]>
Commit | Line | Data |
---|---|---|
1 | /* deflate.c -- compress data using the deflation algorithm | |
2 | * Copyright (C) 1995-2002 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 ftp://ds.internic.net/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 | /* @(#) $Id$ */ | |
51 | ||
52 | #include "deflate.h" | |
53 | ||
54 | const char deflate_copyright[] = | |
55 | " deflate 1.1.4 Copyright 1995-2002 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 | local block_state deflate_slow OF((deflate_state *s, int flush)); | |
80 | local void lm_init OF((deflate_state *s)); | |
81 | local void putShortMSB OF((deflate_state *s, uInt b)); | |
82 | local void flush_pending OF((z_streamp strm)); | |
83 | local int read_buf OF((z_streamp strm, Bytef *buf, unsigned size)); | |
84 | #ifdef ASMV | |
85 | void match_init OF((void)); /* asm code initialization */ | |
86 | uInt longest_match OF((deflate_state *s, IPos cur_match)); | |
87 | #else | |
88 | local uInt longest_match OF((deflate_state *s, IPos cur_match)); | |
89 | #endif | |
90 | ||
91 | #ifdef __WXDEBUG__ | |
92 | local void check_match OF((deflate_state *s, IPos start, IPos match, | |
93 | int length)); | |
94 | #endif | |
95 | ||
96 | /* =========================================================================== | |
97 | * Local data | |
98 | */ | |
99 | ||
100 | #define NIL 0 | |
101 | /* Tail of hash chains */ | |
102 | ||
103 | #ifndef TOO_FAR | |
104 | # define TOO_FAR 4096 | |
105 | #endif | |
106 | /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */ | |
107 | ||
108 | #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1) | |
109 | /* Minimum amount of lookahead, except at the end of the input file. | |
110 | * See deflate.c for comments about the MIN_MATCH+1. | |
111 | */ | |
112 | ||
113 | /* Values for max_lazy_match, good_match and max_chain_length, depending on | |
114 | * the desired pack level (0..9). The values given below have been tuned to | |
115 | * exclude worst case performance for pathological files. Better values may be | |
116 | * found for specific files. | |
117 | */ | |
118 | typedef struct config_s { | |
119 | ush good_length; /* reduce lazy search above this match length */ | |
120 | ush max_lazy; /* do not perform lazy search above this match length */ | |
121 | ush nice_length; /* quit search above this match length */ | |
122 | ush max_chain; | |
123 | compress_func func; | |
124 | } config; | |
125 | ||
126 | local const config configuration_table[10] = { | |
127 | /* good lazy nice chain */ | |
128 | /* 0 */ {0, 0, 0, 0, deflate_stored}, /* store only */ | |
129 | /* 1 */ {4, 4, 8, 4, deflate_fast}, /* maximum speed, no lazy matches */ | |
130 | /* 2 */ {4, 5, 16, 8, deflate_fast}, | |
131 | /* 3 */ {4, 6, 32, 32, deflate_fast}, | |
132 | ||
133 | /* 4 */ {4, 4, 16, 16, deflate_slow}, /* lazy matches */ | |
134 | /* 5 */ {8, 16, 32, 32, deflate_slow}, | |
135 | /* 6 */ {8, 16, 128, 128, deflate_slow}, | |
136 | /* 7 */ {8, 32, 128, 256, deflate_slow}, | |
137 | /* 8 */ {32, 128, 258, 1024, deflate_slow}, | |
138 | /* 9 */ {32, 258, 258, 4096, deflate_slow}}; /* maximum compression */ | |
139 | ||
140 | /* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4 | |
141 | * For deflate_fast() (levels <= 3) good is ignored and lazy has a different | |
142 | * meaning. | |
143 | */ | |
144 | ||
145 | #define EQUAL 0 | |
146 | /* result of memcmp for equal strings */ | |
147 | ||
148 | struct static_tree_desc_s {int dummy;}; /* for buggy compilers */ | |
149 | ||
150 | /* =========================================================================== | |
151 | * Update a hash value with the given input byte | |
152 | * IN assertion: all calls to to UPDATE_HASH are made with consecutive | |
153 | * input characters, so that a running hash key can be computed from the | |
154 | * previous key instead of complete recalculation each time. | |
155 | */ | |
156 | #define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask) | |
157 | ||
158 | ||
159 | /* =========================================================================== | |
160 | * Insert string str in the dictionary and set match_head to the previous head | |
161 | * of the hash chain (the most recent string with same hash key). Return | |
162 | * the previous length of the hash chain. | |
163 | * If this file is compiled with -DFASTEST, the compression level is forced | |
164 | * to 1, and no hash chains are maintained. | |
165 | * IN assertion: all calls to to INSERT_STRING are made with consecutive | |
166 | * input characters and the first MIN_MATCH bytes of str are valid | |
167 | * (except for the last MIN_MATCH-1 bytes of the input file). | |
168 | */ | |
169 | #ifdef FASTEST | |
170 | #define INSERT_STRING(s, str, match_head) \ | |
171 | (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ | |
172 | match_head = s->head[s->ins_h], \ | |
173 | s->head[s->ins_h] = (Pos)(str)) | |
174 | #else | |
175 | #define INSERT_STRING(s, str, match_head) \ | |
176 | (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \ | |
177 | s->prev[(str) & s->w_mask] = match_head = s->head[s->ins_h], \ | |
178 | s->head[s->ins_h] = (Pos)(str)) | |
179 | #endif | |
180 | ||
181 | /* =========================================================================== | |
182 | * Initialize the hash table (avoiding 64K overflow for 16 bit systems). | |
183 | * prev[] will be initialized on the fly. | |
184 | */ | |
185 | #define CLEAR_HASH(s) \ | |
186 | s->head[s->hash_size-1] = NIL; \ | |
187 | zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head)); | |
188 | ||
189 | /* ========================================================================= */ | |
190 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
191 | int ZEXPORT deflateInit_(z_streamp strm, int level, const char* version, int stream_size) | |
192 | #else | |
193 | int ZEXPORT deflateInit_(strm, level, version, stream_size) | |
194 | z_streamp strm; | |
195 | int level; | |
196 | const char *version; | |
197 | int stream_size; | |
198 | #endif | |
199 | { | |
200 | return deflateInit2_(strm, level, Z_DEFLATED, MAX_WBITS, DEF_MEM_LEVEL, | |
201 | Z_DEFAULT_STRATEGY, version, stream_size); | |
202 | /* To do: ignore strm->next_in if we use it as window */ | |
203 | } | |
204 | ||
205 | /* ========================================================================= */ | |
206 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
207 | int ZEXPORT deflateInit2_(z_streamp strm, int level, int method, int windowBits, | |
208 | int memLevel, int strategy, const char* version, int stream_size) | |
209 | #else | |
210 | int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy, | |
211 | version, stream_size) | |
212 | z_streamp strm; | |
213 | int level; | |
214 | int method; | |
215 | int windowBits; | |
216 | int memLevel; | |
217 | int strategy; | |
218 | const char *version; | |
