zlib/inftrees.c

   1 /* inftrees.c -- generate Huffman trees for efficient decoding
   2  * Copyright (C) 1995-2002 Mark Adler
   3  * For conditions of distribution and use, see copyright notice in zlib.h
   4  */
   5
   6 #include "zutil.h"
   7 #include "inftrees.h"
   8
   9 #if !defined(BUILDFIXED) && !defined(STDC)
  10 #  define BUILDFIXED   /* non ANSI compilers may not accept inffixed.h */
  11 #endif
  12
  13 const char inflate_copyright[] =
  14    " inflate 1.1.4 Copyright 1995-2002 Mark Adler ";
  15 /*
  16   If you use the zlib library in a product, an acknowledgment is welcome
  17   in the documentation of your product. If for some reason you cannot
  18   include such an acknowledgment, I would appreciate that you keep this
  19   copyright string in the executable of your product.
  20  */
  21 struct internal_state  {int dummy;}; /* for buggy compilers */
  22
  23 /* simplify the use of the inflate_huft type with some defines */
  24 #define exop word.what.Exop
  25 #define bits word.what.Bits
  26
  27
  28 local int huft_build OF((
  29     uIntf *,            /* code lengths in bits */
  30     uInt,               /* number of codes */
  31     uInt,               /* number of "simple" codes */
  32     const uIntf *,      /* list of base values for non-simple codes */
  33     const uIntf *,      /* list of extra bits for non-simple codes */
  34     inflate_huft * FAR*,/* result: starting table */
  35     uIntf *,            /* maximum lookup bits (returns actual) */
  36     inflate_huft *,     /* space for trees */
  37     uInt *,             /* hufts used in space */
  38     uIntf * ));         /* space for values */
  39
  40 /* Tables for deflate from PKZIP's appnote.txt. */
  41 local const uInt cplens[31] = { /* Copy lengths for literal codes 257..285 */
  42         3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31,
  43         35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0};
  44         /* see note #13 above about 258 */
  45 local const uInt cplext[31] = { /* Extra bits for literal codes 257..285 */
  46         0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2,
  47         3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0, 112, 112}; /* 112==invalid */
  48 local const uInt cpdist[30] = { /* Copy offsets for distance codes 0..29 */
  49         1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193,
  50         257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145,
  51         8193, 12289, 16385, 24577};
  52 local const uInt cpdext[30] = { /* Extra bits for distance codes */
  53         0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6,
  54         7, 7, 8, 8, 9, 9, 10, 10, 11, 11,
  55         12, 12, 13, 13};
  56
  57 /*
  58    Huffman code decoding is performed using a multi-level table lookup.
  59    The fastest way to decode is to simply build a lookup table whose
  60    size is determined by the longest code.  However, the time it takes
  61    to build this table can also be a factor if the data being decoded
  62    is not very long.  The most common codes are necessarily the
  63    shortest codes, so those codes dominate the decoding time, and hence
  64    the speed.  The idea is you can have a shorter table that decodes the
  65    shorter, more probable codes, and then point to subsidiary tables for
  66    the longer codes.  The time it costs to decode the longer codes is
  67    then traded against the time it takes to make longer tables.
  68
  69    This results of this trade are in the variables lbits and dbits
  70    below.  lbits is the number of bits the first level table for literal/
  71    length codes can decode in one step, and dbits is the same thing for
  72    the distance codes.  Subsequent tables are also less than or equal to
  73    those sizes.  These values may be adjusted either when all of the
  74    codes are shorter than that, in which case the longest code length in
  75    bits is used, or when the shortest code is *longer* than the requested
  76    table size, in which case the length of the shortest code in bits is
  77    used.
  78
  79    There are two different values for the two tables, since they code a
  80    different number of possibilities each.  The literal/length table
  81    codes 286 possible values, or in a flat code, a little over eight
  82    bits.  The distance table codes 30 possible values, or a little less
  83    than five bits, flat.  The optimum values for speed end up being
  84    about one bit more than those, so lbits is 8+1 and dbits is 5+1.
  85    The optimum values may differ though from machine to machine, and
  86    possibly even between compilers.  Your mileage may vary.
