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 (const char *name, size_t len,
  93                         unsigned long raw_hash)
  94 {
  95   /* Try to find the symbol in the symbols table */
  96   SCM result = SCM_BOOL_F;
  97   SCM bucket, elt, previous_elt;
  98   unsigned long hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
  99
 100   bucket = SCM_HASHTABLE_BUCKET (symbols, hash);
 101   for (elt = bucket, previous_elt = SCM_BOOL_F;
 102        !scm_is_null (elt);
 103        previous_elt = elt, elt = SCM_CDR (elt))
 104     {
 105       SCM pair, sym;
 106
 107       pair = SCM_CAR (elt);
 108       if (!scm_is_pair (pair))
 109         abort ();
 110
 111       if (SCM_WEAK_PAIR_CAR_DELETED_P (pair))
 112         {
 113           /* PAIR is a weak pair whose key got nullified: remove it from
 114              BUCKET.  */
 115           /* FIXME: Since this is done lazily, i.e., only when a new symbol
 116              is to be inserted in a bucket containing deleted symbols, the
 117              number of items in the hash table may remain erroneous for some
 118              time, thus precluding proper rehashing.  */
 119           if (previous_elt != SCM_BOOL_F)
 120             SCM_SETCDR (previous_elt, SCM_CDR (elt));
 121           else
 122             bucket = SCM_CDR (elt);
 123
 124           SCM_HASHTABLE_DECREMENT (symbols);
 125           continue;
 126         }
 127
 128       sym = SCM_CAR (pair);
 129
 130       if (scm_i_symbol_hash (sym) == raw_hash
 131           && scm_i_symbol_length (sym) == len)
 132         {
 133           const char *chrs = scm_i_symbol_chars (sym);
 134           size_t i = len;
 135
 136           while (i != 0)
 137             {
 138               --i;
 139               if (name[i] != chrs[i])
 140                 goto next_symbol;
 141             }
 142
 143           /* We found it.  */
 144           result = sym;
 145           break;
 146         }
 147     next_symbol:
 148       ;
 149     }
 150
 151   if (SCM_HASHTABLE_N_ITEMS (symbols) < SCM_HASHTABLE_LOWER (symbols))
 152     /* We removed many symbols in this pass so trigger a rehashing.  */
 153     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "lookup_interned_symbol");
 154
 155   return result;
 156 }
 157
 158 /* Intern SYMBOL, an uninterned symbol.  */
 159 static void
 160 intern_symbol (SCM symbol)
 161 {
 162   SCM slot, cell;
 163   unsigned long hash;
 164
 165   hash = scm_i_symbol_hash (symbol) % SCM_HASHTABLE_N_BUCKETS (symbols);
 166   slot = SCM_HASHTABLE_BUCKET (symbols, hash);
 167   cell = scm_cons (symbol, SCM_UNDEFINED);
 168
 169   SCM_SET_HASHTABLE_BUCKET (symbols, hash, scm_cons (cell, slot));
 170   SCM_HASHTABLE_INCREMENT (symbols);
 171
 172   if (SCM_HASHTABLE_N_ITEMS (symbols) > SCM_HASHTABLE_UPPER (symbols))
 173     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "intern_symbol");
 174 }
 175
 176 static SCM
 177 scm_i_c_mem2symbol (const char *name, size_t len)
 178 {
 179   SCM symbol;
 180   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 181
 182   symbol = lookup_interned_symbol (name, len, raw_hash);
 183   if (scm_is_false (symbol))
 184     {
 185       /* The symbol was not found, create it.  */
 186       symbol = scm_i_c_make_symbol (name, len, 0, raw_hash,
 187                                     scm_cons (SCM_BOOL_F, SCM_EOL));
 188       intern_symbol (symbol);
 189     }
 190
 191   return symbol;
 192 }
 193
 194 static SCM
 195 scm_i_mem2symbol (SCM str)
 196 {
 197   SCM symbol;
 198   const char *name = scm_i_string_chars (str);
 199   size_t len = scm_i_string_length (str);
 200   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 201
 202   symbol = lookup_interned_symbol (name, len, raw_hash);