219 | int stream_size; | |
220 | #endif | |
221 | { | |
222 | deflate_state *s; | |
223 | int noheader = 0; | |
224 | static const char* my_version = ZLIB_VERSION; | |
225 | ||
226 | ushf *overlay; | |
227 | /* We overlay pending_buf and d_buf+l_buf. This works since the average | |
228 | * output size for (length,distance) codes is <= 24 bits. | |
229 | */ | |
230 | ||
231 | if (version == Z_NULL || version[0] != my_version[0] || | |
232 | stream_size != sizeof(z_stream)) { | |
233 | return Z_VERSION_ERROR; | |
234 | } | |
235 | if (strm == Z_NULL) return Z_STREAM_ERROR; | |
236 | ||
237 | strm->msg = Z_NULL; | |
238 | if (strm->zalloc == Z_NULL) { | |
239 | strm->zalloc = zcalloc; | |
240 | strm->opaque = (voidpf)0; | |
241 | } | |
242 | if (strm->zfree == Z_NULL) strm->zfree = zcfree; | |
243 | ||
244 | if (level == Z_DEFAULT_COMPRESSION) level = 6; | |
245 | #ifdef FASTEST | |
246 | level = 1; | |
247 | #endif | |
248 | ||
249 | if (windowBits < 0) { /* undocumented feature: suppress zlib header */ | |
250 | noheader = 1; | |
251 | windowBits = -windowBits; | |
252 | } | |
253 | if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED || | |
254 | windowBits < 9 || windowBits > 15 || level < 0 || level > 9 || | |
255 | strategy < 0 || strategy > Z_HUFFMAN_ONLY) { | |
256 | return Z_STREAM_ERROR; | |
257 | } | |
258 | s = (deflate_state *) ZALLOC(strm, 1, sizeof(deflate_state)); | |
259 | if (s == Z_NULL) return Z_MEM_ERROR; | |
260 | strm->state = (struct internal_state FAR *)s; | |
261 | s->strm = strm; | |
262 | ||
263 | s->noheader = noheader; | |
264 | s->w_bits = windowBits; | |
265 | s->w_size = 1 << s->w_bits; | |
266 | s->w_mask = s->w_size - 1; | |
267 | ||
268 | s->hash_bits = memLevel + 7; | |
269 | s->hash_size = 1 << s->hash_bits; | |
270 | s->hash_mask = s->hash_size - 1; | |
271 | s->hash_shift = ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH); | |
272 | ||
273 | s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte)); | |
274 | s->prev = (Posf *) ZALLOC(strm, s->w_size, sizeof(Pos)); | |
275 | s->head = (Posf *) ZALLOC(strm, s->hash_size, sizeof(Pos)); | |
276 | ||
277 | s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */ | |
278 | ||
279 | overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2); | |
280 | s->pending_buf = (uchf *) overlay; | |
281 | s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L); | |
282 | ||
283 | if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL || | |
284 | s->pending_buf == Z_NULL) { | |
285 | strm->msg = (char*)ERR_MSG(Z_MEM_ERROR); | |
286 | deflateEnd (strm); | |
287 | return Z_MEM_ERROR; | |
288 | } | |
289 | s->d_buf = overlay + s->lit_bufsize/sizeof(ush); | |
290 | s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize; | |
291 | ||
292 | s->level = level; | |
293 | s->strategy = strategy; | |
294 | s->method = (Byte)method; | |
295 | ||
296 | return deflateReset(strm); | |
297 | } | |
298 | ||
299 | /* ========================================================================= */ | |
300 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
301 | int ZEXPORT deflateSetDictionary (z_streamp strm, const Bytef* dictionary, uInt dictLength) | |
302 | #else | |
303 | int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength) | |
304 | z_streamp strm; | |
305 | const Bytef *dictionary; | |
306 | uInt dictLength; | |
307 | #endif | |
308 | { | |
309 | deflate_state *s; | |
310 | uInt length = dictLength; | |
311 | uInt n; | |
312 | IPos hash_head = 0; | |
313 | ||
314 | if (strm == Z_NULL || strm->state == Z_NULL || dictionary == Z_NULL || | |
315 | strm->state->status != INIT_STATE) return Z_STREAM_ERROR; | |
316 | ||
317 | s = strm->state; | |
318 | strm->adler = adler32(strm->adler, dictionary, dictLength); | |
319 | ||
320 | if (length < MIN_MATCH) return Z_OK; | |
321 | if (length > MAX_DIST(s)) { | |
322 | length = MAX_DIST(s); | |
323 | #ifndef USE_DICT_HEAD | |
324 | dictionary += dictLength - length; /* use the tail of the dictionary */ | |
325 | #endif | |
326 | } | |
327 | zmemcpy(s->window, dictionary, length); | |
328 | s->strstart = length; | |
329 | s->block_start = (long)length; | |
330 | ||
331 | /* Insert all strings in the hash table (except for the last two bytes). | |
332 | * s->lookahead stays null, so s->ins_h will be recomputed at the next | |
333 | * call of fill_window. | |
334 | */ | |
335 | s->ins_h = s->window[0]; | |
336 | UPDATE_HASH(s, s->ins_h, s->window[1]); | |
337 | for (n = 0; n <= length - MIN_MATCH; n++) { | |
338 | INSERT_STRING(s, n, hash_head); | |
339 | } | |
340 | if (hash_head) hash_head = 0; /* to make compiler happy */ | |
341 | return Z_OK; | |
342 | } | |
343 | ||
344 | /* ========================================================================= */ | |
345 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
346 | int ZEXPORT deflateReset (z_streamp strm) | |
347 | #else | |
348 | int ZEXPORT deflateReset (strm) | |
349 | z_streamp strm; | |
350 | #endif | |
351 | { | |
352 | deflate_state *s; | |
353 | ||
354 | if (strm == Z_NULL || strm->state == Z_NULL || | |
355 | strm->zalloc == Z_NULL || strm->zfree == Z_NULL) return Z_STREAM_ERROR; | |
356 | ||
357 | strm->total_in = strm->total_out = 0; | |
358 | strm->msg = Z_NULL; /* use zfree if we ever allocate msg dynamically */ | |
359 | strm->data_type = Z_UNKNOWN; | |
360 | ||
361 | s = (deflate_state *)strm->state; | |
362 | s->pending = 0; | |
363 | s->pending_out = s->pending_buf; | |
364 | ||
365 | if (s->noheader < 0) { | |
366 | s->noheader = 0; /* was set to -1 by deflate(..., Z_FINISH); */ | |
367 | } | |
368 | s->status = s->noheader ? BUSY_STATE : INIT_STATE; | |
369 | strm->adler = 1; | |
370 | s->last_flush = Z_NO_FLUSH; | |
371 | ||
372 | _tr_init(s); | |
373 | lm_init(s); | |
374 | ||
375 | return Z_OK; | |
376 | } | |
377 | ||
378 | /* ========================================================================= */ | |
379 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
380 | int ZEXPORT deflateParams(z_streamp strm, int level, int strategy) | |
381 | #else | |
382 | int ZEXPORT deflateParams(strm, level, strategy) | |
383 | z_streamp strm; | |
384 | int level; | |
385 | int strategy; | |
386 | #endif | |
387 | { | |
388 | deflate_state *s; | |
389 | compress_func func; | |
390 | int err = Z_OK; | |
391 | ||
392 | if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR; | |
393 | s = strm->state; | |
394 | ||
395 | if (level == Z_DEFAULT_COMPRESSION) { | |
396 | level = 6; | |
397 | } | |
398 | if (level < 0 || level > 9 || strategy < 0 || strategy > Z_HUFFMAN_ONLY) { | |
399 | return Z_STREAM_ERROR; | |
400 | } | |
401 | func = configuration_table[s->level].func; | |
402 | ||
403 | if (func != configuration_table[level].func && strm->total_in != 0) { | |
404 | /* Flush the last buffer: */ | |
405 | err = deflate(strm, Z_PARTIAL_FLUSH); | |
406 | } | |
407 | if (s->level != level) { | |
