Subversion Repositories Kolibri OS

Rev

Go to most recent revision | Blame | Last modification | View Log | Download | RSS feed

  1. /* deflate.c -- compress data using the deflation algorithm
  2.  * Copyright (C) 1995-2010 Jean-loup Gailly and Mark Adler
  3.  * For conditions of distribution and use, see copyright notice in zlib.h
  4.  */
  5.  
  6. /*
  7.  *  ALGORITHM
  8.  *
  9.  *      The "deflation" process depends on being able to identify portions
  10.  *      of the input text which are identical to earlier input (within a
  11.  *      sliding window trailing behind the input currently being processed).
  12.  *
  13.  *      The most straightforward technique turns out to be the fastest for
  14.  *      most input files: try all possible matches and select the longest.
  15.  *      The key feature of this algorithm is that insertions into the string
  16.  *      dictionary are very simple and thus fast, and deletions are avoided
  17.  *      completely. Insertions are performed at each input character, whereas
  18.  *      string matches are performed only when the previous match ends. So it
  19.  *      is preferable to spend more time in matches to allow very fast string
  20.  *      insertions and avoid deletions. The matching algorithm for small
  21.  *      strings is inspired from that of Rabin & Karp. A brute force approach
  22.  *      is used to find longer strings when a small match has been found.
  23.  *      A similar algorithm is used in comic (by Jan-Mark Wams) and freeze
  24.  *      (by Leonid Broukhis).
  25.  *         A previous version of this file used a more sophisticated algorithm
  26.  *      (by Fiala and Greene) which is guaranteed to run in linear amortized
  27.  *      time, but has a larger average cost, uses more memory and is patented.
  28.  *      However the F&G algorithm may be faster for some highly redundant
  29.  *      files if the parameter max_chain_length (described below) is too large.
  30.  *
  31.  *  ACKNOWLEDGEMENTS
  32.  *
  33.  *      The idea of lazy evaluation of matches is due to Jan-Mark Wams, and
  34.  *      I found it in 'freeze' written by Leonid Broukhis.
  35.  *      Thanks to many people for bug reports and testing.
  36.  *
  37.  *  REFERENCES
  38.  *
  39.  *      Deutsch, L.P.,"DEFLATE Compressed Data Format Specification".
  40.  *      Available in http://www.ietf.org/rfc/rfc1951.txt
  41.  *
  42.  *      A description of the Rabin and Karp algorithm is given in the book
  43.  *         "Algorithms" by R. Sedgewick, Addison-Wesley, p252.
  44.  *
  45.  *      Fiala,E.R., and Greene,D.H.
  46.  *         Data Compression with Finite Windows, Comm.ACM, 32,4 (1989) 490-595
  47.  *
  48.  */
  49.  
  50. /* @(#) $Id$ */
  51.  
  52. #include "deflate.h"
  53.  
  54. const char deflate_copyright[] =
  55.    " deflate 1.2.5 Copyright 1995-2010 Jean-loup Gailly and Mark Adler ";
  56. /*
  57.   If you use the zlib library in a product, an acknowledgment is welcome
  58.   in the documentation of your product. If for some reason you cannot
  59.   include such an acknowledgment, I would appreciate that you keep this
  60.   copyright string in the executable of your product.
  61.  */
  62.  
  63. /* ===========================================================================
  64.  *  Function prototypes.
  65.  */
  66. typedef enum {
  67.     need_more,      /* block not completed, need more input or more output */
  68.     block_done,     /* block flush performed */
  69.     finish_started, /* finish started, need only more output at next deflate */
  70.     finish_done     /* finish done, accept no more input or output */
  71. } block_state;
  72.  
  73. typedef block_state (*compress_func) OF((deflate_state *s, int flush));
  74. /* Compression function. Returns the block state after the call. */
  75.  
  76. local void fill_window    OF((deflate_state *s));
  77. local block_state deflate_stored OF((deflate_state *s, int flush));
  78. local block_state deflate_fast   OF((deflate_state *s, int flush));
  79. #ifndef FASTEST
  80. local block_state deflate_slow   OF((deflate_state *s, int flush));
  81. #endif
  82. local block_state deflate_rle    OF((deflate_state *s, int flush));
  83. local block_state deflate_huff   OF((deflate_state *s, int flush));
  84. local void lm_init        OF((deflate_state *s));
  85. local void putShortMSB    OF((deflate_state *s, uInt b));
  86. local void flush_pending  OF((z_streamp strm));
  87. local int read_buf        OF((z_streamp strm, Bytef *buf, unsigned size));
  88. #ifdef ASMV
  89.       void match_init OF((void)); /* asm code initialization */
  90.       uInt longest_match  OF((deflate_state *s, IPos cur_match));
  91. #else
  92. local uInt longest_match  OF((deflate_state *s, IPos cur_match));
  93. #endif
  94.  
  95. #ifdef DEBUG
  96. local  void check_match OF((deflate_state *s, IPos start, IPos match,
  97.                             int length));
  98. #endif
  99.  
  100. /* ===========================================================================
  101.  * Local data
  102.  */
  103.  
  104. #define NIL 0
  105. /* Tail of hash chains */
  106.  
  107. #ifndef TOO_FAR
  108. #  define TOO_FAR 4096
  109. #endif
  110. /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */
  111.  
  112. /* Values for max_lazy_match, good_match and max_chain_length, depending on
  113.  * the desired pack level (0..9). The values given below have been tuned to
  114.  * exclude worst case performance for pathological files. Better values may be
  115.  * found for specific files.
  116.  */
  117. typedef struct config_s {
  118.    ush good_length; /* reduce lazy search above this match length */
  119.    ush max_lazy;    /* do not perform lazy search above this match length */
  120.    ush nice_length; /* quit search above this match length */
  121.    ush max_chain;
  122.    compress_func func;
  123. } config;
  124.  
  125. #ifdef FASTEST
  126. local const config configuration_table[2] = {
  127. /*      good lazy nice chain */
  128. /* 0 */ {0,    0,  0,    0, deflate_stored},  /* store only */
  129. /* 1 */ {4,    4,  8,    4, deflate_fast}}; /* max speed, no lazy matches */
  130. #else
  131. local const config configuration_table[10] = {
  132. /*      good lazy nice chain */
  133. /* 0 */ {0,    0,  0,    0, deflate_stored},  /* store only */
  134. /* 1 */ {4,    4,  8,    4, deflate_fast}, /* max speed, no lazy matches */
  135. /* 2 */ {4,    5, 16,    8, deflate_fast},
  136. /* 3 */ {4,    6, 32,   32, deflate_fast},
  137.  
  138. /* 4 */ {4,    4, 16,   16, deflate_slow},  /* lazy matches */
  139. /* 5 */ {8,   16, 32,   32, deflate_slow},
  140. /* 6 */ {8,   16, 128, 128, deflate_slow},
  141. /* 7 */ {8,   32, 128, 256, deflate_slow},
  142. /* 8 */ {32, 128, 258, 1024, deflate_slow},
  143. /* 9 */ {32, 258, 258, 4096, deflate_slow}}; /* max compression */
  144. #endif
  145.  
  146. /* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4
  147.  * For deflate_fast() (levels <= 3) good is ignored and lazy has a different
  148.  * meaning.
  149.  */
  150.  
  151. #define EQUAL 0
  152. /* result of memcmp for equal strings */
  153.  
  154. #ifndef NO_DUMMY_DECL
  155. struct static_tree_desc_s {int dummy;}; /* for buggy compilers */
  156. #endif
  157.  
  158. /* ===========================================================================
  159.  * Update a hash value with the given input byte
  160.  * IN  assertion: all calls to to UPDATE_HASH are made with consecutive
  161.  *    input characters, so that a running hash key can be computed from the
  162.  *    previous key instead of complete recalculation each time.
  163.  */
  164. #define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask)
  165.  
  166.  
  167. /* ===========================================================================
  168.  * Insert string str in the dictionary and set match_head to the previous head
  169.  * of the hash chain (the most recent string with same hash key). Return
  170.  * the previous length of the hash chain.
  171.  * If this file is compiled with -DFASTEST, the compression level is forced
  172.  * to 1, and no hash chains are maintained.
  173.  * IN  assertion: all calls to to INSERT_STRING are made with consecutive
  174.  *    input characters and the first MIN_MATCH bytes of str are valid
  175.  *    (except for the last MIN_MATCH-1 bytes of the input file).
  176.  */
  177. #ifdef FASTEST
  178. #define INSERT_STRING(s, str, match_head) \
  179.    (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \
  180.     match_head = s->head[s->ins_h], \
  181.     s->head[s->ins_h] = (Pos)(str))
  182. #else
  183. #define INSERT_STRING(s, str, match_head) \
  184.    (UPDATE_HASH(s, s->ins_h, s->window[(str) + (MIN_MATCH-1)]), \
  185.     match_head = s->prev[(str) & s->w_mask] = s->head[s->ins_h], \
  186.     s->head[s->ins_h] = (Pos)(str))
  187. #endif
  188.  
  189. /* ===========================================================================
  190.  * Initialize the hash table (avoiding 64K overflow for 16 bit systems).
  191.  * prev[] will be initialized on the fly.
  192.  */
  193. #define CLEAR_HASH(s) \
  194.     s->head[s->hash_size-1] = NIL; \
  195.     zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head));
  196.  
  197. /* ========================================================================= */
  198. int ZEXPORT deflateInit_(strm, level, version, stream_size)
  199.     z_streamp strm;
  200.     int level;
  201.     const char *version;
  202.     int stream_size;
  203. {
  204.     return deflateInit2_(strm, level, Z_DEFLATED, MAX_WBITS, DEF_MEM_LEVEL,
  205.                          Z_DEFAULT_STRATEGY, version, stream_size);
  206.     /* To do: ignore strm->next_in if we use it as window */
  207. }
  208.  
