libguile/symbols.c

   1 /* Copyright (C) 1995,1996,1997,1998,2000,2001, 2003, 2004, 2006, 2009 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/eval.h"
  28 #include "libguile/hash.h"
  29 #include "libguile/smob.h"
  30 #include "libguile/variable.h"
  31 #include "libguile/alist.h"
  32 #include "libguile/fluids.h"
  33 #include "libguile/strings.h"
  34 #include "libguile/vectors.h"
  35 #include "libguile/hashtab.h"
  36 #include "libguile/weaks.h"
  37 #include "libguile/modules.h"
  38 #include "libguile/read.h"
  39 #include "libguile/srfi-13.h"
  40
  41 #include "libguile/validate.h"
  42 #include "libguile/symbols.h"
  43
  44 #include "libguile/private-options.h"
  45
  46
  47 #ifdef HAVE_STRING_H
  48 #include <string.h>
  49 #endif
  50
  51 \f
  52
  53 static SCM symbols;
  54
  55 #ifdef GUILE_DEBUG
  56 SCM_DEFINE (scm_sys_symbols, "%symbols", 0, 0, 0,
  57             (),
  58             "Return the system symbol obarray.")
  59 #define FUNC_NAME s_scm_sys_symbols
  60 {
  61   return symbols;
  62 }
  63 #undef FUNC_NAME
  64 #endif
  65
  66 \f
  67
  68 /* {Symbols}
  69  */
  70
  71 /* In order to optimize reading speed, this function breaks part of
  72  * the hashtable abstraction.  The optimizations are:
  73  *
  74  * 1. The argument string can be compared directly to symbol objects
  75  *    without first creating an SCM string object.  (This would have
  76  *    been necessary if we had used the hashtable API in hashtab.h.)
  77  *
  78  * 2. We can use the raw hash value stored in scm_i_symbol_hash (sym)
  79  *    to speed up lookup.
  80  *
  81  * Both optimizations might be possible without breaking the
  82  * abstraction if the API in hashtab.c is improved.
  83  */
  84
  85 unsigned long
  86 scm_i_hash_symbol (SCM obj, unsigned long n, void *closure)
  87 {
  88   return scm_i_symbol_hash (obj) % n;
  89 }
  90
  91 static SCM
  92 lookup_interned_symbol (SCM name, unsigned long raw_hash)
  93 {
  94   /* Try to find the symbol in the symbols table */
  95   SCM result = SCM_BOOL_F;
  96   SCM bucket, elt, previous_elt;
  97   size_t len;
  98   unsigned long hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
  99
 100   len = scm_i_string_length (name);
 101   bucket = SCM_HASHTABLE_BUCKET (symbols, hash);
 102
 103   for (elt = bucket, previous_elt = SCM_BOOL_F;
 104        !scm_is_null (elt);
 105        previous_elt = elt, elt = SCM_CDR (elt))
 106     {
 107       SCM pair, sym;
 108
 109       pair = SCM_CAR (elt);
 110       if (!scm_is_pair (pair))
 111         abort ();
 112
 113       if (SCM_WEAK_PAIR_CAR_DELETED_P (pair))
 114         {
 115           /* PAIR is a weak pair whose key got nullified: remove it from
 116              BUCKET.  */
 117           /* FIXME: Since this is done lazily, i.e., only when a new symbol
 118              is to be inserted in a bucket containing deleted symbols, the
 119              number of items in the hash table may remain erroneous for some
 120              time, thus precluding proper rehashing.  */
 121           if (previous_elt != SCM_BOOL_F)
 122             SCM_SETCDR (previous_elt, SCM_CDR (elt));
 123           else
 124             bucket = SCM_CDR (elt);
 125
 126           SCM_HASHTABLE_DECREMENT (symbols);
 127           continue;
 128         }
 129
 130       sym = SCM_CAR (pair);
 131
 132       if (scm_i_symbol_hash (sym) == raw_hash
 133           && scm_i_symbol_length (sym) == len)
 134         {
 135           size_t i = len;
 136
 137           /* Slightly faster path for comparing narrow to narrow.  */
 138           if (scm_i_is_narrow_string (name) && scm_i_is_narrow_symbol (sym))