408 | s->level = level; | |
409 | s->max_lazy_match = configuration_table[level].max_lazy; | |
410 | s->good_match = configuration_table[level].good_length; | |
411 | s->nice_match = configuration_table[level].nice_length; | |
412 | s->max_chain_length = configuration_table[level].max_chain; | |
413 | } | |
414 | s->strategy = strategy; | |
415 | return err; | |
416 | } | |
417 | ||
418 | /* ========================================================================= | |
419 | * Put a short in the pending buffer. The 16-bit value is put in MSB order. | |
420 | * IN assertion: the stream state is correct and there is enough room in | |
421 | * pending_buf. | |
422 | */ | |
423 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
424 | local void putShortMSB (deflate_state* s, uInt b) | |
425 | #else | |
426 | local void putShortMSB (s, b) | |
427 | deflate_state *s; | |
428 | uInt b; | |
429 | #endif | |
430 | { | |
431 | put_byte(s, (Byte)(b >> 8)); | |
432 | put_byte(s, (Byte)(b & 0xff)); | |
433 | } | |
434 | ||
435 | /* ========================================================================= | |
436 | * Flush as much pending output as possible. All deflate() output goes | |
437 | * through this function so some applications may wish to modify it | |
438 | * to avoid allocating a large strm->next_out buffer and copying into it. | |
439 | * (See also read_buf()). | |
440 | */ | |
441 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
442 | local void flush_pending(z_streamp strm) | |
443 | #else | |
444 | local void flush_pending(strm) | |
445 | z_streamp strm; | |
446 | #endif | |
447 | { | |
448 | unsigned len = strm->state->pending; | |
449 | ||
450 | if (len > strm->avail_out) len = strm->avail_out; | |
451 | if (len == 0) return; | |
452 | ||
453 | zmemcpy(strm->next_out, strm->state->pending_out, len); | |
454 | strm->next_out += len; | |
455 | strm->state->pending_out += len; | |
456 | strm->total_out += len; | |
457 | strm->avail_out -= len; | |
458 | strm->state->pending -= len; | |
459 | if (strm->state->pending == 0) { | |
460 | strm->state->pending_out = strm->state->pending_buf; | |
461 | } | |
462 | } | |
463 | ||
464 | /* ========================================================================= */ | |
465 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
466 | int ZEXPORT deflate (z_streamp strm, int flush) | |
467 | #else | |
468 | int ZEXPORT deflate (strm, flush) | |
469 | z_streamp strm; | |
470 | int flush; | |
471 | #endif | |
472 | { | |
473 | int old_flush; /* value of flush param for previous deflate call */ | |
474 | deflate_state *s; | |
475 | ||
476 | if (strm == Z_NULL || strm->state == Z_NULL || | |
477 | flush > Z_FINISH || flush < 0) { | |
478 | return Z_STREAM_ERROR; | |
479 | } | |
480 | s = strm->state; | |
481 | ||
482 | if (strm->next_out == Z_NULL || | |
483 | (strm->next_in == Z_NULL && strm->avail_in != 0) || | |
484 | (s->status == FINISH_STATE && flush != Z_FINISH)) { | |
485 | ERR_RETURN(strm, Z_STREAM_ERROR); | |
486 | } | |
487 | if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR); | |
488 | ||
489 | s->strm = strm; /* just in case */ | |
490 | old_flush = s->last_flush; | |
491 | s->last_flush = flush; | |
492 | ||
493 | /* Write the zlib header */ | |
494 | if (s->status == INIT_STATE) { | |
495 | ||
496 | uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8; | |
497 | uInt level_flags = (s->level-1) >> 1; | |
498 | ||
499 | if (level_flags > 3) level_flags = 3; | |
500 | header |= (level_flags << 6); | |
501 | if (s->strstart != 0) header |= PRESET_DICT; | |
502 | header += 31 - (header % 31); | |
503 | ||
504 | s->status = BUSY_STATE; | |
505 | putShortMSB(s, header); | |
506 | ||
507 | /* Save the adler32 of the preset dictionary: */ | |
508 | if (s->strstart != 0) { | |
509 | putShortMSB(s, (uInt)(strm->adler >> 16)); | |
510 | putShortMSB(s, (uInt)(strm->adler & 0xffff)); | |
511 | } | |
512 | strm->adler = 1L; | |
513 | } | |
514 | ||
515 | /* Flush as much pending output as possible */ | |
516 | if (s->pending != 0) { | |
517 | flush_pending(strm); | |
518 | if (strm->avail_out == 0) { | |
519 | /* Since avail_out is 0, deflate will be called again with | |
520 | * more output space, but possibly with both pending and | |
521 | * avail_in equal to zero. There won't be anything to do, | |
522 | * but this is not an error situation so make sure we | |
523 | * return OK instead of BUF_ERROR at next call of deflate: | |
524 | */ | |
525 | s->last_flush = -1; | |
526 | return Z_OK; | |
527 | } | |
528 | ||
529 | /* Make sure there is something to do and avoid duplicate consecutive | |
530 | * flushes. For repeated and useless calls with Z_FINISH, we keep | |
531 | * returning Z_STREAM_END instead of Z_BUFF_ERROR. | |
532 | */ | |
533 | } else if (strm->avail_in == 0 && flush <= old_flush && | |
534 | flush != Z_FINISH) { | |
535 | ERR_RETURN(strm, Z_BUF_ERROR); | |
536 | } | |
537 | ||
538 | /* User must not provide more input after the first FINISH: */ | |
539 | if (s->status == FINISH_STATE && strm->avail_in != 0) { | |
540 | ERR_RETURN(strm, Z_BUF_ERROR); | |
541 | } | |
542 | ||
543 | /* Start a new block or continue the current one. | |
544 | */ | |
545 | if (strm->avail_in != 0 || s->lookahead != 0 || | |
546 | (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) { | |
547 | block_state bstate; | |
548 | ||
549 | bstate = (*(configuration_table[s->level].func))(s, flush); | |
550 | ||
551 | if (bstate == finish_started || bstate == finish_done) { | |
552 | s->status = FINISH_STATE; | |
553 | } | |
554 | if (bstate == need_more || bstate == finish_started) { | |
555 | if (strm->avail_out == 0) { | |
556 | s->last_flush = -1; /* avoid BUF_ERROR next call, see above */ | |
557 | } | |
558 | return Z_OK; | |
559 | /* If flush != Z_NO_FLUSH && avail_out == 0, the next call | |
560 | * of deflate should use the same flush parameter to make sure | |
561 | * that the flush is complete. So we don't have to output an | |
562 | * empty block here, this will be done at next call. This also | |
563 | * ensures that for a very small output buffer, we emit at most | |
564 | * one empty block. | |
565 | */ | |
566 | } | |
567 | if (bstate == block_done) { | |
568 | if (flush == Z_PARTIAL_FLUSH) { | |
569 | _tr_align(s); | |
570 | } else { /* FULL_FLUSH or SYNC_FLUSH */ | |
571 | _tr_stored_block(s, (char*)0, 0L, 0); | |
572 | /* For a full flush, this empty block will be recognized | |
573 | * as a special marker by inflate_sync(). | |
574 | */ | |
575 | if (flush == Z_FULL_FLUSH) { | |
576 | CLEAR_HASH(s); /* forget history */ | |
577 | } | |
578 | } | |
579 | flush_pending(strm); | |
580 | if (strm->avail_out == 0) { | |
581 | s->last_flush = -1; /* avoid BUF_ERROR at next call, see above */ | |
582 | return Z_OK; | |
583 | } | |
584 | } | |
585 | } | |
586 | Assert(strm->avail_out > 0, "bug2"); | |
587 | ||
