zlib.c

经典的ppp程序

 */
local void fill_window(s)
    deflate_state *s;
{
    register unsigned n, m;
    register Posf *p;
    unsigned more;    /* Amount of free space at the end of the window. */
    uInt wsize = s->w_size;

    do {
        more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart);

        /* Deal with !@#$% 64K limit: */
        if (more == 0 && s->strstart == 0 && s->lookahead == 0) {
            more = wsize;
        } else if (more == (unsigned)(-1)) {
            /* Very unlikely, but possible on 16 bit machine if strstart == 0
             * and lookahead == 1 (input done one byte at time)
             */
            more--;

        /* If the window is almost full and there is insufficient lookahead,
         * move the upper half to the lower one to make room in the upper half.
         */
        } else if (s->strstart >= wsize+MAX_DIST(s)) {

            /* By the IN assertion, the window is not empty so we can't confuse
             * more == 0 with more == 64K on a 16 bit machine.
             */
            zmemcpy((charf *)s->window, (charf *)s->window+wsize,
                   (unsigned)wsize);
            s->match_start -= wsize;
            s->strstart    -= wsize; /* we now have strstart >= MAX_DIST */

            s->block_start -= (long) wsize;

            /* Slide the hash table (could be avoided with 32 bit values
               at the expense of memory usage):
             */
            n = s->hash_size;
            p = &s->head[n];
            do {
                m = *--p;
                *p = (Pos)(m >= wsize ? m-wsize : NIL);
            } while (--n);

            n = wsize;
            p = &s->prev[n];
            do {
                m = *--p;
                *p = (Pos)(m >= wsize ? m-wsize : NIL);
                /* If n is not on any hash chain, prev[n] is garbage but
                 * its value will never be used.
                 */
            } while (--n);

            more += wsize;
        }
        if (s->strm->avail_in == 0) return;

        /* If there was no sliding:
         *    strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
         *    more == window_size - lookahead - strstart
         * => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1)
         * => more >= window_size - 2*WSIZE + 2
         * In the BIG_MEM or MMAP case (not yet supported),
         *   window_size == input_size + MIN_LOOKAHEAD  &&
         *   strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD.
         * Otherwise, window_size == 2*WSIZE so more >= 2.
         * If there was sliding, more >= WSIZE. So in all cases, more >= 2.
         */
        Assert(more >= 2, "more < 2");

        n = read_buf(s->strm, (charf *)s->window + s->strstart + s->lookahead,
                     more);
        s->lookahead += n;

        /* Initialize the hash value now that we have some input: */
        if (s->lookahead >= MIN_MATCH) {
            s->ins_h = s->window[s->strstart];
            UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
#if MIN_MATCH != 3
            Call UPDATE_HASH() MIN_MATCH-3 more times
#endif
        }
        /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage,
         * but this is not important since only literal bytes will be emitted.
         */

    } while (s->lookahead < MIN_LOOKAHEAD && s->strm->avail_in != 0);
}

/* ===========================================================================
 * Flush the current block, with given end-of-file flag.
 * IN assertion: strstart is set to the end of the current match.
 */
#define FLUSH_BLOCK_ONLY(s, flush) { \
   ct_flush_block(s, (s->block_start >= 0L ? \
           (charf *)&s->window[(unsigned)s->block_start] : \
           (charf *)Z_NULL), (long)s->strstart - s->block_start, (flush)); \
   s->block_start = s->strstart; \
   flush_pending(s->strm); \
   Tracev((stderr,"[FLUSH]")); \
}

/* Same but force premature exit if necessary. */
#define FLUSH_BLOCK(s, flush) { \
   FLUSH_BLOCK_ONLY(s, flush); \
   if (s->strm->avail_out == 0) return 1; \
}

