deflate.h Source File

00001 /* deflate.h -- internal compression state
+00002  * Copyright (C) 1995-1998 Jean-loup Gailly
+00003  * For conditions of distribution and use, see copyright notice in zlib.h 
+00004  */
+00005 
+00006 /* WARNING: this file should *not* be used by applications. It is
+00007    part of the implementation of the compression library and is
+00008    subject to change. Applications should only use zlib.h.
+00009  */
+00010 
+00011 /* @(#) $Id: deflate_8h-source.html,v 1.3 2002/06/20 20:23:08 mgruner Exp $ */
+00012 
+00013 #ifndef _DEFLATE_H
+00014 #define _DEFLATE_H
+00015 
+00016 #include "zutil.h"
+00017 
+00018 /* ===========================================================================
+00019  * Internal compression state.
+00020  */
+00021 
+00022 #define LENGTH_CODES 29
+00023 /* number of length codes, not counting the special END_BLOCK code */
+00024 
+00025 #define LITERALS  256
+00026 /* number of literal bytes 0..255 */
+00027 
+00028 #define L_CODES (LITERALS+1+LENGTH_CODES)
+00029 /* number of Literal or Length codes, including the END_BLOCK code */
+00030 
+00031 #define D_CODES   30
+00032 /* number of distance codes */
+00033 
+00034 #define BL_CODES  19
+00035 /* number of codes used to transfer the bit lengths */
+00036 
+00037 #define HEAP_SIZE (2*L_CODES+1)
+00038 /* maximum heap size */
+00039 
+00040 #define MAX_BITS 15
+00041 /* All codes must not exceed MAX_BITS bits */
+00042 
+00043 #define INIT_STATE    42
+00044 #define BUSY_STATE   113
+00045 #define FINISH_STATE 666
+00046 /* Stream status */
+00047 
+00048 
+00049 /* Data structure describing a single value and its code string. */
+00050 typedef struct ct_data_s {
+00051     union {
+00052         ush  freq;       /* frequency count */
+00053         ush  code;       /* bit string */
+00054     } fc;
+00055     union {
+00056         ush  dad;        /* father node in Huffman tree */
+00057         ush  len;        /* length of bit string */
+00058     } dl;
+00059 } FAR ct_data;
+00060 
+00061 #define Freq fc.freq
+00062 #define Code fc.code
+00063 #define Dad  dl.dad
+00064 #define Len  dl.len
+00065 
+00066 typedef struct static_tree_desc_s  static_tree_desc;
+00067 
+00068 typedef struct tree_desc_s {
+00069     ct_data *dyn_tree;           /* the dynamic tree */
+00070     int     max_code;            /* largest code with non zero frequency */
+00071     static_tree_desc *stat_desc; /* the corresponding static tree */
+00072 } FAR tree_desc;
+00073 
+00074 typedef ush Pos;
+00075 typedef Pos FAR Posf;
+00076 typedef unsigned IPos;
+00077 
+00078 /* A Pos is an index in the character window. We use short instead of int to
+00079  * save space in the various tables. IPos is used only for parameter passing.
+00080  */
+00081 
+00082 typedef struct internal_state {
+00083     z_streamp strm;      /* pointer back to this zlib stream */
+00084     int   status;        /* as the name implies */
+00085     Bytef *pending_buf;  /* output still pending */
+00086     ulg   pending_buf_size; /* size of pending_buf */
+00087     Bytef *pending_out;  /* next pending byte to output to the stream */
+00088     int   pending;       /* nb of bytes in the pending buffer */
+00089     int   noheader;      /* suppress zlib header and adler32 */
+00090     Byte  data_type;     /* UNKNOWN, BINARY or ASCII */
+00091     Byte  method;        /* STORED (for zip only) or DEFLATED */
+00092     int   last_flush;    /* value of flush param for previous deflate call */
+00093 
+00094                 /* used by deflate.c: */
+00095 
+00096     uInt  w_size;        /* LZ77 window size (32K by default) */
+00097     uInt  w_bits;        /* log2(w_size)  (8..16) */
+00098     uInt  w_mask;        /* w_size - 1 */
+00099 
+00100     Bytef *window;
+00101     /* Sliding window. Input bytes are read into the second half of the window,
+00102      * and move to the first half later to keep a dictionary of at least wSize
+00103      * bytes. With this organization, matches are limited to a distance of
+00104      * wSize-MAX_MATCH bytes, but this ensures that IO is always
+00105      * performed with a length multiple of the block size. Also, it limits
+00106      * the window size to 64K, which is quite useful on MSDOS.
+00107      * To do: use the user input buffer as sliding window.
+00108      */
+00109 
+00110     ulg window_size;
+00111     /* Actual size of window: 2*wSize, except when the user input buffer
+00112      * is directly used as sliding window.