588 | if (flush != Z_FINISH) return Z_OK; | |
589 | if (s->noheader) return Z_STREAM_END; | |
590 | ||
591 | /* Write the zlib trailer (adler32) */ | |
592 | putShortMSB(s, (uInt)(strm->adler >> 16)); | |
593 | putShortMSB(s, (uInt)(strm->adler & 0xffff)); | |
594 | flush_pending(strm); | |
595 | /* If avail_out is zero, the application will call deflate again | |
596 | * to flush the rest. | |
597 | */ | |
598 | s->noheader = -1; /* write the trailer only once! */ | |
599 | return s->pending != 0 ? Z_OK : Z_STREAM_END; | |
600 | } | |
601 | ||
602 | /* ========================================================================= */ | |
603 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
604 | int ZEXPORT deflateEnd (z_streamp strm) | |
605 | #else | |
606 | int ZEXPORT deflateEnd (strm) | |
607 | z_streamp strm; | |
608 | #endif | |
609 | { | |
610 | int status; | |
611 | ||
612 | if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR; | |
613 | ||
614 | status = strm->state->status; | |
615 | if (status != INIT_STATE && status != BUSY_STATE && | |
616 | status != FINISH_STATE) { | |
617 | return Z_STREAM_ERROR; | |
618 | } | |
619 | ||
620 | /* Deallocate in reverse order of allocations: */ | |
621 | TRY_FREE(strm, strm->state->pending_buf); | |
622 | TRY_FREE(strm, strm->state->head); | |
623 | TRY_FREE(strm, strm->state->prev); | |
624 | TRY_FREE(strm, strm->state->window); | |
625 | ||
626 | ZFREE(strm, strm->state); | |
627 | strm->state = Z_NULL; | |
628 | ||
629 | return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK; | |
630 | } | |
631 | ||
632 | /* ========================================================================= | |
633 | * Copy the source state to the destination state. | |
634 | * To simplify the source, this is not supported for 16-bit MSDOS (which | |
635 | * doesn't have enough memory anyway to duplicate compression states). | |
636 | */ | |
637 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
638 | int ZEXPORT deflateCopy (z_streamp dest, z_streamp source) | |
639 | #else | |
640 | int ZEXPORT deflateCopy (dest, source) | |
641 | z_streamp dest; | |
642 | z_streamp source; | |
643 | #endif | |
644 | { | |
645 | #ifdef MAXSEG_64K | |
646 | return Z_STREAM_ERROR; | |
647 | #else | |
648 | deflate_state *ds; | |
649 | deflate_state *ss; | |
650 | ushf *overlay; | |
651 | ||
652 | ||
653 | if (source == Z_NULL || dest == Z_NULL || source->state == Z_NULL) { | |
654 | return Z_STREAM_ERROR; | |
655 | } | |
656 | ||
657 | ss = source->state; | |
658 | ||
659 | *dest = *source; | |
660 | ||
661 | ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state)); | |
662 | if (ds == Z_NULL) return Z_MEM_ERROR; | |
663 | dest->state = (struct internal_state FAR *) ds; | |
664 | *ds = *ss; | |
665 | ds->strm = dest; | |
666 | ||
667 | ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte)); | |
668 | ds->prev = (Posf *) ZALLOC(dest, ds->w_size, sizeof(Pos)); | |
669 | ds->head = (Posf *) ZALLOC(dest, ds->hash_size, sizeof(Pos)); | |
670 | overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2); | |
671 | ds->pending_buf = (uchf *) overlay; | |
672 | ||
673 | if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL || | |
674 | ds->pending_buf == Z_NULL) { | |
675 | deflateEnd (dest); | |
676 | return Z_MEM_ERROR; | |
677 | } | |
678 | /* following zmemcpy do not work for 16-bit MSDOS */ | |
679 | zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte)); | |
680 | zmemcpy(ds->prev, ss->prev, ds->w_size * sizeof(Pos)); | |
681 | zmemcpy(ds->head, ss->head, ds->hash_size * sizeof(Pos)); | |
682 | zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size); | |
683 | ||
684 | ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf); | |
685 | ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush); | |
686 | ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize; | |
687 | ||
688 | ds->l_desc.dyn_tree = ds->dyn_ltree; | |
689 | ds->d_desc.dyn_tree = ds->dyn_dtree; | |
690 | ds->bl_desc.dyn_tree = ds->bl_tree; | |
691 | ||
692 | return Z_OK; | |
693 | #endif | |
694 | } | |
695 | ||
696 | /* =========================================================================== | |
697 | * Read a new buffer from the current input stream, update the adler32 | |
698 | * and total number of bytes read. All deflate() input goes through | |
699 | * this function so some applications may wish to modify it to avoid | |
700 | * allocating a large strm->next_in buffer and copying from it. | |
701 | * (See also flush_pending()). | |
702 | */ | |
703 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
704 | local int read_buf(z_streamp strm, Bytef* buf, unsigned size) | |
705 | #else | |
706 | local int read_buf(strm, buf, size) | |
707 | z_streamp strm; | |
708 | Bytef *buf; | |
709 | unsigned size; | |
710 | #endif | |
711 | { | |
712 | unsigned len = strm->avail_in; | |
713 | ||
714 | if (len > size) len = size; | |
715 | if (len == 0) return 0; | |
716 | ||
717 | strm->avail_in -= len; | |
718 | ||
719 | if (!strm->state->noheader) { | |
720 | strm->adler = adler32(strm->adler, strm->next_in, len); | |
721 | } | |
722 | zmemcpy(buf, strm->next_in, len); | |
723 | strm->next_in += len; | |
724 | strm->total_in += len; | |
725 | ||
726 | return (int)len; | |
727 | } | |
728 | ||
729 | /* =========================================================================== | |
730 | * Initialize the "longest match" routines for a new zlib stream | |
731 | */ | |
732 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
733 | local void lm_init (deflate_state* s) | |
734 | #else | |
735 | local void lm_init (s) | |
736 | deflate_state *s; | |
737 | #endif | |
738 | { | |
739 | s->window_size = (ulg)2L*s->w_size; | |
740 | ||
741 | CLEAR_HASH(s); | |
742 | ||
743 | /* Set the default configuration parameters: | |
744 | */ | |
745 | s->max_lazy_match = configuration_table[s->level].max_lazy; | |
746 | s->good_match = configuration_table[s->level].good_length; | |
747 | s->nice_match = configuration_table[s->level].nice_length; | |
748 | s->max_chain_length = configuration_table[s->level].max_chain; | |
749 | ||
750 | s->strstart = 0; | |
751 | s->block_start = 0L; | |
752 | s->lookahead = 0; | |
753 | s->match_length = s->prev_length = MIN_MATCH-1; | |
754 | s->match_available = 0; | |
755 | s->ins_h = 0; | |
756 | #ifdef ASMV | |
757 | match_init(); /* initialize the asm code */ | |
758 | #endif | |
759 | } | |
760 | ||
761 | /* =========================================================================== | |
762 | * Set match_start to the longest match starting at the given string and | |
763 | * return its length. Matches shorter or equal to prev_length are discarded, | |
764 | * in which case the result is equal to prev_length and match_start is | |
765 | * garbage. | |
766 | * IN assertions: cur_match is the head of the hash chain for the current | |
767 | * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1 | |
