SecurityServer/MacYarrow/zlib/deflate.h

   1 /*
   2  * Copyright (c) 2000-2001 Apple Computer, Inc. All Rights Reserved.
   3  *
   4  * The contents of this file constitute Original Code as defined in and are
   5  * subject to the Apple Public Source License Version 1.2 (the 'License').
   6  * You may not use this file except in compliance with the License. Please obtain
   7  * a copy of the License at http://www.apple.com/publicsource and read it before
   8  * using this file.
   9  *
  10  * This Original Code and all software distributed under the License are
  11  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS
  12  * OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES, INCLUDING WITHOUT
  13  * LIMITATION, ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
  14  * PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT. Please see the License for the
  15  * specific language governing rights and limitations under the License.
  16  */
  17
  18
  19 /* deflate.h -- internal compression state
  20  * Copyright (C) 1995-1998 Jean-loup Gailly
  21  * For conditions of distribution and use, see copyright notice in zlib.h
  22  */
  23
  24 /* WARNING: this file should *not* be used by applications. It is
  25    part of the implementation of the compression library and is
  26    subject to change. Applications should only use zlib.h.
  27  */
  28
  29 /* @(#) $Id: deflate.h,v 1.1.1.1 2001/05/18 23:14:03 mb Exp $ */
  30
  31 #ifndef _DEFLATE_H
  32 #define _DEFLATE_H
  33
  34 #include "zutil.h"
  35
  36 /* ===========================================================================
  37  * Internal compression state.
  38  */
  39
  40 #define LENGTH_CODES 29
  41 /* number of length codes, not counting the special END_BLOCK code */
  42
  43 #define LITERALS  256
  44 /* number of literal bytes 0..255 */
  45
  46 #define L_CODES (LITERALS+1+LENGTH_CODES)
  47 /* number of Literal or Length codes, including the END_BLOCK code */
  48
  49 #define D_CODES   30
  50 /* number of distance codes */
  51
  52 #define BL_CODES  19
  53 /* number of codes used to transfer the bit lengths */
  54
  55 #define HEAP_SIZE (2*L_CODES+1)
  56 /* maximum heap size */
  57
  58 #define MAX_BITS 15
  59 /* All codes must not exceed MAX_BITS bits */
  60
  61 #define INIT_STATE    42
  62 #define BUSY_STATE   113
  63 #define FINISH_STATE 666
  64 /* Stream status */
  65
  66
  67 /* Data structure describing a single value and its code string. */
  68 typedef struct ct_data_s {
  69     union {
  70         ush  freq;       /* frequency count */
  71         ush  code;       /* bit string */
  72     } fc;
  73     union {
  74         ush  dad;        /* father node in Huffman tree */
  75         ush  len;        /* length of bit string */
  76     } dl;
  77 } FAR ct_data;
  78
  79 #define Freq fc.freq
  80 #define Code fc.code
  81 #define Dad  dl.dad
  82 #define Len  dl.len
  83
  84 typedef struct static_tree_desc_s  static_tree_desc;
  85
  86 typedef struct tree_desc_s {
  87     ct_data *dyn_tree;           /* the dynamic tree */
  88     int     max_code;            /* largest code with non zero frequency */
  89     static_tree_desc *stat_desc; /* the corresponding static tree */
  90 } FAR tree_desc;
  91
  92 typedef ush Pos;
  93 typedef Pos FAR Posf;
  94 typedef unsigned IPos;
  95
  96 /* A Pos is an index in the character window. We use short instead of int to
  97  * save space in the various tables. IPos is used only for parameter passing.
  98  */
  99
 100 typedef struct internal_state {
 101     z_streamp strm;      /* pointer back to this zlib stream */
 102     int   status;        /* as the name implies */
 103     Bytef *pending_buf;  /* output still pending */
 104     ulg   pending_buf_size; /* size of pending_buf */
 105     Bytef *pending_out;  /* next pending byte to output to the stream */
 106     int   pending;       /* nb of bytes in the pending buffer */
 107     int   noheader;      /* suppress zlib header and adler32 */
 108     Byte  data_type;     /* UNKNOWN, BINARY or ASCII */
 109     Byte  method;        /* STORED (for zip only) or DEFLATED */
 110     int   last_flush;    /* value of flush param for previous deflate call */
 111
 112                 /* used by deflate.c: */
 113
 114     uInt  w_size;        /* LZ77 window size (32K by default) */
 115     uInt  w_bits;        /* log2(w_size)  (8..16) */
 116     uInt  w_mask;        /* w_size - 1 */
 117
 118     Bytef *window;
 119     /* Sliding window. Input bytes are read into the second half of the window,
 120      * and move to the first half later to keep a dictionary of at least wSize
 121      * bytes. With this organization, matches are limited to a distance of
 122      * wSize-MAX_MATCH bytes, but this ensures that IO is always
 123      * performed with a length multiple of the block size. Also, it limits
 124      * the window size to 64K, which is quite useful on MSDOS.
