deflate.h revision 17651
117651Speter/* deflate.h -- internal compression state 217651Speter * Copyright (C) 1995-1996 Jean-loup Gailly 317651Speter * For conditions of distribution and use, see copyright notice in zlib.h 417651Speter */ 517651Speter 617651Speter/* WARNING: this file should *not* be used by applications. It is 717651Speter part of the implementation of the compression library and is 817651Speter subject to change. Applications should only use zlib.h. 917651Speter */ 1017651Speter 1117651Speter/* $Id: deflate.h,v 1.10 1996/07/02 12:41:00 me Exp $ */ 1217651Speter 1317651Speter#ifndef _DEFLATE_H 1417651Speter#define _DEFLATE_H 1517651Speter 1617651Speter#include "zutil.h" 1717651Speter 1817651Speter/* =========================================================================== 1917651Speter * Internal compression state. 2017651Speter */ 2117651Speter 2217651Speter#define LENGTH_CODES 29 2317651Speter/* number of length codes, not counting the special END_BLOCK code */ 2417651Speter 2517651Speter#define LITERALS 256 2617651Speter/* number of literal bytes 0..255 */ 2717651Speter 2817651Speter#define L_CODES (LITERALS+1+LENGTH_CODES) 2917651Speter/* number of Literal or Length codes, including the END_BLOCK code */ 3017651Speter 3117651Speter#define D_CODES 30 3217651Speter/* number of distance codes */ 3317651Speter 3417651Speter#define BL_CODES 19 3517651Speter/* number of codes used to transfer the bit lengths */ 3617651Speter 3717651Speter#define HEAP_SIZE (2*L_CODES+1) 3817651Speter/* maximum heap size */ 3917651Speter 4017651Speter#define MAX_BITS 15 4117651Speter/* All codes must not exceed MAX_BITS bits */ 4217651Speter 4317651Speter#define INIT_STATE 42 4417651Speter#define BUSY_STATE 113 4517651Speter#define FINISH_STATE 666 4617651Speter/* Stream status */ 4717651Speter 4817651Speter 4917651Speter/* Data structure describing a single value and its code string. */ 5017651Spetertypedef struct ct_data_s { 5117651Speter union { 5217651Speter ush freq; /* frequency count */ 5317651Speter ush code; /* bit string */ 5417651Speter } fc; 5517651Speter union { 5617651Speter ush dad; /* father node in Huffman tree */ 5717651Speter ush len; /* length of bit string */ 5817651Speter } dl; 5917651Speter} FAR ct_data; 6017651Speter 6117651Speter#define Freq fc.freq 6217651Speter#define Code fc.code 6317651Speter#define Dad dl.dad 6417651Speter#define Len dl.len 6517651Speter 6617651Spetertypedef struct static_tree_desc_s static_tree_desc; 6717651Speter 6817651Spetertypedef struct tree_desc_s { 6917651Speter ct_data *dyn_tree; /* the dynamic tree */ 7017651Speter int max_code; /* largest code with non zero frequency */ 7117651Speter static_tree_desc *stat_desc; /* the corresponding static tree */ 7217651Speter} FAR tree_desc; 7317651Speter 7417651Spetertypedef ush Pos; 7517651Spetertypedef Pos FAR Posf; 7617651Spetertypedef unsigned IPos; 7717651Speter 7817651Speter/* A Pos is an index in the character window. We use short instead of int to 7917651Speter * save space in the various tables. IPos is used only for parameter passing. 8017651Speter */ 8117651Speter 8217651Spetertypedef struct internal_state { 8317651Speter z_streamp strm; /* pointer back to this zlib stream */ 8417651Speter int status; /* as the name implies */ 8517651Speter Bytef *pending_buf; /* output still pending */ 8617651Speter Bytef *pending_out; /* next pending byte to output to the stream */ 8717651Speter int pending; /* nb of bytes in the pending buffer */ 8817651Speter int noheader; /* suppress zlib header and adler32 */ 8917651Speter Byte data_type; /* UNKNOWN, BINARY or ASCII */ 9017651Speter Byte method; /* STORED (for zip only) or DEFLATED */ 9117651Speter int last_flush; /* value of flush param for previous deflate call */ 9217651Speter 9317651Speter /* used by deflate.c: */ 9417651Speter 9517651Speter uInt w_size; /* LZ77 window size (32K by default) */ 9617651Speter uInt w_bits; /* log2(w_size) (8..16) */ 9717651Speter uInt w_mask; /* w_size - 1 */ 9817651Speter 9917651Speter Bytef *window; 10017651Speter /* Sliding window. Input bytes are read into the second half of the window, 10117651Speter * and move to the first half later to keep a dictionary of at least wSize 10217651Speter * bytes. With this organization, matches are limited to a distance of 10317651Speter * wSize-MAX_MATCH bytes, but this ensures that IO is always 10417651Speter * performed with a length multiple of the block size. Also, it limits 10517651Speter * the window size to 64K, which is quite useful on MSDOS. 