768 | * OUT assertion: the match length is not greater than s->lookahead. | |
769 | */ | |
770 | #ifndef ASMV | |
771 | /* For 80x86 and 680x0, an optimized version will be provided in match.asm or | |
772 | * match.S. The code will be functionally equivalent. | |
773 | */ | |
774 | #ifndef FASTEST | |
775 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
776 | local uInt longest_match(deflate_state* s, IPos cur_match) | |
777 | #else | |
778 | local uInt longest_match(s, cur_match) | |
779 | deflate_state *s; | |
780 | IPos cur_match; /* current match */ | |
781 | #endif | |
782 | { | |
783 | unsigned chain_length = s->max_chain_length;/* max hash chain length */ | |
784 | register Bytef *scan = s->window + s->strstart; /* current string */ | |
785 | register Bytef *match; /* matched string */ | |
786 | register int len; /* length of current match */ | |
787 | int best_len = s->prev_length; /* best match length so far */ | |
788 | int nice_match = s->nice_match; /* stop if match long enough */ | |
789 | IPos limit = s->strstart > (IPos)MAX_DIST(s) ? | |
790 | s->strstart - (IPos)MAX_DIST(s) : NIL; | |
791 | /* Stop when cur_match becomes <= limit. To simplify the code, | |
792 | * we prevent matches with the string of window index 0. | |
793 | */ | |
794 | Posf *prev = s->prev; | |
795 | uInt wmask = s->w_mask; | |
796 | ||
797 | #ifdef UNALIGNED_OK | |
798 | /* Compare two bytes at a time. Note: this is not always beneficial. | |
799 | * Try with and without -DUNALIGNED_OK to check. | |
800 | */ | |
801 | register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1; | |
802 | register ush scan_start = *(ushf*)scan; | |
803 | register ush scan_end = *(ushf*)(scan+best_len-1); | |
804 | #else | |
805 | register Bytef *strend = s->window + s->strstart + MAX_MATCH; | |
806 | register Byte scan_end1 = scan[best_len-1]; | |
807 | register Byte scan_end = scan[best_len]; | |
808 | #endif | |
809 | ||
810 | /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. | |
811 | * It is easy to get rid of this optimization if necessary. | |
812 | */ | |
813 | Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever"); | |
814 | ||
815 | /* Do not waste too much time if we already have a good match: */ | |
816 | if (s->prev_length >= s->good_match) { | |
817 | chain_length >>= 2; | |
818 | } | |
819 | /* Do not look for matches beyond the end of the input. This is necessary | |
820 | * to make deflate deterministic. | |
821 | */ | |
822 | if ((uInt)nice_match > s->lookahead) nice_match = s->lookahead; | |
823 | ||
824 | Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead"); | |
825 | ||
826 | do { | |
827 | Assert(cur_match < s->strstart, "no future"); | |
828 | match = s->window + cur_match; | |
829 | ||
830 | /* Skip to next match if the match length cannot increase | |
831 | * or if the match length is less than 2: | |
832 | */ | |
833 | #if (defined(UNALIGNED_OK) && MAX_MATCH == 258) | |
834 | /* This code assumes sizeof(unsigned short) == 2. Do not use | |
835 | * UNALIGNED_OK if your compiler uses a different size. | |
836 | */ | |
837 | if (*(ushf*)(match+best_len-1) != scan_end || | |
838 | *(ushf*)match != scan_start) continue; | |
839 | ||
840 | /* It is not necessary to compare scan[2] and match[2] since they are | |
841 | * always equal when the other bytes match, given that the hash keys | |
842 | * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at | |
843 | * strstart+3, +5, ... up to strstart+257. We check for insufficient | |
844 | * lookahead only every 4th comparison; the 128th check will be made | |
845 | * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is | |
846 | * necessary to put more guard bytes at the end of the window, or | |
847 | * to check more often for insufficient lookahead. | |
848 | */ | |
849 | Assert(scan[2] == match[2], "scan[2]?"); | |
850 | scan++, match++; | |
851 | do { | |
852 | } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) && | |
853 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && | |
854 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && | |
855 | *(ushf*)(scan+=2) == *(ushf*)(match+=2) && | |
856 | scan < strend); | |
857 | /* The funny "do {}" generates better code on most compilers */ | |
858 | ||
859 | /* Here, scan <= window+strstart+257 */ | |
860 | Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan"); | |
861 | if (*scan == *match) scan++; | |
862 | ||
863 | len = (MAX_MATCH - 1) - (int)(strend-scan); | |
864 | scan = strend - (MAX_MATCH-1); | |
865 | ||
866 | #else /* UNALIGNED_OK */ | |
867 | ||
868 | if (match[best_len] != scan_end || | |
869 | match[best_len-1] != scan_end1 || | |
870 | *match != *scan || | |
871 | *++match != scan[1]) continue; | |
872 | ||
873 | /* The check at best_len-1 can be removed because it will be made | |
874 | * again later. (This heuristic is not always a win.) | |
875 | * It is not necessary to compare scan[2] and match[2] since they | |
876 | * are always equal when the other bytes match, given that | |
877 | * the hash keys are equal and that HASH_BITS >= 8. | |
878 | */ | |
879 | scan += 2, match++; | |
880 | Assert(*scan == *match, "match[2]?"); | |
881 | ||
882 | /* We check for insufficient lookahead only every 8th comparison; | |
883 | * the 256th check will be made at strstart+258. | |
884 | */ | |
885 | do { | |
886 | } while (*++scan == *++match && *++scan == *++match && | |
887 | *++scan == *++match && *++scan == *++match && | |
888 | *++scan == *++match && *++scan == *++match && | |
889 | *++scan == *++match && *++scan == *++match && | |
890 | scan < strend); | |
891 | ||
892 | Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan"); | |
893 | ||
894 | len = MAX_MATCH - (int)(strend - scan); | |
895 | scan = strend - MAX_MATCH; | |
896 | ||
897 | #endif /* UNALIGNED_OK */ | |
898 | ||
899 | if (len > best_len) { | |
900 | s->match_start = cur_match; | |
901 | best_len = len; | |
902 | if (len >= nice_match) break; | |
903 | #ifdef UNALIGNED_OK | |
904 | scan_end = *(ushf*)(scan+best_len-1); | |
905 | #else | |
906 | scan_end1 = scan[best_len-1]; | |
907 | scan_end = scan[best_len]; | |
908 | #endif | |
909 | } | |
910 | } while ((cur_match = prev[cur_match & wmask]) > limit | |
911 | && --chain_length != 0); | |
912 | ||
913 | if ((uInt)best_len <= s->lookahead) return (uInt)best_len; | |
914 | return s->lookahead; | |
915 | } | |
916 | ||
917 | #else /* FASTEST */ | |
918 | /* --------------------------------------------------------------------------- | |
919 | * Optimized version for level == 1 only | |
920 | */ | |
921 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
922 | local uInt longest_match(deflate_state* s, IPos cur_match) | |
923 | #else | |
924 | local uInt longest_match(s, cur_match) | |
925 | deflate_state *s; | |
926 | IPos cur_match; /* current match */ | |
927 | #endif | |
928 | { | |