/* ===========================================================================
 * Compress as much as possible from the input stream, return true if
 * processing was terminated prematurely (no more input or output space).
 * This function does not perform lazy evaluationof matches and inserts
 * new strings in the dictionary only for unmatched strings or for short
 * matches. It is used only for the fast compression options.
 */
local int deflate_fast(s, flush)
    deflate_state *s;
    int flush;
{
    IPos hash_head = NIL; /* head of the hash chain */
    int bflush;     /* set if current block must be flushed */

    s->prev_length = MIN_MATCH-1;

    for (;;) {
        /* Make sure that we always have enough lookahead, except
         * at the end of the input file. We need MAX_MATCH bytes
         * for the next match, plus MIN_MATCH bytes to insert the
         * string following the next match.
         */
        if (s->lookahead < MIN_LOOKAHEAD) {
            fill_window(s);
            if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) return 1;

            if (s->lookahead == 0) break; /* flush the current block */
        }

        /* Insert the string window[strstart .. strstart+2] in the
         * dictionary, and set hash_head to the head of the hash chain:
         */
        if (s->lookahead >= MIN_MATCH) {
            INSERT_STRING(s, s->strstart, hash_head);
        }

        /* Find the longest match, discarding those <= prev_length.
         * At this point we have always match_length < MIN_MATCH
         */
        if (hash_head != NIL && s->strstart - hash_head <= MAX_DIST(s)) {
            /* To simplify the code, we prevent matches with the string
             * of window index 0 (in particular we have to avoid a match
             * of the string with itself at the start of the input file).
             */
            if (s->strategy != Z_HUFFMAN_ONLY) {
                s->match_length = longest_match (s, hash_head);
            }
            /* longest_match() sets match_start */

            if (s->match_length > s->lookahead) s->match_length = s->lookahead;
        }
        if (s->match_length >= MIN_MATCH) {
            check_match(s, s->strstart, s->match_start, s->match_length);

            bflush = ct_tally(s, s->strstart - s->match_start,
                              s->match_length - MIN_MATCH);

            s->lookahead -= s->match_length;

            /* Insert new strings in the hash table only if the match length
             * is not too large. This saves time but degrades compression.
             */
            if (s->match_length <= s->max_insert_length &&
                s->lookahead >= MIN_MATCH) {
                s->match_length--; /* string at strstart already in hash table */
                do {
                    s->strstart++;
                    INSERT_STRING(s, s->strstart, hash_head);
                    /* strstart never exceeds WSIZE-MAX_MATCH, so there are
                     * always MIN_MATCH bytes ahead.
                     */
                } while (--s->match_length != 0);
                s->strstart++; 
            } else {
                s->strstart += s->match_length;
                s->match_length = 0;
                s->ins_h = s->window[s->strstart];
                UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
#if MIN_MATCH != 3
                Call UPDATE_HASH() MIN_MATCH-3 more times
#endif
                /* If lookahead < MIN_MATCH, ins_h is garbage, but it does not
                 * matter since it will be recomputed at next deflate call.
                 */
            }
        } else {
            /* No match, output a literal byte */
            Tracevv((stderr,"%c", s->window[s->strstart]));
            bflush = ct_tally (s, 0, s->window[s->strstart]);
            s->lookahead--;
            s->strstart++; 
        }
        if (bflush) FLUSH_BLOCK(s, Z_NO_FLUSH);
    }
    FLUSH_BLOCK(s, flush);
    return 0; /* normal exit */
}

/* ===========================================================================
 * Same as above, but achieves better compression. We use a lazy
 * evaluation for matches: a match is finally adopted only if there is
 * no better match at the next window position.
 */
local int deflate_slow(s, flush)
    deflate_state *s;
    int flush;
{
    IPos hash_head = NIL;    /* head of hash chain */
    int bflush;              /* set if current block must be flushed */

    /* Process the input block. */
    for (;;) {
        /* Make sure that we always have enough lookahead, except
         * at the end of the input file. We need MAX_MATCH bytes
         * for the next match, plus MIN_MATCH bytes to insert the
         * string following the next match.
         */
        if (s->lookahead < MIN_LOOKAHEAD) {
            fill_window(s);
            if (s->lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) return 1;

            if (s->lookahead == 0) break; /* flush the current block */
        }

        /* Insert the string window[strstart .. strstart+2] in the
         * dictionary, and set hash_head to the head of the hash chain:
         */
        if (s->lookahead >= MIN_MATCH) {
            INSERT_STRING(s, s->strstart, hash_head);
        }

