libguile/symbols.c

   1 /* Copyright (C) 1995,1996,1997,1998,2000,2001, 2003, 2004, 2006 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
   5  * License as published by the Free Software Foundation; either
   6  * version 2.1 of the License, or (at your option) any later version.
   7  *
   8  * This library is distributed in the hope that it will be useful,
   9  * but 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 02110-1301 USA
  16  */
  17
  18
  19 \f
  20 #if HAVE_CONFIG_H
  21 #  include <config.h>
  22 #endif
  23
  24 #include "libguile/_scm.h"
  25 #include "libguile/chars.h"
  26 #include "libguile/eval.h"
  27 #include "libguile/hash.h"
  28 #include "libguile/smob.h"
  29 #include "libguile/variable.h"
  30 #include "libguile/alist.h"
  31 #include "libguile/fluids.h"
  32 #include "libguile/strings.h"
  33 #include "libguile/vectors.h"
  34 #include "libguile/hashtab.h"
  35 #include "libguile/weaks.h"
  36 #include "libguile/modules.h"
  37 #include "libguile/read.h"
  38 #include "libguile/srfi-13.h"
  39
  40 #include "libguile/validate.h"
  41 #include "libguile/symbols.h"
  42
  43 #ifdef HAVE_STRING_H
  44 #include <string.h>
  45 #endif
  46
  47 \f
  48
  49 static SCM symbols;
  50
  51 #ifdef GUILE_DEBUG
  52 SCM_DEFINE (scm_sys_symbols, "%symbols", 0, 0, 0,
  53             (),
  54             "Return the system symbol obarray.")
  55 #define FUNC_NAME s_scm_sys_symbols
  56 {
  57   return symbols;
  58 }
  59 #undef FUNC_NAME
  60 #endif
  61
  62 \f
  63
  64 /* {Symbols}
  65  */
  66
  67 /* In order to optimize reading speed, this function breaks part of
  68  * the hashtable abstraction.  The optimizations are:
  69  *
  70  * 1. The argument string can be compared directly to symbol objects
  71  *    without first creating an SCM string object.  (This would have
  72  *    been necessary if we had used the hashtable API in hashtab.h.)
  73  *
  74  * 2. We can use the raw hash value stored in scm_i_symbol_hash (sym)
  75  *    to speed up lookup.
  76  *
  77  * Both optimizations might be possible without breaking the
  78  * abstraction if the API in hashtab.c is improved.
  79  */
  80
  81 unsigned long
  82 scm_i_hash_symbol (SCM obj, unsigned long n, void *closure)
  83 {
  84   return scm_i_symbol_hash (obj) % n;
  85 }
  86
  87 static SCM
  88 lookup_interned_symbol (const char *name, size_t len,
  89                         unsigned long raw_hash)
  90 {
  91   /* Try to find the symbol in the symbols table */
  92   SCM result = SCM_BOOL_F;
  93   SCM bucket, elt, previous_elt;
  94   unsigned long hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
  95
  96   bucket = SCM_HASHTABLE_BUCKET (symbols, hash);
  97   for (elt = bucket, previous_elt = SCM_BOOL_F;
  98        !scm_is_null (elt);
  99        previous_elt = elt, elt = SCM_CDR (elt))
 100     {
 101       SCM pair, sym;
 102
 103       pair = SCM_CAR (elt);
 104       if (!scm_is_pair (pair))
 105         abort ();
 106
 107       if (SCM_WEAK_PAIR_CAR_DELETED_P (pair))
 108         {
 109           /* PAIR is a weak pair whose key got nullified: remove it from
 110              BUCKET.  */
 111           /* FIXME: Since this is done lazily, i.e., only when a new symbol
 112              is to be inserted in a bucket containing deleted symbols, the
 113              number of items in the hash table may remain erroneous for some
 114              time, thus precluding proper rehashing.  */
 115           if (previous_elt != SCM_BOOL_F)
 116             SCM_SETCDR (previous_elt, SCM_CDR (elt));
 117           else
 118             bucket = SCM_CDR (elt);
 119
 120           SCM_HASHTABLE_DECREMENT (symbols);
 121           continue;
 122         }
 123
 124       sym = SCM_CAR (pair);
 125
 126       if (scm_i_symbol_hash (sym) == raw_hash
 127           && scm_i_symbol_length (sym) == len)
 128         {
 129           const char *chrs = scm_i_symbol_chars (sym);
 130           size_t i = len;
 131
 132           while (i != 0)
 133             {