  209. /* ========================================================================= */
  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. {
  221.     deflate_state *s;
  222.     int wrap = 1;
  223.     static const char my_version[] = ZLIB_VERSION;
  224.  
  225.     ushf *overlay;
  226.     /* We overlay pending_buf and d_buf+l_buf. This works since the average
  227.      * output size for (length,distance) codes is <= 24 bits.
  228.      */
  229.  
  230.     if (version == Z_NULL || version[0] != my_version[0] ||
  231.         stream_size != sizeof(z_stream)) {
  232.         return Z_VERSION_ERROR;
  233.     }
  234.     if (strm == Z_NULL) return Z_STREAM_ERROR;
  235.  
  236.     strm->msg = Z_NULL;
  237.     if (strm->zalloc == (alloc_func)0) {
  238.         strm->zalloc = zcalloc;
  239.         strm->opaque = (voidpf)0;
  240.     }
  241.     if (strm->zfree == (free_func)0) strm->zfree = zcfree;
  242.  
  243. #ifdef FASTEST
  244.     if (level != 0) level = 1;
  245. #else
  246.     if (level == Z_DEFAULT_COMPRESSION) level = 6;
  247. #endif
  248.  
  249.     if (windowBits < 0) { /* suppress zlib wrapper */
  250.         wrap = 0;
  251.         windowBits = -windowBits;
  252.     }
  253. #ifdef GZIP
  254.     else if (windowBits > 15) {
  255.         wrap = 2;       /* write gzip wrapper instead */
  256.         windowBits -= 16;
  257.     }
  258. #endif
  259.     if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED ||
  260.         windowBits < 8 || windowBits > 15 || level < 0 || level > 9 ||
  261.         strategy < 0 || strategy > Z_FIXED) {
  262.         return Z_STREAM_ERROR;
  263.     }
  264.     if (windowBits == 8) windowBits = 9;  /* until 256-byte window bug fixed */
  265.     s = (deflate_state *) ZALLOC(strm, 1, sizeof(deflate_state));
  266.     if (s == Z_NULL) return Z_MEM_ERROR;
  267.     strm->state = (struct internal_state FAR *)s;
  268.     s->strm = strm;
  269.  
  270.     s->wrap = wrap;
  271.     s->gzhead = Z_NULL;
  272.     s->w_bits = windowBits;
  273.     s->w_size = 1 << s->w_bits;
  274.     s->w_mask = s->w_size - 1;
  275.  
  276.     s->hash_bits = memLevel + 7;
  277.     s->hash_size = 1 << s->hash_bits;
  278.     s->hash_mask = s->hash_size - 1;
  279.     s->hash_shift =  ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH);
  280.  
  281.     s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte));
  282.     s->prev   = (Posf *)  ZALLOC(strm, s->w_size, sizeof(Pos));
  283.     s->head   = (Posf *)  ZALLOC(strm, s->hash_size, sizeof(Pos));
  284.  
  285.     s->high_water = 0;      /* nothing written to s->window yet */
  286.  
  287.     s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */
  288.  
  289.     overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2);
  290.     s->pending_buf = (uchf *) overlay;
  291.     s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L);
  292.  
  293.     if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL ||
  294.         s->pending_buf == Z_NULL) {
  295.         s->status = FINISH_STATE;
  296.         strm->msg = (char*)ERR_MSG(Z_MEM_ERROR);
  297.         deflateEnd (strm);
  298.         return Z_MEM_ERROR;
  299.     }
  300.     s->d_buf = overlay + s->lit_bufsize/sizeof(ush);
  301.     s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize;
  302.  
  303.     s->level = level;
  304.     s->strategy = strategy;
  305.     s->method = (Byte)method;
  306.  
  307.     return deflateReset(strm);
  308. }
  309.  
  310. /* ========================================================================= */
  311. int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength)
  312.     z_streamp strm;
  313.     const Bytef *dictionary;
  314.     uInt  dictLength;
  315. {
  316.     deflate_state *s;
  317.     uInt length = dictLength;
  318.     uInt n;
  319.     IPos hash_head = 0;
  320.  
  321.     if (strm == Z_NULL || strm->state == Z_NULL || dictionary == Z_NULL ||
  322.         strm->state->wrap == 2 ||
  323.         (strm->state->wrap == 1 && strm->state->status != INIT_STATE))
  324.         return Z_STREAM_ERROR;
  325.  
  326.     s = strm->state;
  327.     if (s->wrap)
  328.         strm->adler = adler32(strm->adler, dictionary, dictLength);
  329.  
  330.     if (length < MIN_MATCH) return Z_OK;
  331.     if (length > s->w_size) {
  332.         length = s->w_size;
  333.         dictionary += dictLength - length; /* use the tail of the dictionary */
  334.     }
  335.     zmemcpy(s->window, dictionary, length);
  336.     s->strstart = length;
  337.     s->block_start = (long)length;
  338.  
  339.     /* Insert all strings in the hash table (except for the last two bytes).
  340.      * s->lookahead stays null, so s->ins_h will be recomputed at the next
  341.      * call of fill_window.
  342.      */
  343.     s->ins_h = s->window[0];
  344.     UPDATE_HASH(s, s->ins_h, s->window[1]);
  345.     for (n = 0; n <= length - MIN_MATCH; n++) {
  346.         INSERT_STRING(s, n, hash_head);
  347.     }
  348.     if (hash_head) hash_head = 0;  /* to make compiler happy */
  349.     return Z_OK;
  350. }
  351.  
  352. /* ========================================================================= */
  353. int ZEXPORT deflateReset (strm)
  354.     z_streamp strm;
  355. {
  356.     deflate_state *s;
  357.  
  358.     if (strm == Z_NULL || strm->state == Z_NULL ||
  359.         strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0) {
  360.         return Z_STREAM_ERROR;
  361.     }
  362.  
  363.     strm->total_in = strm->total_out = 0;
  364.     strm->msg = Z_NULL; /* use zfree if we ever allocate msg dynamically */
  365.     strm->data_type = Z_UNKNOWN;
  366.  
  367.     s = (deflate_state *)strm->state;
  368.     s->pending = 0;
  369.     s->pending_out = s->pending_buf;
  370.  
  371.     if (s->wrap < 0) {
  372.         s->wrap = -s->wrap; /* was made negative by deflate(..., Z_FINISH); */
  373.     }
  374.     s->status = s->wrap ? INIT_STATE : BUSY_STATE;
  375.     strm->adler =
  376. #ifdef GZIP
  377.         s->wrap == 2 ? crc32(0L, Z_NULL, 0) :
  378. #endif
  379.         adler32(0L, Z_NULL, 0);
  380.     s->last_flush = Z_NO_FLUSH;
  381.  
  382.     _tr_init(s);
  383.     lm_init(s);
  384.  
  385.     return Z_OK;
  386. }
  387.  
  388. /* ========================================================================= */
  389. int ZEXPORT deflateSetHeader (strm, head)
  390.     z_streamp strm;
  391.     gz_headerp head;
  392. {
  393.     if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
  394.     if (strm->state->wrap != 2) return Z_STREAM_ERROR;
  395.     strm->state->gzhead = head;
  396.     return Z_OK;
  397. }
  398.  
  399. /* ========================================================================= */
  400. int ZEXPORT deflatePrime (strm, bits, value)
  401.     z_streamp strm;
  402.     int bits;
  403.     int value;
  404. {
  405.     if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
  406.     strm->state->bi_valid = bits;
  407.     strm->state->bi_buf = (ush)(value & ((1 << bits) - 1));
  408.     return Z_OK;
  409. }
  410.  
  411. /* ========================================================================= */
  412. int ZEXPORT deflateParams(strm, level, strategy)
  413.     z_streamp strm;
  414.     int level;
  415.     int strategy;
  416. {
  417.     deflate_state *s;
  418.     compress_func func;
  419.     int err = Z_OK;
  420.  
  421.     if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
  422.     s = strm->state;
  423.  
  424. #ifdef FASTEST
  425.     if (level != 0) level = 1;
  426. #else
  427.     if (level == Z_DEFAULT_COMPRESSION) level = 6;
  428. #endif
  429.     if (level < 0 || level > 9 || strategy < 0 || strategy > Z_FIXED) {
  430.         return Z_STREAM_ERROR;
  431.     }
  432.     func = configuration_table[s->level].func;
  433.  
  434.     if ((strategy != s->strategy || func != configuration_table[level].func) &&
  435.         strm->total_in != 0) {
  436.         /* Flush the last buffer: */
  437.         err = deflate(strm, Z_BLOCK);
  438.     }
  439.     if (s->level != level) {
  440.         s->level = level;
  441.         s->max_lazy_match   = configuration_table[level].max_lazy;
  442.         s->good_match       = configuration_table[level].good_length;
  443.         s->nice_match       = configuration_table[level].nice_length;
  444.         s->max_chain_length = configuration_table[level].max_chain;
  445.     }
  446.     s->strategy = strategy;
  447.     return err;
  448. }
  449.  
  450. /* ========================================================================= */
  451. int ZEXPORT deflateTune(strm, good_length, max_lazy, nice_length, max_chain)
  452.     z_streamp strm;
  453.     int good_length;
  454.     int max_lazy;
  455.     int nice_length;
  456.     int max_chain;
  457. {
  458.     deflate_state *s;
  459.  
  460.     if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
  461.     s = strm->state;
  462.     s->good_match = good_length;
  463.     s->max_lazy_match = max_lazy;
  464.     s->nice_match = nice_length;
  465.     s->max_chain_length = max_chain;
  466.     return Z_OK;
  467. }
  468.  
  469. /* =========================================================================
  470.  * For the default windowBits of 15 and memLevel of 8, this function returns
  471.  * a close to exact, as well as small, upper bound on the compressed size.