        /* Find the longest match, discarding those <= prev_length.
         */
        s->prev_length = s->match_length, s->prev_match = s->match_start;
        s->match_length = MIN_MATCH-1;

        if (hash_head != NIL && s->prev_length < s->max_lazy_match &&
            s->strstart - hash_head <= MAX_DIST(s)) {
            /* To simplify the code, we prevent matches with the string
             * of window index 0 (in particular we have to avoid a match
             * of the string with itself at the start of the input file).
             */
            if (s->strategy != Z_HUFFMAN_ONLY) {
                s->match_length = longest_match (s, hash_head);
            }
            /* longest_match() sets match_start */
            if (s->match_length > s->lookahead) s->match_length = s->lookahead;

            if (s->match_length <= 5 && (s->strategy == Z_FILTERED ||
                 (s->match_length == MIN_MATCH &&
                  s->strstart - s->match_start > TOO_FAR))) {

                /* If prev_match is also MIN_MATCH, match_start is garbage
                 * but we will ignore the current match anyway.
                 */
                s->match_length = MIN_MATCH-1;
            }
        }
        /* If there was a match at the previous step and the current
         * match is not better, output the previous match:
         */
        if (s->prev_length >= MIN_MATCH && s->match_length <= s->prev_length) {
            uInt max_insert = s->strstart + s->lookahead - MIN_MATCH;
            /* Do not insert strings in hash table beyond this. */

            check_match(s, s->strstart-1, s->prev_match, s->prev_length);

            bflush = ct_tally(s, s->strstart -1 - s->prev_match,
                              s->prev_length - MIN_MATCH);

            /* Insert in hash table all strings up to the end of the match.
             * strstart-1 and strstart are already inserted. If there is not
             * enough lookahead, the last two strings are not inserted in
             * the hash table.
             */
            s->lookahead -= s->prev_length-1;
            s->prev_length -= 2;
            do {
                if (++s->strstart <= max_insert) {
                    INSERT_STRING(s, s->strstart, hash_head);
                }
            } while (--s->prev_length != 0);
            s->match_available = 0;
            s->match_length = MIN_MATCH-1;
            s->strstart++;

            if (bflush) FLUSH_BLOCK(s, Z_NO_FLUSH);

        } else if (s->match_available) {
            /* If there was no match at the previous position, output a
             * single literal. If there was a match but the current match
             * is longer, truncate the previous match to a single literal.
             */
            Tracevv((stderr,"%c", s->window[s->strstart-1]));
            if (ct_tally (s, 0, s->window[s->strstart-1])) {
                FLUSH_BLOCK_ONLY(s, Z_NO_FLUSH);
            }
            s->strstart++;
            s->lookahead--;
            if (s->strm->avail_out == 0) return 1;
        } else {
            /* There is no previous match to compare with, wait for
             * the next step to decide.
             */
            s->match_available = 1;
            s->strstart++;
            s->lookahead--;
        }
    }
    Assert (flush != Z_NO_FLUSH, "no flush?");
    if (s->match_available) {
        Tracevv((stderr,"%c", s->window[s->strstart-1]));
        ct_tally (s, 0, s->window[s->strstart-1]);
        s->match_available = 0;
    }
    FLUSH_BLOCK(s, flush);
    return 0;
}


/*+++++*/
/* trees.c -- output deflated data using Huffman coding
 * Copyright (C) 1995 Jean-loup Gailly
 * For conditions of distribution and use, see copyright notice in zlib.h 
 */

/*
 *  ALGORITHM
 *
 *      The "deflation" process uses several Huffman trees. The more
 *      common source values are represented by shorter bit sequences.
 *
 *      Each code tree is stored in a compressed form which is itself
 * a Huffman encoding of the lengths of all the code strings (in
 * ascending order by source values).  The actual code strings are
 * reconstructed from the lengths in the inflate process, as described
 * in the deflate specification.
 *
 *  REFERENCES
 *
 *      Deutsch, L.P.,"'Deflate' Compressed Data Format Specification".
 *      Available in ftp.uu.net:/pub/archiving/zip/doc/deflate-1.1.doc
 *
 *      Storer, James A.
 *          Data Compression:  Methods and Theory, pp. 49-50.
 *          Computer Science Press, 1988.  ISBN 0-7167-8156-5.
 *
 *      Sedgewick, R.
 *          Algorithms, p290.
 *          Addison-Wesley, 1983. ISBN 0-201-06672-6.
 */

/* From: trees.c,v 1.5 1995/05/03 17:27:12 jloup Exp */

#ifdef DEBUG_ZLIB
#  include <ctype.h>
#endif