 125      * To do: use the user input buffer as sliding window.
 126      */
 127
 128     ulg window_size;
 129     /* Actual size of window: 2*wSize, except when the user input buffer
 130      * is directly used as sliding window.
 131      */
 132
 133     Posf *prev;
 134     /* Link to older string with same hash index. To limit the size of this
 135      * array to 64K, this link is maintained only for the last 32K strings.
 136      * An index in this array is thus a window index modulo 32K.
 137      */
 138
 139     Posf *head; /* Heads of the hash chains or NIL. */
 140
 141     uInt  ins_h;          /* hash index of string to be inserted */
 142     uInt  hash_size;      /* number of elements in hash table */
 143     uInt  hash_bits;      /* log2(hash_size) */
 144     uInt  hash_mask;      /* hash_size-1 */
 145
 146     uInt  hash_shift;
 147     /* Number of bits by which ins_h must be shifted at each input
 148      * step. It must be such that after MIN_MATCH steps, the oldest
 149      * byte no longer takes part in the hash key, that is:
 150      *   hash_shift * MIN_MATCH >= hash_bits
 151      */
 152
 153     long block_start;
 154     /* Window position at the beginning of the current output block. Gets
 155      * negative when the window is moved backwards.
 156      */
 157
 158     uInt match_length;           /* length of best match */
 159     IPos prev_match;             /* previous match */
 160     int match_available;         /* set if previous match exists */
 161     uInt strstart;               /* start of string to insert */
 162     uInt match_start;            /* start of matching string */
 163     uInt lookahead;              /* number of valid bytes ahead in window */
 164
 165     uInt prev_length;
 166     /* Length of the best match at previous step. Matches not greater than this
 167      * are discarded. This is used in the lazy match evaluation.
 168      */
 169
 170     uInt max_chain_length;
 171     /* To speed up deflation, hash chains are never searched beyond this
 172      * length.  A higher limit improves compression ratio but degrades the
 173      * speed.
 174      */
 175
 176     uInt max_lazy_match;
 177     /* Attempt to find a better match only when the current match is strictly
 178      * smaller than this value. This mechanism is used only for compression
 179      * levels >= 4.
 180      */
 181 #   define max_insert_length  max_lazy_match
 182     /* Insert new strings in the hash table only if the match length is not
 183      * greater than this length. This saves time but degrades compression.
 184      * max_insert_length is used only for compression levels <= 3.
 185      */
 186
 187     int level;    /* compression level (1..9) */
 188     int strategy; /* favor or force Huffman coding*/
 189
 190     uInt good_match;
 191     /* Use a faster search when the previous match is longer than this */
 192
 193     int nice_match; /* Stop searching when current match exceeds this */
 194
 195                 /* used by trees.c: */
 196     /* Didn't use ct_data typedef below to supress compiler warning */
 197     struct ct_data_s dyn_ltree[HEAP_SIZE];   /* literal and length tree */
 198     struct ct_data_s dyn_dtree[2*D_CODES+1]; /* distance tree */
 199     struct ct_data_s bl_tree[2*BL_CODES+1];  /* Huffman tree for bit lengths */
 200
 201     struct tree_desc_s l_desc;               /* desc. for literal tree */
 202     struct tree_desc_s d_desc;               /* desc. for distance tree */
 203     struct tree_desc_s bl_desc;              /* desc. for bit length tree */
 204
 205     ush bl_count[MAX_BITS+1];
 206     /* number of codes at each bit length for an optimal tree */
 207
 208     int heap[2*L_CODES+1];      /* heap used to build the Huffman trees */
 209     int heap_len;               /* number of elements in the heap */
 210     int heap_max;               /* element of largest frequency */
 211     /* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
 212      * The same heap array is used to build all trees.
 213      */
 214
 215     uch depth[2*L_CODES+1];
 216     /* Depth of each subtree used as tie breaker for trees of equal frequency
 217      */
 218
 219     uchf *l_buf;          /* buffer for literals or lengths */
 220
 221     uInt  lit_bufsize;
 222     /* Size of match buffer for literals/lengths.  There are 4 reasons for
 223      * limiting lit_bufsize to 64K:
 224      *   - frequencies can be kept in 16 bit counters
 225      *   - if compression is not successful for the first block, all input
 226      *     data is still in the window so we can still emit a stored block even
 227      *     when input comes from standard input.  (This can also be done for
 228      *     all blocks if lit_bufsize is not greater than 32K.)
