src/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 #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */
  94 local int huft_build(uIntf* b, uInt n, uInt s, const uIntf* d, const uIntf* e,
  95                      inflate_huft* FAR *t, uIntf* m, inflate_huft* hp, uInt* hn, uIntf* v)
  96 #else
  97 local int huft_build(b, n, s, d, e, t, m, hp, hn, v)
  98 uIntf *b;               /* code lengths in bits (all assumed <= BMAX) */
  99 uInt n;                 /* number of codes (assumed <= 288) */
 100 uInt s;                 /* number of simple-valued codes (0..s-1) */
 101 const uIntf *d;         /* list of base values for non-simple codes */
 102 const uIntf *e;         /* list of extra bits for non-simple codes */
 103 inflate_huft * FAR *t;  /* result: starting table */
 104 uIntf *m;               /* maximum lookup bits, returns actual */
 105 inflate_huft *hp;       /* space for trees */
 106 uInt *hn;               /* hufts used in space */
 107 uIntf *v;               /* working area: values in order of bit length */
 108 #endif
 109 /* Given a list of code lengths and a maximum table size, make a set of
 110    tables to decode that set of codes.  Return Z_OK on success, Z_BUF_ERROR
 111    if the given code set is incomplete (the tables are still built in this
 112    case), or Z_DATA_ERROR if the input is invalid. */
 113 {
 114
 115   uInt a;                       /* counter for codes of length k */
 116   uInt c[BMAX+1];               /* bit length count table */
 117   uInt f;                       /* i repeats in table every f entries */
 118   int g;                        /* maximum code length */
 119   int h;                        /* table level */
 120   register uInt i;              /* counter, current code */
 121   register uInt j;              /* counter */
 122   register int k;               /* number of bits in current code */
 123   int l;                        /* bits per table (returned in m) */
 124   uInt mask;                    /* (1 << w) - 1, to avoid cc -O bug on HP */
 125   register uIntf *p;            /* pointer into c[], b[], or v[] */
 126   inflate_huft *q;              /* points to current table */
 127   struct inflate_huft_s r;      /* table entry for structure assignment */
 128   inflate_huft *u[BMAX];        /* table stack */
 129   register int w;               /* bits before this table == (l * h) */
 130   uInt x[BMAX+1];               /* bit offsets, then code stack */
 131   uIntf *xp;                    /* pointer into x */
 132   int y;                        /* number of dummy codes added */
 133   uInt z;                       /* number of entries in current table */
 134
 135
 136   /* Generate counts for each bit length */
 137   p = c;
 138 #define C0 *p++ = 0;
 139 #define C2 C0 C0 C0 C0
 140 #define C4 C2 C2 C2 C2
 141   C4                            /* clear c[]--assume BMAX+1 is 16 */
 142   p = b;  i = n;
 143   do {
 144     c[*p++]++;                  /* assume all entries <= BMAX */
 145   } while (--i);
 146   if (c[0] == n)                /* null input--all zero length codes */
 147   {
 148     *t = (inflate_huft *)Z_NULL;
 149     *m = 0;
 150     return Z_OK;
 151   }
 152
 153
 154   /* Find minimum and maximum length, bound *m by those */
 155   l = *m;
 156   for (j = 1; j <= BMAX; j++)
 157     if (c[j])
 158       break;
 159   k = j;                        /* minimum code length */
 160   if ((uInt)l < j)
 161     l = j;
 162   for (i = BMAX; i; i--)
 163     if (c[i])
 164       break;
 165   g = i;                        /* maximum code length */
