Home | History | Annotate | Download | only in infback9
      1 /* inftree9.c -- generate Huffman trees for efficient decoding
      2  * Copyright (C) 1995-2010 Mark Adler
      3  * For conditions of distribution and use, see copyright notice in zlib.h
      4  */
      5 
      6 #include "zutil.h"
      7 #include "inftree9.h"
      8 
      9 #define MAXBITS 15
     10 
     11 const char inflate9_copyright[] =
     12    " inflate9 1.2.5 Copyright 1995-2010 Mark Adler ";
     13 /*
     14   If you use the zlib library in a product, an acknowledgment is welcome
     15   in the documentation of your product. If for some reason you cannot
     16   include such an acknowledgment, I would appreciate that you keep this
     17   copyright string in the executable of your product.
     18  */
     19 
     20 /*
     21    Build a set of tables to decode the provided canonical Huffman code.
     22    The code lengths are lens[0..codes-1].  The result starts at *table,
     23    whose indices are 0..2^bits-1.  work is a writable array of at least
     24    lens shorts, which is used as a work area.  type is the type of code
     25    to be generated, CODES, LENS, or DISTS.  On return, zero is success,
     26    -1 is an invalid code, and +1 means that ENOUGH isn't enough.  table
     27    on return points to the next available entry's address.  bits is the
     28    requested root table index bits, and on return it is the actual root
     29    table index bits.  It will differ if the request is greater than the
     30    longest code or if it is less than the shortest code.
     31  */
     32 int inflate_table9(type, lens, codes, table, bits, work)
     33 codetype type;
     34 unsigned short FAR *lens;
     35 unsigned codes;
     36 code FAR * FAR *table;
     37 unsigned FAR *bits;
     38 unsigned short FAR *work;
     39 {
     40     unsigned len;               /* a code's length in bits */
     41     unsigned sym;               /* index of code symbols */
     42     unsigned min, max;          /* minimum and maximum code lengths */
     43     unsigned root;              /* number of index bits for root table */
     44     unsigned curr;              /* number of index bits for current table */
     45     unsigned drop;              /* code bits to drop for sub-table */
     46     int left;                   /* number of prefix codes available */
     47     unsigned used;              /* code entries in table used */
     48     unsigned huff;              /* Huffman code */
     49     unsigned incr;              /* for incrementing code, index */
     50     unsigned fill;              /* index for replicating entries */
     51     unsigned low;               /* low bits for current root entry */
     52     unsigned mask;              /* mask for low root bits */
     53     code this;                  /* table entry for duplication */
     54     code FAR *next;             /* next available space in table */
     55     const unsigned short FAR *base;     /* base value table to use */
     56     const unsigned short FAR *extra;    /* extra bits table to use */
     57     int end;                    /* use base and extra for symbol > end */
     58     unsigned short count[MAXBITS+1];    /* number of codes of each length */
     59     unsigned short offs[MAXBITS+1];     /* offsets in table for each length */
     60     static const unsigned short lbase[31] = { /* Length codes 257..285 base */
     61         3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17,
     62         19, 23, 27, 31, 35, 43, 51, 59, 67, 83, 99, 115,
     63         131, 163, 195, 227, 3, 0, 0};
     64     static const unsigned short lext[31] = { /* Length codes 257..285 extra */
     65         128, 128, 128, 128, 128, 128, 128, 128, 129, 129, 129, 129,
     66         130, 130, 130, 130, 131, 131, 131, 131, 132, 132, 132, 132,
     67         133, 133, 133, 133, 144, 73, 195};
     68     static const unsigned short dbase[32] = { /* Distance codes 0..31 base */
     69         1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49,
     70         65, 97, 129, 193, 257, 385, 513, 769, 1025, 1537, 2049, 3073,
     71         4097, 6145, 8193, 12289, 16385, 24577, 32769, 49153};
     72     static const unsigned short dext[32] = { /* Distance codes 0..31 extra */
     73         128, 128, 128, 128, 129, 129, 130, 130, 131, 131, 132, 132,
     74         133, 133, 134, 134, 135, 135, 136, 136, 137, 137, 138, 138,
     75         139, 139, 140, 140, 141, 141, 142, 142};
     76 
     77     /*
     78        Process a set of code lengths to create a canonical Huffman code.  The
     79        code lengths are lens[0..codes-1].  Each length corresponds to the
     80        symbols 0..codes-1.  The Huffman code is generated by first sorting the
     81        symbols by length from short to long, and retaining the symbol order
     82        for codes with equal lengths.  Then the code starts with all zero bits
     83        for the first code of the shortest length, and the codes are integer
     84        increments for the same length, and zeros are appended as the length
     85        increases.  For the deflate format, these bits are stored backwards
     86        from their more natural integer increment ordering, and so when the
     87        decoding tables are built in the large loop below, the integer codes
     88        are incremented backwards.
