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 l;
  96   size_t len = scm_i_string_length (name);
  97   unsigned long hash = raw_hash % SCM_HASHTABLE_N_BUCKETS (symbols);
  98
  99   for (l = SCM_HASHTABLE_BUCKET (symbols, hash);
 100        !scm_is_null (l);
 101        l = SCM_CDR (l))
 102     {
 103       SCM sym = SCM_CAAR (l);
 104       if (scm_i_symbol_hash (sym) == raw_hash
 105           && scm_i_symbol_length (sym) == len)
 106         {
 107           size_t i = len;
 108
 109           /* Slightly faster path for comparing narrow to narrow.  */
 110           if (scm_i_is_narrow_string (name) && scm_i_is_narrow_symbol (sym))
 111             {
 112               const char *chrs = scm_i_symbol_chars (sym);
 113               const char *str = scm_i_string_chars (name);
 114
 115               while (i != 0)
 116                 {
 117                   --i;
 118                   if (str[i] != chrs[i])
 119                     goto next_symbol;
 120                 }
 121             }
 122           else
 123             {
 124               /* Somewhat slower path for comparing narrow to wide or
 125                  wide to wide.  */
 126               while (i != 0)
 127                 {
 128                   --i;
 129                   if (scm_i_string_ref (name, i) != scm_i_symbol_ref (sym, i))
 130                     goto next_symbol;
 131                 }
 132             }
 133
 134           return sym;
 135         }
 136     next_symbol:
 137       ;
 138     }
 139
 140   return SCM_BOOL_F;
 141 }
 142
 143 /* Intern SYMBOL, an uninterned symbol.  */
 144 static void
 145 intern_symbol (SCM symbol)
 146 {
 147   SCM slot, cell;
 148   unsigned long hash;
 149
 150   hash = scm_i_symbol_hash (symbol) % SCM_HASHTABLE_N_BUCKETS (symbols);
 151   slot = SCM_HASHTABLE_BUCKET (symbols, hash);
 152   cell = scm_cons (symbol, SCM_UNDEFINED);
 153
 154   SCM_SET_HASHTABLE_BUCKET (symbols, hash, scm_cons (cell, slot));
 155   SCM_HASHTABLE_INCREMENT (symbols);
 156
 157   if (SCM_HASHTABLE_N_ITEMS (symbols) > SCM_HASHTABLE_UPPER (symbols))
 158     scm_i_rehash (symbols, scm_i_hash_symbol, 0, "intern_symbol");
 159 }
 160
 161 static SCM
 162 scm_i_str2symbol (SCM str)
 163 {
 164   SCM symbol;
 165   size_t raw_hash = scm_i_string_hash (str);
 166
 167   symbol = lookup_interned_symbol (str, raw_hash);
 168   if (scm_is_false (symbol))
 169     {
 170       /* The symbol was not found, create it.  */
 171       symbol = scm_i_make_symbol (str, 0, raw_hash,
 172                                   scm_cons (SCM_BOOL_F, SCM_EOL));
 173       intern_symbol (symbol);
 174     }
 175
 176   return symbol;
 177 }
 178
 179
 180 static SCM
 181 scm_i_str2uninterned_symbol (SCM str)
 182 {
 183   size_t raw_hash = scm_i_string_hash (str);
 184
 185   return scm_i_make_symbol (str, SCM_I_F_SYMBOL_UNINTERNED,
 186                             raw_hash, scm_cons (SCM_BOOL_F, SCM_EOL));
 187 }
 188
 189 SCM_DEFINE (scm_symbol_p, "symbol?", 1, 0, 0,
 190             (SCM obj),
 191             "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
 192             "@code{#f}.")
 193 #define FUNC_NAME s_scm_symbol_p
 194 {
 195   return scm_from_bool (scm_is_symbol (obj));
 196 }
 197 #undef FUNC_NAME
 198
 199 SCM_DEFINE (scm_symbol_interned_p, "symbol-interned?", 1, 0, 0,
 200             (SCM symbol),
 201             "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
 202             "@code{#f}.")
 203 #define FUNC_NAME s_scm_symbol_interned_p
 204 {
 205   SCM_VALIDATE_SYMBOL (1, symbol);
 206   return scm_from_bool (scm_i_symbol_is_interned (symbol));
 207 }
 208 #undef FUNC_NAME
 209
 210 SCM_DEFINE (scm_make_symbol, "make-symbol", 1, 0, 0,
 211             (SCM name),
 212             "Return a new uninterned symbol with the name @var{name}.  "
 213             "The returned symbol is guaranteed to be unique and future "
 214             "calls to @code{string->symbol} will not return it.")