  87  */
  88
  89
  90 /* If BMAX needs to be larger than 16, then h and x[] should be uLong. */
  91 #define BMAX 15         /* maximum bit length of any code */
  92
  93 local int huft_build(b, n, s, d, e, t, m, hp, hn, v)
  94 uIntf *b;               /* code lengths in bits (all assumed <= BMAX) */
  95 uInt n;                 /* number of codes (assumed <= 288) */
  96 uInt s;                 /* number of simple-valued codes (0..s-1) */
  97 const uIntf *d;         /* list of base values for non-simple codes */
  98 const uIntf *e;         /* list of extra bits for non-simple codes */
  99 inflate_huft * FAR *t;  /* result: starting table */
 100 uIntf *m;               /* maximum lookup bits, returns actual */
 101 inflate_huft *hp;       /* space for trees */
 102 uInt *hn;               /* hufts used in space */
 103 uIntf *v;               /* working area: values in order of bit length */
 104 /* Given a list of code lengths and a maximum table size, make a set of
 105    tables to decode that set of codes.  Return Z_OK on success, Z_BUF_ERROR
 106    if the given code set is incomplete (the tables are still built in this
 107    case), or Z_DATA_ERROR if the input is invalid. */
 108 {
 109
 110   uInt a;                       /* counter for codes of length k */
 111   uInt c[BMAX+1];               /* bit length count table */
 112   uInt f;                       /* i repeats in table every f entries */
 113   int g;                        /* maximum code length */
 114   int h;                        /* table level */
 115   register uInt i;              /* counter, current code */
 116   register uInt j;              /* counter */
 117   register int k;               /* number of bits in current code */
 118   int l;                        /* bits per table (returned in m) */
 119   uInt mask;                    /* (1 << w) - 1, to avoid cc -O bug on HP */
 120   register uIntf *p;            /* pointer into c[], b[], or v[] */
 121   inflate_huft *q;              /* points to current table */
 122   struct inflate_huft_s r;      /* table entry for structure assignment */
 123   inflate_huft *u[BMAX];        /* table stack */
 124   register int w;               /* bits before this table == (l * h) */
 125   uInt x[BMAX+1];               /* bit offsets, then code stack */
 126   uIntf *xp;                    /* pointer into x */
 127   int y;                        /* number of dummy codes added */
 128   uInt z;                       /* number of entries in current table */
 129
 130
 131   /* Generate counts for each bit length */
 132   p = c;
 133 #define C0 *p++ = 0;
 134 #define C2 C0 C0 C0 C0
 135 #define C4 C2 C2 C2 C2
 136   C4                            /* clear c[]--assume BMAX+1 is 16 */
 137   p = b;  i = n;
 138   do {
 139     c[*p++]++;                  /* assume all entries <= BMAX */
 140   } while (--i);
 141   if (c[0] == n)                /* null input--all zero length codes */
 142   {
 143     *t = (inflate_huft *)Z_NULL;
 144     *m = 0;
 145     return Z_OK;
 146   }
 147
 148
 149   /* Find minimum and maximum length, bound *m by those */
 150   l = *m;
 151   for (j = 1; j <= BMAX; j++)
 152     if (c[j])
 153       break;
 154   k = j;                        /* minimum code length */
 155   if ((uInt)l < j)
 156     l = j;
 157   for (i = BMAX; i; i--)
 158     if (c[i])
 159       break;
 160   g = i;                        /* maximum code length */
 161   if ((uInt)l > i)
 162     l = i;
 163   *m = l;
 164
 165
 166   /* Adjust last length count to fill out codes, if needed */
 167   for (y = 1 << j; j < i; j++, y <<= 1)
 168     if ((y -= c[j]) < 0)
 169       return Z_DATA_ERROR;
 170   if ((y -= c[i]) < 0)
 171     return Z_DATA_ERROR;
 172   c[i] += y;
 173
 174
 175   /* Generate starting offsets into the value table for each length */
 176   x[1] = j = 0;
 177   p = c + 1;  xp = x + 2;
 178   while (--i) {                 /* note that i == g from above */
 179     *xp++ = (j += *p++);
 180   }
 181
 182
 183   /* Make a table of values in order of bit lengths */
 184   p = b;  i = 0;
 185   do {
 186     if ((j = *p++) != 0)
 187       v[x[j]++] = i;
 188   } while (++i < n);
 189   n = x[g];                     /* set n to length of v */
 190
 191
 192   /* Generate the Huffman codes and for each, make the table entries */
 193   x[0] = i = 0;                 /* first Huffman code is zero */
 194   p = v;                        /* grab values in bit order */