 139             {
 140               const char *chrs = scm_i_symbol_chars (sym);
 141               const char *str = scm_i_string_chars (name);
 142
 143               while (i != 0)
 144                 {
 145                   --i;
 146                   if (str[i] != chrs[i])
 147                     goto next_symbol;
 148                 }
 149             }
 150           else
 151             {
 152               /* Somewhat slower path for comparing narrow to wide or
 153                  wide to wide.  */
 154               while (i != 0)
 155                 {
 156                   --i;
 157                   if (scm_i_string_ref (name, i) != scm_i_symbol_ref (sym, i))
 158                     goto next_symbol;
 159                 }
 160             }
 161
 162           /* We found it.  */
 163           result = sym;
 164           break;
 165         }
 166     next_symbol:
 167       ;
 168     }
 169
 170   if (SCM_HASHTABLE_N_ITEMS (symbols) < SCM_HASHTABLE_LOWER (symbols))
 171     /* We removed many symbols in this pass so trigger a rehashing.  */
 172     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "lookup_interned_symbol");
 173
 174   return result;
 175 }
 176
 177 /* Intern SYMBOL, an uninterned symbol.  */
 178 static void
 179 intern_symbol (SCM symbol)
 180 {
 181   SCM slot, cell;
 182   unsigned long hash;
 183
 184   hash = scm_i_symbol_hash (symbol) % SCM_HASHTABLE_N_BUCKETS (symbols);
 185   slot = SCM_HASHTABLE_BUCKET (symbols, hash);
 186   cell = scm_cons (symbol, SCM_UNDEFINED);
 187
 188   SCM_SET_HASHTABLE_BUCKET (symbols, hash, scm_cons (cell, slot));
 189   SCM_HASHTABLE_INCREMENT (symbols);
 190
 191   if (SCM_HASHTABLE_N_ITEMS (symbols) > SCM_HASHTABLE_UPPER (symbols))
 192     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "intern_symbol");
 193 }
 194
 195 static SCM
 196 scm_i_str2symbol (SCM str)
 197 {
 198   SCM symbol;
 199   size_t raw_hash = scm_i_string_hash (str);
 200
 201   symbol = lookup_interned_symbol (str, raw_hash);
 202   if (scm_is_false (symbol))
 203     {
 204       /* The symbol was not found, create it.  */
 205       symbol = scm_i_make_symbol (str, 0, raw_hash,
 206                                   scm_cons (SCM_BOOL_F, SCM_EOL));
 207       intern_symbol (symbol);
 208     }
 209
 210   return symbol;
 211 }
 212
 213
 214 static SCM
 215 scm_i_str2uninterned_symbol (SCM str)
 216 {
 217   size_t raw_hash = scm_i_string_hash (str);
 218
 219   return scm_i_make_symbol (str, SCM_I_F_SYMBOL_UNINTERNED,
 220                             raw_hash, scm_cons (SCM_BOOL_F, SCM_EOL));
 221 }
 222
 223 SCM_DEFINE (scm_symbol_p, "symbol?", 1, 0, 0,
 224             (SCM obj),
 225             "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
 226             "@code{#f}.")
 227 #define FUNC_NAME s_scm_symbol_p
 228 {
 229   return scm_from_bool (scm_is_symbol (obj));
 230 }
 231 #undef FUNC_NAME
 232
 233 SCM_DEFINE (scm_symbol_interned_p, "symbol-interned?", 1, 0, 0,
 234             (SCM symbol),
 235             "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
 236             "@code{#f}.")
 237 #define FUNC_NAME s_scm_symbol_interned_p
 238 {
 239   SCM_VALIDATE_SYMBOL (1, symbol);
 240   return scm_from_bool (scm_i_symbol_is_interned (symbol));
 241 }
 242 #undef FUNC_NAME
 243
 244 SCM_DEFINE (scm_make_symbol, "make-symbol", 1, 0, 0,
 245             (SCM name),
 246             "Return a new uninterned symbol with the name @var{name}.  "
 247             "The returned symbol is guaranteed to be unique and future "
 248             "calls to @code{string->symbol} will not return it.")