929 | register Bytef *scan = s->window + s->strstart; /* current string */ | |
930 | register Bytef *match; /* matched string */ | |
931 | register int len; /* length of current match */ | |
932 | register Bytef *strend = s->window + s->strstart + MAX_MATCH; | |
933 | ||
934 | /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. | |
935 | * It is easy to get rid of this optimization if necessary. | |
936 | */ | |
937 | Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever"); | |
938 | ||
939 | Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead"); | |
940 | ||
941 | Assert(cur_match < s->strstart, "no future"); | |
942 | ||
943 | match = s->window + cur_match; | |
944 | ||
945 | /* Return failure if the match length is less than 2: | |
946 | */ | |
947 | if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1; | |
948 | ||
949 | /* The check at best_len-1 can be removed because it will be made | |
950 | * again later. (This heuristic is not always a win.) | |
951 | * It is not necessary to compare scan[2] and match[2] since they | |
952 | * are always equal when the other bytes match, given that | |
953 | * the hash keys are equal and that HASH_BITS >= 8. | |
954 | */ | |
955 | scan += 2, match += 2; | |
956 | Assert(*scan == *match, "match[2]?"); | |
957 | ||
958 | /* We check for insufficient lookahead only every 8th comparison; | |
959 | * the 256th check will be made at strstart+258. | |
960 | */ | |
961 | do { | |
962 | } while (*++scan == *++match && *++scan == *++match && | |
963 | *++scan == *++match && *++scan == *++match && | |
964 | *++scan == *++match && *++scan == *++match && | |
965 | *++scan == *++match && *++scan == *++match && | |
966 | scan < strend); | |
967 | ||
968 | Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan"); | |
969 | ||
970 | len = MAX_MATCH - (int)(strend - scan); | |
971 | ||
972 | if (len < MIN_MATCH) return MIN_MATCH - 1; | |
973 | ||
974 | s->match_start = cur_match; | |
975 | return len <= s->lookahead ? len : s->lookahead; | |
976 | } | |
977 | #endif /* FASTEST */ | |
978 | #endif /* ASMV */ | |
979 | ||
980 | #ifdef __WXDEBUG__ | |
981 | /* =========================================================================== | |
982 | * Check that the match at match_start is indeed a match. | |
983 | */ | |
984 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
985 | local void check_match(deflate_state* s, IPos start, IPos match, int length) | |
986 | #else | |
987 | local void check_match(s, start, match, length) | |
988 | deflate_state *s; | |
989 | IPos start, match; | |
990 | int length; | |
991 | #endif | |
992 | { | |
993 | /* check that the match is indeed a match */ | |
994 | if (zmemcmp(s->window + match, | |
995 | s->window + start, length) != EQUAL) { | |
996 | fprintf(stderr, " start %u, match %u, length %d\n", | |
997 | start, match, length); | |
998 | do { | |
999 | fprintf(stderr, "%c%c", s->window[match++], s->window[start++]); | |
1000 | } while (--length != 0); | |
1001 | z_error("invalid match"); | |
1002 | } | |
1003 | if (z_verbose > 1) { | |
1004 | fprintf(stderr,"\\[%d,%d]", start-match, length); | |
1005 | do { putc(s->window[start++], stderr); } while (--length != 0); | |
1006 | } | |
1007 | } | |
1008 | #else | |
1009 | # define check_match(s, start, match, length) | |
1010 | #endif | |
1011 | ||
1012 | /* =========================================================================== | |
1013 | * Fill the window when the lookahead becomes insufficient. | |
1014 | * Updates strstart and lookahead. | |
1015 | * | |
1016 | * IN assertion: lookahead < MIN_LOOKAHEAD | |
1017 | * OUT assertions: strstart <= window_size-MIN_LOOKAHEAD | |
1018 | * At least one byte has been read, or avail_in == 0; reads are | |
1019 | * performed for at least two bytes (required for the zip translate_eol | |
1020 | * option -- not supported here). | |
1021 | */ | |
1022 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
1023 | local void fill_window(deflate_state* s) | |
1024 | #else | |
1025 | local void fill_window(s) | |
1026 | deflate_state *s; | |
1027 | #endif | |
1028 | { | |
1029 | register unsigned n, m; | |
1030 | register Posf *p; | |
1031 | unsigned more; /* Amount of free space at the end of the window. */ | |
1032 | uInt wsize = s->w_size; | |
1033 | ||
1034 | do { | |
1035 | more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart); | |
1036 | ||
1037 | /* Deal with !@#$% 64K limit: */ | |
1038 | if (more == 0 && s->strstart == 0 && s->lookahead == 0) { | |
1039 | more = wsize; | |
1040 | ||
1041 | } else if (more == (unsigned)(-1)) { | |
1042 | /* Very unlikely, but possible on 16 bit machine if strstart == 0 | |
1043 | * and lookahead == 1 (input done one byte at time) | |
1044 | */ | |
1045 | more--; | |
1046 | ||
1047 | /* If the window is almost full and there is insufficient lookahead, | |
1048 | * move the upper half to the lower one to make room in the upper half. | |
1049 | */ | |
1050 | } else if (s->strstart >= wsize+MAX_DIST(s)) { | |
1051 | ||
1052 | zmemcpy(s->window, s->window+wsize, (unsigned)wsize); | |
1053 | s->match_start -= wsize; | |
1054 | s->strstart -= wsize; /* we now have strstart >= MAX_DIST */ | |
1055 | s->block_start -= (long) wsize; | |
1056 | ||
1057 | /* Slide the hash table (could be avoided with 32 bit values | |
1058 | at the expense of memory usage). We slide even when level == 0 | |
1059 | to keep the hash table consistent if we switch back to level > 0 | |
1060 | later. (Using level 0 permanently is not an optimal usage of | |
1061 | zlib, so we don't care about this pathological case.) | |
1062 | */ | |
1063 | n = s->hash_size; | |
1064 | p = &s->head[n]; | |
1065 | do { | |
1066 | m = *--p; | |
1067 | *p = (Pos)(m >= wsize ? m-wsize : NIL); | |
1068 | } while (--n); | |
1069 | ||
1070 | n = wsize; | |
1071 | #ifndef FASTEST | |
1072 | p = &s->prev[n]; | |
1073 | do { | |
1074 | m = *--p; | |
1075 | *p = (Pos)(m >= wsize ? m-wsize : NIL); | |
1076 | /* If n is not on any hash chain, prev[n] is garbage but | |
1077 | * its value will never be used. | |
1078 | */ | |
1079 | } while (--n); | |
1080 | #endif | |
1081 | more += wsize; | |
1082 | } | |
1083 | if (s->strm->avail_in == 0) return; | |
1084 | ||
1085 | /* If there was no sliding: | |
1086 | * strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 && | |
1087 | * more == window_size - lookahead - strstart | |
1088 | * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1) | |
1089 | * => more >= window_size - 2*WSIZE + 2 | |
1090 | * In the BIG_MEM or MMAP case (not yet supported), | |
1091 | * window_size == input_size + MIN_LOOKAHEAD && | |
1092 | * strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD. | |
1093 | * Otherwise, window_size == 2*WSIZE so more >= 2. | |
1094 | * If there was sliding, more >= WSIZE. So in all cases, more >= 2. | |
1095 | */ | |
1096 | Assert(more >= 2, "more < 2"); | |
1097 | ||
