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 #include "quotearg.h"
  24 #include "hash.h"
  25 #include "complain.h"
  26 #include "symtab.h"
  27 #include "gram.h"
  28
  29 /*------------------------.
  30 | Distinguished symbols.  |
  31 `------------------------*/
  32
  33 symbol_t *errtoken = NULL;
  34 symbol_t *undeftoken = NULL;
  35 symbol_t *eoftoken = NULL;
  36 symbol_t *axiom = NULL;
  37 symbol_t *startsymbol = NULL;
  38 location_t startsymbol_location;
  39
  40 /*---------------------------------.
  41 | Create a new symbol, named TAG.  |
  42 `---------------------------------*/
  43
  44 static symbol_t *
  45 symbol_new (const char *tag, location_t location)
  46 {
  47   symbol_t *res = XMALLOC (symbol_t, 1);
  48
  49   res->tag = xstrdup (tag);
  50   res->type_name = NULL;
  51   res->location = location;
  52   res->number = NUMBER_UNDEFINED;
  53   res->prec = 0;
  54   res->assoc = right_assoc;
  55   res->user_token_number = USER_NUMBER_UNDEFINED;
  56   res->alias = NULL;
  57   res->class = unknown_sym;
  58
  59   nsyms++;
  60   return res;
  61 }
  62
  63
  64 /*-----------------------------------------------------------------.
  65 | Return the tag of this SYMBOL in a printable form.  Warning: use |
  66 | the first QUOTEARG slot: 0.                                      |
  67 `-----------------------------------------------------------------*/
  68
  69 const char *
  70 symbol_tag_get (symbol_t *symbol)
  71 {
  72   return quotearg_style (escape_quoting_style, symbol->tag);
  73 }
  74
  75
  76 /*------------------------------------------------------------.
  77 | Return the tag of this SYMBOL in a printable form.  Use the |
  78 | QUOTEARG slot number N.                                     |
  79 `------------------------------------------------------------*/
  80
  81 const char *
  82 symbol_tag_get_n (symbol_t *symbol, int n)
  83 {
  84   return quotearg_n_style (n, escape_quoting_style, symbol->tag);
  85 }
  86
  87
  88 /*-------------------------------.
  89 | Print the tag of this SYMBOL.  |
  90 `-------------------------------*/
  91
  92 void
  93 symbol_tag_print (symbol_t *symbol, FILE *out)
  94 {
  95   fputs (symbol_tag_get (symbol), out);
  96 }
  97
  98
  99 /*------------------------------------------------------------------.
 100 | Set the TYPE_NAME associated to SYMBOL.  Does nothing if passed 0 |
 101 | as TYPE_NAME.                                                     |
 102 `------------------------------------------------------------------*/
 103
 104 void
 105 symbol_type_set (symbol_t *symbol, char *type_name)
 106 {
 107   if (type_name)
 108     {
 109       if (symbol->type_name)
 110         complain (_("type redeclaration for %s"), symbol->tag);
 111       symbol->type_name = type_name;
 112     }
 113 }
 114
 115
 116 /*------------------------------------------------------------------.
 117 | Set the PRECEDENCE associated to SYMBOL.  Does nothing if invoked |
 118 | with UNDEF_ASSOC as ASSOC.                                        |
 119 `------------------------------------------------------------------*/
 120
 121 void
 122 symbol_precedence_set (symbol_t *symbol,
 123                        int prec, associativity assoc)
 124 {
 125   if (assoc != undef_assoc)
 126     {
 127       if (symbol->prec != 0)
 128         complain (_("redefining precedence of %s"), symbol->tag);
 129       symbol->prec = prec;
 130       symbol->assoc = assoc;
 131     }
 132
 133   /* Only terminals have a precedence. */
 134   symbol_class_set (symbol, token_sym);
 135 }
 136
 137
 138 /*-------------------------------------.
 139 | Set the CLASS associated to SYMBOL.  |
 140 `-------------------------------------*/
 141
 142 void
 143 symbol_class_set (symbol_t *symbol, symbol_class class)
 144 {
 145   if (symbol->class != unknown_sym && symbol->class != class)
 146     complain (_("symbol %s redefined"), symbol->tag);
 147
 148   if (class == nterm_sym && symbol->class != nterm_sym)
 149     symbol->number = nvars++;
 150   else if (class == token_sym && symbol->number == NUMBER_UNDEFINED)
 151     symbol->number = ntokens++;
 152
 153   symbol->class = class;
 154 }
 155
 156
 157 /*-------------------------------------------------.
 158 | Set the USER_TOKEN_NUMBER associated to SYMBOL.  |
 159 `-------------------------------------------------*/
 160
 161 void
 162 symbol_user_token_number_set (symbol_t *symbol, int user_token_number)
 163 {
 164   assert (symbol->class == token_sym);
 165
 166   if (symbol->user_token_number != USER_NUMBER_UNDEFINED
 167       && symbol->user_token_number != user_token_number)
 168     complain (_("redefining user token number of %s"), symbol->tag);
 169
 170   symbol->user_token_number = user_token_number;
 171   /* User defined EOF token? */
 172   if (user_token_number == 0)
 173     {
 174       eoftoken = symbol;
 175       eoftoken->number = 0;
 176       /* It is always mapped to 0, so it was already counted in
 177          NTOKENS.  */
 178       --ntokens;
 179     }
 180 }
 181
 182
 183 /*------------.