 249 #define FUNC_NAME s_scm_make_symbol
 250 {
 251   SCM_VALIDATE_STRING (1, name);
 252   return scm_i_str2uninterned_symbol (name);
 253 }
 254 #undef FUNC_NAME
 255
 256 SCM_DEFINE (scm_symbol_to_string, "symbol->string", 1, 0, 0,
 257            (SCM s),
 258             "Return the name of @var{symbol} as a string.  If the symbol was\n"
 259             "part of an object returned as the value of a literal expression\n"
 260             "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
 261             "Report on Scheme}) or by a call to the @code{read} procedure,\n"
 262             "and its name contains alphabetic characters, then the string\n"
 263             "returned will contain characters in the implementation's\n"
 264             "preferred standard case---some implementations will prefer\n"
 265             "upper case, others lower case.  If the symbol was returned by\n"
 266             "@code{string->symbol}, the case of characters in the string\n"
 267             "returned will be the same as the case in the string that was\n"
 268             "passed to @code{string->symbol}.  It is an error to apply\n"
 269             "mutation procedures like @code{string-set!} to strings returned\n"
 270             "by this procedure.\n"
 271             "\n"
 272             "The following examples assume that the implementation's\n"
 273             "standard case is lower case:\n"
 274             "\n"
 275             "@lisp\n"
 276             "(symbol->string 'flying-fish)    @result{} \"flying-fish\"\n"
 277             "(symbol->string 'Martin)         @result{}  \"martin\"\n"
 278             "(symbol->string\n"
 279             "   (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
 280             "@end lisp")
 281 #define FUNC_NAME s_scm_symbol_to_string
 282 {
 283   SCM_VALIDATE_SYMBOL (1, s);
 284   return scm_i_symbol_substring (s, 0, scm_i_symbol_length (s));
 285 }
 286 #undef FUNC_NAME
 287
 288
 289 SCM_DEFINE (scm_string_to_symbol, "string->symbol", 1, 0, 0,
 290             (SCM string),
 291             "Return the symbol whose name is @var{string}. This procedure\n"
 292             "can create symbols with names containing special characters or\n"
 293             "letters in the non-standard case, but it is usually a bad idea\n"
 294             "to create such symbols because in some implementations of\n"
 295             "Scheme they cannot be read as themselves.  See\n"
 296             "@code{symbol->string}.\n"
 297             "\n"
 298             "The following examples assume that the implementation's\n"
 299             "standard case is lower case:\n"
 300             "\n"
 301             "@lisp\n"
 302             "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
 303             "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
 304             "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
 305             "(eq? 'JollyWog\n"
 306             "  (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
 307             "(string=? \"K. Harper, M.D.\"\n"
 308             "  (symbol->string\n"
 309             "    (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
 310             "@end lisp")
 311 #define FUNC_NAME s_scm_string_to_symbol
 312 {
 313   SCM_VALIDATE_STRING (1, string);
 314   return scm_i_str2symbol (string);
 315 }
 316 #undef FUNC_NAME
 317
 318 SCM_DEFINE (scm_string_ci_to_symbol, "string-ci->symbol", 1, 0, 0,
 319             (SCM str),
 320             "Return the symbol whose name is @var{str}.  @var{str} is\n"
 321             "converted to lowercase before the conversion is done, if Guile\n"
 322             "is currently reading symbols case-insensitively.")
 323 #define FUNC_NAME s_scm_string_ci_to_symbol
 324 {
 325   return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
 326                                ? scm_string_downcase(str)
 327                                : str);
 328 }
 329 #undef FUNC_NAME
 330
 331 #define MAX_PREFIX_LENGTH 30
 332
 333 SCM_DEFINE (scm_gensym, "gensym", 0, 1, 0,
 334             (SCM prefix),
 335             "Create a new symbol with a name constructed from a prefix and\n"
 336             "a counter value. The string @var{prefix} can be specified as\n"
 337             "an optional argument. Default prefix is @code{ g}.  The counter\n"
 338             "is increased by 1 at each call. There is no provision for\n"
 339             "resetting the counter.")
