[562] | 1 | /* deflate.c -- compress data using the deflation algorithm |
---|
| 2 | * Copyright (C) 1995-2004 Jean-loup Gailly. |
---|
| 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://www.ietf.org/rfc/rfc1951.txt |
---|
| 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: deflate.cc,v 1.1 2005/05/12 21:04:53 duns Exp $ */ |
---|
| 51 | |
---|
| 52 | #include "deflate.h" |
---|
| 53 | |
---|
| 54 | const char deflate_copyright[] = |
---|
| 55 | " deflate 1.2.2 Copyright 1995-2004 Jean-loup Gailly "; |
---|
| 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 | |
---|
| 63 | /* =========================================================================== |
---|
| 64 | * Function prototypes. |
---|
| 65 | */ |
---|
| 66 | typedef enum { |
---|
| 67 | need_more, /* block not completed, need more input or more output */ |
---|
| 68 | block_done, /* block flush performed */ |
---|
| 69 | finish_started, /* finish started, need only more output at next deflate */ |
---|
| 70 | finish_done /* finish done, accept no more input or output */ |
---|
| 71 | } block_state; |
---|
| 72 | |
---|
| 73 | typedef block_state (*compress_func) OF((deflate_state *s, int flush)); |
---|
| 74 | /* Compression function. Returns the block state after the call. */ |
---|
| 75 | |
---|
| 76 | local void fill_window OF((deflate_state *s)); |
---|
| 77 | local block_state deflate_stored OF((deflate_state *s, int flush)); |
---|
| 78 | local block_state deflate_fast OF((deflate_state *s, int flush)); |
---|
| 79 | #ifndef FASTEST |
---|
| 80 | local block_state deflate_slow OF((deflate_state *s, int flush)); |
---|
| 81 | #endif |
---|
| 82 | local void lm_init OF((deflate_state *s)); |
---|
| 83 | local void putShortMSB OF((deflate_state *s, uInt b)); |
---|
| 84 | local void flush_pending OF((z_streamp strm)); |
---|
| 85 | local int read_buf OF((z_streamp strm, Bytef *buf, unsigned size)); |
---|
| 86 | #ifndef FASTEST |
---|
| 87 | #ifdef ASMV |
---|
| 88 | void match_init OF((void)); /* asm code initialization */ |
---|
| 89 | uInt longest_match OF((deflate_state *s, IPos cur_match)); |
---|
| 90 | #else |
---|
| 91 | local uInt longest_match OF((deflate_state *s, IPos cur_match)); |
---|
| 92 | #endif |
---|
| 93 | #endif |
---|
| 94 | local uInt longest_match_fast OF((deflate_state *s, IPos cur_match)); |
---|
| 95 | |
---|
| 96 | #ifdef DEBUG |
---|
| 97 | local void check_match OF((deflate_state *s, IPos start, IPos match, |
---|
| 98 | int length)); |
---|
| 99 | #endif |
---|
| 100 | |
---|
| 101 | /* =========================================================================== |
---|
| 102 | * Local data |
---|
| 103 | */ |
---|
| 104 | |
---|
| 105 | #define NIL 0 |
---|
| 106 | /* Tail of hash chains */ |
---|
| 107 | |
---|
| 108 | #ifndef TOO_FAR |
---|
| 109 | # define TOO_FAR 4096 |
---|
| 110 | #endif |
---|
| 111 | /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */ |
---|
| 112 | |
---|
| 113 | #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1) |
---|
| 114 | /* Minimum amount of lookahead, except at the end of the input file. |
---|
| 115 | * See deflate.c for comments about the MIN_MATCH+1. |
---|
| 116 | */ |
---|
| 117 | |
---|
| 118 | /* Values for max_lazy_match, good_match and max_chain_length, depending on |
---|
| 119 | * the desired pack level (0..9). The values given below have been tuned to |
---|
| 120 | * exclude worst case performance for pathological files. Better values may be |
---|
| 121 | * found for specific files. |
---|
| 122 | */ |
---|
| 123 | typedef struct config_s { |
---|
| 124 | ush good_length; /* reduce lazy search above this match length */ |
---|
| 125 | ush max_lazy; /* do not perform lazy search above this match length */ |
---|
| 126 | ush nice_length; /* quit search above this match length */ |
---|
| 127 | ush max_chain; |
---|
| 128 | compress_func func; |
---|
| 129 | } config; |
---|
| 130 | |
---|
| 131 | #ifdef FASTEST |
---|
| 132 | local const config configuration_table[2] = { |
---|
| 133 | /* good lazy nice chain */ |
---|
| 134 | /* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */ |
---|
| 135 | /* 1 */ {4, 4, 8, 4, deflate_fast}}; /* max speed, no lazy matches */ |
---|
| 136 | #else |
---|
| 137 | local const config configuration_table[10] = { |
---|
| 138 | /* good lazy nice chain */ |
---|
| 139 | /* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */ |
---|
| 140 | /* 1 */ {4, 4, 8, 4, deflate_fast}, /* max speed, no lazy matches */ |
---|
| 141 | /* 2 */ {4, 5, 16, 8, deflate_fast}, |
---|
| 142 | /* 3 */ {4, 6, 32, 32, deflate_fast}, |
---|
| 143 | |
---|
| 144 | /* 4 */ {4, 4, 16, 16, deflate_slow}, /* lazy matches */ |
---|
| 145 | /* 5 */ {8, 16, 32, 32, deflate_slow}, |
---|
| 146 | /* 6 */ {8, 16, 128, 128, deflate_slow}, |
---|
| 147 | /* 7 */ {8, 32, 128, 256, deflate_slow}, |
---|
| 148 | /* 8 */ {32, 128, 258, 1024, deflate_slow}, |
---|
| 149 | /* 9 */ {32, 258, 258, 4096, deflate_slow}}; /* max compression */ |
---|
| 150 | #endif |
---|
| 151 | |
---|
| 152 | /* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4 |
---|
| 153 | * For deflate_fast() (levels <= 3) good is ignored and lazy has a different |
---|
| 154 | * meaning. |
---|
| 155 | */ |
---|
| 156 | |
---|
| 157 | #define EQUAL 0 |
---|
| 158 | /* result of memcmp for equal strings */ |
---|
| 159 | |
---|
| 160 | #ifndef NO_DUMMY_DECL |
---|
| 161 | struct static_tree_desc_s {int dummy;}; /* for buggy compilers */ |
---|
| 162 | #endif |
---|
| 163 | |
---|
| 164 | /* =========================================================================== |
---|
| 165 | * Update a hash value with the given input byte |
---|
| 166 | * IN assertion: all calls to to UPDATE_HASH are made with consecutive |
---|
| 167 | * input characters, so that a running hash key can be computed from the |
---|
| 168 | * previous key instead of complete recalculation each time. |
---|
| 169 | */ |
---|
| 170 | #define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask) |
---|
| 171 | |
---|
| 172 | |
---|
| 173 | /* =========================================================================== |
---|
| 174 | * Insert string str in the dictionary and set match_head to the previous head |
---|
| 175 | * of the hash chain (the most recent string with same hash key). Return |
---|
| 176 | * the previous length of the hash chain. |
---|
| 177 | * If this file is compiled with -DFASTEST, the compression level is forced |
---|
| 178 | * to 1, and no hash chains are maintained. |
---|
| 179 | * IN assertion: all calls to to INSERT_STRING are made with consecutive |
---|
| 180 | * input characters and the first MIN_MATCH bytes of str are valid |
---|
| 181 | * (except for the last MIN_MATCH-1 bytes of the input file). |
---|
| 182 | */ |
---|
| 183 | #ifdef FASTEST |
---|
| 184 | #define INSERT_STRING(s, str, match_head) \ |
---|
| 185 | (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ |
---|
| 186 | match_head = s->head[s->ins_h], \ |
---|
| 187 | s->head[s->ins_h] = (Pos)(str)) |
---|
| 188 | #else |
---|
| 189 | #define INSERT_STRING(s, str, match_head) \ |
---|
| 190 | (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ |
---|
| 191 | match_head = s->prev[(str) & s->w_mask] = s->head[s->ins_h], \ |
---|
| 192 | s->head[s->ins_h] = (Pos)(str)) |
---|
| 193 | #endif |
---|
| 194 | |
---|
| 195 | /* =========================================================================== |
---|
| 196 | * Initialize the hash table (avoiding 64K overflow for 16 bit systems). |
---|
| 197 | * prev[] will be initialized on the fly. |
---|
| 198 | */ |
---|
| 199 | #define CLEAR_HASH(s) \ |
---|
| 200 | s->head[s->hash_size-1] = NIL; \ |
---|
| 201 | zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head)); |
---|
| 202 | |
---|
| 203 | /* ========================================================================= */ |
---|
| 204 | int ZEXPORT deflateInit_(z_streamp strm, int level, const char *version, int stream_size) |
---|
| 205 | { |
---|
| 206 | return deflateInit2_(strm, level, Z_DEFLATED, MAX_WBITS, DEF_MEM_LEVEL, |
---|
| 207 | Z_DEFAULT_STRATEGY, version, stream_size); |
---|