  472.  * They are coded as constants here for a reason--if the #define's are
  473.  * changed, then this function needs to be changed as well.  The return
  474.  * value for 15 and 8 only works for those exact settings.
  475.  *
  476.  * For any setting other than those defaults for windowBits and memLevel,
  477.  * the value returned is a conservative worst case for the maximum expansion
  478.  * resulting from using fixed blocks instead of stored blocks, which deflate
  479.  * can emit on compressed data for some combinations of the parameters.
  480.  *
  481.  * This function could be more sophisticated to provide closer upper bounds for
  482.  * every combination of windowBits and memLevel.  But even the conservative
  483.  * upper bound of about 14% expansion does not seem onerous for output buffer
  484.  * allocation.
  485.  */
  486. uLong ZEXPORT deflateBound(strm, sourceLen)
  487.     z_streamp strm;
  488.     uLong sourceLen;
  489. {
  490.     deflate_state *s;
  491.     uLong complen, wraplen;
  492.     Bytef *str;
  493.  
  494.     /* conservative upper bound for compressed data */
  495.     complen = sourceLen +
  496.               ((sourceLen + 7) >> 3) + ((sourceLen + 63) >> 6) + 5;
  497.  
  498.     /* if can't get parameters, return conservative bound plus zlib wrapper */
  499.     if (strm == Z_NULL || strm->state == Z_NULL)
  500.         return complen + 6;
  501.  
  502.     /* compute wrapper length */
  503.     s = strm->state;
  504.     switch (s->wrap) {
  505.     case 0:                                 /* raw deflate */
  506.         wraplen = 0;
  507.         break;
  508.     case 1:                                 /* zlib wrapper */
  509.         wraplen = 6 + (s->strstart ? 4 : 0);
  510.         break;
  511.     case 2:                                 /* gzip wrapper */
  512.         wraplen = 18;
  513.         if (s->gzhead != Z_NULL) {          /* user-supplied gzip header */
  514.             if (s->gzhead->extra != Z_NULL)
  515.                 wraplen += 2 + s->gzhead->extra_len;
  516.             str = s->gzhead->name;
  517.             if (str != Z_NULL)
  518.                 do {
  519.                     wraplen++;
  520.                 } while (*str++);
  521.             str = s->gzhead->comment;
  522.             if (str != Z_NULL)
  523.                 do {
  524.                     wraplen++;
  525.                 } while (*str++);
  526.             if (s->gzhead->hcrc)
  527.                 wraplen += 2;
  528.         }
  529.         break;
  530.     default:                                /* for compiler happiness */
  531.         wraplen = 6;
  532.     }
  533.  
  534.     /* if not default parameters, return conservative bound */
  535.     if (s->w_bits != 15 || s->hash_bits != 8 + 7)
  536.         return complen + wraplen;
  537.  
  538.     /* default settings: return tight bound for that case */
  539.     return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
  540.            (sourceLen >> 25) + 13 - 6 + wraplen;
  541. }
  542.  
  543. /* =========================================================================
  544.  * Put a short in the pending buffer. The 16-bit value is put in MSB order.
  545.  * IN assertion: the stream state is correct and there is enough room in
  546.  * pending_buf.
  547.  */
  548. local void putShortMSB (s, b)
  549.     deflate_state *s;
  550.     uInt b;
  551. {
  552.     put_byte(s, (Byte)(b >> 8));
  553.     put_byte(s, (Byte)(b & 0xff));
  554. }
  555.  
  556. /* =========================================================================
  557.  * Flush as much pending output as possible. All deflate() output goes
  558.  * through this function so some applications may wish to modify it
  559.  * to avoid allocating a large strm->next_out buffer and copying into it.
  560.  * (See also read_buf()).
  561.  */
  562. local void flush_pending(strm)
  563.     z_streamp strm;
  564. {
  565.     unsigned len = strm->state->pending;
  566.  
  567.     if (len > strm->avail_out) len = strm->avail_out;
  568.     if (len == 0) return;
  569.  
  570.     zmemcpy(strm->next_out, strm->state->pending_out, len);
  571.     strm->next_out  += len;
  572.     strm->state->pending_out  += len;
  573.     strm->total_out += len;
  574.     strm->avail_out  -= len;
  575.     strm->state->pending -= len;
  576.     if (strm->state->pending == 0) {
  577.         strm->state->pending_out = strm->state->pending_buf;
  578.     }
  579. }
  580.  
  581. /* ========================================================================= */
  582. int ZEXPORT deflate (strm, flush)
  583.     z_streamp strm;
  584.     int flush;
  585. {
  586.     int old_flush; /* value of flush param for previous deflate call */
  587.     deflate_state *s;
  588.  
  589.     if (strm == Z_NULL || strm->state == Z_NULL ||
  590.         flush > Z_BLOCK || flush < 0) {
  591.         return Z_STREAM_ERROR;
  592.     }
  593.     s = strm->state;
  594.  
  595.     if (strm->next_out == Z_NULL ||
  596.         (strm->next_in == Z_NULL && strm->avail_in != 0) ||
  597.         (s->status == FINISH_STATE && flush != Z_FINISH)) {
  598.         ERR_RETURN(strm, Z_STREAM_ERROR);
  599.     }
  600.     if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR);
  601.  
  602.     s->strm = strm; /* just in case */
  603.     old_flush = s->last_flush;
  604.     s->last_flush = flush;
  605.  
  606.     /* Write the header */
  607.     if (s->status == INIT_STATE) {
  608. #ifdef GZIP
  609.         if (s->wrap == 2) {
  610.             strm->adler = crc32(0L, Z_NULL, 0);
  611.             put_byte(s, 31);
  612.             put_byte(s, 139);
  613.             put_byte(s, 8);
  614.             if (s->gzhead == Z_NULL) {
  615.                 put_byte(s, 0);
  616.                 put_byte(s, 0);
  617.                 put_byte(s, 0);
  618.                 put_byte(s, 0);
  619.                 put_byte(s, 0);
  620.                 put_byte(s, s->level == 9 ? 2 :
  621.                             (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
  622.                              4 : 0));
  623.                 put_byte(s, OS_CODE);
  624.                 s->status = BUSY_STATE;
  625.             }
  626.             else {
  627.                 put_byte(s, (s->gzhead->text ? 1 : 0) +
  628.                             (s->gzhead->hcrc ? 2 : 0) +
  629.                             (s->gzhead->extra == Z_NULL ? 0 : 4) +
  630.                             (s->gzhead->name == Z_NULL ? 0 : 8) +
  631.                             (s->gzhead->comment == Z_NULL ? 0 : 16)
  632.                         );
  633.                 put_byte(s, (Byte)(s->gzhead->time & 0xff));
  634.                 put_byte(s, (Byte)((s->gzhead->time >> 8) & 0xff));
  635.                 put_byte(s, (Byte)((s->gzhead->time >> 16) & 0xff));
  636.                 put_byte(s, (Byte)((s->gzhead->time >> 24) & 0xff));
  637.                 put_byte(s, s->level == 9 ? 2 :
  638.                             (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
  639.                              4 : 0));
  640.                 put_byte(s, s->gzhead->os & 0xff);
  641.                 if (s->gzhead->extra != Z_NULL) {
  642.                     put_byte(s, s->gzhead->extra_len & 0xff);
  643.                     put_byte(s, (s->gzhead->extra_len >> 8) & 0xff);
  644.                 }
  645.                 if (s->gzhead->hcrc)
  646.                     strm->adler = crc32(strm->adler, s->pending_buf,
  647.                                         s->pending);
  648.                 s->gzindex = 0;
  649.                 s->status = EXTRA_STATE;
  650.             }
  651.         }
  652.         else
  653. #endif
  654.         {
  655.             uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8;
  656.             uInt level_flags;
  657.  
  658.             if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2)
  659.                 level_flags = 0;
  660.             else if (s->level < 6)
  661.                 level_flags = 1;
  662.             else if (s->level == 6)
  663.                 level_flags = 2;
  664.             else
  665.                 level_flags = 3;
  666.             header |= (level_flags << 6);
  667.             if (s->strstart != 0) header |= PRESET_DICT;
  668.             header += 31 - (header % 31);
  669.  
  670.             s->status = BUSY_STATE;
  671.             putShortMSB(s, header);
  672.  
  673.             /* Save the adler32 of the preset dictionary: */
  674.             if (s->strstart != 0) {
  675.                 putShortMSB(s, (uInt)(strm->adler >> 16));
  676.                 putShortMSB(s, (uInt)(strm->adler & 0xffff));
  677.             }
  678.             strm->adler = adler32(0L, Z_NULL, 0);
  679.         }
  680.     }
  681. #ifdef GZIP
  682.     if (s->status == EXTRA_STATE) {
  683.         if (s->gzhead->extra != Z_NULL) {
  684.             uInt beg = s->pending;  /* start of bytes to update crc */
  685.  
  686.             while (s->gzindex < (s->gzhead->extra_len & 0xffff)) {
  687.                 if (s->pending == s->pending_buf_size) {
  688.                     if (s->gzhead->hcrc && s->pending > beg)
  689.                         strm->adler = crc32(strm->adler, s->pending_buf + beg,
  690.                                             s->pending - beg);
  691.                     flush_pending(strm);
  692.                     beg = s->pending;
  693.                     if (s->pending == s->pending_buf_size)
  694.                         break;
  695.                 }
  696.                 put_byte(s, s->gzhead->extra[s->gzindex]);
  697.                 s->gzindex++;
  698.             }
  699.             if (s->gzhead->hcrc && s->pending > beg)
  700.                 strm->adler = crc32(strm->adler, s->pending_buf + beg,
  701.                                     s->pending - beg);
  702.             if (s->gzindex == s->gzhead->extra_len) {
  703.                 s->gzindex = 0;
  704.                 s->status = NAME_STATE;
  705.             }
  706.         }
  707.         else
  708.             s->status = NAME_STATE;
  709.     }
  710.     if (s->status == NAME_STATE) {
  711.         if (s->gzhead->name != Z_NULL) {
  712.             uInt beg = s->pending;  /* start of bytes to update crc */
  713.             int val;
  714.  