 195   h = -1;                       /* no tables yet--level -1 */
 196   w = -l;                       /* bits decoded == (l * h) */
 197   u[0] = (inflate_huft *)Z_NULL;        /* just to keep compilers happy */
 198   q = (inflate_huft *)Z_NULL;   /* ditto */
 199   z = 0;                        /* ditto */
 200
 201   /* go through the bit lengths (k already is bits in shortest code) */
 202   for (; k <= g; k++)
 203   {
 204     a = c[k];
 205     while (a--)
 206     {
 207       /* here i is the Huffman code of length k bits for value *p */
 208       /* make tables up to required level */
 209       while (k > w + l)
 210       {
 211         h++;
 212         w += l;                 /* previous table always l bits */
 213
 214         /* compute minimum size table less than or equal to l bits */
 215         z = g - w;
 216         z = z > (uInt)l ? l : z;        /* table size upper limit */
 217         if ((f = 1 << (j = k - w)) > a + 1)     /* try a k-w bit table */
 218         {                       /* too few codes for k-w bit table */
 219           f -= a + 1;           /* deduct codes from patterns left */
 220           xp = c + k;
 221           if (j < z)
 222             while (++j < z)     /* try smaller tables up to z bits */
 223             {
 224               if ((f <<= 1) <= *++xp)
 225                 break;          /* enough codes to use up j bits */
 226               f -= *xp;         /* else deduct codes from patterns */
 227             }
 228         }
 229         z = 1 << j;             /* table entries for j-bit table */
 230
 231         /* allocate new table */
 232         if (*hn + z > MANY)     /* (note: doesn't matter for fixed) */
 233           return Z_DATA_ERROR;  /* overflow of MANY */
 234         u[h] = q = hp + *hn;
 235         *hn += z;
 236
 237         /* connect to last table, if there is one */
 238         if (h)
 239         {
 240           x[h] = i;             /* save pattern for backing up */
 241           r.bits = (Byte)l;     /* bits to dump before this table */
 242           r.exop = (Byte)j;     /* bits in this table */
 243           j = i >> (w - l);
 244           r.base = (uInt)(q - u[h-1] - j);   /* offset to this table */
 245           u[h-1][j] = r;        /* connect to last table */
 246         }
 247         else
 248           *t = q;               /* first table is returned result */
 249       }
 250
 251       /* set up table entry in r */
 252       r.bits = (Byte)(k - w);
 253       if (p >= v + n)
 254         r.exop = 128 + 64;      /* out of values--invalid code */
 255       else if (*p < s)
 256       {
 257         r.exop = (Byte)(*p < 256 ? 0 : 32 + 64);     /* 256 is end-of-block */
 258         r.base = *p++;          /* simple code is just the value */
 259       }
 260       else
 261       {
 262         r.exop = (Byte)(e[*p - s] + 16 + 64);/* non-simple--look up in lists */
 263         r.base = d[*p++ - s];
 264       }
 265
 266       /* fill code-like entries with r */
 267       f = 1 << (k - w);
 268       for (j = i >> w; j < z; j += f)
 269         q[j] = r;
 270
 271       /* backwards increment the k-bit code i */
 272       for (j = 1 << (k - 1); i & j; j >>= 1)
 273         i ^= j;
 274       i ^= j;
 275
 276       /* backup over finished tables */
 277       mask = (1 << w) - 1;      /* needed on HP, cc -O bug */
 278       while ((i & mask) != x[h])
 279       {
 280         h--;                    /* don't need to update q */
 281         w -= l;
 282         mask = (1 << w) - 1;
 283       }
 284     }
 285   }
 286
 287
 288   /* Return Z_BUF_ERROR if we were given an incomplete table */
 289   return y != 0 && g != 1 ? Z_BUF_ERROR : Z_OK;
 290 }
 291
 292
 293 int inflate_trees_bits(c, bb, tb, hp, z)
 294 uIntf *c;               /* 19 code lengths */
 295 uIntf *bb;              /* bits tree desired/actual depth */
 296 inflate_huft * FAR *tb; /* bits tree result */
 297 inflate_huft *hp;       /* space for trees */
 298 z_streamp z;            /* for messages */
 299 {
 300   int r;
 301   uInt hn = 0;          /* hufts used in space */
 302   uIntf *v;             /* work area for huft_build */
 303
 304   if ((v = (uIntf*)ZALLOC(z, 19, sizeof(uInt))) == Z_NULL)
 305     return Z_MEM_ERROR;
 306   r = huft_build(c, 19, 19, (uIntf*)Z_NULL, (uIntf*)Z_NULL,
 307                  tb, bb, hp, &hn, v);
 308   if (r == Z_DATA_ERROR)
 309     z->msg = (char*)"oversubscribed dynamic bit lengths tree";
 310   else if (r == Z_BUF_ERROR || *bb == 0)