 134               --i;
 135               if (name[i] != chrs[i])
 136                 goto next_symbol;
 137             }
 138
 139           /* We found it.  */
 140           result = sym;
 141           break;
 142         }
 143     next_symbol:
 144       ;
 145     }
 146
 147   if (SCM_HASHTABLE_N_ITEMS (symbols) < SCM_HASHTABLE_LOWER (symbols))
 148     /* We removed many symbols in this pass so trigger a rehashing.  */
 149     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "lookup_interned_symbol");
 150
 151   return result;
 152 }
 153
 154 static SCM
 155 scm_i_c_mem2symbol (const char *name, size_t len)
 156 {
 157   SCM symbol;
 158   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 159   size_t hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
 160
 161   symbol = lookup_interned_symbol (name, len, raw_hash);
 162   if (symbol != SCM_BOOL_F)
 163     return symbol;
 164
 165   {
 166     /* The symbol was not found - create it. */
 167     SCM symbol = scm_i_c_make_symbol (name, len, 0, raw_hash,
 168                                       scm_cons (SCM_BOOL_F, SCM_EOL));
 169
 170     SCM slot = SCM_HASHTABLE_BUCKET (symbols, hash);
 171     SCM cell = scm_weak_car_pair (symbol, SCM_UNDEFINED);
 172     SCM_SET_HASHTABLE_BUCKET (symbols, hash, scm_cons (cell, slot));
 173     SCM_HASHTABLE_INCREMENT (symbols);
 174     if (SCM_HASHTABLE_N_ITEMS (symbols) > SCM_HASHTABLE_UPPER (symbols))
 175       scm_i_rehash (symbols, scm_i_hash_symbol, 0, "scm_mem2symbol");
 176
 177     return symbol;
 178   }
 179 }
 180
 181 static SCM
 182 scm_i_mem2symbol (SCM str)
 183 {
 184   SCM symbol;
 185   const char *name = scm_i_string_chars (str);
 186   size_t len = scm_i_string_length (str);
 187   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 188   size_t hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
 189
 190   symbol = lookup_interned_symbol (name, len, raw_hash);
 191   if (symbol != SCM_BOOL_F)
 192     return symbol;
 193
 194   {
 195     /* The symbol was not found - create it. */
 196     SCM symbol = scm_i_make_symbol (str, 0, raw_hash,
 197                                     scm_cons (SCM_BOOL_F, SCM_EOL));
 198
 199     SCM slot = SCM_HASHTABLE_BUCKET (symbols, hash);
 200     SCM cell = scm_weak_car_pair (symbol, SCM_UNDEFINED);
 201     SCM_SET_HASHTABLE_BUCKET (symbols, hash, scm_cons (cell, slot));
 202     SCM_HASHTABLE_INCREMENT (symbols);
 203     if (SCM_HASHTABLE_N_ITEMS (symbols) > SCM_HASHTABLE_UPPER (symbols))
 204       scm_i_rehash (symbols, scm_i_hash_symbol, 0, "scm_mem2symbol");
 205
 206     return symbol;
 207   }
 208 }
 209
 210
 211 static SCM
 212 scm_i_mem2uninterned_symbol (SCM str)
 213 {
 214   const char *name = scm_i_string_chars (str);
 215   size_t len = scm_i_string_length (str);
 216   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 217
 218   return scm_i_make_symbol (str, SCM_I_F_SYMBOL_UNINTERNED,
 219                             raw_hash, scm_cons (SCM_BOOL_F, SCM_EOL));
 220 }
 221
 222 SCM_DEFINE (scm_symbol_p, "symbol?", 1, 0, 0,
 223             (SCM obj),
 224             "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
 225             "@code{#f}.")
 226 #define FUNC_NAME s_scm_symbol_p
 227 {
 228   return scm_from_bool (scm_is_symbol (obj));
 229 }
 230 #undef FUNC_NAME
 231
 232 SCM_DEFINE (scm_symbol_interned_p, "symbol-interned?", 1, 0, 0,
 233             (SCM symbol),
 234             "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
 235             "@code{#f}.")
 236 #define FUNC_NAME s_scm_symbol_interned_p
 237 {
 238   SCM_VALIDATE_SYMBOL (1, symbol);
 239   return scm_from_bool (scm_i_symbol_is_interned (symbol));
 240 }
 241 #undef FUNC_NAME
 242
 243 SCM_DEFINE (scm_make_symbol, "make-symbol", 1, 0, 0,
 244             (SCM name),
 245             "Return a new uninterned symbol with the name @var{name}.  "
 246             "The returned symbol is guaranteed to be unique and future "
 247             "calls to @code{string->symbol} will not return it.")