  715.             do {
  716.                 if (s->pending == s->pending_buf_size) {
  717.                     if (s->gzhead->hcrc && s->pending > beg)
  718.                         strm->adler = crc32(strm->adler, s->pending_buf + beg,
  719.                                             s->pending - beg);
  720.                     flush_pending(strm);
  721.                     beg = s->pending;
  722.                     if (s->pending == s->pending_buf_size) {
  723.                         val = 1;
  724.                         break;
  725.                     }
  726.                 }
  727.                 val = s->gzhead->name[s->gzindex++];
  728.                 put_byte(s, val);
  729.             } while (val != 0);
  730.             if (s->gzhead->hcrc && s->pending > beg)
  731.                 strm->adler = crc32(strm->adler, s->pending_buf + beg,
  732.                                     s->pending - beg);
  733.             if (val == 0) {
  734.                 s->gzindex = 0;
  735.                 s->status = COMMENT_STATE;
  736.             }
  737.         }
  738.         else
  739.             s->status = COMMENT_STATE;
  740.     }
  741.     if (s->status == COMMENT_STATE) {
  742.         if (s->gzhead->comment != Z_NULL) {
  743.             uInt beg = s->pending;  /* start of bytes to update crc */
  744.             int val;
  745.  
  746.             do {
  747.                 if (s->pending == s->pending_buf_size) {
  748.                     if (s->gzhead->hcrc && s->pending > beg)
  749.                         strm->adler = crc32(strm->adler, s->pending_buf + beg,
  750.                                             s->pending - beg);
  751.                     flush_pending(strm);
  752.                     beg = s->pending;
  753.                     if (s->pending == s->pending_buf_size) {
  754.                         val = 1;
  755.                         break;
  756.                     }
  757.                 }
  758.                 val = s->gzhead->comment[s->gzindex++];
  759.                 put_byte(s, val);
  760.             } while (val != 0);
  761.             if (s->gzhead->hcrc && s->pending > beg)
  762.                 strm->adler = crc32(strm->adler, s->pending_buf + beg,
  763.                                     s->pending - beg);
  764.             if (val == 0)
  765.                 s->status = HCRC_STATE;
  766.         }
  767.         else
  768.             s->status = HCRC_STATE;
  769.     }
  770.     if (s->status == HCRC_STATE) {
  771.         if (s->gzhead->hcrc) {
  772.             if (s->pending + 2 > s->pending_buf_size)
  773.                 flush_pending(strm);
  774.             if (s->pending + 2 <= s->pending_buf_size) {
  775.                 put_byte(s, (Byte)(strm->adler & 0xff));
  776.                 put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
  777.                 strm->adler = crc32(0L, Z_NULL, 0);
  778.                 s->status = BUSY_STATE;
  779.             }
  780.         }
  781.         else
  782.             s->status = BUSY_STATE;
  783.     }
  784. #endif
  785.  
  786.     /* Flush as much pending output as possible */
  787.     if (s->pending != 0) {
  788.         flush_pending(strm);
  789.         if (strm->avail_out == 0) {
  790.             /* Since avail_out is 0, deflate will be called again with
  791.              * more output space, but possibly with both pending and
  792.              * avail_in equal to zero. There won't be anything to do,
  793.              * but this is not an error situation so make sure we
  794.              * return OK instead of BUF_ERROR at next call of deflate:
  795.              */
  796.             s->last_flush = -1;
  797.             return Z_OK;
  798.         }
  799.  
  800.     /* Make sure there is something to do and avoid duplicate consecutive
  801.      * flushes. For repeated and useless calls with Z_FINISH, we keep
  802.      * returning Z_STREAM_END instead of Z_BUF_ERROR.
  803.      */
  804.     } else if (strm->avail_in == 0 && flush <= old_flush &&
  805.                flush != Z_FINISH) {
  806.         ERR_RETURN(strm, Z_BUF_ERROR);
  807.     }
  808.  
  809.     /* User must not provide more input after the first FINISH: */
  810.     if (s->status == FINISH_STATE && strm->avail_in != 0) {
  811.         ERR_RETURN(strm, Z_BUF_ERROR);
  812.     }
  813.  
  814.     /* Start a new block or continue the current one.
  815.      */
  816.     if (strm->avail_in != 0 || s->lookahead != 0 ||
  817.         (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) {
  818.         block_state bstate;
  819.  
  820.         bstate = s->strategy == Z_HUFFMAN_ONLY ? deflate_huff(s, flush) :
  821.                     (s->strategy == Z_RLE ? deflate_rle(s, flush) :
  822.                         (*(configuration_table[s->level].func))(s, flush));
  823.  
  824.         if (bstate == finish_started || bstate == finish_done) {
  825.             s->status = FINISH_STATE;
  826.         }
  827.         if (bstate == need_more || bstate == finish_started) {
  828.             if (strm->avail_out == 0) {
  829.                 s->last_flush = -1; /* avoid BUF_ERROR next call, see above */
  830.             }
  831.             return Z_OK;
  832.             /* If flush != Z_NO_FLUSH && avail_out == 0, the next call
  833.              * of deflate should use the same flush parameter to make sure
  834.              * that the flush is complete. So we don't have to output an
  835.              * empty block here, this will be done at next call. This also
  836.              * ensures that for a very small output buffer, we emit at most
  837.              * one empty block.
  838.              */
  839.         }
  840.         if (bstate == block_done) {
  841.             if (flush == Z_PARTIAL_FLUSH) {
  842.                 _tr_align(s);
  843.             } else if (flush != Z_BLOCK) { /* FULL_FLUSH or SYNC_FLUSH */
  844.                 _tr_stored_block(s, (char*)0, 0L, 0);
  845.                 /* For a full flush, this empty block will be recognized
  846.                  * as a special marker by inflate_sync().
  847.                  */
  848.                 if (flush == Z_FULL_FLUSH) {
  849.                     CLEAR_HASH(s);             /* forget history */
  850.                     if (s->lookahead == 0) {
  851.                         s->strstart = 0;
  852.                         s->block_start = 0L;
  853.                     }
  854.                 }
  855.             }
  856.             flush_pending(strm);
  857.             if (strm->avail_out == 0) {
  858.               s->last_flush = -1; /* avoid BUF_ERROR at next call, see above */
  859.               return Z_OK;
  860.             }
  861.         }
  862.     }
  863.     Assert(strm->avail_out > 0, "bug2");
  864.  
  865.     if (flush != Z_FINISH) return Z_OK;
  866.     if (s->wrap <= 0) return Z_STREAM_END;
  867.  
  868.     /* Write the trailer */
  869. #ifdef GZIP
  870.     if (s->wrap == 2) {
  871.         put_byte(s, (Byte)(strm->adler & 0xff));
  872.         put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
  873.         put_byte(s, (Byte)((strm->adler >> 16) & 0xff));
  874.         put_byte(s, (Byte)((strm->adler >> 24) & 0xff));
  875.         put_byte(s, (Byte)(strm->total_in & 0xff));
  876.         put_byte(s, (Byte)((strm->total_in >> 8) & 0xff));
  877.         put_byte(s, (Byte)((strm->total_in >> 16) & 0xff));
  878.         put_byte(s, (Byte)((strm->total_in >> 24) & 0xff));
  879.     }
  880.     else
  881. #endif
  882.     {
  883.         putShortMSB(s, (uInt)(strm->adler >> 16));
  884.         putShortMSB(s, (uInt)(strm->adler & 0xffff));
  885.     }
  886.     flush_pending(strm);
  887.     /* If avail_out is zero, the application will call deflate again
  888.      * to flush the rest.
  889.      */
  890.     if (s->wrap > 0) s->wrap = -s->wrap; /* write the trailer only once! */
  891.     return s->pending != 0 ? Z_OK : Z_STREAM_END;
  892. }
  893.  
  894. /* ========================================================================= */
  895. int ZEXPORT deflateEnd (strm)
  896.     z_streamp strm;
  897. {
  898.     int status;
  899.  
  900.     if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
  901.  
  902.     status = strm->state->status;
  903.     if (status != INIT_STATE &&
  904.         status != EXTRA_STATE &&
  905.         status != NAME_STATE &&
  906.         status != COMMENT_STATE &&
  907.         status != HCRC_STATE &&
  908.         status != BUSY_STATE &&
  909.         status != FINISH_STATE) {
  910.       return Z_STREAM_ERROR;
  911.     }
  912.  
  913.     /* Deallocate in reverse order of allocations: */
  914.     TRY_FREE(strm, strm->state->pending_buf);
  915.     TRY_FREE(strm, strm->state->head);
  916.     TRY_FREE(strm, strm->state->prev);
  917.     TRY_FREE(strm, strm->state->window);
  918.  
  919.     ZFREE(strm, strm->state);
  920.     strm->state = Z_NULL;
  921.  
  922.     return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK;
  923. }
  924.  
  925. /* =========================================================================
  926.  * Copy the source state to the destination state.
  927.  * To simplify the source, this is not supported for 16-bit MSDOS (which
  928.  * doesn't have enough memory anyway to duplicate compression states).
  929.  */
  930. int ZEXPORT deflateCopy (dest, source)
  931.     z_streamp dest;
  932.     z_streamp source;
  933. {
  934. #ifdef MAXSEG_64K
  935.     return Z_STREAM_ERROR;
  936. #else
  937.     deflate_state *ds;
  938.     deflate_state *ss;
  939.     ushf *overlay;
  940.  