| 208 | /* To do: ignore strm->next_in if we use it as window */ |
---|
| 209 | } |
---|
| 210 | |
---|
| 211 | /* ========================================================================= */ |
---|
| 212 | int ZEXPORT deflateInit2_(z_streamp strm, int level, int method, int windowBits, int memLevel, int strategy, |
---|
| 213 | const char *version, int stream_size) |
---|
| 214 | { |
---|
| 215 | deflate_state *s; |
---|
| 216 | int wrap = 1; |
---|
| 217 | static const char my_version[] = ZLIB_VERSION; |
---|
| 218 | |
---|
| 219 | ushf *overlay; |
---|
| 220 | /* We overlay pending_buf and d_buf+l_buf. This works since the average |
---|
| 221 | * output size for (length,distance) codes is <= 24 bits. |
---|
| 222 | */ |
---|
| 223 | |
---|
| 224 | if (version == Z_NULL || version[0] != my_version[0] || |
---|
| 225 | stream_size != sizeof(z_stream)) { |
---|
| 226 | return Z_VERSION_ERROR; |
---|
| 227 | } |
---|
| 228 | if (strm == Z_NULL) return Z_STREAM_ERROR; |
---|
| 229 | |
---|
| 230 | strm->msg = Z_NULL; |
---|
| 231 | if (strm->zalloc == (alloc_func)0) { |
---|
| 232 | strm->zalloc = zcalloc; |
---|
| 233 | strm->opaque = (voidpf)0; |
---|
| 234 | } |
---|
| 235 | if (strm->zfree == (free_func)0) strm->zfree = zcfree; |
---|
| 236 | |
---|
| 237 | #ifdef FASTEST |
---|
| 238 | if (level != 0) level = 1; |
---|
| 239 | #else |
---|
| 240 | if (level == Z_DEFAULT_COMPRESSION) level = 6; |
---|
| 241 | #endif |
---|
| 242 | |
---|
| 243 | if (windowBits < 0) { /* suppress zlib wrapper */ |
---|
| 244 | wrap = 0; |
---|
| 245 | windowBits = -windowBits; |
---|
| 246 | } |
---|
| 247 | #ifdef GZIP |
---|
| 248 | else if (windowBits > 15) { |
---|
| 249 | wrap = 2; /* write gzip wrapper instead */ |
---|
| 250 | windowBits -= 16; |
---|
| 251 | } |
---|
| 252 | #endif |
---|
| 253 | if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED || |
---|
| 254 | windowBits < 8 || windowBits > 15 || level < 0 || level > 9 || |
---|
| 255 | strategy < 0 || strategy > Z_RLE) { |
---|
| 256 | return Z_STREAM_ERROR; |
---|
| 257 | } |
---|
| 258 | if (windowBits == 8) windowBits = 9; /* until 256-byte window bug fixed */ |
---|
| 259 | s = (deflate_state *) ZALLOC(strm, 1, sizeof(deflate_state)); |
---|
| 260 | if (s == Z_NULL) return Z_MEM_ERROR; |
---|
| 261 | strm->state = (struct internal_state FAR *)s; |
---|
| 262 | s->strm = strm; |
---|
| 263 | |
---|
| 264 | s->wrap = wrap; |
---|
| 265 | s->w_bits = windowBits; |
---|
| 266 | s->w_size = 1 << s->w_bits; |
---|
| 267 | s->w_mask = s->w_size - 1; |
---|
| 268 | |
---|
| 269 | s->hash_bits = memLevel + 7; |
---|
| 270 | s->hash_size = 1 << s->hash_bits; |
---|
| 271 | s->hash_mask = s->hash_size - 1; |
---|
| 272 | s->hash_shift = ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH); |
---|
| 273 | |
---|
| 274 | s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte)); |
---|
| 275 | s->prev = (Posf *) ZALLOC(strm, s->w_size, sizeof(Pos)); |
---|
| 276 | s->head = (Posf *) ZALLOC(strm, s->hash_size, sizeof(Pos)); |
---|
| 277 | |
---|
| 278 | s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */ |
---|
| 279 | |
---|
| 280 | overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2); |
---|
| 281 | s->pending_buf = (uchf *) overlay; |
---|
| 282 | s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L); |
---|
| 283 | |
---|
| 284 | if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL || |
---|
| 285 | s->pending_buf == Z_NULL) { |
---|
| 286 | s->status = FINISH_STATE; |
---|
| 287 | strm->msg = (char*)ERR_MSG(Z_MEM_ERROR); |
---|
| 288 | deflateEnd (strm); |
---|
| 289 | return Z_MEM_ERROR; |
---|
| 290 | } |
---|
| 291 | s->d_buf = overlay + s->lit_bufsize/sizeof(ush); |
---|
| 292 | s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize; |
---|
| 293 | |
---|
| 294 | s->level = level; |
---|
| 295 | s->strategy = strategy; |
---|
| 296 | s->method = (Byte)method; |
---|
| 297 | |
---|
| 298 | return deflateReset(strm); |
---|
| 299 | } |
---|
| 300 | |
---|
| 301 | /* ========================================================================= */ |
---|
| 302 | int ZEXPORT deflateSetDictionary (z_streamp strm, const Bytef *dictionary, uInt dictLength) |
---|
| 303 | { |
---|
| 304 | deflate_state *s; |
---|
| 305 | uInt length = dictLength; |
---|
| 306 | uInt n; |
---|
| 307 | IPos hash_head = 0; |
---|
| 308 | |
---|
| 309 | if (strm == Z_NULL || strm->state == Z_NULL || dictionary == Z_NULL || |
---|
| 310 | strm->state->wrap == 2 || |
---|
| 311 | (strm->state->wrap == 1 && strm->state->status != INIT_STATE)) |
---|
| 312 | return Z_STREAM_ERROR; |
---|
| 313 | |
---|
| 314 | s = strm->state; |
---|
| 315 | if (s->wrap) |
---|
| 316 | strm->adler = adler32(strm->adler, dictionary, dictLength); |
---|
| 317 | |
---|
| 318 | if (length < MIN_MATCH) return Z_OK; |
---|
| 319 | if (length > MAX_DIST(s)) { |
---|
| 320 | length = MAX_DIST(s); |
---|
| 321 | #ifndef USE_DICT_HEAD |
---|
| 322 | dictionary += dictLength - length; /* use the tail of the dictionary */ |
---|
| 323 | #endif |
---|
| 324 | } |
---|
| 325 | zmemcpy(s->window, dictionary, length); |
---|
| 326 | s->strstart = length; |
---|
| 327 | s->block_start = (long)length; |
---|
| 328 | |
---|
| 329 | /* Insert all strings in the hash table (except for the last two bytes). |
---|
| 330 | * s->lookahead stays null, so s->ins_h will be recomputed at the next |
---|
| 331 | * call of fill_window. |
---|
| 332 | */ |
---|
| 333 | s->ins_h = s->window[0]; |
---|
| 334 | UPDATE_HASH(s, s->ins_h, s->window[1]); |
---|
| 335 | for (n = 0; n <= length - MIN_MATCH; n++) { |
---|
| 336 | INSERT_STRING(s, n, hash_head); |
---|
| 337 | } |
---|
| 338 | if (hash_head) hash_head = 0; /* to make compiler happy */ |
---|
| 339 | return Z_OK; |
---|
| 340 | } |
---|
| 341 | |
---|
| 342 | /* ========================================================================= */ |
---|
| 343 | int ZEXPORT deflateReset (z_streamp strm) |
---|
| 344 | { |
---|
| 345 | deflate_state *s; |
---|
| 346 | |
---|
| 347 | if (strm == Z_NULL || strm->state == Z_NULL || |
---|
| 348 | strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0) { |
---|
| 349 | return Z_STREAM_ERROR; |
---|
| 350 | } |
---|
| 351 | |
---|
| 352 | strm->total_in = strm->total_out = 0; |
---|
| 353 | strm->msg = Z_NULL; /* use zfree if we ever allocate msg dynamically */ |
---|
| 354 | strm->data_type = Z_UNKNOWN; |
---|
| 355 | |
---|
| 356 | s = (deflate_state *)strm->state; |
---|
| 357 | s->pending = 0; |
---|
| 358 | s->pending_out = s->pending_buf; |
---|
| 359 | |
---|
| 360 | if (s->wrap < 0) { |
---|
| 361 | s->wrap = -s->wrap; /* was made negative by deflate(..., Z_FINISH); */ |
---|
| 362 | } |
---|
| 363 | s->status = s->wrap ? INIT_STATE : BUSY_STATE; |
---|
| 364 | strm->adler = |
---|
| 365 | #ifdef GZIP |
---|
| 366 | s->wrap == 2 ? crc32(0L, Z_NULL, 0) : |
---|
| 367 | #endif |
---|
| 368 | adler32(0L, Z_NULL, 0); |
---|
| 369 | s->last_flush = Z_NO_FLUSH; |
---|
| 370 | |
---|
| 371 | _tr_init(s); |
---|
| 372 | lm_init(s); |
---|
| 373 | |
---|
| 374 | return Z_OK; |
---|
| 375 | } |
---|
| 376 | |
---|
| 377 | /* ========================================================================= */ |
---|
| 378 | int ZEXPORT deflatePrime (z_streamp strm, int bits, int value) |
---|
| 379 | { |
---|
| 380 | if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR; |
---|
| 381 | strm->state->bi_valid = bits; |
---|
| 382 | strm->state->bi_buf = (ush)(value & ((1 << bits) - 1)); |
---|
| 383 | return Z_OK; |
---|
| 384 | } |
---|
| 385 | |
---|
| 386 | /* ========================================================================= */ |
---|
| 387 | int ZEXPORT deflateParams(z_streamp strm, int level, int strategy) |
---|
| 388 | { |
---|
| 389 | deflate_state *s; |
---|
| 390 | compress_func func; |
---|
| 391 | int err = Z_OK; |
---|
| 392 | |
---|
| 393 | if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR; |
---|
| 394 | s = strm->state; |
---|
| 395 | |
---|
| 396 | #ifdef FASTEST |
---|
| 397 | if (level != 0) level = 1; |
---|
| 398 | #else |
---|
| 399 | if (level == Z_DEFAULT_COMPRESSION) level = 6; |
---|
| 400 | #endif |
---|