+00113      */
+00114 
+00115     Posf *prev;
+00116     /* Link to older string with same hash index. To limit the size of this
+00117      * array to 64K, this link is maintained only for the last 32K strings.
+00118      * An index in this array is thus a window index modulo 32K.
+00119      */
+00120 
+00121     Posf *head; /* Heads of the hash chains or NIL. */
+00122 
+00123     uInt  ins_h;          /* hash index of string to be inserted */
+00124     uInt  hash_size;      /* number of elements in hash table */
+00125     uInt  hash_bits;      /* log2(hash_size) */
+00126     uInt  hash_mask;      /* hash_size-1 */
+00127 
+00128     uInt  hash_shift;
+00129     /* Number of bits by which ins_h must be shifted at each input
+00130      * step. It must be such that after MIN_MATCH steps, the oldest
+00131      * byte no longer takes part in the hash key, that is:
+00132      *   hash_shift * MIN_MATCH >= hash_bits
+00133      */
+00134 
+00135     long block_start;
+00136     /* Window position at the beginning of the current output block. Gets
+00137      * negative when the window is moved backwards.
+00138      */
+00139 
+00140     uInt match_length;           /* length of best match */
+00141     IPos prev_match;             /* previous match */
+00142     int match_available;         /* set if previous match exists */
+00143     uInt strstart;               /* start of string to insert */
+00144     uInt match_start;            /* start of matching string */
+00145     uInt lookahead;              /* number of valid bytes ahead in window */
+00146 
+00147     uInt prev_length;
+00148     /* Length of the best match at previous step. Matches not greater than this
+00149      * are discarded. This is used in the lazy match evaluation.
+00150      */
+00151 
+00152     uInt max_chain_length;
+00153     /* To speed up deflation, hash chains are never searched beyond this
+00154      * length.  A higher limit improves compression ratio but degrades the
+00155      * speed.
+00156      */
+00157 
+00158     uInt max_lazy_match;
+00159     /* Attempt to find a better match only when the current match is strictly
+00160      * smaller than this value. This mechanism is used only for compression
+00161      * levels >= 4.
+00162      */
+00163 #   define max_insert_length  max_lazy_match
+00164     /* Insert new strings in the hash table only if the match length is not
+00165      * greater than this length. This saves time but degrades compression.
+00166      * max_insert_length is used only for compression levels <= 3.
+00167      */
+00168 
+00169     int level;    /* compression level (1..9) */
+00170     int strategy; /* favor or force Huffman coding*/
+00171 
+00172     uInt good_match;
+00173     /* Use a faster search when the previous match is longer than this */
+00174 
+00175     int nice_match; /* Stop searching when current match exceeds this */
+00176 
+00177                 /* used by trees.c: */
+00178     /* Didn't use ct_data typedef below to supress compiler warning */
+00179     struct ct_data_s dyn_ltree[HEAP_SIZE];   /* literal and length tree */
+00180     struct ct_data_s dyn_dtree[2*D_CODES+1]; /* distance tree */
+00181     struct ct_data_s bl_tree[2*BL_CODES+1];  /* Huffman tree for bit lengths */
+00182 
+00183     struct tree_desc_s l_desc;               /* desc. for literal tree */
+00184     struct tree_desc_s d_desc;               /* desc. for distance tree */
+00185     struct tree_desc_s bl_desc;              /* desc. for bit length tree */
+00186 
+00187     ush bl_count[MAX_BITS+1];
+00188     /* number of codes at each bit length for an optimal tree */
+00189 
+00190     int heap[2*L_CODES+1];      /* heap used to build the Huffman trees */
+00191     int heap_len;               /* number of elements in the heap */
+00192     int heap_max;               /* element of largest frequency */
+00193     /* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
+00194      * The same heap array is used to build all trees.
+00195      */
+00196 
+00197     uch depth[2*L_CODES+1];
+00198     /* Depth of each subtree used as tie breaker for trees of equal frequency
+00199      */
+00200 
+00201     uchf *l_buf;          /* buffer for literals or lengths */
+00202 
+00203     uInt  lit_bufsize;
+00204     /* Size of match buffer for literals/lengths.  There are 4 reasons for
+00205      * limiting lit_bufsize to 64K:
+00206      *   - frequencies can be kept in 16 bit counters
+00207      *   - if compression is not successful for the first block, all input
+00208      *     data is still in the window so we can still emit a stored block even
+00209      *     when input comes from standard input.  (This can also be done for
+00210      *     all blocks if lit_bufsize is not greater than 32K.)
+00211      *   - if compression is not successful for a file smaller than 64K, we can
+00212      *     even emit a stored file instead of a stored block (saving 5 bytes).
+00213      *     This is applicable only for zip (not gzip or zlib).