  941.  
  942.     if (source == Z_NULL || dest == Z_NULL || source->state == Z_NULL) {
  943.         return Z_STREAM_ERROR;
  944.     }
  945.  
  946.     ss = source->state;
  947.  
  948.     zmemcpy(dest, source, sizeof(z_stream));
  949.  
  950.     ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state));
  951.     if (ds == Z_NULL) return Z_MEM_ERROR;
  952.     dest->state = (struct internal_state FAR *) ds;
  953.     zmemcpy(ds, ss, sizeof(deflate_state));
  954.     ds->strm = dest;
  955.  
  956.     ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte));
  957.     ds->prev   = (Posf *)  ZALLOC(dest, ds->w_size, sizeof(Pos));
  958.     ds->head   = (Posf *)  ZALLOC(dest, ds->hash_size, sizeof(Pos));
  959.     overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2);
  960.     ds->pending_buf = (uchf *) overlay;
  961.  
  962.     if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL ||
  963.         ds->pending_buf == Z_NULL) {
  964.         deflateEnd (dest);
  965.         return Z_MEM_ERROR;
  966.     }
  967.     /* following zmemcpy do not work for 16-bit MSDOS */
  968.     zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte));
  969.     zmemcpy(ds->prev, ss->prev, ds->w_size * sizeof(Pos));
  970.     zmemcpy(ds->head, ss->head, ds->hash_size * sizeof(Pos));
  971.     zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size);
  972.  
  973.     ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf);
  974.     ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush);
  975.     ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize;
  976.  
  977.     ds->l_desc.dyn_tree = ds->dyn_ltree;
  978.     ds->d_desc.dyn_tree = ds->dyn_dtree;
  979.     ds->bl_desc.dyn_tree = ds->bl_tree;
  980.  
  981.     return Z_OK;
  982. #endif /* MAXSEG_64K */
  983. }
  984.  
  985. /* ===========================================================================
  986.  * Read a new buffer from the current input stream, update the adler32
  987.  * and total number of bytes read.  All deflate() input goes through
  988.  * this function so some applications may wish to modify it to avoid
  989.  * allocating a large strm->next_in buffer and copying from it.
  990.  * (See also flush_pending()).
  991.  */
  992. local int read_buf(strm, buf, size)
  993.     z_streamp strm;
  994.     Bytef *buf;
  995.     unsigned size;
  996. {
  997.     unsigned len = strm->avail_in;
  998.  
  999.     if (len > size) len = size;
  1000.     if (len == 0) return 0;
  1001.  
  1002.     strm->avail_in  -= len;
  1003.  
  1004.     if (strm->state->wrap == 1) {
  1005.         strm->adler = adler32(strm->adler, strm->next_in, len);
  1006.     }
  1007. #ifdef GZIP
  1008.     else if (strm->state->wrap == 2) {
  1009.         strm->adler = crc32(strm->adler, strm->next_in, len);
  1010.     }
  1011. #endif
  1012.     zmemcpy(buf, strm->next_in, len);
  1013.     strm->next_in  += len;
  1014.     strm->total_in += len;
  1015.  
  1016.     return (int)len;
  1017. }
  1018.  
  1019. /* ===========================================================================
  1020.  * Initialize the "longest match" routines for a new zlib stream
  1021.  */
  1022. local void lm_init (s)
  1023.     deflate_state *s;
  1024. {
  1025.     s->window_size = (ulg)2L*s->w_size;
  1026.  
  1027.     CLEAR_HASH(s);
  1028.  
  1029.     /* Set the default configuration parameters:
  1030.      */
  1031.     s->max_lazy_match   = configuration_table[s->level].max_lazy;
  1032.     s->good_match       = configuration_table[s->level].good_length;
  1033.     s->nice_match       = configuration_table[s->level].nice_length;
  1034.     s->max_chain_length = configuration_table[s->level].max_chain;
  1035.  
  1036.     s->strstart = 0;
  1037.     s->block_start = 0L;
  1038.     s->lookahead = 0;
  1039.     s->match_length = s->prev_length = MIN_MATCH-1;
  1040.     s->match_available = 0;
  1041.     s->ins_h = 0;
  1042. #ifndef FASTEST
  1043. #ifdef ASMV
  1044.     match_init(); /* initialize the asm code */
  1045. #endif
  1046. #endif
  1047. }
  1048.  
  1049. #ifndef FASTEST
  1050. /* ===========================================================================
  1051.  * Set match_start to the longest match starting at the given string and
  1052.  * return its length. Matches shorter or equal to prev_length are discarded,
  1053.  * in which case the result is equal to prev_length and match_start is
  1054.  * garbage.
  1055.  * IN assertions: cur_match is the head of the hash chain for the current
  1056.  *   string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
  1057.  * OUT assertion: the match length is not greater than s->lookahead.
  1058.  */
  1059. #ifndef ASMV
  1060. /* For 80x86 and 680x0, an optimized version will be provided in match.asm or
  1061.  * match.S. The code will be functionally equivalent.
  1062.  */
  1063. local uInt longest_match(s, cur_match)
  1064.     deflate_state *s;
  1065.     IPos cur_match;                             /* current match */
  1066. {
  1067.     unsigned chain_length = s->max_chain_length;/* max hash chain length */
  1068.     register Bytef *scan = s->window + s->strstart; /* current string */
  1069.     register Bytef *match;                       /* matched string */
  1070.     register int len;                           /* length of current match */
  1071.     int best_len = s->prev_length;              /* best match length so far */
  1072.     int nice_match = s->nice_match;             /* stop if match long enough */
  1073.     IPos limit = s->strstart > (IPos)MAX_DIST(s) ?
  1074.         s->strstart - (IPos)MAX_DIST(s) : NIL;
  1075.     /* Stop when cur_match becomes <= limit. To simplify the code,
  1076.      * we prevent matches with the string of window index 0.
  1077.      */
  1078.     Posf *prev = s->prev;
  1079.     uInt wmask = s->w_mask;
  1080.  
  1081. #ifdef UNALIGNED_OK
  1082.     /* Compare two bytes at a time. Note: this is not always beneficial.
  1083.      * Try with and without -DUNALIGNED_OK to check.
  1084.      */
  1085.     register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1;
  1086.     register ush scan_start = *(ushf*)scan;
  1087.     register ush scan_end   = *(ushf*)(scan+best_len-1);
  1088. #else
  1089.     register Bytef *strend = s->window + s->strstart + MAX_MATCH;
  1090.     register Byte scan_end1  = scan[best_len-1];
  1091.     register Byte scan_end   = scan[best_len];
  1092. #endif
  1093.  
  1094.     /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
  1095.      * It is easy to get rid of this optimization if necessary.
  1096.      */
  1097.     Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");
  1098.  
  1099.     /* Do not waste too much time if we already have a good match: */
  1100.     if (s->prev_length >= s->good_match) {
  1101.         chain_length >>= 2;
  1102.     }
  1103.     /* Do not look for matches beyond the end of the input. This is necessary
  1104.      * to make deflate deterministic.
  1105.      */
  1106.     if ((uInt)nice_match > s->lookahead) nice_match = s->lookahead;
  1107.  
  1108.     Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
  1109.  
  1110.     do {
  1111.         Assert(cur_match < s->strstart, "no future");
  1112.         match = s->window + cur_match;
  1113.  
  1114.         /* Skip to next match if the match length cannot increase
  1115.          * or if the match length is less than 2.  Note that the checks below
  1116.          * for insufficient lookahead only occur occasionally for performance
  1117.          * reasons.  Therefore uninitialized memory will be accessed, and
  1118.          * conditional jumps will be made that depend on those values.
  1119.          * However the length of the match is limited to the lookahead, so
  1120.          * the output of deflate is not affected by the uninitialized values.
  1121.          */
  1122. #if (defined(UNALIGNED_OK) && MAX_MATCH == 258)
  1123.         /* This code assumes sizeof(unsigned short) == 2. Do not use
  1124.          * UNALIGNED_OK if your compiler uses a different size.
  1125.          */
  1126.         if (*(ushf*)(match+best_len-1) != scan_end ||
  1127.             *(ushf*)match != scan_start) continue;
  1128.  
  1129.         /* It is not necessary to compare scan[2] and match[2] since they are
  1130.          * always equal when the other bytes match, given that the hash keys
  1131.          * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at
  1132.          * strstart+3, +5, ... up to strstart+257. We check for insufficient
  1133.          * lookahead only every 4th comparison; the 128th check will be made
  1134.          * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is
  1135.          * necessary to put more guard bytes at the end of the window, or
  1136.          * to check more often for insufficient lookahead.
  1137.          */
  1138.         Assert(scan[2] == match[2], "scan[2]?");
  1139.         scan++, match++;
  1140.         do {
  1141.         } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
  1142.                  *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
  1143.                  *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
  1144.                  *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
  1145.                  scan < strend);
  1146.         /* The funny "do {}" generates better code on most compilers */
  1147.  
  1148.         /* Here, scan <= window+strstart+257 */
  1149.         Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
  1150.         if (*scan == *match) scan++;
  1151.  
  1152.         len = (MAX_MATCH - 1) - (int)(strend-scan);
  1153.         scan = strend - (MAX_MATCH-1);
  1154.  
  1155. #else /* UNALIGNED_OK */
  1156.  
  1157.         if (match[best_len]   != scan_end  ||
  1158.             match[best_len-1] != scan_end1 ||
  1159.             *match            != *scan     ||
  1160.             *++match          != scan[1])      continue;
  1161.  
  1162.         /* The check at best_len-1 can be removed because it will be made
  1163.          * again later. (This heuristic is not always a win.)