 166   if ((uInt)l > i)
 167     l = i;
 168   *m = l;
 169
 170
 171   /* Adjust last length count to fill out codes, if needed */
 172   for (y = 1 << j; j < i; j++, y <<= 1)
 173     if ((y -= c[j]) < 0)
 174       return Z_DATA_ERROR;
 175   if ((y -= c[i]) < 0)
 176     return Z_DATA_ERROR;
 177   c[i] += y;
 178
 179
 180   /* Generate starting offsets into the value table for each length */
 181   x[1] = j = 0;
 182   p = c + 1;  xp = x + 2;
 183   while (--i) {                 /* note that i == g from above */
 184     *xp++ = (j += *p++);
 185   }
 186
 187
 188   /* Make a table of values in order of bit lengths */
 189   p = b;  i = 0;
 190   do {
 191     if ((j = *p++) != 0)
 192       v[x[j]++] = i;
 193   } while (++i < n);
 194   n = x[g];                     /* set n to length of v */
 195
 196
 197   /* Generate the Huffman codes and for each, make the table entries */
 198   x[0] = i = 0;                 /* first Huffman code is zero */
 199   p = v;                        /* grab values in bit order */
 200   h = -1;                       /* no tables yet--level -1 */
 201   w = -l;                       /* bits decoded == (l * h) */
 202   u[0] = (inflate_huft *)Z_NULL;        /* just to keep compilers happy */
 203   q = (inflate_huft *)Z_NULL;   /* ditto */
 204   z = 0;                        /* ditto */
 205
 206   /* go through the bit lengths (k already is bits in shortest code) */
 207   for (; k <= g; k++)
 208   {
 209     a = c[k];
 210     while (a--)
 211     {
 212       /* here i is the Huffman code of length k bits for value *p */
 213       /* make tables up to required level */
 214       while (k > w + l)
 215       {
 216         h++;
 217         w += l;                 /* previous table always l bits */
 218
 219         /* compute minimum size table less than or equal to l bits */
 220         z = g - w;
 221         z = z > (uInt)l ? l : z;        /* table size upper limit */
 222         if ((f = 1 << (j = k - w)) > a + 1)     /* try a k-w bit table */
 223         {                       /* too few codes for k-w bit table */
 224           f -= a + 1;           /* deduct codes from patterns left */
 225           xp = c + k;
 226           if (j < z)
 227             while (++j < z)     /* try smaller tables up to z bits */
 228             {
 229               if ((f <<= 1) <= *++xp)
 230                 break;          /* enough codes to use up j bits */
 231               f -= *xp;         /* else deduct codes from patterns */
 232             }
 233         }
 234         z = 1 << j;             /* table entries for j-bit table */
 235
 236         /* allocate new table */
 237         if (*hn + z > MANY)     /* (note: doesn't matter for fixed) */
 238           return Z_DATA_ERROR;  /* overflow of MANY */
 239         u[h] = q = hp + *hn;
 240         *hn += z;
 241
 242         /* connect to last table, if there is one */
 243         if (h)
 244         {
 245           x[h] = i;             /* save pattern for backing up */
 246           r.bits = (Byte)l;     /* bits to dump before this table */
 247           r.exop = (Byte)j;     /* bits in this table */
 248           j = i >> (w - l);
 249           r.base = (uInt)(q - u[h-1] - j);   /* offset to this table */
 250           u[h-1][j] = r;        /* connect to last table */
 251         }
 252         else
 253           *t = q;               /* first table is returned result */
 254       }
 255
 256       /* set up table entry in r */
 257       r.bits = (Byte)(k - w);
 258       if (p >= v + n)