| 401 | if (level < 0 || level > 9 || strategy < 0 || strategy > Z_RLE) { |
---|
| 402 | return Z_STREAM_ERROR; |
---|
| 403 | } |
---|
| 404 | func = configuration_table[s->level].func; |
---|
| 405 | |
---|
| 406 | if (func != configuration_table[level].func && strm->total_in != 0) { |
---|
| 407 | /* Flush the last buffer: */ |
---|
| 408 | err = deflate(strm, Z_PARTIAL_FLUSH); |
---|
| 409 | } |
---|
| 410 | if (s->level != level) { |
---|
| 411 | s->level = level; |
---|
| 412 | s->max_lazy_match = configuration_table[level].max_lazy; |
---|
| 413 | s->good_match = configuration_table[level].good_length; |
---|
| 414 | s->nice_match = configuration_table[level].nice_length; |
---|
| 415 | s->max_chain_length = configuration_table[level].max_chain; |
---|
| 416 | } |
---|
| 417 | s->strategy = strategy; |
---|
| 418 | return err; |
---|
| 419 | } |
---|
| 420 | |
---|
| 421 | /* ========================================================================= |
---|
| 422 | * For the default windowBits of 15 and memLevel of 8, this function returns |
---|
| 423 | * a close to exact, as well as small, upper bound on the compressed size. |
---|
| 424 | * They are coded as constants here for a reason--if the #define's are |
---|
| 425 | * changed, then this function needs to be changed as well. The return |
---|
| 426 | * value for 15 and 8 only works for those exact settings. |
---|
| 427 | * |
---|
| 428 | * For any setting other than those defaults for windowBits and memLevel, |
---|
| 429 | * the value returned is a conservative worst case for the maximum expansion |
---|
| 430 | * resulting from using fixed blocks instead of stored blocks, which deflate |
---|
| 431 | * can emit on compressed data for some combinations of the parameters. |
---|
| 432 | * |
---|
| 433 | * This function could be more sophisticated to provide closer upper bounds |
---|
| 434 | * for every combination of windowBits and memLevel, as well as wrap. |
---|
| 435 | * But even the conservative upper bound of about 14% expansion does not |
---|
| 436 | * seem onerous for output buffer allocation. |
---|
| 437 | */ |
---|
| 438 | uLong ZEXPORT deflateBound(z_streamp strm, uLong sourceLen) |
---|
| 439 | { |
---|
| 440 | deflate_state *s; |
---|
| 441 | uLong destLen; |
---|
| 442 | |
---|
| 443 | /* conservative upper bound */ |
---|
| 444 | destLen = sourceLen + |
---|
| 445 | ((sourceLen + 7) >> 3) + ((sourceLen + 63) >> 6) + 11; |
---|
| 446 | |
---|
| 447 | /* if can't get parameters, return conservative bound */ |
---|
| 448 | if (strm == Z_NULL || strm->state == Z_NULL) |
---|
| 449 | return destLen; |
---|
| 450 | |
---|
| 451 | /* if not default parameters, return conservative bound */ |
---|
| 452 | s = strm->state; |
---|
| 453 | if (s->w_bits != 15 || s->hash_bits != 8 + 7) |
---|
| 454 | return destLen; |
---|
| 455 | |
---|
| 456 | /* default settings: return tight bound for that case */ |
---|
| 457 | return compressBound(sourceLen); |
---|
| 458 | } |
---|
| 459 | |
---|
| 460 | /* ========================================================================= |
---|
| 461 | * Put a short in the pending buffer. The 16-bit value is put in MSB order. |
---|
| 462 | * IN assertion: the stream state is correct and there is enough room in |
---|
| 463 | * pending_buf. |
---|
| 464 | */ |
---|
| 465 | local void putShortMSB (deflate_state *s, uInt b) |
---|
| 466 | { |
---|
| 467 | put_byte(s, (Byte)(b >> 8)); |
---|
| 468 | put_byte(s, (Byte)(b & 0xff)); |
---|
| 469 | } |
---|
| 470 | |
---|
| 471 | /* ========================================================================= |
---|
| 472 | * Flush as much pending output as possible. All deflate() output goes |
---|
| 473 | * through this function so some applications may wish to modify it |
---|
| 474 | * to avoid allocating a large strm->next_out buffer and copying into it. |
---|
| 475 | * (See also read_buf()). |
---|
| 476 | */ |
---|
| 477 | local void flush_pending(z_streamp strm) |
---|
| 478 | { |
---|
| 479 | unsigned len = strm->state->pending; |
---|
| 480 | |
---|
| 481 | if (len > strm->avail_out) len = strm->avail_out; |
---|
| 482 | if (len == 0) return; |
---|
| 483 | |
---|
| 484 | zmemcpy(strm->next_out, strm->state->pending_out, len); |
---|
| 485 | strm->next_out += len; |
---|
| 486 | strm->state->pending_out += len; |
---|
| 487 | strm->total_out += len; |
---|
| 488 | strm->avail_out -= len; |
---|
| 489 | strm->state->pending -= len; |
---|
| 490 | if (strm->state->pending == 0) { |
---|
| 491 | strm->state->pending_out = strm->state->pending_buf; |
---|
| 492 | } |
---|
| 493 | } |
---|
| 494 | |
---|
| 495 | /* ========================================================================= */ |
---|
| 496 | int ZEXPORT deflate (z_streamp strm, int flush) |
---|
| 497 | { |
---|
| 498 | int old_flush; /* value of flush param for previous deflate call */ |
---|
| 499 | deflate_state *s; |
---|
| 500 | |
---|
| 501 | if (strm == Z_NULL || strm->state == Z_NULL || |
---|
| 502 | flush > Z_FINISH || flush < 0) { |
---|
| 503 | return Z_STREAM_ERROR; |
---|
| 504 | } |
---|
| 505 | s = strm->state; |
---|
| 506 | |
---|
| 507 | if (strm->next_out == Z_NULL || |
---|
| 508 | (strm->next_in == Z_NULL && strm->avail_in != 0) || |
---|
| 509 | (s->status == FINISH_STATE && flush != Z_FINISH)) { |
---|
| 510 | ERR_RETURN(strm, Z_STREAM_ERROR); |
---|
| 511 | } |
---|
| 512 | if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR); |
---|
| 513 | |
---|
| 514 | s->strm = strm; /* just in case */ |
---|
| 515 | old_flush = s->last_flush; |
---|
| 516 | s->last_flush = flush; |
---|
| 517 | |
---|
| 518 | /* Write the header */ |
---|
| 519 | if (s->status == INIT_STATE) { |
---|
| 520 | #ifdef GZIP |
---|
| 521 | if (s->wrap == 2) { |
---|
| 522 | put_byte(s, 31); |
---|
| 523 | put_byte(s, 139); |
---|
| 524 | put_byte(s, 8); |
---|
| 525 | put_byte(s, 0); |
---|
| 526 | put_byte(s, 0); |
---|
| 527 | put_byte(s, 0); |
---|
| 528 | put_byte(s, 0); |
---|
| 529 | put_byte(s, 0); |
---|
| 530 | put_byte(s, s->level == 9 ? 2 : |
---|
| 531 | (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ? |
---|
| 532 | 4 : 0)); |
---|
| 533 | put_byte(s, 255); |
---|
| 534 | s->status = BUSY_STATE; |
---|
| 535 | strm->adler = crc32(0L, Z_NULL, 0); |
---|
| 536 | } |
---|
| 537 | else |
---|
| 538 | #endif |
---|
| 539 | { |
---|
| 540 | uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8; |
---|
| 541 | uInt level_flags; |
---|
| 542 | |
---|
| 543 | if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2) |
---|
| 544 | level_flags = 0; |
---|
| 545 | else if (s->level < 6) |
---|
| 546 | level_flags = 1; |
---|
| 547 | else if (s->level == 6) |
---|
| 548 | level_flags = 2; |
---|
| 549 | else |
---|
| 550 | level_flags = 3; |
---|
| 551 | header |= (level_flags << 6); |
---|
| 552 | if (s->strstart != 0) header |= PRESET_DICT; |
---|
| 553 | header += 31 - (header % 31); |
---|
| 554 | |
---|
| 555 | s->status = BUSY_STATE; |
---|
| 556 | putShortMSB(s, header); |
---|
| 557 | |
---|
| 558 | /* Save the adler32 of the preset dictionary: */ |
---|
| 559 | if (s->strstart != 0) { |
---|
| 560 | putShortMSB(s, (uInt)(strm->adler >> 16)); |
---|
| 561 | putShortMSB(s, (uInt)(strm->adler & 0xffff)); |
---|
| 562 | } |
---|
| 563 | strm->adler = adler32(0L, Z_NULL, 0); |
---|
| 564 | } |
---|
| 565 | } |
---|
| 566 | |
---|
| 567 | /* Flush as much pending output as possible */ |
---|
| 568 | if (s->pending != 0) { |
---|
| 569 | flush_pending(strm); |
---|
| 570 | if (strm->avail_out == 0) { |
---|
| 571 | /* Since avail_out is 0, deflate will be called again with |
---|
| 572 | * more output space, but possibly with both pending and |
---|
| 573 | * avail_in equal to zero. There won't be anything to do, |
---|
| 574 | * but this is not an error situation so make sure we |
---|
| 575 | * return OK instead of BUF_ERROR at next call of deflate: |
---|
| 576 | */ |
---|
| 577 | s->last_flush = -1; |
---|
| 578 | return Z_OK; |
---|
| 579 | } |
---|
| 580 | |
---|
| 581 | /* Make sure there is something to do and avoid duplicate consecutive |
---|
| 582 | * flushes. For repeated and useless calls with Z_FINISH, we keep |