 203   if (scm_is_false (symbol))
 204     {
 205       /* The symbol was not found, create it.  */
 206       symbol = scm_i_make_symbol (str, 0, raw_hash,
 207                                   scm_cons (SCM_BOOL_F, SCM_EOL));
 208       intern_symbol (symbol);
 209     }
 210
 211   return symbol;
 212 }
 213
 214
 215 static SCM
 216 scm_i_mem2uninterned_symbol (SCM str)
 217 {
 218   const char *name = scm_i_string_chars (str);
 219   size_t len = scm_i_string_length (str);
 220   size_t raw_hash = scm_string_hash ((const unsigned char *) name, len);
 221
 222   return scm_i_make_symbol (str, SCM_I_F_SYMBOL_UNINTERNED,
 223                             raw_hash, scm_cons (SCM_BOOL_F, SCM_EOL));
 224 }
 225
 226 SCM_DEFINE (scm_symbol_p, "symbol?", 1, 0, 0,
 227             (SCM obj),
 228             "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
 229             "@code{#f}.")
 230 #define FUNC_NAME s_scm_symbol_p
 231 {
 232   return scm_from_bool (scm_is_symbol (obj));
 233 }
 234 #undef FUNC_NAME
 235
 236 SCM_DEFINE (scm_symbol_interned_p, "symbol-interned?", 1, 0, 0,
 237             (SCM symbol),
 238             "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
 239             "@code{#f}.")
 240 #define FUNC_NAME s_scm_symbol_interned_p
 241 {
 242   SCM_VALIDATE_SYMBOL (1, symbol);
 243   return scm_from_bool (scm_i_symbol_is_interned (symbol));
 244 }
 245 #undef FUNC_NAME
 246
 247 SCM_DEFINE (scm_make_symbol, "make-symbol", 1, 0, 0,
 248             (SCM name),
 249             "Return a new uninterned symbol with the name @var{name}.  "
 250             "The returned symbol is guaranteed to be unique and future "
 251             "calls to @code{string->symbol} will not return it.")
 252 #define FUNC_NAME s_scm_make_symbol
 253 {
 254   SCM_VALIDATE_STRING (1, name);
 255   return scm_i_mem2uninterned_symbol (name);
 256 }
 257 #undef FUNC_NAME
 258
 259 SCM_DEFINE (scm_symbol_to_string, "symbol->string", 1, 0, 0,
 260            (SCM s),
 261             "Return the name of @var{symbol} as a string.  If the symbol was\n"
 262             "part of an object returned as the value of a literal expression\n"
 263             "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
 264             "Report on Scheme}) or by a call to the @code{read} procedure,\n"
 265             "and its name contains alphabetic characters, then the string\n"
 266             "returned will contain characters in the implementation's\n"
 267             "preferred standard case---some implementations will prefer\n"
 268             "upper case, others lower case.  If the symbol was returned by\n"
 269             "@code{string->symbol}, the case of characters in the string\n"
 270             "returned will be the same as the case in the string that was\n"
 271             "passed to @code{string->symbol}.  It is an error to apply\n"
 272             "mutation procedures like @code{string-set!} to strings returned\n"
 273             "by this procedure.\n"
 274             "\n"
 275             "The following examples assume that the implementation's\n"
 276             "standard case is lower case:\n"
 277             "\n"
 278             "@lisp\n"
 279             "(symbol->string 'flying-fish)    @result{} \"flying-fish\"\n"
 280             "(symbol->string 'Martin)         @result{}  \"martin\"\n"
 281             "(symbol->string\n"
 282             "   (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
 283             "@end lisp")
 284 #define FUNC_NAME s_scm_symbol_to_string
 285 {
 286   SCM_VALIDATE_SYMBOL (1, s);
 287   return scm_i_symbol_substring (s, 0, scm_i_symbol_length (s));
 288 }
 289 #undef FUNC_NAME
 290
 291