 248 #define FUNC_NAME s_scm_make_symbol
 249 {
 250   SCM_VALIDATE_STRING (1, name);
 251   return scm_i_mem2uninterned_symbol (name);
 252 }
 253 #undef FUNC_NAME
 254
 255 SCM_DEFINE (scm_symbol_to_string, "symbol->string", 1, 0, 0,
 256            (SCM s),
 257             "Return the name of @var{symbol} as a string.  If the symbol was\n"
 258             "part of an object returned as the value of a literal expression\n"
 259             "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
 260             "Report on Scheme}) or by a call to the @code{read} procedure,\n"
 261             "and its name contains alphabetic characters, then the string\n"
 262             "returned will contain characters in the implementation's\n"
 263             "preferred standard case---some implementations will prefer\n"
 264             "upper case, others lower case.  If the symbol was returned by\n"
 265             "@code{string->symbol}, the case of characters in the string\n"
 266             "returned will be the same as the case in the string that was\n"
 267             "passed to @code{string->symbol}.  It is an error to apply\n"
 268             "mutation procedures like @code{string-set!} to strings returned\n"
 269             "by this procedure.\n"
 270             "\n"
 271             "The following examples assume that the implementation's\n"
 272             "standard case is lower case:\n"
 273             "\n"
 274             "@lisp\n"
 275             "(symbol->string 'flying-fish)    @result{} \"flying-fish\"\n"
 276             "(symbol->string 'Martin)         @result{}  \"martin\"\n"
 277             "(symbol->string\n"
 278             "   (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
 279             "@end lisp")
 280 #define FUNC_NAME s_scm_symbol_to_string
 281 {
 282   SCM_VALIDATE_SYMBOL (1, s);
 283   return scm_i_symbol_substring (s, 0, scm_i_symbol_length (s));
 284 }
 285 #undef FUNC_NAME
 286
 287
 288 SCM_DEFINE (scm_string_to_symbol, "string->symbol", 1, 0, 0,
 289             (SCM string),
 290             "Return the symbol whose name is @var{string}. This procedure\n"
 291             "can create symbols with names containing special characters or\n"
 292             "letters in the non-standard case, but it is usually a bad idea\n"
 293             "to create such symbols because in some implementations of\n"
 294             "Scheme they cannot be read as themselves.  See\n"
 295             "@code{symbol->string}.\n"
 296             "\n"
 297             "The following examples assume that the implementation's\n"
 298             "standard case is lower case:\n"
 299             "\n"
 300             "@lisp\n"
 301             "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
 302             "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
 303             "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
 304             "(eq? 'JollyWog\n"
 305             "  (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
 306             "(string=? \"K. Harper, M.D.\"\n"
 307             "  (symbol->string\n"
 308             "    (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
 309             "@end lisp")
 310 #define FUNC_NAME s_scm_string_to_symbol
 311 {
 312   SCM_VALIDATE_STRING (1, string);
 313   return scm_i_mem2symbol (string);
 314 }
 315 #undef FUNC_NAME
 316
 317 SCM_DEFINE (scm_string_ci_to_symbol, "string-ci->symbol", 1, 0, 0,
 318             (SCM str),
 319             "Return the symbol whose name is @var{str}.  @var{str} is\n"
 320             "converted to lowercase before the conversion is done, if Guile\n"
 321             "is currently reading symbols case-insensitively.")
 322 #define FUNC_NAME s_scm_string_ci_to_symbol
 323 {
 324   return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
 325                                ? scm_string_downcase(str)
 326                                : str);
 327 }
 328 #undef FUNC_NAME
 329
 330 #define MAX_PREFIX_LENGTH 30
 331
 332 SCM_DEFINE (scm_gensym, "gensym", 0, 1, 0,
 333             (SCM prefix),
 334             "Create a new symbol with a name constructed from a prefix and\n"
 335             "a counter value. The string @var{prefix} can be specified as\n"
 336             "an optional argument. Default prefix is @code{ g}.  The counter\n"
 337             "is increased by 1 at each call. There is no provision for\n"
 338             "resetting the counter.")
 339 #define FUNC_NAME s_scm_gensym
 340 {
 341   static int gensym_counter = 0;
 342
 343   SCM suffix, name;
 344   int n, n_digits;
 345   char buf[SCM_INTBUFLEN];
 346
 347   if (SCM_UNBNDP (prefix))
 348     prefix = scm_from_locale_string (" g");
 349
 350   /* mutex in case another thread looks and incs at the exact same moment */
 351   scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex);
 352   n = gensym_counter++;
 353   scm_i_pthread_mutex_unlock (&scm_i_misc_mutex);
 354
 355   n_digits = scm_iint2str (n, 10, buf);
 356   suffix = scm_from_locale_stringn (buf, n_digits);
 357   name = scm_string_append (scm_list_2 (prefix, suffix));
 358   return scm_string_to_symbol (name);
 359 }
 360 #undef FUNC_NAME
 361
 362 SCM_DEFINE (scm_symbol_hash, "symbol-hash", 1, 0, 0,
 363             (SCM symbol),
 364             "Return a hash value for @var{symbol}.")