1098 | n = read_buf(s->strm, s->window + s->strstart + s->lookahead, more); | |
1099 | s->lookahead += n; | |
1100 | ||
1101 | /* Initialize the hash value now that we have some input: */ | |
1102 | if (s->lookahead >= MIN_MATCH) { | |
1103 | s->ins_h = s->window[s->strstart]; | |
1104 | UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]); | |
1105 | #if MIN_MATCH != 3 | |
1106 | Call UPDATE_HASH() MIN_MATCH-3 more times | |
1107 | #endif | |
1108 | } | |
1109 | /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage, | |
1110 | * but this is not important since only literal bytes will be emitted. | |
1111 | */ | |
1112 | ||
1113 | } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0); | |
1114 | } | |
1115 | ||
1116 | /* =========================================================================== | |
1117 | * Flush the current block, with given end-of-file flag. | |
1118 | * IN assertion: strstart is set to the end of the current match. | |
1119 | */ | |
1120 | #define FLUSH_BLOCK_ONLY(s, eof) { \ | |
1121 | _tr_flush_block(s, (s->block_start >= 0L ? \ | |
1122 | (charf *)&s->window[(unsigned)s->block_start] : \ | |
1123 | (charf *)Z_NULL), \ | |
1124 | (ulg)((long)s->strstart - s->block_start), \ | |
1125 | (eof)); \ | |
1126 | s->block_start = s->strstart; \ | |
1127 | flush_pending(s->strm); \ | |
1128 | Tracev((stderr,"[FLUSH]")); \ | |
1129 | } | |
1130 | ||
1131 | /* Same but force premature exit if necessary. */ | |
1132 | #define FLUSH_BLOCK(s, eof) { \ | |
1133 | FLUSH_BLOCK_ONLY(s, eof); \ | |
1134 | if (s->strm->avail_out == 0) return (eof) ? finish_started : need_more; \ | |
1135 | } | |
1136 | ||
1137 | /* =========================================================================== | |
1138 | * Copy without compression as much as possible from the input stream, return | |
1139 | * the current block state. | |
1140 | * This function does not insert new strings in the dictionary since | |
1141 | * uncompressible data is probably not useful. This function is used | |
1142 | * only for the level=0 compression option. | |
1143 | * NOTE: this function should be optimized to avoid extra copying from | |
1144 | * window to pending_buf. | |
1145 | */ | |
1146 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
1147 | local block_state deflate_stored(deflate_state* s, int flush) | |
1148 | #else | |
1149 | local block_state deflate_stored(s, flush) | |
1150 | deflate_state *s; | |
1151 | int flush; | |
1152 | #endif | |
1153 | { | |
1154 | /* Stored blocks are limited to 0xffff bytes, pending_buf is limited | |
1155 | * to pending_buf_size, and each stored block has a 5 byte header: | |
1156 | */ | |
1157 | ulg max_block_size = 0xffff; | |
1158 | ulg max_start; | |
1159 | ||
1160 | if (max_block_size > s->pending_buf_size - 5) { | |
1161 | max_block_size = s->pending_buf_size - 5; | |
1162 | } | |
1163 | ||
1164 | /* Copy as much as possible from input to output: */ | |
1165 | for (;;) { | |
1166 | /* Fill the window as much as possible: */ | |
1167 | if (s->lookahead <= 1) { | |
1168 | ||
1169 | Assert(s->strstart < s->w_size+MAX_DIST(s) || | |
1170 | s->block_start >= (long)s->w_size, "slide too late"); | |
1171 | ||
1172 | fill_window(s); | |
1173 | if (s->lookahead == 0 && flush == Z_NO_FLUSH) return need_more; | |
1174 | ||
1175 | if (s->lookahead == 0) break; /* flush the current block */ | |
1176 | } | |
1177 | Assert(s->block_start >= 0L, "block gone"); | |
1178 | ||
1179 | s->strstart += s->lookahead; | |
1180 | s->lookahead = 0; | |
1181 | ||
1182 | /* Emit a stored block if pending_buf will be full: */ | |
1183 | max_start = s->block_start + max_block_size; | |
1184 | if (s->strstart == 0 || (ulg)s->strstart >= max_start) { | |
1185 | /* strstart == 0 is possible when wraparound on 16-bit machine */ | |
1186 | s->lookahead = (uInt)(s->strstart - max_start); | |
1187 | s->strstart = (uInt)max_start; | |
1188 | FLUSH_BLOCK(s, 0); | |
1189 | } | |
1190 | /* Flush if we may have to slide, otherwise block_start may become | |
1191 | * negative and the data will be gone: | |
1192 | */ | |
1193 | if (s->strstart - (uInt)s->block_start >= MAX_DIST(s)) { | |
1194 | FLUSH_BLOCK(s, 0); | |
1195 | } | |
1196 | } | |
1197 | FLUSH_BLOCK(s, flush == Z_FINISH); | |
1198 | return flush == Z_FINISH ? finish_done : block_done; | |
1199 | } | |
1200 | ||
1201 | /* =========================================================================== | |
1202 | * Compress as much as possible from the input stream, return the current | |
1203 | * block state. | |
1204 | * This function does not perform lazy evaluation of matches and inserts | |
1205 | * new strings in the dictionary only for unmatched strings or for short | |
1206 | * matches. It is used only for the fast compression options. | |
1207 | */ | |
1208 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
1209 | local block_state deflate_fast(deflate_state* s, int flush) | |
1210 | #else | |
1211 | local block_state deflate_fast(s, flush) | |
1212 | deflate_state *s; | |
1213 | int flush; | |
1214 | #endif | |
1215 | { | |
1216 | IPos hash_head = NIL; /* head of the hash chain */ | |
1217 | int bflush; /* set if current block must be flushed */ | |
1218 | ||
1219 | for (;;) { | |
1220 | /* Make sure that we always have enough lookahead, except | |
1221 | * at the end of the input file. We need MAX_MATCH bytes | |
1222 | * for the next match, plus MIN_MATCH bytes to insert the | |
1223 | * string following the next match. | |
1224 | */ | |
1225 | if (s->lookahead < MIN_LOOKAHEAD) { | |
1226 | fill_window(s); | |
1227 | if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { | |
1228 | return need_more; | |
1229 | } | |
1230 | if (s->lookahead == 0) break; /* flush the current block */ | |
1231 | } | |
1232 | ||
1233 | /* Insert the string window[strstart .. strstart+2] in the | |
1234 | * dictionary, and set hash_head to the head of the hash chain: | |
1235 | */ | |
1236 | if (s->lookahead >= MIN_MATCH) { | |
1237 | INSERT_STRING(s, s->strstart, hash_head); | |
1238 | } | |
1239 | ||
1240 | /* Find the longest match, discarding those <= prev_length. | |
1241 | * At this point we have always match_length < MIN_MATCH | |
1242 | */ | |
1243 | if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) { | |
1244 | /* To simplify the code, we prevent matches with the string | |
1245 | * of window index 0 (in particular we have to avoid a match | |
1246 | * of the string with itself at the start of the input file). | |
1247 | */ | |
1248 | if (s->strategy != Z_HUFFMAN_ONLY) { | |
1249 | s->match_length = longest_match (s, hash_head); | |
1250 | } | |
1251 | /* longest_match() sets match_start */ | |
1252 | } | |
1253 | if (s->match_length >= MIN_MATCH) { | |
1254 | check_match(s, s->strstart, s->match_start, s->match_length); | |
1255 | ||
1256 | _tr_tally_dist(s, s->strstart - s->match_start, | |
1257 | s->match_length - MIN_MATCH, bflush); | |