 292 SCM_DEFINE (scm_string_to_symbol, "string->symbol", 1, 0, 0,
 293             (SCM string),
 294             "Return the symbol whose name is @var{string}. This procedure\n"
 295             "can create symbols with names containing special characters or\n"
 296             "letters in the non-standard case, but it is usually a bad idea\n"
 297             "to create such symbols because in some implementations of\n"
 298             "Scheme they cannot be read as themselves.  See\n"
 299             "@code{symbol->string}.\n"
 300             "\n"
 301             "The following examples assume that the implementation's\n"
 302             "standard case is lower case:\n"
 303             "\n"
 304             "@lisp\n"
 305             "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
 306             "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
 307             "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
 308             "(eq? 'JollyWog\n"
 309             "  (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
 310             "(string=? \"K. Harper, M.D.\"\n"
 311             "  (symbol->string\n"
 312             "    (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
 313             "@end lisp")
 314 #define FUNC_NAME s_scm_string_to_symbol
 315 {
 316   SCM_VALIDATE_STRING (1, string);
 317   return scm_i_mem2symbol (string);
 318 }
 319 #undef FUNC_NAME
 320
 321 SCM_DEFINE (scm_string_ci_to_symbol, "string-ci->symbol", 1, 0, 0,
 322             (SCM str),
 323             "Return the symbol whose name is @var{str}.  @var{str} is\n"
 324             "converted to lowercase before the conversion is done, if Guile\n"
 325             "is currently reading symbols case-insensitively.")
 326 #define FUNC_NAME s_scm_string_ci_to_symbol
 327 {
 328   return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
 329                                ? scm_string_downcase(str)
 330                                : str);
 331 }
 332 #undef FUNC_NAME
 333
 334 #define MAX_PREFIX_LENGTH 30
 335
 336 SCM_DEFINE (scm_gensym, "gensym", 0, 1, 0,
 337             (SCM prefix),
 338             "Create a new symbol with a name constructed from a prefix and\n"
 339             "a counter value. The string @var{prefix} can be specified as\n"
 340             "an optional argument. Default prefix is @code{ g}.  The counter\n"
 341             "is increased by 1 at each call. There is no provision for\n"
 342             "resetting the counter.")
 343 #define FUNC_NAME s_scm_gensym
 344 {
 345   static int gensym_counter = 0;
 346
 347   SCM suffix, name;
 348   int n, n_digits;
 349   char buf[SCM_INTBUFLEN];
 350
 351   if (SCM_UNBNDP (prefix))
 352     prefix = scm_from_locale_string (" g");
 353
 354   /* mutex in case another thread looks and incs at the exact same moment */
 355   scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex);
 356   n = gensym_counter++;
 357   scm_i_pthread_mutex_unlock (&scm_i_misc_mutex);
 358
 359   n_digits = scm_iint2str (n, 10, buf);
 360   suffix = scm_from_locale_stringn (buf, n_digits);
 361   name = scm_string_append (scm_list_2 (prefix, suffix));
 362   return scm_string_to_symbol (name);
 363 }
 364 #undef FUNC_NAME
 365
 366 SCM_DEFINE (scm_symbol_hash, "symbol-hash", 1, 0, 0,
 367             (SCM symbol),
 368             "Return a hash value for @var{symbol}.")
 369 #define FUNC_NAME s_scm_symbol_hash
 370 {
 371   SCM_VALIDATE_SYMBOL (1, symbol);
 372   return scm_from_ulong (scm_i_symbol_hash (symbol));
 373 }
 374 #undef FUNC_NAME
 375
 376 SCM_DEFINE (scm_symbol_fref, "symbol-fref", 1, 0, 0,
 377            (SCM s),
 378             "Return the contents of @var{symbol}'s @dfn{function slot}.")