 311   {
 312     z->msg = (char*)"incomplete dynamic bit lengths tree";
 313     r = Z_DATA_ERROR;
 314   }
 315   ZFREE(z, v);
 316   return r;
 317 }
 318
 319
 320 int inflate_trees_dynamic(nl, nd, c, bl, bd, tl, td, hp, z)
 321 uInt nl;                /* number of literal/length codes */
 322 uInt nd;                /* number of distance codes */
 323 uIntf *c;               /* that many (total) code lengths */
 324 uIntf *bl;              /* literal desired/actual bit depth */
 325 uIntf *bd;              /* distance desired/actual bit depth */
 326 inflate_huft * FAR *tl; /* literal/length tree result */
 327 inflate_huft * FAR *td; /* distance tree result */
 328 inflate_huft *hp;       /* space for trees */
 329 z_streamp z;            /* for messages */
 330 {
 331   int r;
 332   uInt hn = 0;          /* hufts used in space */
 333   uIntf *v;             /* work area for huft_build */
 334
 335   /* allocate work area */
 336   if ((v = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 337     return Z_MEM_ERROR;
 338
 339   /* build literal/length tree */
 340   r = huft_build(c, nl, 257, cplens, cplext, tl, bl, hp, &hn, v);
 341   if (r != Z_OK || *bl == 0)
 342   {
 343     if (r == Z_DATA_ERROR)
 344       z->msg = (char*)"oversubscribed literal/length tree";
 345     else if (r != Z_MEM_ERROR)
 346     {
 347       z->msg = (char*)"incomplete literal/length tree";
 348       r = Z_DATA_ERROR;
 349     }
 350     ZFREE(z, v);
 351     return r;
 352   }
 353
 354   /* build distance tree */
 355   r = huft_build(c + nl, nd, 0, cpdist, cpdext, td, bd, hp, &hn, v);
 356   if (r != Z_OK || (*bd == 0 && nl > 257))
 357   {
 358     if (r == Z_DATA_ERROR)
 359       z->msg = (char*)"oversubscribed distance tree";
 360     else if (r == Z_BUF_ERROR) {
 361 #ifdef PKZIP_BUG_WORKAROUND
 362       r = Z_OK;
 363     }
 364 #else
 365       z->msg = (char*)"incomplete distance tree";
 366       r = Z_DATA_ERROR;
 367     }
 368     else if (r != Z_MEM_ERROR)
 369     {
 370       z->msg = (char*)"empty distance tree with lengths";
 371       r = Z_DATA_ERROR;
 372     }
 373     ZFREE(z, v);
 374     return r;
 375 #endif
 376   }
 377
 378   /* done */
 379   ZFREE(z, v);
 380   return Z_OK;
 381 }
 382
 383
 384 /* build fixed tables only once--keep them here */
 385 #ifdef BUILDFIXED
 386 local int fixed_built = 0;
 387 #define FIXEDH 544      /* number of hufts used by fixed tables */
 388 local inflate_huft fixed_mem[FIXEDH];
 389 local uInt fixed_bl;
 390 local uInt fixed_bd;
 391 local inflate_huft *fixed_tl;
 392 local inflate_huft *fixed_td;
 393 #else
 394 #include "inffixed.h"
 395 #endif
 396
 397
 398 int inflate_trees_fixed(bl, bd, tl, td, z)
 399 uIntf *bl;               /* literal desired/actual bit depth */
 400 uIntf *bd;               /* distance desired/actual bit depth */
 401 inflate_huft * FAR *tl;  /* literal/length tree result */
 402 inflate_huft * FAR *td;  /* distance tree result */
 403 z_streamp z;             /* for memory allocation */
 404 {
 405 #ifdef BUILDFIXED
 406   /* build fixed tables if not already */
 407   if (!fixed_built)
 408   {
 409     int k;              /* temporary variable */
 410     uInt f = 0;         /* number of hufts used in fixed_mem */
 411     uIntf *c;           /* length list for huft_build */
 412     uIntf *v;           /* work area for huft_build */
 413
 414     /* allocate memory */
 415     if ((c = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 416       return Z_MEM_ERROR;
 417     if ((v = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 418     {
 419       ZFREE(z, c);
 420       return Z_MEM_ERROR;
 421     }
 422
 423     /* literal table */
 424     for (k = 0; k < 144; k++)
 425       c[k] = 8;
 426     for (; k < 256; k++)
 427       c[k] = 9;
 428     for (; k < 280; k++)
 429       c[k] = 7;
 430     for (; k < 288; k++)
 431       c[k] = 8;
 432     fixed_bl = 9;
 433     huft_build(c, 288, 257, cplens, cplext, &fixed_tl, &fixed_bl,
 434                fixed_mem, &f, v);
 435
 436     /* distance table */
 437     for (k = 0; k < 30; k++)
 438       c[k] = 5;
 439     fixed_bd = 5;
 440     huft_build(c, 30, 0, cpdist, cpdext, &fixed_td, &fixed_bd,
 441                fixed_mem, &f, v);
 442
 443     /* done */
 444     ZFREE(z, v);
 445     ZFREE(z, c);
 446     fixed_built = 1;
 447   }
 448 #endif
 449   *bl = fixed_bl;
 450   *bd = fixed_bd;
 451   *tl = fixed_tl;
 452   *td = fixed_td;
 453   return Z_OK;
 454 }