 1715 * than s->w_size bytes were copied. Also update the number of bytes to 1716 * insert in the hash tables, in the event that deflateParams() switches to 1717 * a non-zero compression level. 1718 */ 1719 used -= s->strm->avail_in; /* number of input bytes directly copied */ 1720 if (used) { 1721 /* If any input was used, then no unused input remains in the window, 1722 * therefore s->block_start == s->strstart. 1723 */ 1724 if (used >= s->w_size) { /* supplant the previous history */ 1725 s->matches = 2; /* clear hash */ 1726 zmemcpy(s->window, s->strm->next_in - s->w_size, s->w_size); 1727 s->strstart = s->w_size; 1728 s->insert = s->strstart; 1729 } 1730 else { 1731 if (s->window_size - s->strstart <= used) { 1732 /* Slide the window down. */ 1733 s->strstart -= s->w_size; 1734 zmemcpy(s->window, s->window + s->w_size, s->strstart); 1735 if (s->matches < 2) 1736 s->matches++; /* add a pending slide_hash() */ 1737 if (s->insert > s->strstart) 1738 s->insert = s->strstart; 1739 } 1740 zmemcpy(s->window + s->strstart, s->strm->next_in - used, used); 1741 s->strstart += used; 1742 s->insert += MIN(used, s->w_size - s->insert); 1743 } 1744 s->block_start = s->strstart; 1745 } 1746 if (s->high_water < s->strstart) 1747 s->high_water = s->strstart; 1748 1749 /* If the last block was written to next_out, then done. */ 1750 if (last) 1751 return finish_done; 1752 1753 /* If flushing and all input has been consumed, then done. */ 1754 if (flush != Z_NO_FLUSH && flush != Z_FINISH && 1755 s->strm->avail_in == 0 && (long)s->strstart == s->block_start) 1756 return block_done; 1757 1758 /* Fill the window with any remaining input. */ 1759 have = s->window_size - s->strstart; 1760 if (s->strm->avail_in > have && s->block_start >= (long)s->w_size) { 1761 /* Slide the window down. */ 1762 s->block_start -= s->w_size; 1763 s->strstart -= s->w_size; 1764 zmemcpy(s->window, s->window + s->w_size, s->strstart); 1765 if (s->matches < 2) 1766 s->matches++; /* add a pending slide_hash() */ 1767 have += s->w_size; /* more space now */ 1768 if (s->insert > s->strstart) 1769 s->insert = s->strstart; 1770 } 1771 if (have > s->strm->avail_in) 1772 have = s->strm->avail_in; 1773 if (have) { 1774 read_buf(s->strm, s->window + s->strstart, have); 1775 s->strstart += have; 1776 s->insert += MIN(have, s->w_size - s->insert); 1777 } 1778 if (s->high_water < s->strstart) 1779 s->high_water = s->strstart; 1780 1781 /* There was not enough avail_out to write a complete worthy or flushed 1782 * stored block to next_out. Write a stored block to pending instead, if we 1783 * have enough input for a worthy block, or if flushing and there is enough 1784 * room for the remaining input as a stored block in the pending buffer. 1785 */ 1786 have = (s->bi_valid + 42) >> 3; /* number of header bytes */ 1787 /* maximum stored block length that will fit in pending: */ 1788 have = MIN(s->pending_buf_size - have, MAX_STORED); 1789 min_block = MIN(have, s->w_size); 1790 left = s->strstart - s->block_start; 1791 if (left >= min_block || 1792 ((left || flush == Z_FINISH) && flush != Z_NO_FLUSH && 1793 s->strm->avail_in == 0 && left <= have)) { 1794 len = MIN(left, have); 1795 last = flush == Z_FINISH && s->strm->avail_in == 0 && 1796 len == left ? 1 : 0; 1797 _tr_stored_block(s, (charf *)s->window + s->block_start, len, last); 1798 s->block_start += len; 1799 flush_pending(s->strm); 1800 } 1801 1802 /* We've done all we can with the available input and output. */ 1803 return last ? finish_started : need_more; 1804} 1805 1806/* =========================================================================== 1807 * Compress as much as possible from the input stream, return the current 1808 * block state. 