+00214      *   - creating new Huffman trees less frequently may not provide fast
+00215      *     adaptation to changes in the input data statistics. (Take for
+00216      *     example a binary file with poorly compressible code followed by
+00217      *     a highly compressible string table.) Smaller buffer sizes give
+00218      *     fast adaptation but have of course the overhead of transmitting
+00219      *     trees more frequently.
+00220      *   - I can't count above 4
+00221      */
+00222 
+00223     uInt last_lit;      /* running index in l_buf */
+00224 
+00225     ushf *d_buf;
+00226     /* Buffer for distances. To simplify the code, d_buf and l_buf have
+00227      * the same number of elements. To use different lengths, an extra flag
+00228      * array would be necessary.
+00229      */
+00230 
+00231     ulg opt_len;        /* bit length of current block with optimal trees */
+00232     ulg static_len;     /* bit length of current block with static trees */
+00233     uInt matches;       /* number of string matches in current block */
+00234     int last_eob_len;   /* bit length of EOB code for last block */
+00235 
+00236 #ifdef DEBUG
+00237     ulg compressed_len; /* total bit length of compressed file mod 2^32 */
+00238     ulg bits_sent;      /* bit length of compressed data sent mod 2^32 */
+00239 #endif
+00240 
+00241     ush bi_buf;
+00242     /* Output buffer. bits are inserted starting at the bottom (least
+00243      * significant bits).
+00244      */
+00245     int bi_valid;
+00246     /* Number of valid bits in bi_buf.  All bits above the last valid bit
+00247      * are always zero.
+00248      */
+00249 
+00250 } FAR deflate_state;
+00251 
+00252 /* Output a byte on the stream.
+00253  * IN assertion: there is enough room in pending_buf.
+00254  */
+00255 #define put_byte(s, c) {s->pending_buf[s->pending++] = (c);}
+00256 
+00257 
+00258 #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
+00259 /* Minimum amount of lookahead, except at the end of the input file.
+00260  * See deflate.c for comments about the MIN_MATCH+1.
+00261  */
+00262 
+00263 #define MAX_DIST(s)  ((s)->w_size-MIN_LOOKAHEAD)
+00264 /* In order to simplify the code, particularly on 16 bit machines, match
+00265  * distances are limited to MAX_DIST instead of WSIZE.
+00266  */
+00267 
+00268         /* in trees.c */
+00269 void _tr_init         OF((deflate_state *s));
+00270 int  _tr_tally        OF((deflate_state *s, unsigned dist, unsigned lc));
+00271 void _tr_flush_block  OF((deflate_state *s, charf *buf, ulg stored_len,
+00272                           int eof));
+00273 void _tr_align        OF((deflate_state *s));
+00274 void _tr_stored_block OF((deflate_state *s, charf *buf, ulg stored_len,
+00275                           int eof));
+00276 
+00277 #define d_code(dist) \
+00278    ((dist) < 256 ? _dist_code[dist] : _dist_code[256+((dist)>>7)])
+00279 /* Mapping from a distance to a distance code. dist is the distance - 1 and
+00280  * must not have side effects. _dist_code[256] and _dist_code[257] are never
+00281  * used.
+00282  */
+00283 
+00284 #ifndef DEBUG
+00285 /* Inline versions of _tr_tally for speed: */
+00286 
+00287 #if defined(GEN_TREES_H) || !defined(STDC)
+00288   extern uch _length_code[];
+00289   extern uch _dist_code[];
+00290 #else
+00291   extern const uch _length_code[];
+00292   extern const uch _dist_code[];
+00293 #endif
+00294 
+00295 # define _tr_tally_lit(s, c, flush) \
+00296   { uch cc = (c); \
+00297     s->d_buf[s->last_lit] = 0; \
+00298     s->l_buf[s->last_lit++] = cc; \
+00299     s->dyn_ltree[cc].Freq++; \
+00300     flush = (s->last_lit == s->lit_bufsize-1); \
+00301    }
+00302 # define _tr_tally_dist(s, distance, length, flush) \
+00303   { uch len = (length); \
+00304     ush dist = (distance); \
+00305     s->d_buf[s->last_lit] = dist; \
+00306     s->l_buf[s->last_lit++] = len; \
+00307     dist--; \
+00308     s->dyn_ltree[_length_code[len]+LITERALS+1].Freq++; \
+00309     s->dyn_dtree[d_code(dist)].Freq++; \
+00310     flush = (s->last_lit == s->lit_bufsize-1); \
+00311   }
+00312 #else
+00313 # define _tr_tally_lit(s, c, flush) flush = _tr_tally(s, 0, c)
+00314 # define _tr_tally_dist(s, distance, length, flush) \
+00315               flush = _tr_tally(s, distance, length) 
+00316 #endif
+00317 
+00318 #endif
+