 215 #define FUNC_NAME s_scm_make_symbol
 216 {
 217   SCM_VALIDATE_STRING (1, name);
 218   return scm_i_str2uninterned_symbol (name);
 219 }
 220 #undef FUNC_NAME
 221
 222 SCM_DEFINE (scm_symbol_to_string, "symbol->string", 1, 0, 0,
 223            (SCM s),
 224             "Return the name of @var{symbol} as a string.  If the symbol was\n"
 225             "part of an object returned as the value of a literal expression\n"
 226             "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
 227             "Report on Scheme}) or by a call to the @code{read} procedure,\n"
 228             "and its name contains alphabetic characters, then the string\n"
 229             "returned will contain characters in the implementation's\n"
 230             "preferred standard case---some implementations will prefer\n"
 231             "upper case, others lower case.  If the symbol was returned by\n"
 232             "@code{string->symbol}, the case of characters in the string\n"
 233             "returned will be the same as the case in the string that was\n"
 234             "passed to @code{string->symbol}.  It is an error to apply\n"
 235             "mutation procedures like @code{string-set!} to strings returned\n"
 236             "by this procedure.\n"
 237             "\n"
 238             "The following examples assume that the implementation's\n"
 239             "standard case is lower case:\n"
 240             "\n"
 241             "@lisp\n"
 242             "(symbol->string 'flying-fish)    @result{} \"flying-fish\"\n"
 243             "(symbol->string 'Martin)         @result{}  \"martin\"\n"
 244             "(symbol->string\n"
 245             "   (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
 246             "@end lisp")
 247 #define FUNC_NAME s_scm_symbol_to_string
 248 {
 249   SCM_VALIDATE_SYMBOL (1, s);
 250   return scm_i_symbol_substring (s, 0, scm_i_symbol_length (s));
 251 }
 252 #undef FUNC_NAME
 253
 254
 255 SCM_DEFINE (scm_string_to_symbol, "string->symbol", 1, 0, 0,
 256             (SCM string),
 257             "Return the symbol whose name is @var{string}. This procedure\n"
 258             "can create symbols with names containing special characters or\n"
 259             "letters in the non-standard case, but it is usually a bad idea\n"
 260             "to create such symbols because in some implementations of\n"
 261             "Scheme they cannot be read as themselves.  See\n"
 262             "@code{symbol->string}.\n"
 263             "\n"
 264             "The following examples assume that the implementation's\n"
 265             "standard case is lower case:\n"
 266             "\n"
 267             "@lisp\n"
 268             "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
 269             "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
 270             "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
 271             "(eq? 'JollyWog\n"
 272             "  (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
 273             "(string=? \"K. Harper, M.D.\"\n"
 274             "  (symbol->string\n"
 275             "    (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
 276             "@end lisp")
 277 #define FUNC_NAME s_scm_string_to_symbol
 278 {
 279   SCM_VALIDATE_STRING (1, string);
 280   return scm_i_str2symbol (string);
 281 }
 282 #undef FUNC_NAME
 283
 284 SCM_DEFINE (scm_string_ci_to_symbol, "string-ci->symbol", 1, 0, 0,
 285             (SCM str),
 286             "Return the symbol whose name is @var{str}.  @var{str} is\n"
 287             "converted to lowercase before the conversion is done, if Guile\n"
 288             "is currently reading symbols case-insensitively.")
 289 #define FUNC_NAME s_scm_string_ci_to_symbol
 290 {
 291   return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
 292                                ? scm_string_downcase(str)
 293                                : str);
 294 }
 295 #undef FUNC_NAME
 296
 297 #define MAX_PREFIX_LENGTH 30
 298
 299 SCM_DEFINE (scm_gensym, "gensym", 0, 1, 0,
 300             (SCM prefix),
 301             "Create a new symbol with a name constructed from a prefix and\n"
 302             "a counter value. The string @var{prefix} can be specified as\n"
 303             "an optional argument. Default prefix is @code{ g}.  The counter\n"
 304             "is increased by 1 at each call. There is no provision for\n"
 305             "resetting the counter.")