10617651Speter * To do: use the user input buffer as sliding window. 10717651Speter */ 10817651Speter 10917651Speter ulg window_size; 11017651Speter /* Actual size of window: 2*wSize, except when the user input buffer 11117651Speter * is directly used as sliding window. 11217651Speter */ 11317651Speter 11417651Speter Posf *prev; 11517651Speter /* Link to older string with same hash index. To limit the size of this 11617651Speter * array to 64K, this link is maintained only for the last 32K strings. 11717651Speter * An index in this array is thus a window index modulo 32K. 11817651Speter */ 11917651Speter 12017651Speter Posf *head; /* Heads of the hash chains or NIL. */ 12117651Speter 12217651Speter uInt ins_h; /* hash index of string to be inserted */ 12317651Speter uInt hash_size; /* number of elements in hash table */ 12417651Speter uInt hash_bits; /* log2(hash_size) */ 12517651Speter uInt hash_mask; /* hash_size-1 */ 12617651Speter 12717651Speter uInt hash_shift; 12817651Speter /* Number of bits by which ins_h must be shifted at each input 12917651Speter * step. It must be such that after MIN_MATCH steps, the oldest 13017651Speter * byte no longer takes part in the hash key, that is: 13117651Speter * hash_shift * MIN_MATCH >= hash_bits 13217651Speter */ 13317651Speter 13417651Speter long block_start; 13517651Speter /* Window position at the beginning of the current output block. Gets 13617651Speter * negative when the window is moved backwards. 13717651Speter */ 13817651Speter 13917651Speter uInt match_length; /* length of best match */ 14017651Speter IPos prev_match; /* previous match */ 14117651Speter int match_available; /* set if previous match exists */ 14217651Speter uInt strstart; /* start of string to insert */ 14317651Speter uInt match_start; /* start of matching string */ 14417651Speter uInt lookahead; /* number of valid bytes ahead in window */ 14517651Speter 14617651Speter uInt prev_length; 14717651Speter /* Length of the best match at previous step. Matches not greater than this 14817651Speter * are discarded. This is used in the lazy match evaluation. 14917651Speter */ 15017651Speter 15117651Speter uInt max_chain_length; 15217651Speter /* To speed up deflation, hash chains are never searched beyond this 15317651Speter * length. A higher limit improves compression ratio but degrades the 15417651Speter * speed. 15517651Speter */ 15617651Speter 15717651Speter uInt max_lazy_match; 15817651Speter /* Attempt to find a better match only when the current match is strictly 15917651Speter * smaller than this value. This mechanism is used only for compression 16017651Speter * levels >= 4. 16117651Speter */ 16217651Speter# define max_insert_length max_lazy_match 16317651Speter /* Insert new strings in the hash table only if the match length is not 16417651Speter * greater than this length. This saves time but degrades compression. 16517651Speter * max_insert_length is used only for compression levels <= 3. 16617651Speter */ 16717651Speter 16817651Speter int level; /* compression level (1..9) */ 16917651Speter int strategy; /* favor or force Huffman coding*/ 17017651Speter 17117651Speter uInt good_match; 17217651Speter /* Use a faster search when the previous match is longer than this */ 17317651Speter 17417651Speter int nice_match; /* Stop searching when current match exceeds this */ 17517651Speter 17617651Speter /* used by trees.c: */ 17717651Speter /* Didn't use ct_data typedef below to supress compiler warning */ 17817651Speter struct ct_data_s dyn_ltree[HEAP_SIZE]; /* literal and length tree */ 17917651Speter struct ct_data_s dyn_dtree[2*D_CODES+1]; /* distance tree */ 18017651Speter struct ct_data_s bl_tree[2*BL_CODES+1]; /* Huffman tree for bit lengths */ 18117651Speter 18217651Speter struct tree_desc_s l_desc; /* desc. for literal tree */ 18317651Speter struct tree_desc_s d_desc; /* desc. for distance tree */ 18417651Speter struct tree_desc_s bl_desc; /* desc. for bit length tree */ 18517651Speter 18617651Speter ush bl_count[MAX_BITS+1]; 18717651Speter /* number of codes at each bit length for an optimal tree */ 18817651Speter 18917651Speter int heap[2*L_CODES+1]; /* heap used to build the Huffman trees */ 19017651Speter int heap_len; /* number of elements in the heap */ 19117651Speter int heap_max; /* element of largest frequency */ 19217651Speter /* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used. 19317651Speter * The same heap array is used to build all trees. 