     89 
     90        This routine assumes, but does not check, that all of the entries in
     91        lens[] are in the range 0..MAXBITS.  The caller must assure this.
     92        1..MAXBITS is interpreted as that code length.  zero means that that
     93        symbol does not occur in this code.
     94 
     95        The codes are sorted by computing a count of codes for each length,
     96        creating from that a table of starting indices for each length in the
     97        sorted table, and then entering the symbols in order in the sorted
     98        table.  The sorted table is work[], with that space being provided by
     99        the caller.
    100 
    101        The length counts are used for other purposes as well, i.e. finding
    102        the minimum and maximum length codes, determining if there are any
    103        codes at all, checking for a valid set of lengths, and looking ahead
    104        at length counts to determine sub-table sizes when building the
    105        decoding tables.
    106      */
    107 
    108     /* accumulate lengths for codes (assumes lens[] all in 0..MAXBITS) */
    109     for (len = 0; len <= MAXBITS; len++)
    110         count[len] = 0;
    111     for (sym = 0; sym < codes; sym++)
    112         count[lens[sym]]++;
    113 
    114     /* bound code lengths, force root to be within code lengths */
    115     root = *bits;
    116     for (max = MAXBITS; max >= 1; max--)
    117         if (count[max] != 0) break;
    118     if (root > max) root = max;
    119     if (max == 0) return -1;            /* no codes! */
    120     for (min = 1; min <= MAXBITS; min++)
    121         if (count[min] != 0) break;
    122     if (root < min) root = min;
    123 
    124     /* check for an over-subscribed or incomplete set of lengths */
    125     left = 1;
    126     for (len = 1; len <= MAXBITS; len++) {
    127         left <<= 1;
    128         left -= count[len];
    129         if (left < 0) return -1;        /* over-subscribed */
    130     }
    131     if (left > 0 && (type == CODES || max != 1))
    132         return -1;                      /* incomplete set */
    133 
    134     /* generate offsets into symbol table for each length for sorting */
    135     offs[1] = 0;
    136     for (len = 1; len < MAXBITS; len++)
    137         offs[len + 1] = offs[len] + count[len];
    138 
    139     /* sort symbols by length, by symbol order within each length */
    140     for (sym = 0; sym < codes; sym++)
    141         if (lens[sym] != 0) work[offs[lens[sym]]++] = (unsigned short)sym;
    142 
    143     /*
    144        Create and fill in decoding tables.  In this loop, the table being
    145        filled is at next and has curr index bits.  The code being used is huff
    146        with length len.  That code is converted to an index by dropping drop
    147        bits off of the bottom.  For codes where len is less than drop + curr,
    148        those top drop + curr - len bits are incremented through all values to
    149        fill the table with replicated entries.
    150 
    151        root is the number of index bits for the root table.  When len exceeds
    152        root, sub-tables are created pointed to by the root entry with an index
    153        of the low root bits of huff.  This is saved in low to check for when a
    154        new sub-table should be started.  drop is zero when the root table is
    155        being filled, and drop is root when sub-tables are being filled.
    156 
    157        When a new sub-table is needed, it is necessary to look ahead in the
    158        code lengths to determine what size sub-table is needed.  The length
    159        counts are used for this, and so count[] is decremented as codes are
    160        entered in the tables.
    161 
    162        used keeps track of how many table entries have been allocated from the
    163        provided *table space.  It is checked for LENS and DIST tables against
    164        the constants ENOUGH_LENS and ENOUGH_DISTS to guard against changes in
    165        the initial root table size constants.  See the comments in inftree9.h
    166        for more information.
    167 
    168        sym increments through all symbols, and the loop terminates when
    169        all codes of length max, i.e. all codes, have been processed.  This
    170        routine permits incomplete codes, so another loop after this one fills
    171        in the rest of the decoding tables with invalid code markers.
