src/symtab.c

   1 /* Symbol table manager for Bison,
   2    Copyright (C) 1984, 1989, 2000, 2001, 2002 Free Software Foundation, Inc.
   3
   4    This file is part of Bison, the GNU Compiler Compiler.
   5
   6    Bison is free software; you can redistribute it and/or modify
   7    it under the terms of the GNU General Public License as published by
   8    the Free Software Foundation; either version 2, or (at your option)
   9    any later version.
  10
  11    Bison is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14    GNU General Public License for more details.
  15
  16    You should have received a copy of the GNU General Public License
  17    along with Bison; see the file COPYING.  If not, write to
  18    the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  19    Boston, MA 02111-1307, USA.  */
  20
  21
  22 #include "system.h"
  23
  24 #include <hash.h>
  25 #include <quotearg.h>
  26
  27 #include "complain.h"
  28 #include "gram.h"
  29 #include "symtab.h"
  30
  31 /*------------------------.
  32 | Distinguished symbols.  |
  33 `------------------------*/
  34
  35 symbol *errtoken = NULL;
  36 symbol *undeftoken = NULL;
  37 symbol *endtoken = NULL;
  38 symbol *accept = NULL;
  39 symbol *startsymbol = NULL;
  40 location startsymbol_location;
  41
  42 /*---------------------------------.
  43 | Create a new symbol, named TAG.  |
  44 `---------------------------------*/
  45
  46 static symbol *
  47 symbol_new (uniqstr tag, location loc)
  48 {
  49   symbol *res = XMALLOC (symbol, 1);
  50
  51   uniqstr_assert (tag);
  52   res->tag = tag;
  53   res->location = loc;
  54
  55   res->type_name = NULL;
  56   res->destructor = NULL;
  57   res->printer = NULL;
  58
  59   res->number = NUMBER_UNDEFINED;
  60   res->prec = 0;
  61   res->assoc = undef_assoc;
  62   res->user_token_number = USER_NUMBER_UNDEFINED;
  63
  64   res->alias = NULL;
  65   res->class = unknown_sym;
  66
  67   nsyms++;
  68   return res;
  69 }
  70
  71
  72 /*-----------------------------------------------------------------.
  73 | Set the TYPE_NAME associated with SYM.  Does nothing if passed 0 |
  74 | as TYPE_NAME.                                                    |
  75 `-----------------------------------------------------------------*/
  76
  77 void
  78 symbol_type_set (symbol *sym, uniqstr type_name, location loc)
  79 {
  80   if (type_name)
  81     {
  82       if (sym->type_name)
  83         complain_at (loc, _("type redeclaration for %s"), sym->tag);
  84       uniqstr_assert (type_name);
  85       sym->type_name = type_name;
  86     }
  87 }
  88
  89
  90 /*------------------------------------------------------------------.
  91 | Set the DESTRUCTOR associated with SYM.  Do nothing if passed 0.  |
  92 `------------------------------------------------------------------*/
  93
  94 void
  95 symbol_destructor_set (symbol *sym, char *destructor, location loc)
  96 {
  97   if (destructor)
  98     {
  99       if (sym->destructor)
 100         complain_at (loc, _("%s redeclaration for %s"),
 101                      "%destructor", sym->tag);
 102       sym->destructor = destructor;
 103       sym->destructor_location = loc;
 104     }
 105 }
 106
 107
 108 /*---------------------------------------------------------------.
 109 | Set the PRINTER associated with SYM.  Do nothing if passed 0.  |
 110 `---------------------------------------------------------------*/
 111
 112 void
 113 symbol_printer_set (symbol *sym, char *printer, location loc)
 114 {
 115   if (printer)
 116     {
 117       if (sym->printer)
 118         complain_at (loc, _("%s redeclaration for %s"),
 119                      "%printer", sym->tag);
 120       sym->printer = printer;
 121       sym->printer_location = loc;
 122     }
 123 }
 124
 125
 126 /*-----------------------------------------------------------------.
 127 | Set the PRECEDENCE associated with SYM.  Does nothing if invoked |
 128 | with UNDEF_ASSOC as ASSOC.                                       |
 129 `-----------------------------------------------------------------*/
 130
 131 void
 132 symbol_precedence_set (symbol *sym, int prec, assoc a, location loc)
 133 {
 134   if (a != undef_assoc)
 135     {
 136       if (sym->prec != 0)
 137         complain_at (loc, _("redefining precedence of %s"), sym->tag);
 138       sym->prec = prec;
 139       sym->assoc = a;
 140     }
 141
 142   /* Only terminals have a precedence. */
 143   symbol_class_set (sym, token_sym, loc);
 144 }
 145
 146
 147 /*------------------------------------.