1809 * This function does not perform lazy evaluation of matches and inserts 1810 * new strings in the dictionary only for unmatched strings or for short 1811 * matches. It is used only for the fast compression options. 1812 */ 1813local block_state deflate_fast(deflate_state *s, int flush) { 1814 IPos hash_head; /* head of the hash chain */ 1815 int bflush; /* set if current block must be flushed */ 1816 1817 for (;;) { 1818 /* Make sure that we always have enough lookahead, except 1819 * at the end of the input file. We need MAX_MATCH bytes 1820 * for the next match, plus MIN_MATCH bytes to insert the 1821 * string following the next match. 1822 */ 1823 if (s->lookahead < MIN_LOOKAHEAD) { 1824 fill_window(s); 1825 if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { 1826 return need_more; 1827 } 1828 if (s->lookahead == 0) break; /* flush the current block */ 1829 } 1830 1831 /* Insert the string window[strstart .. strstart + 2] in the 1832 * dictionary, and set hash_head to the head of the hash chain: 1833 */ 1834 hash_head = NIL; 1835 if (s->lookahead >= MIN_MATCH) { 1836 INSERT_STRING(s, s->strstart, hash_head); 1837 } 1838 1839 /* Find the longest match, discarding those <= prev_length. 1840 * At this point we have always match_length < MIN_MATCH 1841 */ 1842 if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) { 1843 /* To simplify the code, we prevent matches with the string 1844 * of window index 0 (in particular we have to avoid a match 1845 * of the string with itself at the start of the input file). 1846 */ 1847 s->match_length = longest_match (s, hash_head); 1848 /* longest_match() sets match_start */ 1849 } 1850 if (s->match_length >= MIN_MATCH) { 1851 check_match(s, s->strstart, s->match_start, s->match_length); 1852 1853 _tr_tally_dist(s, s->strstart - s->match_start, 1854 s->match_length - MIN_MATCH, bflush); 1855 1856 s->lookahead -= s->match_length; 1857 1858 /* Insert new strings in the hash table only if the match length 1859 * is not too large. This saves time but degrades compression. 1860 */ 1861#ifndef FASTEST 1862 if (s->match_length <= s->max_insert_length && 1863 s->lookahead >= MIN_MATCH) { 1864 s->match_length--; /* string at strstart already in table */ 1865 do { 1866 s->strstart++; 1867 INSERT_STRING(s, s->strstart, hash_head); 1868 /* strstart never exceeds WSIZE-MAX_MATCH, so there are 1869 * always MIN_MATCH bytes ahead. 1870 */ 1871 } while (--s->match_length != 0); 1872 s->strstart++; 1873 } else 1874#endif 1875 { 1876 s->strstart += s->match_length; 1877 s->match_length = 0; 1878 s->ins_h = s->window[s->strstart]; 1879 UPDATE_HASH(s, s->ins_h, s->window[s->strstart + 1]); 1880#if MIN_MATCH != 3 1881 Call UPDATE_HASH() MIN_MATCH-3 more times 1882#endif 1883 /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not 1884 * matter since it will be recomputed at next deflate call. 1885 */ 1886 } 1887 } else { 1888 /* No match, output a literal byte */ 1889 Tracevv((stderr,"%c", s->window[s->strstart])); 1890 _tr_tally_lit(s, s->window[s->strstart], bflush); 1891 s->lookahead--; 1892 s->strstart++; 1893 } 1894 if (bflush) FLUSH_BLOCK(s, 0); 1895 } 1896 s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1; 1897 if (flush == Z_FINISH) { 1898 FLUSH_BLOCK(s, 1); 1899 return finish_done; 1900 } 1901 if (s->sym_next) 1902 FLUSH_BLOCK(s, 0); 1903 return block_done; 1904} 1905 1906#ifndef FASTEST 1907/* =========================================================================== 1908 * Same as above, but achieves better compression. We use a lazy 1909 * evaluation for matches: a match is finally adopted only if there is 1910 * no better match at the next window position. 