 229      *   - if compression is not successful for a file smaller than 64K, we can
 230      *     even emit a stored file instead of a stored block (saving 5 bytes).
 231      *     This is applicable only for zip (not gzip or zlib).
 232      *   - creating new Huffman trees less frequently may not provide fast
 233      *     adaptation to changes in the input data statistics. (Take for
 234      *     example a binary file with poorly compressible code followed by
 235      *     a highly compressible string table.) Smaller buffer sizes give
 236      *     fast adaptation but have of course the overhead of transmitting
 237      *     trees more frequently.
 238      *   - I can't count above 4
 239      */
 240
 241     uInt last_lit;      /* running index in l_buf */
 242
 243     ushf *d_buf;
 244     /* Buffer for distances. To simplify the code, d_buf and l_buf have
 245      * the same number of elements. To use different lengths, an extra flag
 246      * array would be necessary.
 247      */
 248
 249     ulg opt_len;        /* bit length of current block with optimal trees */
 250     ulg static_len;     /* bit length of current block with static trees */
 251     uInt matches;       /* number of string matches in current block */
 252     int last_eob_len;   /* bit length of EOB code for last block */
 253
 254 #ifdef DEBUG
 255     ulg compressed_len; /* total bit length of compressed file mod 2^32 */
 256     ulg bits_sent;      /* bit length of compressed data sent mod 2^32 */
 257 #endif
 258
 259     ush bi_buf;
 260     /* Output buffer. bits are inserted starting at the bottom (least
 261      * significant bits).
 262      */
 263     int bi_valid;
 264     /* Number of valid bits in bi_buf.  All bits above the last valid bit
 265      * are always zero.
 266      */
 267
 268 } FAR deflate_state;
 269
 270 /* Output a byte on the stream.
 271  * IN assertion: there is enough room in pending_buf.
 272  */
 273 #define put_byte(s, c) {s->pending_buf[s->pending++] = (c);}
 274
 275
 276 #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
 277 /* Minimum amount of lookahead, except at the end of the input file.
 278  * See deflate.c for comments about the MIN_MATCH+1.
 279  */
 280
 281 #define MAX_DIST(s)  ((s)->w_size-MIN_LOOKAHEAD)
 282 /* In order to simplify the code, particularly on 16 bit machines, match
 283  * distances are limited to MAX_DIST instead of WSIZE.
 284  */
 285
 286         /* in trees.c */
 287 void _tr_init         OF((deflate_state *s));
 288 int  _tr_tally        OF((deflate_state *s, unsigned dist, unsigned lc));
 289 void _tr_flush_block  OF((deflate_state *s, charf *buf, ulg stored_len,
 290                           int eof));
 291 void _tr_align        OF((deflate_state *s));
 292 void _tr_stored_block OF((deflate_state *s, charf *buf, ulg stored_len,
 293                           int eof));
 294
 295 #define d_code(dist) \
 296    ((dist) < 256 ? _dist_code[dist] : _dist_code[256+((dist)>>7)])
 297 /* Mapping from a distance to a distance code. dist is the distance - 1 and
 298  * must not have side effects. _dist_code[256] and _dist_code[257] are never
 299  * used.
 300  */
 301
 302 #ifndef DEBUG
 303 /* Inline versions of _tr_tally for speed: */
 304
 305 #if defined(GEN_TREES_H) || !defined(STDC)
 306   extern uch _length_code[];
 307   extern uch _dist_code[];
 308 #else
 309   extern const uch _length_code[];
 310   extern const uch _dist_code[];
 311 #endif
 312
 313 # define _tr_tally_lit(s, c, flush) \
 314   { uch cc = (c); \
 315     s->d_buf[s->last_lit] = 0; \
 316     s->l_buf[s->last_lit++] = cc; \
 317     s->dyn_ltree[cc].Freq++; \
 318     flush = (s->last_lit == s->lit_bufsize-1); \
 319    }
 320 # define _tr_tally_dist(s, distance, length, flush) \
 321   { uch len = (length); \
 322     ush dist = (distance); \
 323     s->d_buf[s->last_lit] = dist; \
 324     s->l_buf[s->last_lit++] = len; \
 325     dist--; \
 326     s->dyn_ltree[_length_code[len]+LITERALS+1].Freq++; \
 327     s->dyn_dtree[d_code(dist)].Freq++; \
 328     flush = (s->last_lit == s->lit_bufsize-1); \
 329   }
 330 #else
 331 # define _tr_tally_lit(s, c, flush) flush = _tr_tally(s, 0, c)
 332 # define _tr_tally_dist(s, distance, length, flush) \
 333               flush = _tr_tally(s, distance, length)
 334 #endif
 335
 336 #endif