 148 | Set the CLASS associated with SYM.  |
 149 `------------------------------------*/
 150
 151 void
 152 symbol_class_set (symbol *sym, symbol_class class, location loc)
 153 {
 154   if (sym->class != unknown_sym && sym->class != class)
 155     complain_at (loc, _("symbol %s redefined"), sym->tag);
 156
 157   if (class == nterm_sym && sym->class != nterm_sym)
 158     sym->number = nvars++;
 159   else if (class == token_sym && sym->number == NUMBER_UNDEFINED)
 160     sym->number = ntokens++;
 161
 162   sym->class = class;
 163 }
 164
 165
 166 /*------------------------------------------------.
 167 | Set the USER_TOKEN_NUMBER associated with SYM.  |
 168 `------------------------------------------------*/
 169
 170 void
 171 symbol_user_token_number_set (symbol *sym, int user_token_number, location loc)
 172 {
 173   if (sym->class != token_sym)
 174     abort ();
 175
 176   if (sym->user_token_number != USER_NUMBER_UNDEFINED
 177       && sym->user_token_number != user_token_number)
 178     complain_at (loc, _("redefining user token number of %s"), sym->tag);
 179
 180   sym->user_token_number = user_token_number;
 181   /* User defined $end token? */
 182   if (user_token_number == 0)
 183     {
 184       endtoken = sym;
 185       endtoken->number = 0;
 186       /* It is always mapped to 0, so it was already counted in
 187          NTOKENS.  */
 188       --ntokens;
 189     }
 190 }
 191
 192
 193 /*-----------.
 194 | Free SYM.  |
 195 `-----------*/
 196
 197 static void
 198 symbol_free (symbol *sym)
 199 {
 200   free (sym);
 201 }
 202
 203
 204 /*----------------------------------------------------------.
 205 | If SYM is not defined, report an error, and consider it a |
 206 | nonterminal.                                              |
 207 `----------------------------------------------------------*/
 208
 209 static bool
 210 symbol_check_defined (symbol *sym)
 211 {
 212   if (sym->class == unknown_sym)
 213     {
 214       complain_at
 215         (sym->location,
 216          _("symbol %s is used, but is not defined as a token and has no rules"),
 217          sym->tag);
 218       sym->class = nterm_sym;
 219       sym->number = nvars++;
 220     }
 221
 222   return true;
 223 }
 224
 225
 226 /*------------------------------------------------------------------.
 227 | Declare the new symbol SYM.  Make it an alias of SYMVAL, and type |
 228 | them with TYPENAME.                                               |
 229 `------------------------------------------------------------------*/
 230
 231 void
 232 symbol_make_alias (symbol *sym, symbol *symval, location loc)
 233 {
 234   if (symval->alias)
 235     warn_at (loc, _("symbol `%s' used more than once as a literal string"),
 236           symval->tag);
 237   else if (sym->alias)
 238     warn_at (loc, _("symbol `%s' given more than one literal string"),
 239           sym->tag);
 240   else
 241     {
 242       symval->class = token_sym;
 243       symval->user_token_number = sym->user_token_number;
 244       sym->user_token_number = USER_NUMBER_ALIAS;
 245       symval->alias = sym;
 246       sym->alias = symval;
 247       /* sym and symval combined are only one symbol.  */
 248       nsyms--;
 249       ntokens--;
 250       if (ntokens != sym->number && ntokens != symval->number)
 251         abort ();
 252       sym->number = symval->number =
 253         (symval->number < sym->number) ? symval->number : sym->number;
 254     }
 255 }
 256
 257
 258 /*---------------------------------------------------------.
 259 | Check that THIS, and its alias, have same precedence and |
 260 | associativity.                                           |
 261 `---------------------------------------------------------*/
 262
 263 static bool
 264 symbol_check_alias_consistence (symbol *this)
 265 {
 266   /* Check only those who _are_ the aliases. */
 267   if (this->alias && this->user_token_number == USER_NUMBER_ALIAS)
 268     {
 269       if (this->prec != this->alias->prec)
 270         {
 271           if (this->prec != 0 && this->alias->prec != 0)
 272             complain_at (this->alias->location,
 273                          _("conflicting precedences for %s and %s"),
 274                          this->tag, this->alias->tag);
 275           if (this->prec != 0)
 276             this->alias->prec = this->prec;
 277           else
 278             this->prec = this->alias->prec;
 279         }
 280
 281       if (this->assoc != this->alias->assoc)
 282         {
 283           if (this->assoc != undef_assoc && this->alias->assoc != undef_assoc)
 284             complain_at (this->alias->location,
 285                          _("conflicting associativities for %s (%s) and %s (%s)"),
 286                          this->tag, assoc_to_string (this->assoc),
 287                          this->alias->tag, assoc_to_string (this->alias->assoc));
 288           if (this->assoc != undef_assoc)
 289             this->alias->assoc = this->assoc;
 290           else
 291             this->assoc = this->alias->assoc;
 292         }
 293     }
 294   return true;
 295 }
 296
 297
 298 /*-------------------------------------------------------------------.