19417651Speter */ 19517651Speter 19617651Speter uch depth[2*L_CODES+1]; 19717651Speter /* Depth of each subtree used as tie breaker for trees of equal frequency 19817651Speter */ 19917651Speter 20017651Speter uchf *l_buf; /* buffer for literals or lengths */ 20117651Speter 20217651Speter uInt lit_bufsize; 20317651Speter /* Size of match buffer for literals/lengths. There are 4 reasons for 20417651Speter * limiting lit_bufsize to 64K: 20517651Speter * - frequencies can be kept in 16 bit counters 20617651Speter * - if compression is not successful for the first block, all input 20717651Speter * data is still in the window so we can still emit a stored block even 20817651Speter * when input comes from standard input. (This can also be done for 20917651Speter * all blocks if lit_bufsize is not greater than 32K.) 21017651Speter * - if compression is not successful for a file smaller than 64K, we can 21117651Speter * even emit a stored file instead of a stored block (saving 5 bytes). 21217651Speter * This is applicable only for zip (not gzip or zlib). 21317651Speter * - creating new Huffman trees less frequently may not provide fast 21417651Speter * adaptation to changes in the input data statistics. (Take for 21517651Speter * example a binary file with poorly compressible code followed by 21617651Speter * a highly compressible string table.) Smaller buffer sizes give 21717651Speter * fast adaptation but have of course the overhead of transmitting 21817651Speter * trees more frequently. 21917651Speter * - I can't count above 4 22017651Speter */ 22117651Speter 22217651Speter uInt last_lit; /* running index in l_buf */ 22317651Speter 22417651Speter ushf *d_buf; 22517651Speter /* Buffer for distances. To simplify the code, d_buf and l_buf have 22617651Speter * the same number of elements. To use different lengths, an extra flag 22717651Speter * array would be necessary. 22817651Speter */ 22917651Speter 23017651Speter ulg opt_len; /* bit length of current block with optimal trees */ 23117651Speter ulg static_len; /* bit length of current block with static trees */ 23217651Speter ulg compressed_len; /* total bit length of compressed file */ 23317651Speter uInt matches; /* number of string matches in current block */ 23417651Speter int last_eob_len; /* bit length of EOB code for last block */ 23517651Speter 23617651Speter#ifdef DEBUG 23717651Speter ulg bits_sent; /* bit length of the compressed data */ 23817651Speter#endif 23917651Speter 24017651Speter ush bi_buf; 24117651Speter /* Output buffer. bits are inserted starting at the bottom (least 24217651Speter * significant bits). 24317651Speter */ 24417651Speter int bi_valid; 24517651Speter /* Number of valid bits in bi_buf. All bits above the last valid bit 24617651Speter * are always zero. 24717651Speter */ 24817651Speter 24917651Speter} FAR deflate_state; 25017651Speter 25117651Speter/* Output a byte on the stream. 25217651Speter * IN assertion: there is enough room in pending_buf. 25317651Speter */ 25417651Speter#define put_byte(s, c) {s->pending_buf[s->pending++] = (c);} 25517651Speter 25617651Speter 25717651Speter#define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1) 25817651Speter/* Minimum amount of lookahead, except at the end of the input file. 25917651Speter * See deflate.c for comments about the MIN_MATCH+1. 26017651Speter */ 26117651Speter 26217651Speter#define MAX_DIST(s) ((s)->w_size-MIN_LOOKAHEAD) 26317651Speter/* In order to simplify the code, particularly on 16 bit machines, match 26417651Speter * distances are limited to MAX_DIST instead of WSIZE. 26517651Speter */ 26617651Speter 26717651Speter /* in trees.c */ 26817651Spetervoid _tr_init OF((deflate_state *s)); 26917651Speterint _tr_tally OF((deflate_state *s, unsigned dist, unsigned lc)); 27017651Speterulg _tr_flush_block OF((deflate_state *s, charf *buf, ulg stored_len, 27117651Speter int eof)); 27217651Spetervoid _tr_align OF((deflate_state *s)); 27317651Spetervoid _tr_stored_block OF((deflate_state *s, charf *buf, ulg stored_len, 27417651Speter int eof)); 27517651Speter#endif 276