libguile/hash.c

   1 /*      Copyright (C) 1995,1996,1997, 2000, 2001, 2003, 2004, 2006, 2008, 2009, 2010 Free Software Foundation, Inc.
   2  *
   3  * This library is free software; you can redistribute it and/or
   4  * modify it under the terms of the GNU Lesser General Public License
   5  * as published by the Free Software Foundation; either version 3 of
   6  * the License, or (at your option) any later version.
   7  *
   8  * This library is distributed in the hope that it will be useful, but
   9  * WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  11  * Lesser General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU Lesser General Public
  14  * License along with this library; if not, write to the Free Software
  15  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  16  * 02110-1301 USA
  17  */
  18
  19
  20 \f
  21 #ifdef HAVE_CONFIG_H
  22 # include <config.h>
  23 #endif
  24
  25 #include "libguile/_scm.h"
  26 #include "libguile/chars.h"
  27 #include "libguile/ports.h"
  28 #include "libguile/strings.h"
  29 #include "libguile/symbols.h"
  30 #include "libguile/vectors.h"
  31
  32 #include "libguile/validate.h"
  33 #include "libguile/hash.h"
  34 \f
  35
  36 #ifndef floor
  37 extern double floor();
  38 #endif
  39
  40
  41 unsigned long
  42 scm_string_hash (const unsigned char *str, size_t len)
  43 {
  44   /* from suggestion at: */
  45   /* http://srfi.schemers.org/srfi-13/mail-archive/msg00112.html */
  46
  47   unsigned long h = 0;
  48   while (len-- > 0)
  49     h = *str++ + h*37;
  50   return h;
  51 }
  52
  53 unsigned long
  54 scm_i_string_hash (SCM str)
  55 {
  56   size_t len = scm_i_string_length (str);
  57   size_t i = 0;
  58
  59   unsigned long h = 0;
  60   while (len-- > 0)
  61     h = (unsigned long) scm_i_string_ref (str, i++) + h * 37;
  62
  63   scm_remember_upto_here_1 (str);
  64   return h;
  65 }
  66
  67
  68 /* Dirk:FIXME:: why downcase for characters? (2x: scm_hasher, scm_ihashv) */
  69 /* Dirk:FIXME:: scm_hasher could be made static. */
  70
  71
  72 unsigned long
  73 scm_hasher(SCM obj, unsigned long n, size_t d)
  74 {
  75   switch (SCM_ITAG3 (obj)) {
  76   case scm_tc3_int_1:
  77   case scm_tc3_int_2:
  78     return SCM_I_INUM(obj) % n;   /* SCM_INUMP(obj) */
  79   case scm_tc3_imm24:
  80     if (SCM_CHARP(obj))
  81       return (unsigned)(scm_c_downcase(SCM_CHAR(obj))) % n;
  82     switch (SCM_UNPACK (obj)) {
  83 #ifndef SICP
  84     case SCM_UNPACK(SCM_EOL):
  85       d = 256;
  86       break;
  87 #endif
  88     case SCM_UNPACK(SCM_BOOL_T):
  89       d = 257;
  90       break;
  91     case SCM_UNPACK(SCM_BOOL_F):
  92       d = 258;
  93       break;
  94     case SCM_UNPACK(SCM_EOF_VAL):
  95       d = 259;
  96       break;
  97     default:
  98       d = 263;          /* perhaps should be error */
  99     }
 100     return d % n;
 101   default:
 102     return 263 % n;     /* perhaps should be error */
 103   case scm_tc3_cons:
 104     switch SCM_TYP7(obj) {
 105     default:
 106       return 263 % n;
 107     case scm_tc7_smob:
 108       return 263 % n;
 109     case scm_tc7_number:
 110       switch SCM_TYP16 (obj) {
 111       case scm_tc16_big:
 112         return scm_to_ulong (scm_modulo (obj, scm_from_ulong (n)));
 113       case scm_tc16_real:
 114         {
 115           double r = SCM_REAL_VALUE (obj);
 116           if (floor (r) == r)
 117             {
 118               obj = scm_inexact_to_exact (obj);
 119               return scm_to_ulong (scm_modulo (obj, scm_from_ulong (n)));
 120             }
 121         }
 122         /* Fall through */
 123       case scm_tc16_complex:
 124       case scm_tc16_fraction:
 125         obj = scm_number_to_string (obj, scm_from_int (10));
 126         /* Fall through */
 127       }
 128       /* Fall through */
 129     case scm_tc7_string:
 130       {
 131         unsigned long hash =
 132           scm_i_string_hash (obj) % n;
 133         scm_remember_upto_here_1 (obj);
 134         return hash;
 135       }
 136     case scm_tc7_symbol:
 137       return scm_i_symbol_hash (obj) % n;
 138     case scm_tc7_wvect:
 139     case scm_tc7_vector:
 140       {
 141         size_t len = SCM_SIMPLE_VECTOR_LENGTH (obj);
 142         if (len > 5)
 143           {
 144             size_t i = d/2;
 145             unsigned long h = 1;
 146             while (i--)
 147               {
 148                 SCM elt = SCM_SIMPLE_VECTOR_REF (obj, h % len);