---|
| 583 | * returning Z_STREAM_END instead of Z_BUF_ERROR. |
---|
| 584 | */ |
---|
| 585 | } else if (strm->avail_in == 0 && flush <= old_flush && |
---|
| 586 | flush != Z_FINISH) { |
---|
| 587 | ERR_RETURN(strm, Z_BUF_ERROR); |
---|
| 588 | } |
---|
| 589 | |
---|
| 590 | /* User must not provide more input after the first FINISH: */ |
---|
| 591 | if (s->status == FINISH_STATE && strm->avail_in != 0) { |
---|
| 592 | ERR_RETURN(strm, Z_BUF_ERROR); |
---|
| 593 | } |
---|
| 594 | |
---|
| 595 | /* Start a new block or continue the current one. |
---|
| 596 | */ |
---|
| 597 | if (strm->avail_in != 0 || s->lookahead != 0 || |
---|
| 598 | (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) { |
---|
| 599 | block_state bstate; |
---|
| 600 | |
---|
| 601 | bstate = (*(configuration_table[s->level].func))(s, flush); |
---|
| 602 | |
---|
| 603 | if (bstate == finish_started || bstate == finish_done) { |
---|
| 604 | s->status = FINISH_STATE; |
---|
| 605 | } |
---|
| 606 | if (bstate == need_more || bstate == finish_started) { |
---|
| 607 | if (strm->avail_out == 0) { |
---|
| 608 | s->last_flush = -1; /* avoid BUF_ERROR next call, see above */ |
---|
| 609 | } |
---|
| 610 | return Z_OK; |
---|
| 611 | /* If flush != Z_NO_FLUSH && avail_out == 0, the next call |
---|
| 612 | * of deflate should use the same flush parameter to make sure |
---|
| 613 | * that the flush is complete. So we don't have to output an |
---|
| 614 | * empty block here, this will be done at next call. This also |
---|
| 615 | * ensures that for a very small output buffer, we emit at most |
---|
| 616 | * one empty block. |
---|
| 617 | */ |
---|
| 618 | } |
---|
| 619 | if (bstate == block_done) { |
---|
| 620 | if (flush == Z_PARTIAL_FLUSH) { |
---|
| 621 | _tr_align(s); |
---|
| 622 | } else { /* FULL_FLUSH or SYNC_FLUSH */ |
---|
| 623 | _tr_stored_block(s, (char*)0, 0L, 0); |
---|
| 624 | /* For a full flush, this empty block will be recognized |
---|
| 625 | * as a special marker by inflate_sync(). |
---|
| 626 | */ |
---|
| 627 | if (flush == Z_FULL_FLUSH) { |
---|
| 628 | CLEAR_HASH(s); /* forget history */ |
---|
| 629 | } |
---|
| 630 | } |
---|
| 631 | flush_pending(strm); |
---|
| 632 | if (strm->avail_out == 0) { |
---|
| 633 | s->last_flush = -1; /* avoid BUF_ERROR at next call, see above */ |
---|
| 634 | return Z_OK; |
---|
| 635 | } |
---|
| 636 | } |
---|
| 637 | } |
---|
| 638 | Assert(strm->avail_out > 0, (char*)"bug2"); |
---|
| 639 | |
---|
| 640 | if (flush != Z_FINISH) return Z_OK; |
---|
| 641 | if (s->wrap <= 0) return Z_STREAM_END; |
---|
| 642 | |
---|
| 643 | /* Write the trailer */ |
---|
| 644 | #ifdef GZIP |
---|
| 645 | if (s->wrap == 2) { |
---|
| 646 | put_byte(s, (Byte)(strm->adler & 0xff)); |
---|
| 647 | put_byte(s, (Byte)((strm->adler >> 8) & 0xff)); |
---|
| 648 | put_byte(s, (Byte)((strm->adler >> 16) & 0xff)); |
---|
| 649 | put_byte(s, (Byte)((strm->adler >> 24) & 0xff)); |
---|
| 650 | put_byte(s, (Byte)(strm->total_in & 0xff)); |
---|
| 651 | put_byte(s, (Byte)((strm->total_in >> 8) & 0xff)); |
---|
| 652 | put_byte(s, (Byte)((strm->total_in >> 16) & 0xff)); |
---|
| 653 | put_byte(s, (Byte)((strm->total_in >> 24) & 0xff)); |
---|
| 654 | } |
---|
| 655 | else |
---|
| 656 | #endif |
---|
| 657 | { |
---|
| 658 | putShortMSB(s, (uInt)(strm->adler >> 16)); |
---|
| 659 | putShortMSB(s, (uInt)(strm->adler & 0xffff)); |
---|
| 660 | } |
---|
| 661 | flush_pending(strm); |
---|
| 662 | /* If avail_out is zero, the application will call deflate again |
---|
| 663 | * to flush the rest. |
---|
| 664 | */ |
---|
| 665 | if (s->wrap > 0) s->wrap = -s->wrap; /* write the trailer only once! */ |
---|
| 666 | return s->pending != 0 ? Z_OK : Z_STREAM_END; |
---|
| 667 | } |
---|
| 668 | |
---|
| 669 | /* ========================================================================= */ |
---|
| 670 | int ZEXPORT deflateEnd (z_streamp strm) |
---|
| 671 | { |
---|
| 672 | int status; |
---|
| 673 | |
---|
| 674 | if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR; |
---|
| 675 | |
---|
| 676 | status = strm->state->status; |
---|
| 677 | if (status != INIT_STATE && status != BUSY_STATE && |
---|
| 678 | status != FINISH_STATE) { |
---|
| 679 | return Z_STREAM_ERROR; |
---|
| 680 | } |
---|
| 681 | |
---|
| 682 | /* Deallocate in reverse order of allocations: */ |
---|
| 683 | TRY_FREE(strm, strm->state->pending_buf); |
---|
| 684 | TRY_FREE(strm, strm->state->head); |
---|
| 685 | TRY_FREE(strm, strm->state->prev); |
---|
| 686 | TRY_FREE(strm, strm->state->window); |
---|
| 687 | |
---|
| 688 | ZFREE(strm, strm->state); |
---|
| 689 | strm->state = Z_NULL; |
---|
| 690 | |
---|
| 691 | return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK; |
---|
| 692 | } |
---|
| 693 | |
---|
| 694 | /* ========================================================================= |
---|
| 695 | * Copy the source state to the destination state. |
---|
| 696 | * To simplify the source, this is not supported for 16-bit MSDOS (which |
---|
| 697 | * doesn't have enough memory anyway to duplicate compression states). |
---|
| 698 | */ |
---|
| 699 | int ZEXPORT deflateCopy (z_streamp dest, z_streamp source) |
---|
| 700 | { |
---|
| 701 | #ifdef MAXSEG_64K |
---|
| 702 | return Z_STREAM_ERROR; |
---|
| 703 | #else |
---|
| 704 | deflate_state *ds; |
---|
| 705 | deflate_state *ss; |
---|
| 706 | ushf *overlay; |
---|
| 707 | |
---|
| 708 | |
---|
| 709 | if (source == Z_NULL || dest == Z_NULL || source->state == Z_NULL) { |
---|
| 710 | return Z_STREAM_ERROR; |
---|
| 711 | } |
---|
| 712 | |
---|
| 713 | ss = source->state; |
---|
| 714 | |
---|
| 715 | *dest = *source; |
---|
| 716 | |
---|
| 717 | ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state)); |
---|
| 718 | if (ds == Z_NULL) return Z_MEM_ERROR; |
---|
| 719 | dest->state = (struct internal_state FAR *) ds; |
---|
| 720 | *ds = *ss; |
---|
| 721 | ds->strm = dest; |
---|
| 722 | |
---|
| 723 | ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte)); |
---|
| 724 | ds->prev = (Posf *) ZALLOC(dest, ds->w_size, sizeof(Pos)); |
---|
| 725 | ds->head = (Posf *) ZALLOC(dest, ds->hash_size, sizeof(Pos)); |
---|
| 726 | overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2); |
---|
| 727 | ds->pending_buf = (uchf *) overlay; |
---|
| 728 | |
---|
| 729 | if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL || |
---|
| 730 | ds->pending_buf == Z_NULL) { |
---|
| 731 | deflateEnd (dest); |
---|
| 732 | return Z_MEM_ERROR; |
---|
| 733 | } |
---|
| 734 | /* following zmemcpy do not work for 16-bit MSDOS */ |
---|
| 735 | zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte)); |
---|
| 736 | zmemcpy(ds->prev, ss->prev, ds->w_size * sizeof(Pos)); |
---|
| 737 | zmemcpy(ds->head, ss->head, ds->hash_size * sizeof(Pos)); |
---|
| 738 | zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size); |
---|
| 739 | |
---|
| 740 | ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf); |
---|
| 741 | ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush); |
---|
| 742 | ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize; |
---|
| 743 | |
---|
| 744 | ds->l_desc.dyn_tree = ds->dyn_ltree; |
---|
| 745 | ds->d_desc.dyn_tree = ds->dyn_dtree; |
---|
| 746 | ds->bl_desc.dyn_tree = ds->bl_tree; |
---|
| 747 | |
---|
| 748 | return Z_OK; |
---|
| 749 | #endif /* MAXSEG_64K */ |
---|
| 750 | } |
---|
| 751 | |
---|
| 752 | /* =========================================================================== |
---|
| 753 | * Read a new buffer from the current input stream, update the adler32 |
---|
| 754 | * and total number of bytes read. All deflate() input goes through |
---|
| 755 | * this function so some applications may wish to modify it to avoid |
---|
| 756 | * allocating a large strm->next_in buffer and copying from it. |
---|
| 757 | * (See also flush_pending()). |
---|
| 758 | */ |
---|
| 759 | local int read_buf(z_streamp strm, Bytef *buf, unsigned size) |
---|
| 760 | { |
---|
| 761 | unsigned len = strm->avail_in; |
---|
| 762 | |
---|
| 763 | if (len > size) len = size; |
---|
| 764 | if (len == 0) return 0; |
---|
| 765 | |
---|
| 766 | strm->avail_in -= len; |