 365 #define FUNC_NAME s_scm_symbol_hash
 366 {
 367   SCM_VALIDATE_SYMBOL (1, symbol);
 368   return scm_from_ulong (scm_i_symbol_hash (symbol));
 369 }
 370 #undef FUNC_NAME
 371
 372 SCM_DEFINE (scm_symbol_fref, "symbol-fref", 1, 0, 0,
 373            (SCM s),
 374             "Return the contents of @var{symbol}'s @dfn{function slot}.")
 375 #define FUNC_NAME s_scm_symbol_fref
 376 {
 377   SCM_VALIDATE_SYMBOL (1, s);
 378   return SCM_CAR (SCM_CELL_OBJECT_3 (s));
 379 }
 380 #undef FUNC_NAME
 381
 382
 383 SCM_DEFINE (scm_symbol_pref, "symbol-pref", 1, 0, 0,
 384            (SCM s),
 385             "Return the @dfn{property list} currently associated with @var{symbol}.")
 386 #define FUNC_NAME s_scm_symbol_pref
 387 {
 388   SCM_VALIDATE_SYMBOL (1, s);
 389   return SCM_CDR (SCM_CELL_OBJECT_3 (s));
 390 }
 391 #undef FUNC_NAME
 392
 393
 394 SCM_DEFINE (scm_symbol_fset_x, "symbol-fset!", 2, 0, 0,
 395            (SCM s, SCM val),
 396             "Change the binding of @var{symbol}'s function slot.")
 397 #define FUNC_NAME s_scm_symbol_fset_x
 398 {
 399   SCM_VALIDATE_SYMBOL (1, s);
 400   SCM_SETCAR (SCM_CELL_OBJECT_3 (s), val);
 401   return SCM_UNSPECIFIED;
 402 }
 403 #undef FUNC_NAME
 404
 405
 406 SCM_DEFINE (scm_symbol_pset_x, "symbol-pset!", 2, 0, 0,
 407            (SCM s, SCM val),
 408             "Change the binding of @var{symbol}'s property slot.")
 409 #define FUNC_NAME s_scm_symbol_pset_x
 410 {
 411   SCM_VALIDATE_SYMBOL (1, s);
 412   SCM_SETCDR (SCM_CELL_OBJECT_3 (s), val);
 413   return SCM_UNSPECIFIED;
 414 }
 415 #undef FUNC_NAME
 416
 417 SCM
 418 scm_from_locale_symbol (const char *sym)
 419 {
 420   return scm_i_c_mem2symbol (sym, strlen (sym));
 421 }
 422
 423 SCM
 424 scm_from_locale_symboln (const char *sym, size_t len)
 425 {
 426   return scm_i_c_mem2symbol (sym, len);
 427 }
 428
 429 SCM
 430 scm_take_locale_symboln (char *sym, size_t len)
 431 {
 432   SCM res;
 433   unsigned long raw_hash;
 434
 435   if (len == (size_t)-1)
 436     len = strlen (sym);
 437   else
 438     {
 439       /* Ensure STR is null terminated.  A realloc for 1 extra byte should
 440          often be satisfied from the alignment padding after the block, with
 441          no actual data movement.  */
 442       sym = scm_realloc (sym, len+1);
 443       sym[len] = '\0';
 444     }
 445
 446   raw_hash = scm_string_hash ((unsigned char *)sym, len);
 447   res = lookup_interned_symbol (sym, len, raw_hash);
 448   if (res != SCM_BOOL_F)
 449     {
 450       free (sym);
 451       return res;
 452     }
 453
 454   res = scm_i_c_take_symbol (sym, len, 0, raw_hash,
 455                              scm_cons (SCM_BOOL_F, SCM_EOL));
 456
 457   return res;
 458 }
 459
 460 SCM
 461 scm_take_locale_symbol (char *sym)
 462 {
 463   return scm_take_locale_symboln (sym, (size_t)-1);
 464 }
 465
 466 void
 467 scm_symbols_prehistory ()
 468 {
 469   symbols = scm_make_weak_key_hash_table (scm_from_int (2139));
 470 }
 471
 472
 473 void
 474 scm_init_symbols ()
 475 {
 476 #include "libguile/symbols.x"
 477 }
 478
 479 /*
 480   Local Variables:
 481   c-file-style: "gnu"
 482   End:
 483 */