  1164.          * It is not necessary to compare scan[2] and match[2] since they
  1165.          * are always equal when the other bytes match, given that
  1166.          * the hash keys are equal and that HASH_BITS >= 8.
  1167.          */
  1168.         scan += 2, match++;
  1169.         Assert(*scan == *match, "match[2]?");
  1170.  
  1171.         /* We check for insufficient lookahead only every 8th comparison;
  1172.          * the 256th check will be made at strstart+258.
  1173.          */
  1174.         do {
  1175.         } while (*++scan == *++match && *++scan == *++match &&
  1176.                  *++scan == *++match && *++scan == *++match &&
  1177.                  *++scan == *++match && *++scan == *++match &&
  1178.                  *++scan == *++match && *++scan == *++match &&
  1179.                  scan < strend);
  1180.  
  1181.         Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
  1182.  
  1183.         len = MAX_MATCH - (int)(strend - scan);
  1184.         scan = strend - MAX_MATCH;
  1185.  
  1186. #endif /* UNALIGNED_OK */
  1187.  
  1188.         if (len > best_len) {
  1189.             s->match_start = cur_match;
  1190.             best_len = len;
  1191.             if (len >= nice_match) break;
  1192. #ifdef UNALIGNED_OK
  1193.             scan_end = *(ushf*)(scan+best_len-1);
  1194. #else
  1195.             scan_end1  = scan[best_len-1];
  1196.             scan_end   = scan[best_len];
  1197. #endif
  1198.         }
  1199.     } while ((cur_match = prev[cur_match & wmask]) > limit
  1200.              && --chain_length != 0);
  1201.  
  1202.     if ((uInt)best_len <= s->lookahead) return (uInt)best_len;
  1203.     return s->lookahead;
  1204. }
  1205. #endif /* ASMV */
  1206.  
  1207. #else /* FASTEST */
  1208.  
  1209. /* ---------------------------------------------------------------------------
  1210.  * Optimized version for FASTEST only
  1211.  */
  1212. local uInt longest_match(s, cur_match)
  1213.     deflate_state *s;
  1214.     IPos cur_match;                             /* current match */
  1215. {
  1216.     register Bytef *scan = s->window + s->strstart; /* current string */
  1217.     register Bytef *match;                       /* matched string */
  1218.     register int len;                           /* length of current match */
  1219.     register Bytef *strend = s->window + s->strstart + MAX_MATCH;
  1220.  
  1221.     /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
  1222.      * It is easy to get rid of this optimization if necessary.
  1223.      */
  1224.     Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");
  1225.  
  1226.     Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
  1227.  
  1228.     Assert(cur_match < s->strstart, "no future");
  1229.  
  1230.     match = s->window + cur_match;
  1231.  
  1232.     /* Return failure if the match length is less than 2:
  1233.      */
  1234.     if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1;
  1235.  
  1236.     /* The check at best_len-1 can be removed because it will be made
  1237.      * again later. (This heuristic is not always a win.)
  1238.      * It is not necessary to compare scan[2] and match[2] since they
  1239.      * are always equal when the other bytes match, given that
  1240.      * the hash keys are equal and that HASH_BITS >= 8.
  1241.      */
  1242.     scan += 2, match += 2;
  1243.     Assert(*scan == *match, "match[2]?");
  1244.  
  1245.     /* We check for insufficient lookahead only every 8th comparison;
  1246.      * the 256th check will be made at strstart+258.
  1247.      */
  1248.     do {
  1249.     } while (*++scan == *++match && *++scan == *++match &&
  1250.              *++scan == *++match && *++scan == *++match &&
  1251.              *++scan == *++match && *++scan == *++match &&
  1252.              *++scan == *++match && *++scan == *++match &&
  1253.              scan < strend);
  1254.  
  1255.     Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
  1256.  
  1257.     len = MAX_MATCH - (int)(strend - scan);
  1258.  
  1259.     if (len < MIN_MATCH) return MIN_MATCH - 1;
  1260.  
  1261.     s->match_start = cur_match;
  1262.     return (uInt)len <= s->lookahead ? (uInt)len : s->lookahead;
  1263. }
  1264.  
  1265. #endif /* FASTEST */
  1266.  
  1267. #ifdef DEBUG
  1268. /* ===========================================================================
  1269.  * Check that the match at match_start is indeed a match.
  1270.  */
  1271. local void check_match(s, start, match, length)
  1272.     deflate_state *s;
  1273.     IPos start, match;
  1274.     int length;
  1275. {
  1276.     /* check that the match is indeed a match */
  1277.     if (zmemcmp(s->window + match,
  1278.                 s->window + start, length) != EQUAL) {
  1279.         fprintf(stderr, " start %u, match %u, length %d\n",
  1280.                 start, match, length);
  1281.         do {
  1282.             fprintf(stderr, "%c%c", s->window[match++], s->window[start++]);
  1283.         } while (--length != 0);
  1284.         z_error("invalid match");
  1285.     }
  1286.     if (z_verbose > 1) {
  1287.         fprintf(stderr,"\\[%d,%d]", start-match, length);
  1288.         do { putc(s->window[start++], stderr); } while (--length != 0);
  1289.     }
  1290. }
  1291. #else
  1292. #  define check_match(s, start, match, length)
  1293. #endif /* DEBUG */
  1294.  
  1295. /* ===========================================================================
  1296.  * Fill the window when the lookahead becomes insufficient.
  1297.  * Updates strstart and lookahead.
  1298.  *
  1299.  * IN assertion: lookahead < MIN_LOOKAHEAD
  1300.  * OUT assertions: strstart <= window_size-MIN_LOOKAHEAD
  1301.  *    At least one byte has been read, or avail_in == 0; reads are
  1302.  *    performed for at least two bytes (required for the zip translate_eol
  1303.  *    option -- not supported here).
  1304.  */
  1305. local void fill_window(s)
  1306.     deflate_state *s;
  1307. {
  1308.     register unsigned n, m;
  1309.     register Posf *p;
  1310.     unsigned more;    /* Amount of free space at the end of the window. */
  1311.     uInt wsize = s->w_size;
  1312.  
  1313.     do {
  1314.         more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart);
  1315.  
  1316.         /* Deal with !@#$% 64K limit: */
  1317.         if (sizeof(int) <= 2) {
  1318.             if (more == 0 && s->strstart == 0 && s->lookahead == 0) {
  1319.                 more = wsize;
  1320.  
  1321.             } else if (more == (unsigned)(-1)) {
  1322.                 /* Very unlikely, but possible on 16 bit machine if
  1323.                  * strstart == 0 && lookahead == 1 (input done a byte at time)
  1324.                  */
  1325.                 more--;
  1326.             }
  1327.         }
  1328.  
  1329.         /* If the window is almost full and there is insufficient lookahead,
  1330.          * move the upper half to the lower one to make room in the upper half.
  1331.          */
  1332.         if (s->strstart >= wsize+MAX_DIST(s)) {
  1333.  
  1334.             zmemcpy(s->window, s->window+wsize, (unsigned)wsize);
  1335.             s->match_start -= wsize;
  1336.             s->strstart    -= wsize; /* we now have strstart >= MAX_DIST */
  1337.             s->block_start -= (long) wsize;
  1338.  
  1339.             /* Slide the hash table (could be avoided with 32 bit values
  1340.                at the expense of memory usage). We slide even when level == 0
  1341.                to keep the hash table consistent if we switch back to level > 0
  1342.                later. (Using level 0 permanently is not an optimal usage of
  1343.                zlib, so we don't care about this pathological case.)
  1344.              */
  1345.             n = s->hash_size;
  1346.             p = &s->head[n];
  1347.             do {
  1348.                 m = *--p;
  1349.                 *p = (Pos)(m >= wsize ? m-wsize : NIL);
  1350.             } while (--n);
  1351.  
  1352.             n = wsize;
  1353. #ifndef FASTEST
  1354.             p = &s->prev[n];
  1355.             do {
  1356.                 m = *--p;
  1357.                 *p = (Pos)(m >= wsize ? m-wsize : NIL);
  1358.                 /* If n is not on any hash chain, prev[n] is garbage but
  1359.                  * its value will never be used.
  1360.                  */
  1361.             } while (--n);
  1362. #endif
  1363.             more += wsize;
  1364.         }
  1365.         if (s->strm->avail_in == 0) return;
  1366.  
  1367.         /* If there was no sliding:
  1368.          *    strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
  1369.          *    more == window_size - lookahead - strstart
  1370.          * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1)
  1371.          * => more >= window_size - 2*WSIZE + 2
  1372.          * In the BIG_MEM or MMAP case (not yet supported),
  1373.          *   window_size == input_size + MIN_LOOKAHEAD  &&
  1374.          *   strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD.
  1375.          * Otherwise, window_size == 2*WSIZE so more >= 2.
  1376.          * If there was sliding, more >= WSIZE. So in all cases, more >= 2.
  1377.          */
  1378.         Assert(more >= 2, "more < 2");
  1379.  
  1380.         n = read_buf(s->strm, s->window + s->strstart + s->lookahead, more);
  1381.         s->lookahead += n;
  1382.  
  1383.         /* Initialize the hash value now that we have some input: */
  1384.         if (s->lookahead >= MIN_MATCH) {
  1385.             s->ins_h = s->window[s->strstart];
  1386.             UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
  1387. #if MIN_MATCH != 3
  1388.             Call UPDATE_HASH() MIN_MATCH-3 more times
  1389. #endif
  1390.         }
  1391.         /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage,
  1392.          * but this is not important since only literal bytes will be emitted.
  1393.          */
  1394.  
  1395.     } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0);
  1396.  
  1397.     /* If the WIN_INIT bytes after the end of the current data have never been
  1398.      * written, then zero those bytes in order to avoid memory check reports of
  1399.      * the use of uninitialized (or uninitialised as Julian writes) bytes by
  1400.      * the longest match routines.  Update the high water mark for the next
  1401.      * time through here.  WIN_INIT is set to MAX_MATCH since the longest match
  1402.      * routines allow scanning to strstart + MAX_MATCH, ignoring lookahead.