---|
| 767 | |
---|
| 768 | if (strm->state->wrap == 1) { |
---|
| 769 | strm->adler = adler32(strm->adler, strm->next_in, len); |
---|
| 770 | } |
---|
| 771 | #ifdef GZIP |
---|
| 772 | else if (strm->state->wrap == 2) { |
---|
| 773 | strm->adler = crc32(strm->adler, strm->next_in, len); |
---|
| 774 | } |
---|
| 775 | #endif |
---|
| 776 | zmemcpy(buf, strm->next_in, len); |
---|
| 777 | strm->next_in += len; |
---|
| 778 | strm->total_in += len; |
---|
| 779 | |
---|
| 780 | return (int)len; |
---|
| 781 | } |
---|
| 782 | |
---|
| 783 | /* =========================================================================== |
---|
| 784 | * Initialize the "longest match" routines for a new zlib stream |
---|
| 785 | */ |
---|
| 786 | local void lm_init (deflate_state *s) |
---|
| 787 | { |
---|
| 788 | s->window_size = (ulg)2L*s->w_size; |
---|
| 789 | |
---|
| 790 | CLEAR_HASH(s); |
---|
| 791 | |
---|
| 792 | /* Set the default configuration parameters: |
---|
| 793 | */ |
---|
| 794 | s->max_lazy_match = configuration_table[s->level].max_lazy; |
---|
| 795 | s->good_match = configuration_table[s->level].good_length; |
---|
| 796 | s->nice_match = configuration_table[s->level].nice_length; |
---|
| 797 | s->max_chain_length = configuration_table[s->level].max_chain; |
---|
| 798 | |
---|
| 799 | s->strstart = 0; |
---|
| 800 | s->block_start = 0L; |
---|
| 801 | s->lookahead = 0; |
---|
| 802 | s->match_length = s->prev_length = MIN_MATCH-1; |
---|
| 803 | s->match_available = 0; |
---|
| 804 | s->ins_h = 0; |
---|
| 805 | #ifdef ASMV |
---|
| 806 | match_init(); /* initialize the asm code */ |
---|
| 807 | #endif |
---|
| 808 | } |
---|
| 809 | |
---|
| 810 | #ifndef FASTEST |
---|
| 811 | /* =========================================================================== |
---|
| 812 | * Set match_start to the longest match starting at the given string and |
---|
| 813 | * return its length. Matches shorter or equal to prev_length are discarded, |
---|
| 814 | * in which case the result is equal to prev_length and match_start is |
---|
| 815 | * garbage. |
---|
| 816 | * IN assertions: cur_match is the head of the hash chain for the current |
---|
| 817 | * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1 |
---|
| 818 | * OUT assertion: the match length is not greater than s->lookahead. |
---|
| 819 | */ |
---|
| 820 | #ifndef ASMV |
---|
| 821 | /* For 80x86 and 680x0, an optimized version will be provided in match.asm or |
---|
| 822 | * match.S. The code will be functionally equivalent. |
---|
| 823 | */ |
---|
| 824 | local uInt longest_match(deflate_state *s, IPos cur_match) |
---|
| 825 | { |
---|
| 826 | unsigned chain_length = s->max_chain_length;/* max hash chain length */ |
---|
| 827 | register Bytef *scan = s->window + s->strstart; /* current string */ |
---|
| 828 | register Bytef *match; /* matched string */ |
---|
| 829 | register int len; /* length of current match */ |
---|
| 830 | int best_len = s->prev_length; /* best match length so far */ |
---|
| 831 | int nice_match = s->nice_match; /* stop if match long enough */ |
---|
| 832 | IPos limit = s->strstart > (IPos)MAX_DIST(s) ? |
---|
| 833 | s->strstart - (IPos)MAX_DIST(s) : NIL; |
---|
| 834 | /* Stop when cur_match becomes <= limit. To simplify the code, |
---|
| 835 | * we prevent matches with the string of window index 0. |
---|
| 836 | */ |
---|
| 837 | Posf *prev = s->prev; |
---|
| 838 | uInt wmask = s->w_mask; |
---|
| 839 | |
---|
| 840 | #ifdef UNALIGNED_OK |
---|
| 841 | /* Compare two bytes at a time. Note: this is not always beneficial. |
---|
| 842 | * Try with and without -DUNALIGNED_OK to check. |
---|
| 843 | */ |
---|
| 844 | register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1; |
---|
| 845 | register ush scan_start = *(ushf*)scan; |
---|
| 846 | register ush scan_end = *(ushf*)(scan+best_len-1); |
---|
| 847 | #else |
---|
| 848 | register Bytef *strend = s->window + s->strstart + MAX_MATCH; |
---|
| 849 | register Byte scan_end1 = scan[best_len-1]; |
---|
| 850 | register Byte scan_end = scan[best_len]; |
---|
| 851 | #endif |
---|
| 852 | |
---|
| 853 | /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. |
---|
| 854 | * It is easy to get rid of this optimization if necessary. |
---|
| 855 | */ |
---|
| 856 | Assert(s->hash_bits >= 8 && MAX_MATCH == 258, (char*)"Code too clever"); |
---|
| 857 | |
---|
| 858 | /* Do not waste too much time if we already have a good match: */ |
---|
| 859 | if (s->prev_length >= s->good_match) { |
---|
| 860 | chain_length >>= 2; |
---|
| 861 | } |
---|
| 862 | /* Do not look for matches beyond the end of the input. This is necessary |
---|
| 863 | * to make deflate deterministic. |
---|
| 864 | */ |
---|
| 865 | if ((uInt)nice_match > s->lookahead) nice_match = s->lookahead; |
---|
| 866 | |
---|
| 867 | Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, (char*)"need lookahead"); |
---|
| 868 | |
---|
| 869 | do { |
---|
| 870 | Assert(cur_match < s->strstart, (char*)"no future"); |
---|
| 871 | match = s->window + cur_match; |
---|
| 872 | |
---|
| 873 | /* Skip to next match if the match length cannot increase |
---|
| 874 | * or if the match length is less than 2: |
---|
| 875 | */ |
---|
| 876 | #if (defined(UNALIGNED_OK) && MAX_MATCH == 258) |
---|
| 877 | /* This code assumes sizeof(unsigned short) == 2. Do not use |
---|
| 878 | * UNALIGNED_OK if your compiler uses a different size. |
---|
| 879 | */ |
---|
| 880 | if (*(ushf*)(match+best_len-1) != scan_end || |
---|
| 881 | *(ushf*)match != scan_start) continue; |
---|
| 882 | |
---|
| 883 | /* It is not necessary to compare scan[2] and match[2] since they are |
---|
| 884 | * always equal when the other bytes match, given that the hash keys |
---|
| 885 | * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at |
---|
| 886 | * strstart+3, +5, ... up to strstart+257. We check for insufficient |
---|
| 887 | * lookahead only every 4th comparison; the 128th check will be made |
---|
| 888 | * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is |
---|
| 889 | * necessary to put more guard bytes at the end of the window, or |
---|
| 890 | * to check more often for insufficient lookahead. |
---|
| 891 | */ |
---|
| 892 | Assert(scan[2] == match[2], (char*)"scan[2]?"); |
---|
| 893 | scan++, match++; |
---|
| 894 | do { |
---|
| 895 | } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) && |
---|
| 896 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && |
---|
| 897 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && |
---|
| 898 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && |
---|
| 899 | scan < strend); |
---|
| 900 | /* The funny "do {}" generates better code on most compilers */ |
---|
| 901 | |
---|
| 902 | /* Here, scan <= window+strstart+257 */ |
---|
| 903 | Assert(scan <= s->window+(unsigned)(s->window_size-1), (char*)"wild scan"); |
---|
| 904 | if (*scan == *match) scan++; |
---|
| 905 | |
---|
| 906 | len = (MAX_MATCH - 1) - (int)(strend-scan); |
---|
| 907 | scan = strend - (MAX_MATCH-1); |
---|
| 908 | |
---|
| 909 | #else /* UNALIGNED_OK */ |
---|
| 910 | |
---|
| 911 | if (match[best_len] != scan_end || |
---|
| 912 | match[best_len-1] != scan_end1 || |
---|
| 913 | *match != *scan || |
---|
| 914 | *++match != scan[1]) continue; |
---|
| 915 | |
---|
| 916 | /* The check at best_len-1 can be removed because it will be made |
---|
| 917 | * again later. (This heuristic is not always a win.) |
---|
| 918 | * It is not necessary to compare scan[2] and match[2] since they |
---|
| 919 | * are always equal when the other bytes match, given that |
---|
| 920 | * the hash keys are equal and that HASH_BITS >= 8. |
---|
| 921 | */ |
---|
| 922 | scan += 2, match++; |
---|
| 923 | Assert(*scan == *match, (char*)"match[2]?"); |
---|
| 924 | |
---|
| 925 | /* We check for insufficient lookahead only every 8th comparison; |
---|
| 926 | * the 256th check will be made at strstart+258. |
---|
| 927 | */ |
---|
| 928 | do { |
---|
| 929 | } while (*++scan == *++match && *++scan == *++match && |
---|
| 930 | *++scan == *++match && *++scan == *++match && |
---|
| 931 | *++scan == *++match && *++scan == *++match && |