 306 #define FUNC_NAME s_scm_gensym
 307 {
 308   static int gensym_counter = 0;
 309
 310   SCM suffix, name;
 311   int n, n_digits;
 312   char buf[SCM_INTBUFLEN];
 313
 314   if (SCM_UNBNDP (prefix))
 315     prefix = scm_from_locale_string (" g");
 316
 317   /* mutex in case another thread looks and incs at the exact same moment */
 318   scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex);
 319   n = gensym_counter++;
 320   scm_i_pthread_mutex_unlock (&scm_i_misc_mutex);
 321
 322   n_digits = scm_iint2str (n, 10, buf);
 323   suffix = scm_from_locale_stringn (buf, n_digits);
 324   name = scm_string_append (scm_list_2 (prefix, suffix));
 325   return scm_string_to_symbol (name);
 326 }
 327 #undef FUNC_NAME
 328
 329 SCM_DEFINE (scm_symbol_hash, "symbol-hash", 1, 0, 0,
 330             (SCM symbol),
 331             "Return a hash value for @var{symbol}.")
 332 #define FUNC_NAME s_scm_symbol_hash
 333 {
 334   SCM_VALIDATE_SYMBOL (1, symbol);
 335   return scm_from_ulong (scm_i_symbol_hash (symbol));
 336 }
 337 #undef FUNC_NAME
 338
 339 SCM_DEFINE (scm_symbol_fref, "symbol-fref", 1, 0, 0,
 340            (SCM s),
 341             "Return the contents of @var{symbol}'s @dfn{function slot}.")
 342 #define FUNC_NAME s_scm_symbol_fref
 343 {
 344   SCM_VALIDATE_SYMBOL (1, s);
 345   return SCM_CAR (SCM_CELL_OBJECT_3 (s));
 346 }
 347 #undef FUNC_NAME
 348
 349
 350 SCM_DEFINE (scm_symbol_pref, "symbol-pref", 1, 0, 0,
 351            (SCM s),
 352             "Return the @dfn{property list} currently associated with @var{symbol}.")
 353 #define FUNC_NAME s_scm_symbol_pref
 354 {
 355   SCM_VALIDATE_SYMBOL (1, s);
 356   return SCM_CDR (SCM_CELL_OBJECT_3 (s));
 357 }
 358 #undef FUNC_NAME
 359
 360
 361 SCM_DEFINE (scm_symbol_fset_x, "symbol-fset!", 2, 0, 0,
 362            (SCM s, SCM val),
 363             "Change the binding of @var{symbol}'s function slot.")
 364 #define FUNC_NAME s_scm_symbol_fset_x
 365 {
 366   SCM_VALIDATE_SYMBOL (1, s);
 367   SCM_SETCAR (SCM_CELL_OBJECT_3 (s), val);
 368   return SCM_UNSPECIFIED;
 369 }
 370 #undef FUNC_NAME
 371
 372
 373 SCM_DEFINE (scm_symbol_pset_x, "symbol-pset!", 2, 0, 0,
 374            (SCM s, SCM val),
 375             "Change the binding of @var{symbol}'s property slot.")
 376 #define FUNC_NAME s_scm_symbol_pset_x
 377 {
 378   SCM_VALIDATE_SYMBOL (1, s);
 379   SCM_SETCDR (SCM_CELL_OBJECT_3 (s), val);
 380   return SCM_UNSPECIFIED;
 381 }
 382 #undef FUNC_NAME
 383
 384 SCM
 385 scm_from_locale_symbol (const char *sym)
 386 {
 387   return scm_from_locale_symboln (sym, -1);
 388 }
 389
 390 SCM
 391 scm_from_locale_symboln (const char *sym, size_t len)
 392 {
 393   SCM str = scm_from_locale_stringn (sym, len);
 394   return scm_i_str2symbol (str);
 395 }
 396
 397 SCM
 398 scm_take_locale_symboln (char *sym, size_t len)
 399 {
 400   SCM str;
 401
 402   str = scm_take_locale_stringn (sym, len);
 403   return scm_i_str2symbol (str);
 404 }
 405
 406 SCM
 407 scm_take_locale_symbol (char *sym)
 408 {
 409   return scm_take_locale_symboln (sym, (size_t)-1);
 410 }
 411
 412 void
 413 scm_symbols_prehistory ()
 414 {
 415   symbols = scm_make_weak_key_hash_table (scm_from_int (2139));
 416   scm_permanent_object (symbols);
 417 }
 418
 419
 420 void
 421 scm_init_symbols ()
 422 {
 423 #include "libguile/symbols.x"
 424 }
 425
 426 /*
 427   Local Variables:
 428   c-file-style: "gnu"
 429   End:
 430 */