 149                 h = ((h << 8) + (scm_hasher (elt, n, 2))) % n;
 150               }
 151             return h;
 152           }
 153         else
 154           {
 155             size_t i = len;
 156             unsigned long h = (n)-1;
 157             while (i--)
 158               {
 159                 SCM elt = SCM_SIMPLE_VECTOR_REF (obj, h % len);
 160                 h = ((h << 8) + (scm_hasher (elt, n, d/len))) % n;
 161               }
 162             return h;
 163           }
 164       }
 165     case scm_tcs_cons_imcar:
 166     case scm_tcs_cons_nimcar:
 167       if (d) return (scm_hasher (SCM_CAR (obj), n, d/2)
 168                      + scm_hasher (SCM_CDR (obj), n, d/2)) % n;
 169       else return 1;
 170     case scm_tc7_port:
 171       return ((SCM_RDNG & SCM_CELL_WORD_0 (obj)) ? 260 : 261) % n;
 172     case scm_tc7_program:
 173       return 262 % n;
 174     }
 175   }
 176 }
 177
 178
 179 \f
 180
 181
 182 unsigned long
 183 scm_ihashq (SCM obj, unsigned long n)
 184 {
 185   return (SCM_UNPACK (obj) >> 1) % n;
 186 }
 187
 188
 189 SCM_DEFINE (scm_hashq, "hashq", 2, 0, 0,
 190            (SCM key, SCM size),
 191             "Determine a hash value for @var{key} that is suitable for\n"
 192             "lookups in a hashtable of size @var{size}, where @code{eq?} is\n"
 193             "used as the equality predicate.  The function returns an\n"
 194             "integer in the range 0 to @var{size} - 1.  Note that\n"
 195             "@code{hashq} may use internal addresses.  Thus two calls to\n"
 196             "hashq where the keys are @code{eq?} are not guaranteed to\n"
 197             "deliver the same value if the key object gets garbage collected\n"
 198             "in between.  This can happen, for example with symbols:\n"
 199             "@code{(hashq 'foo n) (gc) (hashq 'foo n)} may produce two\n"
 200             "different values, since @code{foo} will be garbage collected.")
 201 #define FUNC_NAME s_scm_hashq
 202 {
 203   unsigned long sz = scm_to_unsigned_integer (size, 1, ULONG_MAX);
 204   return scm_from_ulong (scm_ihashq (key, sz));
 205 }
 206 #undef FUNC_NAME
 207
 208
 209 \f
 210
 211
 212 unsigned long
 213 scm_ihashv (SCM obj, unsigned long n)
 214 {
 215   if (SCM_CHARP(obj))
 216     return ((unsigned long) (scm_c_downcase (SCM_CHAR (obj)))) % n; /* downcase!?!! */
 217
 218   if (SCM_NUMP(obj))
 219     return (unsigned long) scm_hasher(obj, n, 10);
 220   else
 221     return SCM_UNPACK (obj) % n;
 222 }
 223
 224
 225 SCM_DEFINE (scm_hashv, "hashv", 2, 0, 0,
 226            (SCM key, SCM size),
 227             "Determine a hash value for @var{key} that is suitable for\n"
 228             "lookups in a hashtable of size @var{size}, where @code{eqv?} is\n"
 229             "used as the equality predicate.  The function returns an\n"
 230             "integer in the range 0 to @var{size} - 1.  Note that\n"
 231             "@code{(hashv key)} may use internal addresses.  Thus two calls\n"
 232             "to hashv where the keys are @code{eqv?} are not guaranteed to\n"
 233             "deliver the same value if the key object gets garbage collected\n"
 234             "in between.  This can happen, for example with symbols:\n"
 235             "@code{(hashv 'foo n) (gc) (hashv 'foo n)} may produce two\n"
 236             "different values, since @code{foo} will be garbage collected.")
 237 #define FUNC_NAME s_scm_hashv
 238 {
 239   unsigned long sz = scm_to_unsigned_integer (size, 1, ULONG_MAX);
 240   return scm_from_ulong (scm_ihashv (key, sz));
 241 }
 242 #undef FUNC_NAME
 243
 244
 245 \f
 246
 247
 248 unsigned long
 249 scm_ihash (SCM obj, unsigned long n)
 250 {
 251   return (unsigned long) scm_hasher (obj, n, 10);
 252 }
 253
 254 SCM_DEFINE (scm_hash, "hash", 2, 0, 0,
 255            (SCM key, SCM size),
 256             "Determine a hash value for @var{key} that is suitable for\n"
 257             "lookups in a hashtable of size @var{size}, where @code{equal?}\n"
 258             "is used as the equality predicate.  The function returns an\n"
 259             "integer in the range 0 to @var{size} - 1.")
 260 #define FUNC_NAME s_scm_hash
 261 {
 262   unsigned long sz = scm_to_unsigned_integer (size, 1, ULONG_MAX);
 263   return scm_from_ulong (scm_ihash (key, sz));
 264 }
 265 #undef FUNC_NAME
 266
 267
 268 \f
 269
 270
 271 void
 272 scm_init_hash ()
 273 {
 274 #include "libguile/hash.x"
 275 }
 276
 277
 278 /*
 279   Local Variables:
 280   c-file-style: "gnu"
 281   End:
 282 */