 299 | Assign a symbol number, and write the definition of the token name |
 300 | into FDEFINES.  Put in SYMBOLS.                                    |
 301 `-------------------------------------------------------------------*/
 302
 303 static bool
 304 symbol_pack (symbol *this)
 305 {
 306   if (this->class == nterm_sym)
 307     {
 308       this->number += ntokens;
 309     }
 310   else if (this->alias)
 311     {
 312       /* This symbol and its alias are a single token defn.
 313          Allocate a tokno, and assign to both check agreement of
 314          prec and assoc fields and make both the same */
 315       if (this->number == NUMBER_UNDEFINED)
 316         {
 317           if (this == endtoken || this->alias == endtoken)
 318             this->number = this->alias->number = 0;
 319           else
 320             {
 321               if (this->alias->number == NUMBER_UNDEFINED)
 322                 abort ();
 323               this->number = this->alias->number;
 324             }
 325         }
 326       /* Do not do processing below for USER_NUMBER_ALIASes.  */
 327       if (this->user_token_number == USER_NUMBER_ALIAS)
 328         return true;
 329     }
 330   else /* this->class == token_sym */
 331     {
 332       if (this->number == NUMBER_UNDEFINED)
 333         abort ();
 334     }
 335
 336   symbols[this->number] = this;
 337   return true;
 338 }
 339
 340
 341
 342
 343 /*--------------------------------------------------.
 344 | Put THIS in TOKEN_TRANSLATIONS if it is a token.  |
 345 `--------------------------------------------------*/
 346
 347 static bool
 348 symbol_translation (symbol *this)
 349 {
 350   /* Non-terminal? */
 351   if (this->class == token_sym
 352       && this->user_token_number != USER_NUMBER_ALIAS)
 353     {
 354       /* A token which translation has already been set? */
 355       if (token_translations[this->user_token_number] != undeftoken->number)
 356         complain_at (this->location,
 357                      _("tokens %s and %s both assigned number %d"),
 358                      symbols[token_translations[this->user_token_number]]->tag,
 359                      this->tag, this->user_token_number);
 360
 361       token_translations[this->user_token_number] = this->number;
 362     }
 363
 364   return true;
 365 }
 366
 367
 368 /*----------------------.
 369 | A symbol hash table.  |
 370 `----------------------*/
 371
 372 /* Initial capacity of symbols hash table.  */
 373 #define HT_INITIAL_CAPACITY 257
 374
 375 static struct hash_table *symbol_table = NULL;
 376
 377 static bool
 378 hash_compare_symbol (const symbol *m1, const symbol *m2)
 379 {
 380   /* Since tags are unique, we can compare the pointers themselves.  */
 381   return UNIQSTR_EQ (m1->tag, m2->tag);
 382 }
 383
 384 static unsigned int
 385 hash_symbol (const symbol *m, unsigned int tablesize)
 386 {
 387   /* Since tags are unique, we can hash the pointer itself.  */
 388   return ((size_t) m->tag) % tablesize;
 389 }
 390
 391
 392 /*-------------------------------.
 393 | Create the symbol hash table.  |
 394 `-------------------------------*/
 395
 396 void
 397 symbols_new (void)
 398 {
 399   symbol_table = hash_initialize (HT_INITIAL_CAPACITY,
 400                                   NULL,
 401                                   (Hash_hasher) hash_symbol,
 402                                   (Hash_comparator) hash_compare_symbol,
 403                                   (Hash_data_freer) symbol_free);
 404 }
 405
 406
 407 /*----------------------------------------------------------------.
 408 | Find the symbol named KEY, and return it.  If it does not exist |
 409 | yet, create it.                                                 |
 410 `----------------------------------------------------------------*/
 411
 412 symbol *
 413 symbol_get (const char *key, location loc)
 414 {
 415   symbol probe;
 416   symbol *entry;
 417
 418   /* Keep the symbol in a printable form.  */
 419   key = uniqstr_new (quotearg_style (escape_quoting_style, key));
 420   *(char const **) &probe.tag = key;
 421   entry = hash_lookup (symbol_table, &probe);
 422
 423   if (!entry)
 424     {
 425       /* First insertion in the hash. */
 426       entry = symbol_new (key, loc);
 427       hash_insert (symbol_table, entry);
 428     }
 429   return entry;
 430 }
 431
 432
 433 /*------------------------------------------------------------------.