---|
| 932 | *++scan == *++match && *++scan == *++match && |
---|
| 933 | scan < strend); |
---|
| 934 | |
---|
| 935 | Assert(scan <= s->window+(unsigned)(s->window_size-1), (char*)"wild scan"); |
---|
| 936 | |
---|
| 937 | len = MAX_MATCH - (int)(strend - scan); |
---|
| 938 | scan = strend - MAX_MATCH; |
---|
| 939 | |
---|
| 940 | #endif /* UNALIGNED_OK */ |
---|
| 941 | |
---|
| 942 | if (len > best_len) { |
---|
| 943 | s->match_start = cur_match; |
---|
| 944 | best_len = len; |
---|
| 945 | if (len >= nice_match) break; |
---|
| 946 | #ifdef UNALIGNED_OK |
---|
| 947 | scan_end = *(ushf*)(scan+best_len-1); |
---|
| 948 | #else |
---|
| 949 | scan_end1 = scan[best_len-1]; |
---|
| 950 | scan_end = scan[best_len]; |
---|
| 951 | #endif |
---|
| 952 | } |
---|
| 953 | } while ((cur_match = prev[cur_match & wmask]) > limit |
---|
| 954 | && --chain_length != 0); |
---|
| 955 | |
---|
| 956 | if ((uInt)best_len <= s->lookahead) return (uInt)best_len; |
---|
| 957 | return s->lookahead; |
---|
| 958 | } |
---|
| 959 | #endif /* ASMV */ |
---|
| 960 | #endif /* FASTEST */ |
---|
| 961 | |
---|
| 962 | /* --------------------------------------------------------------------------- |
---|
| 963 | * Optimized version for level == 1 or strategy == Z_RLE only |
---|
| 964 | */ |
---|
| 965 | local uInt longest_match_fast(deflate_state *s, IPos cur_match) |
---|
| 966 | { |
---|
| 967 | register Bytef *scan = s->window + s->strstart; /* current string */ |
---|
| 968 | register Bytef *match; /* matched string */ |
---|
| 969 | register int len; /* length of current match */ |
---|
| 970 | register Bytef *strend = s->window + s->strstart + MAX_MATCH; |
---|
| 971 | |
---|
| 972 | /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. |
---|
| 973 | * It is easy to get rid of this optimization if necessary. |
---|
| 974 | */ |
---|
| 975 | Assert(s->hash_bits >= 8 && MAX_MATCH == 258, (char*)"Code too clever"); |
---|
| 976 | |
---|
| 977 | Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, (char*)"need lookahead"); |
---|
| 978 | |
---|
| 979 | Assert(cur_match < s->strstart, (char*)"no future"); |
---|
| 980 | |
---|
| 981 | match = s->window + cur_match; |
---|
| 982 | |
---|
| 983 | /* Return failure if the match length is less than 2: |
---|
| 984 | */ |
---|
| 985 | if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1; |
---|
| 986 | |
---|
| 987 | /* The check at best_len-1 can be removed because it will be made |
---|
| 988 | * again later. (This heuristic is not always a win.) |
---|
| 989 | * It is not necessary to compare scan[2] and match[2] since they |
---|
| 990 | * are always equal when the other bytes match, given that |
---|
| 991 | * the hash keys are equal and that HASH_BITS >= 8. |
---|
| 992 | */ |
---|
| 993 | scan += 2, match += 2; |
---|
| 994 | Assert(*scan == *match, (char*)"match[2]?"); |
---|
| 995 | |
---|
| 996 | /* We check for insufficient lookahead only every 8th comparison; |
---|
| 997 | * the 256th check will be made at strstart+258. |
---|
| 998 | */ |
---|
| 999 | do { |
---|
| 1000 | } while (*++scan == *++match && *++scan == *++match && |
---|
| 1001 | *++scan == *++match && *++scan == *++match && |
---|
| 1002 | *++scan == *++match && *++scan == *++match && |
---|
| 1003 | *++scan == *++match && *++scan == *++match && |
---|
| 1004 | scan < strend); |
---|
| 1005 | |
---|
| 1006 | Assert(scan <= s->window+(unsigned)(s->window_size-1), (char*)"wild scan"); |
---|
| 1007 | |
---|
| 1008 | len = MAX_MATCH - (int)(strend - scan); |
---|
| 1009 | |
---|
| 1010 | if (len < MIN_MATCH) return MIN_MATCH - 1; |
---|
| 1011 | |
---|
| 1012 | s->match_start = cur_match; |
---|
| 1013 | return (uInt)len <= s->lookahead ? (uInt)len : s->lookahead; |
---|
| 1014 | } |
---|
| 1015 | |
---|
| 1016 | #ifdef DEBUG |
---|
| 1017 | /* =========================================================================== |
---|
| 1018 | * Check that the match at match_start is indeed a match. |
---|
| 1019 | */ |
---|
| 1020 | local void check_match(deflate_state *s, IPos start, IPos match, int length) |
---|
| 1021 | { |
---|
| 1022 | /* check that the match is indeed a match */ |
---|
| 1023 | if (zmemcmp(s->window + match, |
---|
| 1024 | s->window + start, length) != EQUAL) { |
---|
| 1025 | fprintf(stderr, (char*)" start %u, match %u, length %d\n", |
---|
| 1026 | start, match, length); |
---|
| 1027 | do { |
---|
| 1028 | fprintf(stderr, (char*)"%c%c", s->window[match++], s->window[start++]); |
---|
| 1029 | } while (--length != 0); |
---|
| 1030 | z_error((char*)"invalid match"); |
---|
| 1031 | } |
---|
| 1032 | if (z_verbose > 1) { |
---|
| 1033 | fprintf(stderr,(char*)"\\[%d,%d]", start-match, length); |
---|
| 1034 | do { putc(s->window[start++], stderr); } while (--length != 0); |
---|
| 1035 | } |
---|
| 1036 | } |
---|
| 1037 | #else |
---|
| 1038 | # define check_match(s, start, match, length) |
---|
| 1039 | #endif /* DEBUG */ |
---|
| 1040 | |
---|
| 1041 | /* =========================================================================== |
---|
| 1042 | * Fill the window when the lookahead becomes insufficient. |
---|
| 1043 | * Updates strstart and lookahead. |
---|
| 1044 | * |
---|
| 1045 | * IN assertion: lookahead < MIN_LOOKAHEAD |
---|
| 1046 | * OUT assertions: strstart <= window_size-MIN_LOOKAHEAD |
---|
| 1047 | * At least one byte has been read, or avail_in == 0; reads are |
---|
| 1048 | * performed for at least two bytes (required for the zip translate_eol |
---|
| 1049 | * option -- not supported here). |
---|
| 1050 | */ |
---|
| 1051 | local void fill_window(deflate_state *s) |
---|
| 1052 | { |
---|
| 1053 | register unsigned n, m; |
---|
| 1054 | register Posf *p; |
---|
| 1055 | unsigned more; /* Amount of free space at the end of the window. */ |
---|
| 1056 | uInt wsize = s->w_size; |
---|
| 1057 | |
---|
| 1058 | do { |
---|
| 1059 | more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart); |
---|
| 1060 | |
---|
| 1061 | /* Deal with !@#$% 64K limit: */ |
---|
| 1062 | if (sizeof(int) <= 2) { |
---|
| 1063 | if (more == 0 && s->strstart == 0 && s->lookahead == 0) { |
---|
| 1064 | more = wsize; |
---|
| 1065 | |
---|
| 1066 | } else if (more == (unsigned)(-1)) { |
---|
| 1067 | /* Very unlikely, but possible on 16 bit machine if |
---|
| 1068 | * strstart == 0 && lookahead == 1 (input done a byte at time) |
---|
| 1069 | */ |
---|
| 1070 | more--; |
---|
| 1071 | } |
---|
| 1072 | } |
---|
| 1073 | |
---|
| 1074 | /* If the window is almost full and there is insufficient lookahead, |
---|
| 1075 | * move the upper half to the lower one to make room in the upper half. |
---|
| 1076 | */ |
---|
| 1077 | if (s->strstart >= wsize+MAX_DIST(s)) { |
---|
| 1078 | |
---|
| 1079 | zmemcpy(s->window, s->window+wsize, (unsigned)wsize); |
---|
| 1080 | s->match_start -= wsize; |
---|
| 1081 | s->strstart -= wsize; /* we now have strstart >= MAX_DIST */ |
---|
| 1082 | s->block_start -= (long) wsize; |
---|
| 1083 | |
---|
| 1084 | /* Slide the hash table (could be avoided with 32 bit values |
---|
| 1085 | at the expense of memory usage). We slide even when level == 0 |
---|
| 1086 | to keep the hash table consistent if we switch back to level > 0 |
---|
| 1087 | later. (Using level 0 permanently is not an optimal usage of |
---|
| 1088 | zlib, so we don't care about this pathological case.) |
---|
| 1089 | */ |
---|
| 1090 | n = s->hash_size; |
---|
| 1091 | p = &s->head[n]; |
---|
| 1092 | do { |
---|
| 1093 | m = *--p; |
---|
| 1094 | *p = (Pos)(m >= wsize ? m-wsize : NIL); |
---|
| 1095 | } while (--n); |
---|
| 1096 | |
---|
| 1097 | n = wsize; |
---|
| 1098 | #ifndef FASTEST |
---|
| 1099 | p = &s->prev[n]; |
---|
| 1100 | do { |
---|
| 1101 | m = *--p; |
---|
| 1102 | *p = (Pos)(m >= wsize ? m-wsize : NIL); |
---|
| 1103 | /* If n is not on any hash chain, prev[n] is garbage but |
---|
| 1104 | * its value will never be used. |
---|
| 1105 | */ |
---|
| 1106 | } while (--n); |
---|
| 1107 | #endif |
---|
| 1108 | more += wsize; |
---|
| 1109 | } |
---|
| 1110 | if (s->strm->avail_in == 0) return; |
---|
| 1111 | |
---|
| 1112 | /* If there was no sliding: |