    172      */
    173 
    174     /* set up for code type */
    175     switch (type) {
    176     case CODES:
    177         base = extra = work;    /* dummy value--not used */
    178         end = 19;
    179         break;
    180     case LENS:
    181         base = lbase;
    182         base -= 257;
    183         extra = lext;
    184         extra -= 257;
    185         end = 256;
    186         break;
    187     default:            /* DISTS */
    188         base = dbase;
    189         extra = dext;
    190         end = -1;
    191     }
    192 
    193     /* initialize state for loop */
    194     huff = 0;                   /* starting code */
    195     sym = 0;                    /* starting code symbol */
    196     len = min;                  /* starting code length */
    197     next = *table;              /* current table to fill in */
    198     curr = root;                /* current table index bits */
    199     drop = 0;                   /* current bits to drop from code for index */
    200     low = (unsigned)(-1);       /* trigger new sub-table when len > root */
    201     used = 1U << root;          /* use root table entries */
    202     mask = used - 1;            /* mask for comparing low */
    203 
    204     /* check available table space */
    205     if ((type == LENS && used >= ENOUGH_LENS) ||
    206         (type == DISTS && used >= ENOUGH_DISTS))
    207         return 1;
    208 
    209     /* process all codes and make table entries */
    210     for (;;) {
    211         /* create table entry */
    212         this.bits = (unsigned char)(len - drop);
    213         if ((int)(work[sym]) < end) {
    214             this.op = (unsigned char)0;
    215             this.val = work[sym];
    216         }
    217         else if ((int)(work[sym]) > end) {
    218             this.op = (unsigned char)(extra[work[sym]]);
    219             this.val = base[work[sym]];
    220         }
    221         else {
    222             this.op = (unsigned char)(32 + 64);         /* end of block */
    223             this.val = 0;
    224         }
    225 
    226         /* replicate for those indices with low len bits equal to huff */
    227         incr = 1U << (len - drop);
    228         fill = 1U << curr;
    229         do {
    230             fill -= incr;
    231             next[(huff >> drop) + fill] = this;
    232         } while (fill != 0);
    233 
    234         /* backwards increment the len-bit code huff */
    235         incr = 1U << (len - 1);
    236         while (huff & incr)
    237             incr >>= 1;
    238         if (incr != 0) {
    239             huff &= incr - 1;
    240             huff += incr;
    241         }
    242         else
    243             huff = 0;
    244 
    245         /* go to next symbol, update count, len */
    246         sym++;
    247         if (--(count[len]) == 0) {
    248             if (len == max) break;
    249             len = lens[work[sym]];
    250         }
    251 
    252         /* create new sub-table if needed */
    253         if (len > root && (huff & mask) != low) {
    254             /* if first time, transition to sub-tables */
    255             if (drop == 0)
    256                 drop = root;
    257 
    258             /* increment past last table */
    259             next += 1U << curr;
    260 
    261             /* determine length of next table */
    262             curr = len - drop;
    263             left = (int)(1 << curr);
    264             while (curr + drop < max) {
    265                 left -= count[curr + drop];
    266                 if (left <= 0) break;
    267                 curr++;
    268                 left <<= 1;
    269             }
    270 
    271             /* check for enough space */
    272             used += 1U << curr;
    273             if ((type == LENS && used >= ENOUGH_LENS) ||
    274                 (type == DISTS && used >= ENOUGH_DISTS))
    275                 return 1;
    276 
    277             /* point entry in root table to sub-table */
    278             low = huff & mask;
    279             (*table)[low].op = (unsigned char)curr;
    280             (*table)[low].bits = (unsigned char)root;
    281             (*table)[low].val = (unsigned short)(next - *table);
    282         }
    283     }
    284 
    285     /*
    286        Fill in rest of table for incomplete codes.  This loop is similar to the
    287        loop above in incrementing huff for table indices.  It is assumed that
    288        len is equal to curr + drop, so there is no loop needed to increment
    289        through high index bits.  When the current sub-table is filled, the loop
    290        drops back to the root table to fill in any remaining entries there.
    291      */
    292     this.op = (unsigned char)64;                /* invalid code marker */
    293     this.bits = (unsigned char)(len - drop);
    294     this.val = (unsigned short)0;
    295     while (huff != 0) {
    296         /* when done with sub-table, drop back to root table */
    297         if (drop != 0 && (huff & mask) != low) {
    298             drop = 0;
    299             len = root;
    300             next = *table;
    301             curr = root;
    302             this.bits = (unsigned char)len;
    303         }
    304 
    305         /* put invalid code marker in table */
    306         next[huff >> drop] = this;
    307 
    308         /* backwards increment the len-bit code huff */
    309         incr = 1U << (len - 1);
    310         while (huff & incr)
    311             incr >>= 1;
    312         if (incr != 0) {
    313             huff &= incr - 1;
    314             huff += incr;
    315         }
    316         else
    317             huff = 0;
    318     }
    319 
    320     /* set return parameters */
    321     *table += used;
    322     *bits = root;
    323     return 0;
    324 }
    325