 434 | Generate a dummy nonterminal, whose name cannot conflict with the |
 435 | user's names.                                                     |
 436 `------------------------------------------------------------------*/
 437
 438 symbol *
 439 dummy_symbol_get (location loc)
 440 {
 441   /* Incremented for each generated symbol.  */
 442   static int dummy_count = 0;
 443   static char buf[256];
 444
 445   symbol *sym;
 446
 447   sprintf (buf, "@%d", ++dummy_count);
 448   sym = symbol_get (buf, loc);
 449   sym->class = nterm_sym;
 450   sym->number = nvars++;
 451   return sym;
 452 }
 453
 454
 455 /*-------------------.
 456 | Free the symbols.  |
 457 `-------------------*/
 458
 459 void
 460 symbols_free (void)
 461 {
 462   hash_free (symbol_table);
 463   free (symbols);
 464 }
 465
 466
 467 /*---------------------------------------------------------------.
 468 | Look for undefined symbols, report an error, and consider them |
 469 | terminals.                                                     |
 470 `---------------------------------------------------------------*/
 471
 472 void
 473 symbols_do (symbol_processor processor, void *processor_data)
 474 {
 475   hash_do_for_each (symbol_table,
 476                     (Hash_processor) processor,
 477                     processor_data);
 478 }
 479
 480
 481 /*--------------------------------------------------------------.
 482 | Check that all the symbols are defined.  Report any undefined |
 483 | symbols and consider them nonterminals.                       |
 484 `--------------------------------------------------------------*/
 485
 486 void
 487 symbols_check_defined (void)
 488 {
 489   symbols_do (symbol_check_defined, NULL);
 490 }
 491
 492 /*------------------------------------------------------------------.
 493 | Set TOKEN_TRANSLATIONS.  Check that no two symbols share the same |
 494 | number.                                                           |
 495 `------------------------------------------------------------------*/
 496
 497 static void
 498 symbols_token_translations_init (void)
 499 {
 500   bool num_256_available_p = true;
 501   int i;
 502
 503   /* Find the highest user token number, and whether 256, the POSIX
 504      preferred user token number for the error token, is used.  */
 505   max_user_token_number = 0;
 506   for (i = 0; i < ntokens; ++i)
 507     {
 508       symbol *this = symbols[i];
 509       if (this->user_token_number != USER_NUMBER_UNDEFINED)
 510         {
 511           if (this->user_token_number > max_user_token_number)
 512             max_user_token_number = this->user_token_number;
 513           if (this->user_token_number == 256)
 514             num_256_available_p = false;
 515         }
 516     }
 517
 518   /* If 256 is not used, assign it to error, to follow POSIX.  */
 519   if (num_256_available_p
 520       && errtoken->user_token_number == USER_NUMBER_UNDEFINED)
 521     errtoken->user_token_number = 256;
 522
 523   /* Set the missing user numbers. */
 524   if (max_user_token_number < 256)
 525     max_user_token_number = 256;
 526
 527   for (i = 0; i < ntokens; ++i)
 528     {
 529       symbol *this = symbols[i];
 530       if (this->user_token_number == USER_NUMBER_UNDEFINED)
 531         this->user_token_number = ++max_user_token_number;
 532       if (this->user_token_number > max_user_token_number)
 533         max_user_token_number = this->user_token_number;
 534     }
 535
 536   token_translations = XCALLOC (symbol_number, max_user_token_number + 1);
 537
 538   /* Initialize all entries for literal tokens to 2, the internal
 539      token number for $undefined, which represents all invalid inputs.
 540      */
 541   for (i = 0; i < max_user_token_number + 1; i++)
 542     token_translations[i] = undeftoken->number;
 543   symbols_do (symbol_translation, NULL);
 544 }
 545
 546
 547 /*----------------------------------------------------------------.
 548 | Assign symbol numbers, and write definition of token names into |
 549 | FDEFINES.  Set up vectors SYMBOL_TABLE, TAGS of symbols.        |
 550 `----------------------------------------------------------------*/
 551
 552 void
 553 symbols_pack (void)
 554 {
 555   symbols = XCALLOC (symbol *, nsyms);
 556
 557   symbols_do (symbol_check_alias_consistence, NULL);
 558   symbols_do (symbol_pack, NULL);
 559
 560   symbols_token_translations_init ();
 561
 562   if (startsymbol->class == unknown_sym)
 563     fatal_at (startsymbol_location,
 564               _("the start symbol %s is undefined"),
 565               startsymbol->tag);
 566   else if (startsymbol->class == token_sym)
 567     fatal_at (startsymbol_location,
 568               _("the start symbol %s is a token"),
 569               startsymbol->tag);
 570 }