 340 #define FUNC_NAME s_scm_gensym
 341 {
 342   static int gensym_counter = 0;
 343
 344   SCM suffix, name;
 345   int n, n_digits;
 346   char buf[SCM_INTBUFLEN];
 347
 348   if (SCM_UNBNDP (prefix))
 349     prefix = scm_from_locale_string (" g");
 350
 351   /* mutex in case another thread looks and incs at the exact same moment */
 352   scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex);
 353   n = gensym_counter++;
 354   scm_i_pthread_mutex_unlock (&scm_i_misc_mutex);
 355
 356   n_digits = scm_iint2str (n, 10, buf);
 357   suffix = scm_from_locale_stringn (buf, n_digits);
 358   name = scm_string_append (scm_list_2 (prefix, suffix));
 359   return scm_string_to_symbol (name);
 360 }
 361 #undef FUNC_NAME
 362
 363 SCM_DEFINE (scm_symbol_hash, "symbol-hash", 1, 0, 0,
 364             (SCM symbol),
 365             "Return a hash value for @var{symbol}.")
 366 #define FUNC_NAME s_scm_symbol_hash
 367 {
 368   SCM_VALIDATE_SYMBOL (1, symbol);
 369   return scm_from_ulong (scm_i_symbol_hash (symbol));
 370 }
 371 #undef FUNC_NAME
 372
 373 SCM_DEFINE (scm_symbol_fref, "symbol-fref", 1, 0, 0,
 374            (SCM s),
 375             "Return the contents of @var{symbol}'s @dfn{function slot}.")
 376 #define FUNC_NAME s_scm_symbol_fref
 377 {
 378   SCM_VALIDATE_SYMBOL (1, s);
 379   return SCM_CAR (SCM_CELL_OBJECT_3 (s));
 380 }
 381 #undef FUNC_NAME
 382
 383
 384 SCM_DEFINE (scm_symbol_pref, "symbol-pref", 1, 0, 0,
 385            (SCM s),
 386             "Return the @dfn{property list} currently associated with @var{symbol}.")
 387 #define FUNC_NAME s_scm_symbol_pref
 388 {
 389   SCM_VALIDATE_SYMBOL (1, s);
 390   return SCM_CDR (SCM_CELL_OBJECT_3 (s));
 391 }
 392 #undef FUNC_NAME
 393
 394
 395 SCM_DEFINE (scm_symbol_fset_x, "symbol-fset!", 2, 0, 0,
 396            (SCM s, SCM val),
 397             "Change the binding of @var{symbol}'s function slot.")
 398 #define FUNC_NAME s_scm_symbol_fset_x
 399 {
 400   SCM_VALIDATE_SYMBOL (1, s);
 401   SCM_SETCAR (SCM_CELL_OBJECT_3 (s), val);
 402   return SCM_UNSPECIFIED;
 403 }
 404 #undef FUNC_NAME
 405
 406
 407 SCM_DEFINE (scm_symbol_pset_x, "symbol-pset!", 2, 0, 0,
 408            (SCM s, SCM val),
 409             "Change the binding of @var{symbol}'s property slot.")
 410 #define FUNC_NAME s_scm_symbol_pset_x
 411 {
 412   SCM_VALIDATE_SYMBOL (1, s);
 413   SCM_SETCDR (SCM_CELL_OBJECT_3 (s), val);
 414   return SCM_UNSPECIFIED;
 415 }
 416 #undef FUNC_NAME
 417
 418 SCM
 419 scm_from_locale_symbol (const char *sym)
 420 {
 421   return scm_from_locale_symboln (sym, -1);
 422 }
 423
 424 SCM
 425 scm_from_locale_symboln (const char *sym, size_t len)
 426 {
 427   SCM str = scm_from_locale_stringn (sym, len);
 428   return scm_i_str2symbol (str);
 429 }
 430
 431 SCM
 432 scm_take_locale_symboln (char *sym, size_t len)
 433 {
 434   SCM str;
 435
 436   str = scm_take_locale_stringn (sym, len);
 437   return scm_i_str2symbol (str);
 438 }
 439
 440 SCM
 441 scm_take_locale_symbol (char *sym)
 442 {
 443   return scm_take_locale_symboln (sym, (size_t)-1);
 444 }
 445
 446 void
 447 scm_symbols_prehistory ()
 448 {
 449   symbols = scm_make_weak_key_hash_table (scm_from_int (2139));
 450 }
 451
 452
 453 void
 454 scm_init_symbols ()
 455 {
 456 #include "libguile/symbols.x"
 457 }
 458
 459 /*
 460   Local Variables:
 461   c-file-style: "gnu"
 462   End:
 463 */