  1403.      */
  1404.     if (s->high_water < s->window_size) {
  1405.         ulg curr = s->strstart + (ulg)(s->lookahead);
  1406.         ulg init;
  1407.  
  1408.         if (s->high_water < curr) {
  1409.             /* Previous high water mark below current data -- zero WIN_INIT
  1410.              * bytes or up to end of window, whichever is less.
  1411.              */
  1412.             init = s->window_size - curr;
  1413.             if (init > WIN_INIT)
  1414.                 init = WIN_INIT;
  1415.             zmemzero(s->window + curr, (unsigned)init);
  1416.             s->high_water = curr + init;
  1417.         }
  1418.         else if (s->high_water < (ulg)curr + WIN_INIT) {
  1419.             /* High water mark at or above current data, but below current data
  1420.              * plus WIN_INIT -- zero out to current data plus WIN_INIT, or up
  1421.              * to end of window, whichever is less.
  1422.              */
  1423.             init = (ulg)curr + WIN_INIT - s->high_water;
  1424.             if (init > s->window_size - s->high_water)
  1425.                 init = s->window_size - s->high_water;
  1426.             zmemzero(s->window + s->high_water, (unsigned)init);
  1427.             s->high_water += init;
  1428.         }
  1429.     }
  1430. }
  1431.  
  1432. /* ===========================================================================
  1433.  * Flush the current block, with given end-of-file flag.
  1434.  * IN assertion: strstart is set to the end of the current match.
  1435.  */
  1436. #define FLUSH_BLOCK_ONLY(s, last) { \
  1437.    _tr_flush_block(s, (s->block_start >= 0L ? \
  1438.                    (charf *)&s->window[(unsigned)s->block_start] : \
  1439.                    (charf *)Z_NULL), \
  1440.                 (ulg)((long)s->strstart - s->block_start), \
  1441.                 (last)); \
  1442.    s->block_start = s->strstart; \
  1443.    flush_pending(s->strm); \
  1444.    Tracev((stderr,"[FLUSH]")); \
  1445. }
  1446.  
  1447. /* Same but force premature exit if necessary. */
  1448. #define FLUSH_BLOCK(s, last) { \
  1449.    FLUSH_BLOCK_ONLY(s, last); \
  1450.    if (s->strm->avail_out == 0) return (last) ? finish_started : need_more; \
  1451. }
  1452.  
  1453. /* ===========================================================================
  1454.  * Copy without compression as much as possible from the input stream, return
  1455.  * the current block state.
  1456.  * This function does not insert new strings in the dictionary since
  1457.  * uncompressible data is probably not useful. This function is used
  1458.  * only for the level=0 compression option.
  1459.  * NOTE: this function should be optimized to avoid extra copying from
  1460.  * window to pending_buf.
  1461.  */
  1462. local block_state deflate_stored(s, flush)
  1463.     deflate_state *s;
  1464.     int flush;
  1465. {
  1466.     /* Stored blocks are limited to 0xffff bytes, pending_buf is limited
  1467.      * to pending_buf_size, and each stored block has a 5 byte header:
  1468.      */
  1469.     ulg max_block_size = 0xffff;
  1470.     ulg max_start;
  1471.  
  1472.     if (max_block_size > s->pending_buf_size - 5) {
  1473.         max_block_size = s->pending_buf_size - 5;
  1474.     }
  1475.  
  1476.     /* Copy as much as possible from input to output: */
  1477.     for (;;) {
  1478.         /* Fill the window as much as possible: */
  1479.         if (s->lookahead <= 1) {
  1480.  
  1481.             Assert(s->strstart < s->w_size+MAX_DIST(s) ||
  1482.                    s->block_start >= (long)s->w_size, "slide too late");
  1483.  
  1484.             fill_window(s);
  1485.             if (s->lookahead == 0 && flush == Z_NO_FLUSH) return need_more;
  1486.  
  1487.             if (s->lookahead == 0) break; /* flush the current block */
  1488.         }
  1489.         Assert(s->block_start >= 0L, "block gone");
  1490.  
  1491.         s->strstart += s->lookahead;
  1492.         s->lookahead = 0;
  1493.  
  1494.         /* Emit a stored block if pending_buf will be full: */
  1495.         max_start = s->block_start + max_block_size;
  1496.         if (s->strstart == 0 || (ulg)s->strstart >= max_start) {
  1497.             /* strstart == 0 is possible when wraparound on 16-bit machine */
  1498.             s->lookahead = (uInt)(s->strstart - max_start);
  1499.             s->strstart = (uInt)max_start;
  1500.             FLUSH_BLOCK(s, 0);
  1501.         }
  1502.         /* Flush if we may have to slide, otherwise block_start may become
  1503.          * negative and the data will be gone:
  1504.          */
  1505.         if (s->strstart - (uInt)s->block_start >= MAX_DIST(s)) {
  1506.             FLUSH_BLOCK(s, 0);
  1507.         }
  1508.     }
  1509.     FLUSH_BLOCK(s, flush == Z_FINISH);
  1510.     return flush == Z_FINISH ? finish_done : block_done;
  1511. }
  1512.  
  1513. /* ===========================================================================
  1514.  * Compress as much as possible from the input stream, return the current
  1515.  * block state.
  1516.  * This function does not perform lazy evaluation of matches and inserts
  1517.  * new strings in the dictionary only for unmatched strings or for short
  1518.  * matches. It is used only for the fast compression options.
  1519.  */
  1520. local block_state deflate_fast(s, flush)
  1521.     deflate_state *s;
  1522.     int flush;
  1523. {
  1524.     IPos hash_head;       /* head of the hash chain */
  1525.     int bflush;           /* set if current block must be flushed */
  1526.  
  1527.     for (;;) {
  1528.         /* Make sure that we always have enough lookahead, except
  1529.          * at the end of the input file. We need MAX_MATCH bytes
  1530.          * for the next match, plus MIN_MATCH bytes to insert the
  1531.          * string following the next match.
  1532.          */
  1533.         if (s->lookahead < MIN_LOOKAHEAD) {
  1534.             fill_window(s);
  1535.             if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) {
  1536.                 return need_more;
  1537.             }
  1538.             if (s->lookahead == 0) break; /* flush the current block */
  1539.         }
  1540.  
  1541.         /* Insert the string window[strstart .. strstart+2] in the
  1542.          * dictionary, and set hash_head to the head of the hash chain:
  1543.          */
  1544.         hash_head = NIL;
  1545.         if (s->lookahead >= MIN_MATCH) {
  1546.             INSERT_STRING(s, s->strstart, hash_head);
  1547.         }
  1548.  
  1549.         /* Find the longest match, discarding those <= prev_length.
  1550.          * At this point we have always match_length < MIN_MATCH
  1551.          */
  1552.         if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) {
  1553.             /* To simplify the code, we prevent matches with the string
  1554.              * of window index 0 (in particular we have to avoid a match
  1555.              * of the string with itself at the start of the input file).
  1556.              */
  1557.             s->match_length = longest_match (s, hash_head);
  1558.             /* longest_match() sets match_start */
  1559.         }
  1560.         if (s->match_length >= MIN_MATCH) {
  1561.             check_match(s, s->strstart, s->match_start, s->match_length);
  1562.  
  1563.             _tr_tally_dist(s, s->strstart - s->match_start,
  1564.                            s->match_length - MIN_MATCH, bflush);
  1565.  
  1566.             s->lookahead -= s->match_length;
  1567.  
  1568.             /* Insert new strings in the hash table only if the match length
  1569.              * is not too large. This saves time but degrades compression.
  1570.              */
  1571. #ifndef FASTEST
  1572.             if (s->match_length <= s->max_insert_length &&
  1573.                 s->lookahead >= MIN_MATCH) {
  1574.                 s->match_length--; /* string at strstart already in table */
  1575.                 do {
  1576.                     s->strstart++;
  1577.                     INSERT_STRING(s, s->strstart, hash_head);
  1578.                     /* strstart never exceeds WSIZE-MAX_MATCH, so there are
  1579.                      * always MIN_MATCH bytes ahead.
  1580.                      */
  1581.                 } while (--s->match_length != 0);
  1582.                 s->strstart++;
  1583.             } else
  1584. #endif
  1585.             {
  1586.                 s->strstart += s->match_length;
  1587.                 s->match_length = 0;
  1588.                 s->ins_h = s->window[s->strstart];
  1589.                 UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
  1590. #if MIN_MATCH != 3
  1591.                 Call UPDATE_HASH() MIN_MATCH-3 more times
  1592. #endif
  1593.                 /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not
  1594.                  * matter since it will be recomputed at next deflate call.
  1595.                  */
  1596.             }
  1597.         } else {
  1598.             /* No match, output a literal byte */
  1599.             Tracevv((stderr,"%c", s->window[s->strstart]));
  1600.             _tr_tally_lit (s, s->window[s->strstart], bflush);
  1601.             s->lookahead--;
  1602.             s->strstart++;
  1603.         }
  1604.         if (bflush) FLUSH_BLOCK(s, 0);
  1605.     }
  1606.     FLUSH_BLOCK(s, flush == Z_FINISH);
  1607.     return flush == Z_FINISH ? finish_done : block_done;
  1608. }
  1609.  
  1610. #ifndef FASTEST
  1611. /* ===========================================================================
  1612.  * Same as above, but achieves better compression. We use a lazy
  1613.  * evaluation for matches: a match is finally adopted only if there is
  1614.  * no better match at the next window position.