 379 #define FUNC_NAME s_scm_symbol_fref
 380 {
 381   SCM_VALIDATE_SYMBOL (1, s);
 382   return SCM_CAR (SCM_CELL_OBJECT_3 (s));
 383 }
 384 #undef FUNC_NAME
 385
 386
 387 SCM_DEFINE (scm_symbol_pref, "symbol-pref", 1, 0, 0,
 388            (SCM s),
 389             "Return the @dfn{property list} currently associated with @var{symbol}.")
 390 #define FUNC_NAME s_scm_symbol_pref
 391 {
 392   SCM_VALIDATE_SYMBOL (1, s);
 393   return SCM_CDR (SCM_CELL_OBJECT_3 (s));
 394 }
 395 #undef FUNC_NAME
 396
 397
 398 SCM_DEFINE (scm_symbol_fset_x, "symbol-fset!", 2, 0, 0,
 399            (SCM s, SCM val),
 400             "Change the binding of @var{symbol}'s function slot.")
 401 #define FUNC_NAME s_scm_symbol_fset_x
 402 {
 403   SCM_VALIDATE_SYMBOL (1, s);
 404   SCM_SETCAR (SCM_CELL_OBJECT_3 (s), val);
 405   return SCM_UNSPECIFIED;
 406 }
 407 #undef FUNC_NAME
 408
 409
 410 SCM_DEFINE (scm_symbol_pset_x, "symbol-pset!", 2, 0, 0,
 411            (SCM s, SCM val),
 412             "Change the binding of @var{symbol}'s property slot.")
 413 #define FUNC_NAME s_scm_symbol_pset_x
 414 {
 415   SCM_VALIDATE_SYMBOL (1, s);
 416   SCM_SETCDR (SCM_CELL_OBJECT_3 (s), val);
 417   return SCM_UNSPECIFIED;
 418 }
 419 #undef FUNC_NAME
 420
 421 SCM
 422 scm_from_locale_symbol (const char *sym)
 423 {
 424   return scm_i_c_mem2symbol (sym, strlen (sym));
 425 }
 426
 427 SCM
 428 scm_from_locale_symboln (const char *sym, size_t len)
 429 {
 430   return scm_i_c_mem2symbol (sym, len);
 431 }
 432
 433 SCM
 434 scm_take_locale_symboln (char *sym, size_t len)
 435 {
 436   SCM res;
 437   unsigned long raw_hash;
 438
 439   if (len == (size_t)-1)
 440     len = strlen (sym);
 441   else
 442     {
 443       /* Ensure STR is null terminated.  A realloc for 1 extra byte should
 444          often be satisfied from the alignment padding after the block, with
 445          no actual data movement.  */
 446       sym = scm_realloc (sym, len+1);
 447       sym[len] = '\0';
 448     }
 449
 450   raw_hash = scm_string_hash ((unsigned char *)sym, len);
 451   res = lookup_interned_symbol (sym, len, raw_hash);
 452   if (scm_is_false (res))
 453     {
 454       res = scm_i_c_take_symbol (sym, len, 0, raw_hash,
 455                                  scm_cons (SCM_BOOL_F, SCM_EOL));
 456       intern_symbol (res);
 457     }
 458   else
 459     free (sym);
 460
 461   return res;
 462 }
 463
 464 SCM
 465 scm_take_locale_symbol (char *sym)
 466 {
 467   return scm_take_locale_symboln (sym, (size_t)-1);
 468 }
 469
 470 void
 471 scm_symbols_prehistory ()
 472 {
 473   symbols = scm_make_weak_key_hash_table (scm_from_int (2139));
 474 }
 475
 476
 477 void
 478 scm_init_symbols ()
 479 {
 480 #include "libguile/symbols.x"
 481 }
 482
 483 /*
 484   Local Variables:
 485   c-file-style: "gnu"
 486   End:
 487 */