1911 */ 1912local block_state deflate_slow(deflate_state *s, int flush) { 1913 IPos hash_head; /* head of hash chain */ 1914 int bflush; /* set if current block must be flushed */ 1915 1916 /* Process the input block. */ 1917 for (;;) { 1918 /* Make sure that we always have enough lookahead, except 1919 * at the end of the input file. We need MAX_MATCH bytes 1920 * for the next match, plus MIN_MATCH bytes to insert the 1921 * string following the next match. 1922 */ 1923 if (s->lookahead < MIN_LOOKAHEAD) { 1924 fill_window(s); 1925 if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { 1926 return need_more; 1927 } 1928 if (s->lookahead == 0) break; /* flush the current block */ 1929 } 1930 1931 /* Insert the string window[strstart .. strstart + 2] in the 1932 * dictionary, and set hash_head to the head of the hash chain: 1933 */ 1934 hash_head = NIL; 1935 if (s->lookahead >= MIN_MATCH) { 1936 INSERT_STRING(s, s->strstart, hash_head); 1937 } 1938 1939 /* Find the longest match, discarding those <= prev_length. 1940 */ 1941 s->prev_length = s->match_length, s->prev_match = s->match_start; 1942 s->match_length = MIN_MATCH-1; 1943 1944 if (hash_head != NIL && s->prev_length < s->max_lazy_match && 1945 s->strstart - hash_head <= MAX_DIST(s)) { 1946 /* To simplify the code, we prevent matches with the string 1947 * of window index 0 (in particular we have to avoid a match 1948 * of the string with itself at the start of the input file). 1949 */ 1950 s->match_length = longest_match (s, hash_head); 1951 /* longest_match() sets match_start */ 1952 1953 if (s->match_length <= 5 && (s->strategy == Z_FILTERED 1954#if TOO_FAR <= 32767 1955 || (s->match_length == MIN_MATCH && 1956 s->strstart - s->match_start > TOO_FAR) 1957#endif 1958 )) { 1959 1960 /* If prev_match is also MIN_MATCH, match_start is garbage 1961 * but we will ignore the current match anyway. 1962 */ 1963 s->match_length = MIN_MATCH-1; 1964 } 1965 } 1966 /* If there was a match at the previous step and the current 1967 * match is not better, output the previous match: 1968 */ 1969 if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) { 1970 uInt max_insert = s->strstart + s->lookahead - MIN_MATCH; 1971 /* Do not insert strings in hash table beyond this. */ 1972 1973 check_match(s, s->strstart - 1, s->prev_match, s->prev_length); 1974 1975 _tr_tally_dist(s, s->strstart - 1 - s->prev_match, 1976 s->prev_length - MIN_MATCH, bflush); 1977 1978 /* Insert in hash table all strings up to the end of the match. 1979 * strstart - 1 and strstart are already inserted. If there is not 1980 * enough lookahead, the last two strings are not inserted in 1981 * the hash table. 1982 */ 1983 s->lookahead -= s->prev_length - 1; 1984 s->prev_length -= 2; 1985 do { 1986 if (++s->strstart <= max_insert) { 1987 INSERT_STRING(s, s->strstart, hash_head); 1988 } 1989 } while (--s->prev_length != 0); 1990 s->match_available = 0; 1991 s->match_length = MIN_MATCH-1; 1992 s->strstart++; 1993 1994 if (bflush) FLUSH_BLOCK(s, 0); 1995 1996 } else if (s->match_available) { 1997 /* If there was no match at the previous position, output a 1998 * single literal. If there was a match but the current match 1999 * is longer, truncate the previous match to a single literal. 2000 */ 2001 Tracevv((stderr,"%c", s->window[s->strstart - 1])); 2002 _tr_tally_lit(s, s->window[s->strstart - 1], bflush); 2003 if (bflush) { 2004 FLUSH_BLOCK_ONLY(s, 0); 2005 } 2006 s->strstart++; 2007 s->lookahead--; 2008 if (s->strm->avail_out == 0) return need_more; 2009 } else { 2010 /* There is no previous match to compare with, wait for 2011 * the next step to decide. 