 259         r.exop = 128 + 64;      /* out of values--invalid code */
 260       else if (*p < s)
 261       {
 262         r.exop = (Byte)(*p < 256 ? 0 : 32 + 64);     /* 256 is end-of-block */
 263         r.base = *p++;          /* simple code is just the value */
 264       }
 265       else
 266       {
 267         r.exop = (Byte)(e[*p - s] + 16 + 64);/* non-simple--look up in lists */
 268         r.base = d[*p++ - s];
 269       }
 270
 271       /* fill code-like entries with r */
 272       f = 1 << (k - w);
 273       for (j = i >> w; j < z; j += f)
 274         q[j] = r;
 275
 276       /* backwards increment the k-bit code i */
 277       for (j = 1 << (k - 1); i & j; j >>= 1)
 278         i ^= j;
 279       i ^= j;
 280
 281       /* backup over finished tables */
 282       mask = (1 << w) - 1;      /* needed on HP, cc -O bug */
 283       while ((i & mask) != x[h])
 284       {
 285         h--;                    /* don't need to update q */
 286         w -= l;
 287         mask = (1 << w) - 1;
 288       }
 289     }
 290   }
 291
 292
 293   /* Return Z_BUF_ERROR if we were given an incomplete table */
 294   return y != 0 && g != 1 ? Z_BUF_ERROR : Z_OK;
 295 }
 296
 297 #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */
 298 int inflate_trees_bits(uIntf* c, uIntf* bb, inflate_huft* FAR *tb, inflate_huft* hp, z_streamp z)
 299 #else
 300 int inflate_trees_bits(c, bb, tb, hp, z)
 301 uIntf *c;               /* 19 code lengths */
 302 uIntf *bb;              /* bits tree desired/actual depth */
 303 inflate_huft * FAR *tb; /* bits tree result */
 304 inflate_huft *hp;       /* space for trees */
 305 z_streamp z;            /* for messages */
 306 #endif
 307 {
 308   int r;
 309   uInt hn = 0;          /* hufts used in space */
 310   uIntf *v;             /* work area for huft_build */
 311
 312   if ((v = (uIntf*)ZALLOC(z, 19, sizeof(uInt))) == Z_NULL)
 313     return Z_MEM_ERROR;
 314   r = huft_build(c, 19, 19, (uIntf*)Z_NULL, (uIntf*)Z_NULL,
 315                  tb, bb, hp, &hn, v);
 316   if (r == Z_DATA_ERROR)
 317     z->msg = (char*)"oversubscribed dynamic bit lengths tree";
 318   else if (r == Z_BUF_ERROR || *bb == 0)
 319   {
 320     z->msg = (char*)"incomplete dynamic bit lengths tree";
 321     r = Z_DATA_ERROR;
 322   }
 323   ZFREE(z, v);
 324   return r;
 325 }
 326
 327 #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */
 328 int inflate_trees_dynamic(uInt nl, uInt nd, uInt* c, uInt* bl, uInt *bd, inflate_huft* FAR *tl,
 329                           inflate_huft* FAR *td, inflate_huft* hp, z_streamp z)
 330 #else
 331 int inflate_trees_dynamic(nl, nd, c, bl, bd, tl, td, hp, z)
 332 uInt nl;                /* number of literal/length codes */
 333 uInt nd;                /* number of distance codes */
 334 uIntf *c;               /* that many (total) code lengths */
 335 uIntf *bl;              /* literal desired/actual bit depth */
 336 uIntf *bd;              /* distance desired/actual bit depth */
 337 inflate_huft * FAR *tl; /* literal/length tree result */
 338 inflate_huft * FAR *td; /* distance tree result */
 339 inflate_huft *hp;       /* space for trees */
 340 z_streamp z;            /* for messages */
 341 #endif
 342 {
 343   int r;
 344   uInt hn = 0;          /* hufts used in space */
 345   uIntf *v;             /* work area for huft_build */
 346
 347   /* allocate work area */
 348   if ((v = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 349     return Z_MEM_ERROR;
 350
 351   /* build literal/length tree */
 352   r = huft_build(c, nl, 257, cplens, cplext, tl, bl, hp, &hn, v);