---|
| 1113 | * strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 && |
---|
| 1114 | * more == window_size - lookahead - strstart |
---|
| 1115 | * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1) |
---|
| 1116 | * => more >= window_size - 2*WSIZE + 2 |
---|
| 1117 | * In the BIG_MEM or MMAP case (not yet supported), |
---|
| 1118 | * window_size == input_size + MIN_LOOKAHEAD && |
---|
| 1119 | * strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD. |
---|
| 1120 | * Otherwise, window_size == 2*WSIZE so more >= 2. |
---|
| 1121 | * If there was sliding, more >= WSIZE. So in all cases, more >= 2. |
---|
| 1122 | */ |
---|
| 1123 | Assert(more >= 2, (char*)"more < 2"); |
---|
| 1124 | |
---|
| 1125 | n = read_buf(s->strm, s->window + s->strstart + s->lookahead, more); |
---|
| 1126 | s->lookahead += n; |
---|
| 1127 | |
---|
| 1128 | /* Initialize the hash value now that we have some input: */ |
---|
| 1129 | if (s->lookahead >= MIN_MATCH) { |
---|
| 1130 | s->ins_h = s->window[s->strstart]; |
---|
| 1131 | UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]); |
---|
| 1132 | #if MIN_MATCH != 3 |
---|
| 1133 | Call UPDATE_HASH() MIN_MATCH-3 more times |
---|
| 1134 | #endif |
---|
| 1135 | } |
---|
| 1136 | /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage, |
---|
| 1137 | * but this is not important since only literal bytes will be emitted. |
---|
| 1138 | */ |
---|
| 1139 | |
---|
| 1140 | } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0); |
---|
| 1141 | } |
---|
| 1142 | |
---|
| 1143 | /* =========================================================================== |
---|
| 1144 | * Flush the current block, with given end-of-file flag. |
---|
| 1145 | * IN assertion: strstart is set to the end of the current match. |
---|
| 1146 | */ |
---|
| 1147 | #define FLUSH_BLOCK_ONLY(s, eof) { \ |
---|
| 1148 | _tr_flush_block(s, (s->block_start >= 0L ? \ |
---|
| 1149 | (charf *)&s->window[(unsigned)s->block_start] : \ |
---|
| 1150 | (charf *)Z_NULL), \ |
---|
| 1151 | (ulg)((long)s->strstart - s->block_start), \ |
---|
| 1152 | (eof)); \ |
---|
| 1153 | s->block_start = s->strstart; \ |
---|
| 1154 | flush_pending(s->strm); \ |
---|
| 1155 | Tracev((stderr,(char*)"[FLUSH]")); \ |
---|
| 1156 | } |
---|
| 1157 | |
---|
| 1158 | /* Same but force premature exit if necessary. */ |
---|
| 1159 | #define FLUSH_BLOCK(s, eof) { \ |
---|
| 1160 | FLUSH_BLOCK_ONLY(s, eof); \ |
---|
| 1161 | if (s->strm->avail_out == 0) return (eof) ? finish_started : need_more; \ |
---|
| 1162 | } |
---|
| 1163 | |
---|
| 1164 | /* =========================================================================== |
---|
| 1165 | * Copy without compression as much as possible from the input stream, return |
---|
| 1166 | * the current block state. |
---|
| 1167 | * This function does not insert new strings in the dictionary since |
---|
| 1168 | * uncompressible data is probably not useful. This function is used |
---|
| 1169 | * only for the level=0 compression option. |
---|
| 1170 | * NOTE: this function should be optimized to avoid extra copying from |
---|
| 1171 | * window to pending_buf. |
---|
| 1172 | */ |
---|
| 1173 | local block_state deflate_stored(deflate_state *s, int flush) |
---|
| 1174 | { |
---|
| 1175 | /* Stored blocks are limited to 0xffff bytes, pending_buf is limited |
---|
| 1176 | * to pending_buf_size, and each stored block has a 5 byte header: |
---|
| 1177 | */ |
---|
| 1178 | ulg max_block_size = 0xffff; |
---|
| 1179 | ulg max_start; |
---|
| 1180 | |
---|
| 1181 | if (max_block_size > s->pending_buf_size - 5) { |
---|
| 1182 | max_block_size = s->pending_buf_size - 5; |
---|
| 1183 | } |
---|
| 1184 | |
---|
| 1185 | /* Copy as much as possible from input to output: */ |
---|
| 1186 | for (;;) { |
---|
| 1187 | /* Fill the window as much as possible: */ |
---|
| 1188 | if (s->lookahead <= 1) { |
---|
| 1189 | |
---|
| 1190 | Assert(s->strstart < s->w_size+MAX_DIST(s) || |
---|
| 1191 | s->block_start >= (long)s->w_size, (char*)"slide too late"); |
---|
| 1192 | |
---|
| 1193 | fill_window(s); |
---|
| 1194 | if (s->lookahead == 0 && flush == Z_NO_FLUSH) return need_more; |
---|
| 1195 | |
---|
| 1196 | if (s->lookahead == 0) break; /* flush the current block */ |
---|
| 1197 | } |
---|
| 1198 | Assert(s->block_start >= 0L, (char*)"block gone"); |
---|
| 1199 | |
---|
| 1200 | s->strstart += s->lookahead; |
---|
| 1201 | s->lookahead = 0; |
---|
| 1202 | |
---|
| 1203 | /* Emit a stored block if pending_buf will be full: */ |
---|
| 1204 | max_start = s->block_start + max_block_size; |
---|
| 1205 | if (s->strstart == 0 || (ulg)s->strstart >= max_start) { |
---|
| 1206 | /* strstart == 0 is possible when wraparound on 16-bit machine */ |
---|
| 1207 | s->lookahead = (uInt)(s->strstart - max_start); |
---|
| 1208 | s->strstart = (uInt)max_start; |
---|
| 1209 | FLUSH_BLOCK(s, 0); |
---|
| 1210 | } |
---|
| 1211 | /* Flush if we may have to slide, otherwise block_start may become |
---|
| 1212 | * negative and the data will be gone: |
---|
| 1213 | */ |
---|
| 1214 | if (s->strstart - (uInt)s->block_start >= MAX_DIST(s)) { |
---|
| 1215 | FLUSH_BLOCK(s, 0); |
---|
| 1216 | } |
---|
| 1217 | } |
---|
| 1218 | FLUSH_BLOCK(s, flush == Z_FINISH); |
---|
| 1219 | return flush == Z_FINISH ? finish_done : block_done; |
---|
| 1220 | } |
---|
| 1221 | |
---|
| 1222 | /* =========================================================================== |
---|
| 1223 | * Compress as much as possible from the input stream, return the current |
---|
| 1224 | * block state. |
---|
| 1225 | * This function does not perform lazy evaluation of matches and inserts |
---|
| 1226 | * new strings in the dictionary only for unmatched strings or for short |
---|
| 1227 | * matches. It is used only for the fast compression options. |
---|
| 1228 | */ |
---|
| 1229 | local block_state deflate_fast(deflate_state *s, int flush) |
---|
| 1230 | { |
---|
| 1231 | IPos hash_head = NIL; /* head of the hash chain */ |
---|
| 1232 | int bflush; /* set if current block must be flushed */ |
---|
| 1233 | |
---|
| 1234 | for (;;) { |
---|
| 1235 | /* Make sure that we always have enough lookahead, except |
---|
| 1236 | * at the end of the input file. We need MAX_MATCH bytes |
---|
| 1237 | * for the next match, plus MIN_MATCH bytes to insert the |
---|
| 1238 | * string following the next match. |
---|
| 1239 | */ |
---|
| 1240 | if (s->lookahead < MIN_LOOKAHEAD) { |
---|
| 1241 | fill_window(s); |
---|
| 1242 | if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { |
---|
| 1243 | return need_more; |
---|
| 1244 | } |
---|
| 1245 | if (s->lookahead == 0) break; /* flush the current block */ |
---|
| 1246 | } |
---|
| 1247 | |
---|
| 1248 | /* Insert the string window[strstart .. strstart+2] in the |
---|
| 1249 | * dictionary, and set hash_head to the head of the hash chain: |
---|
| 1250 | */ |
---|
| 1251 | if (s->lookahead >= MIN_MATCH) { |
---|
| 1252 | INSERT_STRING(s, s->strstart, hash_head); |
---|
| 1253 | } |
---|
| 1254 | |
---|
| 1255 | /* Find the longest match, discarding those <= prev_length. |
---|
| 1256 | * At this point we have always match_length < MIN_MATCH |
---|
| 1257 | */ |
---|
| 1258 | if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) { |
---|
| 1259 | /* To simplify the code, we prevent matches with the string |
---|
| 1260 | * of window index 0 (in particular we have to avoid a match |
---|
| 1261 | * of the string with itself at the start of the input file). |
---|
| 1262 | */ |
---|
| 1263 | #ifdef FASTEST |
---|
| 1264 | if ((s->strategy < Z_HUFFMAN_ONLY) || |
---|
| 1265 | (s->strategy == Z_RLE && s->strstart - hash_head == 1)) { |
---|
| 1266 | s->match_length = longest_match_fast (s, hash_head); |
---|
| 1267 | } |
---|
| 1268 | #else |
---|
| 1269 | if (s->strategy < Z_HUFFMAN_ONLY) { |
---|
| 1270 | s->match_length = longest_match (s, hash_head); |
---|
| 1271 | } else if (s->strategy == Z_RLE && s->strstart - hash_head == 1) { |
---|
| 1272 | s->match_length = longest_match_fast (s, hash_head); |
---|
| 1273 | } |
---|
| 1274 | #endif |
---|
| 1275 | /* longest_match() or longest_match_fast() sets match_start */ |