2012 */ 2013 s->match_available = 1; 2014 s->strstart++; 2015 s->lookahead--; 2016 } 2017 } 2018 Assert (flush != Z_NO_FLUSH, "no flush?"); 2019 if (s->match_available) { 2020 Tracevv((stderr,"%c", s->window[s->strstart - 1])); 2021 _tr_tally_lit(s, s->window[s->strstart - 1], bflush); 2022 s->match_available = 0; 2023 } 2024 s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1; 2025 if (flush == Z_FINISH) { 2026 FLUSH_BLOCK(s, 1); 2027 return finish_done; 2028 } 2029 if (s->sym_next) 2030 FLUSH_BLOCK(s, 0); 2031 return block_done; 2032} 2033#endif /* FASTEST */ 2034 2035/* =========================================================================== 2036 * For Z_RLE, simply look for runs of bytes, generate matches only of distance 2037 * one. Do not maintain a hash table. (It will be regenerated if this run of 2038 * deflate switches away from Z_RLE.) 2039 */ 2040local block_state deflate_rle(deflate_state *s, int flush) { 2041 int bflush; /* set if current block must be flushed */ 2042 uInt prev; /* byte at distance one to match */ 2043 Bytef *scan, *strend; /* scan goes up to strend for length of run */ 2044 2045 for (;;) { 2046 /* Make sure that we always have enough lookahead, except 2047 * at the end of the input file. We need MAX_MATCH bytes 2048 * for the longest run, plus one for the unrolled loop. 2049 */ 2050 if (s->lookahead <= MAX_MATCH) { 2051 fill_window(s); 2052 if (s->lookahead <= MAX_MATCH && flush == Z_NO_FLUSH) { 2053 return need_more; 2054 } 2055 if (s->lookahead == 0) break; /* flush the current block */ 2056 } 2057 2058 /* See how many times the previous byte repeats */ 2059 s->match_length = 0; 2060 if (s->lookahead >= MIN_MATCH && s->strstart > 0) { 2061 scan = s->window + s->strstart - 1; 2062 prev = *scan; 2063 if (prev == *++scan && prev == *++scan && prev == *++scan) { 2064 strend = s->window + s->strstart + MAX_MATCH; 2065 do { 2066 } while (prev == *++scan && prev == *++scan && 2067 prev == *++scan && prev == *++scan && 2068 prev == *++scan && prev == *++scan && 2069 prev == *++scan && prev == *++scan && 2070 scan < strend); 2071 s->match_length = MAX_MATCH - (uInt)(strend - scan); 2072 if (s->match_length > s->lookahead) 2073 s->match_length = s->lookahead; 2074 } 2075 Assert(scan <= s->window + (uInt)(s->window_size - 1), 2076 "wild scan"); 2077 } 2078 2079 /* Emit match if have run of MIN_MATCH or longer, else emit literal */ 2080 if (s->match_length >= MIN_MATCH) { 2081 check_match(s, s->strstart, s->strstart - 1, s->match_length); 2082 2083 _tr_tally_dist(s, 1, s->match_length - MIN_MATCH, bflush); 2084 2085 s->lookahead -= s->match_length; 2086 s->strstart += s->match_length; 2087 s->match_length = 0; 2088 } else { 2089 /* No match, output a literal byte */ 2090 Tracevv((stderr,"%c", s->window[s->strstart])); 2091 _tr_tally_lit(s, s->window[s->strstart], bflush); 2092 s->lookahead--; 2093 s->strstart++; 2094 } 2095 if (bflush) FLUSH_BLOCK(s, 0); 2096 } 2097 s->insert = 0; 2098 if (flush == Z_FINISH) { 2099 FLUSH_BLOCK(s, 1); 2100 return finish_done; 2101 } 2102 if (s->sym_next) 2103 FLUSH_BLOCK(s, 0); 2104 return block_done; 2105} 2106 2107/* =========================================================================== 2108 * For Z_HUFFMAN_ONLY, do not look for matches. Do not maintain a hash table. 2109 * (It will be regenerated if this run of deflate switches away from Huffman.) 2110 */ 2111local block_state deflate_huff(deflate_state *s, int flush) { 2112 int bflush; /* set if current block must be flushed */ 2113 2114 for (;;) { 2115 /* Make sure that we have a literal to write. */ 2116 if (s->lookahead == 0) { 2117 fill_window(s); 2118 if (s->lookahead == 0) { 2119 if (flush == Z_NO_FLUSH) 2120 return need_more; 2121 break; /* flush the current block */ 2122 } 2123 } 2124 2125 /* Output a literal byte */ 2126 s->match_length = 0; 2127 Tracevv((stderr,"%c", s->window[s->strstart])); 2128 _tr_tally_lit(s, s->window[s->strstart], bflush); 2129 s->lookahead--; 2130 s->strstart++; 2131 if (bflush) FLUSH_BLOCK(s, 0); 2132 } 2133 s->insert = 0; 2134 if (flush == Z_FINISH) { 2135 FLUSH_BLOCK(s, 1); 2136 return finish_done; 2137 } 2138 if (s->sym_next) 2139 FLUSH_BLOCK(s, 0); 2140 return block_done; 2141} 2142