  1615.  */
  1616. local block_state deflate_slow(s, flush)
  1617.     deflate_state *s;
  1618.     int flush;
  1619. {
  1620.     IPos hash_head;          /* head of hash chain */
  1621.     int bflush;              /* set if current block must be flushed */
  1622.  
  1623.     /* Process the input block. */
  1624.     for (;;) {
  1625.         /* Make sure that we always have enough lookahead, except
  1626.          * at the end of the input file. We need MAX_MATCH bytes
  1627.          * for the next match, plus MIN_MATCH bytes to insert the
  1628.          * string following the next match.
  1629.          */
  1630.         if (s->lookahead < MIN_LOOKAHEAD) {
  1631.             fill_window(s);
  1632.             if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) {
  1633.                 return need_more;
  1634.             }
  1635.             if (s->lookahead == 0) break; /* flush the current block */
  1636.         }
  1637.  
  1638.         /* Insert the string window[strstart .. strstart+2] in the
  1639.          * dictionary, and set hash_head to the head of the hash chain:
  1640.          */
  1641.         hash_head = NIL;
  1642.         if (s->lookahead >= MIN_MATCH) {
  1643.             INSERT_STRING(s, s->strstart, hash_head);
  1644.         }
  1645.  
  1646.         /* Find the longest match, discarding those <= prev_length.
  1647.          */
  1648.         s->prev_length = s->match_length, s->prev_match = s->match_start;
  1649.         s->match_length = MIN_MATCH-1;
  1650.  
  1651.         if (hash_head != NIL && s->prev_length < s->max_lazy_match &&
  1652.             s->strstart - hash_head <= MAX_DIST(s)) {
  1653.             /* To simplify the code, we prevent matches with the string
  1654.              * of window index 0 (in particular we have to avoid a match
  1655.              * of the string with itself at the start of the input file).
  1656.              */
  1657.             s->match_length = longest_match (s, hash_head);
  1658.             /* longest_match() sets match_start */
  1659.  
  1660.             if (s->match_length <= 5 && (s->strategy == Z_FILTERED
  1661. #if TOO_FAR <= 32767
  1662.                 || (s->match_length == MIN_MATCH &&
  1663.                     s->strstart - s->match_start > TOO_FAR)
  1664. #endif
  1665.                 )) {
  1666.  
  1667.                 /* If prev_match is also MIN_MATCH, match_start is garbage
  1668.                  * but we will ignore the current match anyway.
  1669.                  */
  1670.                 s->match_length = MIN_MATCH-1;
  1671.             }
  1672.         }
  1673.         /* If there was a match at the previous step and the current
  1674.          * match is not better, output the previous match:
  1675.          */
  1676.         if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) {
  1677.             uInt max_insert = s->strstart + s->lookahead - MIN_MATCH;
  1678.             /* Do not insert strings in hash table beyond this. */
  1679.  
  1680.             check_match(s, s->strstart-1, s->prev_match, s->prev_length);
  1681.  
  1682.             _tr_tally_dist(s, s->strstart -1 - s->prev_match,
  1683.                            s->prev_length - MIN_MATCH, bflush);
  1684.  
  1685.             /* Insert in hash table all strings up to the end of the match.
  1686.              * strstart-1 and strstart are already inserted. If there is not
  1687.              * enough lookahead, the last two strings are not inserted in
  1688.              * the hash table.
  1689.              */
  1690.             s->lookahead -= s->prev_length-1;
  1691.             s->prev_length -= 2;
  1692.             do {
  1693.                 if (++s->strstart <= max_insert) {
  1694.                     INSERT_STRING(s, s->strstart, hash_head);
  1695.                 }
  1696.             } while (--s->prev_length != 0);
  1697.             s->match_available = 0;
  1698.             s->match_length = MIN_MATCH-1;
  1699.             s->strstart++;
  1700.  
  1701.             if (bflush) FLUSH_BLOCK(s, 0);
  1702.  
  1703.         } else if (s->match_available) {
  1704.             /* If there was no match at the previous position, output a
  1705.              * single literal. If there was a match but the current match
  1706.              * is longer, truncate the previous match to a single literal.
  1707.              */
  1708.             Tracevv((stderr,"%c", s->window[s->strstart-1]));
  1709.             _tr_tally_lit(s, s->window[s->strstart-1], bflush);
  1710.             if (bflush) {
  1711.                 FLUSH_BLOCK_ONLY(s, 0);
  1712.             }
  1713.             s->strstart++;
  1714.             s->lookahead--;
  1715.             if (s->strm->avail_out == 0) return need_more;
  1716.         } else {
  1717.             /* There is no previous match to compare with, wait for
  1718.              * the next step to decide.
  1719.              */
  1720.             s->match_available = 1;
  1721.             s->strstart++;
  1722.             s->lookahead--;
  1723.         }
  1724.     }
  1725.     Assert (flush != Z_NO_FLUSH, "no flush?");
  1726.     if (s->match_available) {
  1727.         Tracevv((stderr,"%c", s->window[s->strstart-1]));
  1728.         _tr_tally_lit(s, s->window[s->strstart-1], bflush);
  1729.         s->match_available = 0;
  1730.     }
  1731.     FLUSH_BLOCK(s, flush == Z_FINISH);
  1732.     return flush == Z_FINISH ? finish_done : block_done;
  1733. }
  1734. #endif /* FASTEST */
  1735.  
  1736. /* ===========================================================================
  1737.  * For Z_RLE, simply look for runs of bytes, generate matches only of distance
  1738.  * one.  Do not maintain a hash table.  (It will be regenerated if this run of
  1739.  * deflate switches away from Z_RLE.)
  1740.  */
  1741. local block_state deflate_rle(s, flush)
  1742.     deflate_state *s;
  1743.     int flush;
  1744. {
  1745.     int bflush;             /* set if current block must be flushed */
  1746.     uInt prev;              /* byte at distance one to match */
  1747.     Bytef *scan, *strend;   /* scan goes up to strend for length of run */
  1748.  
  1749.     for (;;) {
  1750.         /* Make sure that we always have enough lookahead, except
  1751.          * at the end of the input file. We need MAX_MATCH bytes
  1752.          * for the longest encodable run.
  1753.          */
  1754.         if (s->lookahead < MAX_MATCH) {
  1755.             fill_window(s);
  1756.             if (s->lookahead < MAX_MATCH && flush == Z_NO_FLUSH) {
  1757.                 return need_more;
  1758.             }
  1759.             if (s->lookahead == 0) break; /* flush the current block */
  1760.         }
  1761.  
  1762.         /* See how many times the previous byte repeats */
  1763.         s->match_length = 0;
  1764.         if (s->lookahead >= MIN_MATCH && s->strstart > 0) {
  1765.             scan = s->window + s->strstart - 1;
  1766.             prev = *scan;
  1767.             if (prev == *++scan && prev == *++scan && prev == *++scan) {
  1768.                 strend = s->window + s->strstart + MAX_MATCH;
  1769.                 do {
  1770.                 } while (prev == *++scan && prev == *++scan &&
  1771.                          prev == *++scan && prev == *++scan &&
  1772.                          prev == *++scan && prev == *++scan &&
  1773.                          prev == *++scan && prev == *++scan &&
  1774.                          scan < strend);
  1775.                 s->match_length = MAX_MATCH - (int)(strend - scan);
  1776.                 if (s->match_length > s->lookahead)
  1777.                     s->match_length = s->lookahead;
  1778.             }
  1779.         }
  1780.  
  1781.         /* Emit match if have run of MIN_MATCH or longer, else emit literal */
  1782.         if (s->match_length >= MIN_MATCH) {
  1783.             check_match(s, s->strstart, s->strstart - 1, s->match_length);
  1784.  
  1785.             _tr_tally_dist(s, 1, s->match_length - MIN_MATCH, bflush);
  1786.  
  1787.             s->lookahead -= s->match_length;
  1788.             s->strstart += s->match_length;
  1789.             s->match_length = 0;
  1790.         } else {
  1791.             /* No match, output a literal byte */
  1792.             Tracevv((stderr,"%c", s->window[s->strstart]));
  1793.             _tr_tally_lit (s, s->window[s->strstart], bflush);
  1794.             s->lookahead--;
  1795.             s->strstart++;
  1796.         }
  1797.         if (bflush) FLUSH_BLOCK(s, 0);
  1798.     }
  1799.     FLUSH_BLOCK(s, flush == Z_FINISH);
  1800.     return flush == Z_FINISH ? finish_done : block_done;
  1801. }
  1802.  
  1803. /* ===========================================================================
  1804.  * For Z_HUFFMAN_ONLY, do not look for matches.  Do not maintain a hash table.
  1805.  * (It will be regenerated if this run of deflate switches away from Huffman.)
  1806.  */
  1807. local block_state deflate_huff(s, flush)
  1808.     deflate_state *s;
  1809.     int flush;
  1810. {
  1811.     int bflush;             /* set if current block must be flushed */
  1812.  
  1813.     for (;;) {
  1814.         /* Make sure that we have a literal to write. */
  1815.         if (s->lookahead == 0) {
  1816.             fill_window(s);
  1817.             if (s->lookahead == 0) {
  1818.                 if (flush == Z_NO_FLUSH)
  1819.                     return need_more;
  1820.                 break;      /* flush the current block */
  1821.             }
  1822.         }
  1823.  
  1824.         /* Output a literal byte */
  1825.         s->match_length = 0;
  1826.         Tracevv((stderr,"%c", s->window[s->strstart]));
  1827.         _tr_tally_lit (s, s->window[s->strstart], bflush);
  1828.         s->lookahead--;
  1829.         s->strstart++;
  1830.         if (bflush) FLUSH_BLOCK(s, 0);
  1831.     }
  1832.     FLUSH_BLOCK(s, flush == Z_FINISH);
  1833.     return flush == Z_FINISH ? finish_done : block_done;
  1834. }
  1835.