---|
| 1276 | } |
---|
| 1277 | if (s->match_length >= MIN_MATCH) { |
---|
| 1278 | check_match(s, s->strstart, s->match_start, s->match_length); |
---|
| 1279 | |
---|
| 1280 | _tr_tally_dist(s, s->strstart - s->match_start, |
---|
| 1281 | s->match_length - MIN_MATCH, bflush); |
---|
| 1282 | |
---|
| 1283 | s->lookahead -= s->match_length; |
---|
| 1284 | |
---|
| 1285 | /* Insert new strings in the hash table only if the match length |
---|
| 1286 | * is not too large. This saves time but degrades compression. |
---|
| 1287 | */ |
---|
| 1288 | #ifndef FASTEST |
---|
| 1289 | if (s->match_length <= s->max_insert_length && |
---|
| 1290 | s->lookahead >= MIN_MATCH) { |
---|
| 1291 | s->match_length--; /* string at strstart already in table */ |
---|
| 1292 | do { |
---|
| 1293 | s->strstart++; |
---|
| 1294 | INSERT_STRING(s, s->strstart, hash_head); |
---|
| 1295 | /* strstart never exceeds WSIZE-MAX_MATCH, so there are |
---|
| 1296 | * always MIN_MATCH bytes ahead. |
---|
| 1297 | */ |
---|
| 1298 | } while (--s->match_length != 0); |
---|
| 1299 | s->strstart++; |
---|
| 1300 | } else |
---|
| 1301 | #endif |
---|
| 1302 | { |
---|
| 1303 | s->strstart += s->match_length; |
---|
| 1304 | s->match_length = 0; |
---|
| 1305 | s->ins_h = s->window[s->strstart]; |
---|
| 1306 | UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]); |
---|
| 1307 | #if MIN_MATCH != 3 |
---|
| 1308 | Call UPDATE_HASH() MIN_MATCH-3 more times |
---|
| 1309 | #endif |
---|
| 1310 | /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not |
---|
| 1311 | * matter since it will be recomputed at next deflate call. |
---|
| 1312 | */ |
---|
| 1313 | } |
---|
| 1314 | } else { |
---|
| 1315 | /* No match, output a literal byte */ |
---|
| 1316 | Tracevv((stderr,"%c", s->window[s->strstart])); |
---|
| 1317 | _tr_tally_lit (s, s->window[s->strstart], bflush); |
---|
| 1318 | s->lookahead--; |
---|
| 1319 | s->strstart++; |
---|
| 1320 | } |
---|
| 1321 | if (bflush) FLUSH_BLOCK(s, 0); |
---|
| 1322 | } |
---|
| 1323 | FLUSH_BLOCK(s, flush == Z_FINISH); |
---|
| 1324 | return flush == Z_FINISH ? finish_done : block_done; |
---|
| 1325 | } |
---|
| 1326 | |
---|
| 1327 | #ifndef FASTEST |
---|
| 1328 | /* =========================================================================== |
---|
| 1329 | * Same as above, but achieves better compression. We use a lazy |
---|
| 1330 | * evaluation for matches: a match is finally adopted only if there is |
---|
| 1331 | * no better match at the next window position. |
---|
| 1332 | */ |
---|
| 1333 | local block_state deflate_slow(deflate_state *s, int flush) |
---|
| 1334 | { |
---|
| 1335 | IPos hash_head = NIL; /* head of hash chain */ |
---|
| 1336 | int bflush; /* set if current block must be flushed */ |
---|
| 1337 | |
---|
| 1338 | /* Process the input block. */ |
---|
| 1339 | for (;;) { |
---|
| 1340 | /* Make sure that we always have enough lookahead, except |
---|
| 1341 | * at the end of the input file. We need MAX_MATCH bytes |
---|
| 1342 | * for the next match, plus MIN_MATCH bytes to insert the |
---|
| 1343 | * string following the next match. |
---|
| 1344 | */ |
---|
| 1345 | if (s->lookahead < MIN_LOOKAHEAD) { |
---|
| 1346 | fill_window(s); |
---|
| 1347 | if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { |
---|
| 1348 | return need_more; |
---|
| 1349 | } |
---|
| 1350 | if (s->lookahead == 0) break; /* flush the current block */ |
---|
| 1351 | } |
---|
| 1352 | |
---|
| 1353 | /* Insert the string window[strstart .. strstart+2] in the |
---|
| 1354 | * dictionary, and set hash_head to the head of the hash chain: |
---|
| 1355 | */ |
---|
| 1356 | if (s->lookahead >= MIN_MATCH) { |
---|
| 1357 | INSERT_STRING(s, s->strstart, hash_head); |
---|
| 1358 | } |
---|
| 1359 | |
---|
| 1360 | /* Find the longest match, discarding those <= prev_length. |
---|
| 1361 | */ |
---|
| 1362 | s->prev_length = s->match_length, s->prev_match = s->match_start; |
---|
| 1363 | s->match_length = MIN_MATCH-1; |
---|
| 1364 | |
---|
| 1365 | if (hash_head != NIL && s->prev_length < s->max_lazy_match && |
---|
| 1366 | s->strstart - hash_head <= MAX_DIST(s)) { |
---|
| 1367 | /* To simplify the code, we prevent matches with the string |
---|
| 1368 | * of window index 0 (in particular we have to avoid a match |
---|
| 1369 | * of the string with itself at the start of the input file). |
---|
| 1370 | */ |
---|
| 1371 | if (s->strategy < Z_HUFFMAN_ONLY) { |
---|
| 1372 | s->match_length = longest_match (s, hash_head); |
---|
| 1373 | } else if (s->strategy == Z_RLE && s->strstart - hash_head == 1) { |
---|
| 1374 | s->match_length = longest_match_fast (s, hash_head); |
---|
| 1375 | } |
---|
| 1376 | /* longest_match() or longest_match_fast() sets match_start */ |
---|
| 1377 | |
---|
| 1378 | if (s->match_length <= 5 && (s->strategy == Z_FILTERED |
---|
| 1379 | #if TOO_FAR <= 32767 |
---|
| 1380 | || (s->match_length == MIN_MATCH && |
---|
| 1381 | s->strstart - s->match_start > TOO_FAR) |
---|
| 1382 | #endif |
---|
| 1383 | )) { |
---|
| 1384 | |
---|
| 1385 | /* If prev_match is also MIN_MATCH, match_start is garbage |
---|
| 1386 | * but we will ignore the current match anyway. |
---|
| 1387 | */ |
---|
| 1388 | s->match_length = MIN_MATCH-1; |
---|
| 1389 | } |
---|
| 1390 | } |
---|
| 1391 | /* If there was a match at the previous step and the current |
---|
| 1392 | * match is not better, output the previous match: |
---|
| 1393 | */ |
---|
| 1394 | if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) { |
---|
| 1395 | uInt max_insert = s->strstart + s->lookahead - MIN_MATCH; |
---|
| 1396 | /* Do not insert strings in hash table beyond this. */ |
---|
| 1397 | |
---|
| 1398 | check_match(s, s->strstart-1, s->prev_match, s->prev_length); |
---|
| 1399 | |
---|
| 1400 | _tr_tally_dist(s, s->strstart -1 - s->prev_match, |
---|
| 1401 | s->prev_length - MIN_MATCH, bflush); |
---|
| 1402 | |
---|
| 1403 | /* Insert in hash table all strings up to the end of the match. |
---|
| 1404 | * strstart-1 and strstart are already inserted. If there is not |
---|
| 1405 | * enough lookahead, the last two strings are not inserted in |
---|
| 1406 | * the hash table. |
---|
| 1407 | */ |
---|
| 1408 | s->lookahead -= s->prev_length-1; |
---|
| 1409 | s->prev_length -= 2; |
---|
| 1410 | do { |
---|
| 1411 | if (++s->strstart <= max_insert) { |
---|
| 1412 | INSERT_STRING(s, s->strstart, hash_head); |
---|
| 1413 | } |
---|
| 1414 | } while (--s->prev_length != 0); |
---|
| 1415 | s->match_available = 0; |
---|
| 1416 | s->match_length = MIN_MATCH-1; |
---|
| 1417 | s->strstart++; |
---|
| 1418 | |
---|
| 1419 | if (bflush) FLUSH_BLOCK(s, 0); |
---|
| 1420 | |
---|
| 1421 | } else if (s->match_available) { |
---|
| 1422 | /* If there was no match at the previous position, output a |
---|
| 1423 | * single literal. If there was a match but the current match |
---|
| 1424 | * is longer, truncate the previous match to a single literal. |
---|
| 1425 | */ |
---|
| 1426 | Tracevv((stderr,(char*)"%c", s->window[s->strstart-1])); |
---|
| 1427 | _tr_tally_lit(s, s->window[s->strstart-1], bflush); |
---|
| 1428 | if (bflush) { |
---|
| 1429 | FLUSH_BLOCK_ONLY(s, 0); |
---|
| 1430 | } |
---|
| 1431 | s->strstart++; |
---|
| 1432 | s->lookahead--; |
---|
| 1433 | if (s->strm->avail_out == 0) return need_more; |
---|
| 1434 | } else { |
---|
| 1435 | /* There is no previous match to compare with, wait for |
---|
| 1436 | * the next step to decide. |
---|
| 1437 | */ |
---|
| 1438 | s->match_available = 1; |
---|
| 1439 | s->strstart++; |
---|
| 1440 | s->lookahead--; |
---|
| 1441 | } |
---|
| 1442 | } |
---|
| 1443 | Assert (flush != Z_NO_FLUSH, (char*)"no flush?"); |
---|
| 1444 | if (s->match_available) { |
---|
| 1445 | Tracevv((stderr,(char*)"%c", s->window[s->strstart-1])); |
---|
| 1446 | _tr_tally_lit(s, s->window[s->strstart-1], bflush); |
---|
| 1447 | s->match_available = 0; |
---|
| 1448 | } |
---|
| 1449 | FLUSH_BLOCK(s, flush == Z_FINISH); |
---|
| 1450 | return flush == Z_FINISH ? finish_done : block_done; |
---|
| 1451 | } |
---|
| 1452 | #endif /* FASTEST */ |
---|