 353   if (r != Z_OK || *bl == 0)
 354   {
 355     if (r == Z_DATA_ERROR)
 356       z->msg = (char*)"oversubscribed literal/length tree";
 357     else if (r != Z_MEM_ERROR)
 358     {
 359       z->msg = (char*)"incomplete literal/length tree";
 360       r = Z_DATA_ERROR;
 361     }
 362     ZFREE(z, v);
 363     return r;
 364   }
 365
 366   /* build distance tree */
 367   r = huft_build(c + nl, nd, 0, cpdist, cpdext, td, bd, hp, &hn, v);
 368   if (r != Z_OK || (*bd == 0 && nl > 257))
 369   {
 370     if (r == Z_DATA_ERROR)
 371       z->msg = (char*)"oversubscribed distance tree";
 372     else if (r == Z_BUF_ERROR) {
 373 #ifdef PKZIP_BUG_WORKAROUND
 374       r = Z_OK;
 375     }
 376 #else
 377       z->msg = (char*)"incomplete distance tree";
 378       r = Z_DATA_ERROR;
 379     }
 380     else if (r != Z_MEM_ERROR)
 381     {
 382       z->msg = (char*)"empty distance tree with lengths";
 383       r = Z_DATA_ERROR;
 384     }
 385     ZFREE(z, v);
 386     return r;
 387 #endif
 388   }
 389
 390   /* done */
 391   ZFREE(z, v);
 392   return Z_OK;
 393 }
 394
 395 /* build fixed tables only once--keep them here */
 396 #ifdef BUILDFIXED
 397 local int fixed_built = 0;
 398 #define FIXEDH 544      /* number of hufts used by fixed tables */
 399 local inflate_huft fixed_mem[FIXEDH];
 400 local uInt fixed_bl;
 401 local uInt fixed_bd;
 402 local inflate_huft *fixed_tl;
 403 local inflate_huft *fixed_td;
 404 #else
 405 #include "inffixed.h"
 406 #endif
 407
 408 #if defined(__VISAGECPP__) /* Visualage can't handle this antiquated interface */
 409 int inflate_trees_fixed(uIntf* bl, uIntf *bd, inflate_huft* FAR *tl,
 410                         inflate_huft* FAR *td, z_streamp z)
 411 #else
 412 int inflate_trees_fixed(bl, bd, tl, td, z)
 413 uIntf *bl;               /* literal desired/actual bit depth */
 414 uIntf *bd;               /* distance desired/actual bit depth */
 415 inflate_huft * FAR *tl;  /* literal/length tree result */
 416 inflate_huft * FAR *td;  /* distance tree result */
 417 z_streamp z;             /* for memory allocation */
 418 #endif
 419 {
 420 #ifdef BUILDFIXED
 421   /* build fixed tables if not already */
 422   if (!fixed_built)
 423   {
 424     int k;              /* temporary variable */
 425     uInt f = 0;         /* number of hufts used in fixed_mem */
 426     uIntf *c;           /* length list for huft_build */
 427     uIntf *v;           /* work area for huft_build */
 428
 429     /* allocate memory */
 430     if ((c = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 431       return Z_MEM_ERROR;
 432     if ((v = (uIntf*)ZALLOC(z, 288, sizeof(uInt))) == Z_NULL)
 433     {
 434       ZFREE(z, c);
 435       return Z_MEM_ERROR;
 436     }
 437
 438     /* literal table */
 439     for (k = 0; k < 144; k++)
 440       c[k] = 8;
 441     for (; k < 256; k++)
 442       c[k] = 9;
 443     for (; k < 280; k++)
 444       c[k] = 7;
 445     for (; k < 288; k++)
 446       c[k] = 8;
 447     fixed_bl = 9;
 448     huft_build(c, 288, 257, cplens, cplext, &fixed_tl, &fixed_bl,
 449                fixed_mem, &f, v);
 450
 451     /* distance table */
 452     for (k = 0; k < 30; k++)
 453       c[k] = 5;
 454     fixed_bd = 5;
 455     huft_build(c, 30, 0, cpdist, cpdext, &fixed_td, &fixed_bd,
 456                fixed_mem, &f, v);
 457
 458     /* done */
 459     ZFREE(z, v);
 460     ZFREE(z, c);
 461     fixed_built = 1;
 462   }
 463 #endif
 464   *bl = fixed_bl;
 465   *bd = fixed_bd;
 466   *tl = fixed_tl;
 467   *td = fixed_td;
 468   return Z_OK;
 469 }