1258 | ||
1259 | s->lookahead -= s->match_length; | |
1260 | ||
1261 | /* Insert new strings in the hash table only if the match length | |
1262 | * is not too large. This saves time but degrades compression. | |
1263 | */ | |
1264 | #ifndef FASTEST | |
1265 | if (s->match_length <= s->max_insert_length && | |
1266 | s->lookahead >= MIN_MATCH) { | |
1267 | s->match_length--; /* string at strstart already in hash table */ | |
1268 | do { | |
1269 | s->strstart++; | |
1270 | INSERT_STRING(s, s->strstart, hash_head); | |
1271 | /* strstart never exceeds WSIZE-MAX_MATCH, so there are | |
1272 | * always MIN_MATCH bytes ahead. | |
1273 | */ | |
1274 | } while (--s->match_length != 0); | |
1275 | s->strstart++; | |
1276 | } else | |
1277 | #endif | |
1278 | { | |
1279 | s->strstart += s->match_length; | |
1280 | s->match_length = 0; | |
1281 | s->ins_h = s->window[s->strstart]; | |
1282 | UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]); | |
1283 | #if MIN_MATCH != 3 | |
1284 | Call UPDATE_HASH() MIN_MATCH-3 more times | |
1285 | #endif | |
1286 | /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not | |
1287 | * matter since it will be recomputed at next deflate call. | |
1288 | */ | |
1289 | } | |
1290 | } else { | |
1291 | /* No match, output a literal byte */ | |
1292 | Tracevv((stderr,"%c", s->window[s->strstart])); | |
1293 | _tr_tally_lit (s, s->window[s->strstart], bflush); | |
1294 | s->lookahead--; | |
1295 | s->strstart++; | |
1296 | } | |
1297 | if (bflush) FLUSH_BLOCK(s, 0); | |
1298 | } | |
1299 | FLUSH_BLOCK(s, flush == Z_FINISH); | |
1300 | return flush == Z_FINISH ? finish_done : block_done; | |
1301 | } | |
1302 | ||
1303 | /* =========================================================================== | |
1304 | * Same as above, but achieves better compression. We use a lazy | |
1305 | * evaluation for matches: a match is finally adopted only if there is | |
1306 | * no better match at the next window position. | |
1307 | */ | |
1308 | #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */ | |
1309 | local block_state deflate_slow(deflate_state* s, int flush) | |
1310 | #else | |
1311 | local block_state deflate_slow(s, flush) | |
1312 | deflate_state *s; | |
1313 | int flush; | |
1314 | #endif | |
1315 | { | |
1316 | IPos hash_head = NIL; /* head of hash chain */ | |
1317 | int bflush; /* set if current block must be flushed */ | |
1318 | ||
1319 | /* Process the input block. */ | |
1320 | for (;;) { | |
1321 | /* Make sure that we always have enough lookahead, except | |
1322 | * at the end of the input file. We need MAX_MATCH bytes | |
1323 | * for the next match, plus MIN_MATCH bytes to insert the | |
1324 | * string following the next match. | |
1325 | */ | |
1326 | if (s->lookahead < MIN_LOOKAHEAD) { | |
1327 | fill_window(s); | |
1328 | if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) { | |
1329 | return need_more; | |
1330 | } | |
1331 | if (s->lookahead == 0) break; /* flush the current block */ | |
1332 | } | |
1333 | ||
1334 | /* Insert the string window[strstart .. strstart+2] in the | |
1335 | * dictionary, and set hash_head to the head of the hash chain: | |
1336 | */ | |
1337 | if (s->lookahead >= MIN_MATCH) { | |
1338 | INSERT_STRING(s, s->strstart, hash_head); | |
1339 | } | |
1340 | ||
1341 | /* Find the longest match, discarding those <= prev_length. | |
1342 | */ | |
1343 | s->prev_length = s->match_length, s->prev_match = s->match_start; | |
1344 | s->match_length = MIN_MATCH-1; | |
1345 | ||
1346 | if (hash_head != NIL && s->prev_length < s->max_lazy_match && | |
1347 | s->strstart - hash_head <= MAX_DIST(s)) { | |
1348 | /* To simplify the code, we prevent matches with the string | |
1349 | * of window index 0 (in particular we have to avoid a match | |
1350 | * of the string with itself at the start of the input file). | |
1351 | */ | |
1352 | if (s->strategy != Z_HUFFMAN_ONLY) { | |
1353 | s->match_length = longest_match (s, hash_head); | |
1354 | } | |
1355 | /* longest_match() sets match_start */ | |
1356 | ||
1357 | if (s->match_length <= 5 && (s->strategy == Z_FILTERED || | |
1358 | (s->match_length == MIN_MATCH && | |
1359 | s->strstart - s->match_start > TOO_FAR))) { | |
1360 | ||
1361 | /* If prev_match is also MIN_MATCH, match_start is garbage | |
1362 | * but we will ignore the current match anyway. | |
1363 | */ | |
1364 | s->match_length = MIN_MATCH-1; | |
1365 | } | |
1366 | } | |
1367 | /* If there was a match at the previous step and the current | |
1368 | * match is not better, output the previous match: | |
1369 | */ | |
1370 | if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) { | |
1371 | uInt max_insert = s->strstart + s->lookahead - MIN_MATCH; | |
1372 | /* Do not insert strings in hash table beyond this. */ | |
1373 | ||
1374 | check_match(s, s->strstart-1, s->prev_match, s->prev_length); | |
1375 | ||
1376 | _tr_tally_dist(s, s->strstart -1 - s->prev_match, | |
1377 | s->prev_length - MIN_MATCH, bflush); | |
1378 | ||
1379 | /* Insert in hash table all strings up to the end of the match. | |
1380 | * strstart-1 and strstart are already inserted. If there is not | |
1381 | * enough lookahead, the last two strings are not inserted in | |
1382 | * the hash table. | |
1383 | */ | |
1384 | s->lookahead -= s->prev_length-1; | |
1385 | s->prev_length -= 2; | |
1386 | do { | |
1387 | if (++s->strstart <= max_insert) { | |
1388 | INSERT_STRING(s, s->strstart, hash_head); | |
1389 | } | |
1390 | } while (--s->prev_length != 0); | |
1391 | s->match_available = 0; | |
1392 | s->match_length = MIN_MATCH-1; | |
1393 | s->strstart++; | |
1394 | ||
1395 | if (bflush) FLUSH_BLOCK(s, 0); | |
1396 | ||
1397 | } else if (s->match_available) { | |
1398 | /* If there was no match at the previous position, output a | |
1399 | * single literal. If there was a match but the current match | |
1400 | * is longer, truncate the previous match to a single literal. | |
1401 | */ | |
1402 | Tracevv((stderr,"%c", s->window[s->strstart-1])); | |
1403 | _tr_tally_lit(s, s->window[s->strstart-1], bflush); | |
1404 | if (bflush) { | |
1405 | FLUSH_BLOCK_ONLY(s, 0); | |
1406 | } | |
1407 | s->strstart++; | |
1408 | s->lookahead--; | |
1409 | if (s->strm->avail_out == 0) return need_more; | |
1410 | } else { | |
1411 | /* There is no previous match to compare with, wait for | |
1412 | * the next step to decide. | |
1413 | */ | |
1414 | s->match_available = 1; | |
1415 | s->strstart++; | |
1416 | s->lookahead--; | |
1417 | } | |
1418 | } | |
1419 | Assert (flush != Z_NO_FLUSH, "no flush?"); | |
1420 | if (s->match_available) { | |
1421 | Tracevv((stderr,"%c", s->window[s->strstart-1])); | |
1422 | _tr_tally_lit(s, s->window[s->strstart-1], bflush); | |
1423 | s->match_available = 0; | |
1424 | } | |
1425 | FLUSH_BLOCK(s, flush == Z_FINISH); | |
1426 | return flush == Z_FINISH ? finish_done : block_done; | |
1427 | } |