 184 | Free THIS.  |
 185 `------------*/
 186
 187 static void
 188 symbol_free (symbol_t *this)
 189 {
 190 #if 0
 191   /* This causes crashes because one string can appear more
 192      than once.  */
 193   XFREE (this->type_name);
 194 #endif
 195   XFREE (this->tag);
 196   XFREE (this);
 197 }
 198
 199
 200 /*-----------------------------------------------------------.
 201 | If THIS is not defined, report an error, and consider it a |
 202 | nonterminal.                                               |
 203 `-----------------------------------------------------------*/
 204
 205 static bool
 206 symbol_check_defined (symbol_t *this)
 207 {
 208   if (this->class == unknown_sym)
 209     {
 210       complain_at
 211         (this->location,
 212          _("symbol %s is used, but is not defined as a token and has no rules"),
 213          this->tag);
 214       this->class = nterm_sym;
 215       this->number = nvars++;
 216     }
 217
 218   return TRUE;
 219 }
 220
 221
 222 /*-------------------------------------------------------------------.
 223 | Declare the new SYMBOL.  Make it an alias of SYMVAL, and type them |
 224 | with TYPENAME.                                                     |
 225 `-------------------------------------------------------------------*/
 226
 227 void
 228 symbol_make_alias (symbol_t *symbol, symbol_t *symval)
 229 {
 230   if (symval->alias)
 231     warn (_("symbol `%s' used more than once as a literal string"),
 232           symval->tag);
 233   else if (symbol->alias)
 234     warn (_("symbol `%s' given more than one literal string"),
 235           symbol->tag);
 236   else
 237     {
 238       symval->class = token_sym;
 239       symval->user_token_number = symbol->user_token_number;
 240       symbol->user_token_number = USER_NUMBER_ALIAS;
 241       symval->alias = symbol;
 242       symbol->alias = symval;
 243       /* symbol and symval combined are only one symbol */
 244       nsyms--;
 245       ntokens--;
 246       assert (ntokens == symbol->number || ntokens == symval->number);
 247       symbol->number = symval->number =
 248         (symval->number < symbol->number) ? symval->number : symbol->number;
 249     }
 250 }
 251
 252
 253 /*---------------------------------------------------------.
 254 | Check that THIS, and its alias, have same precedence and |
 255 | associativity.                                           |
 256 `---------------------------------------------------------*/
 257
 258 static bool
 259 symbol_check_alias_consistence (symbol_t *this)
 260 {
 261   /* Check only those who _are_ the aliases. */
 262   if (this->alias && this->user_token_number == USER_NUMBER_ALIAS)
 263     {
 264       if (this->prec != this->alias->prec)
 265         {
 266           if (this->prec != 0 && this->alias->prec != 0)
 267             complain (_("conflicting precedences for %s and %s"),
 268                       this->tag, this->alias->tag);
 269           if (this->prec != 0)
 270             this->alias->prec = this->prec;
 271           else
 272             this->prec = this->alias->prec;
 273         }
 274
 275       if (this->assoc != this->alias->assoc)
 276         {
 277           /* FIXME: For some reason (probably the S/R => keep the S),
 278              the right assoc is chosen has the ``not set''.  This is
 279              not nice, fix this!  */
 280           if (this->assoc != right_assoc
 281               && this->alias->assoc != right_assoc)
 282             complain (_("conflicting associativities for %s and %s"),
 283                       this->tag, this->alias->tag);
 284           if (this->assoc != 0)
 285             this->alias->assoc = this->assoc;
 286           else
 287             this->assoc = this->alias->assoc;
 288         }
 289     }
 290   return TRUE;
 291 }
 292
 293
 294 /*-------------------------------------------------------------------.
 295 | Assign a symbol number, and write the definition of the token name |
 296 | into FDEFINES.  Put in SYMBOLS.                                    |
 297 `-------------------------------------------------------------------*/
 298
 299 static bool
 300 symbol_pack (symbol_t *this)
 301 {
 302   if (this->class == nterm_sym)
 303     {
 304       this->number += ntokens;
 305     }
 306   else if (this->alias)
 307     {
 308       /* This symbol and its alias are a single token defn.
 309          Allocate a tokno, and assign to both check agreement of
 310          prec and assoc fields and make both the same */
 311       if (this->number == NUMBER_UNDEFINED)
 312         {
 313           if (this == eoftoken || this->alias == eoftoken)
 314             this->number = this->alias->number = 0;
 315           else
 316             {
 317               assert (this->alias->number != NUMBER_UNDEFINED);
 318               this->number = this->alias->number;
 319             }
 320         }
 321       /* Do not do processing below for USER_NUMBER_ALIASs.  */
 322       if (this->user_token_number == USER_NUMBER_ALIAS)
 323         return TRUE;
 324     }
 325   else /* this->class == token_sym */
 326     {
 327       assert (this->number != NUMBER_UNDEFINED);
 328     }
 329
 330   symbols[this->number] = this;
 331   return TRUE;
 332 }
 333
 334
 335
 336
 337 /*--------------------------------------------------.
 338 | Put THIS in TOKEN_TRANSLATIONS if it is a token.  |
 339 `--------------------------------------------------*/
 340
 341 static bool
 342 symbol_translation (symbol_t *this)
 343 {
 344   /* Non-terminal? */
 345   if (this->class == token_sym
 346       && this->user_token_number != USER_NUMBER_ALIAS)
 347     {
 348       /* A token which translation has already been set? */
 349       if (token_translations[this->user_token_number] != undeftoken->number)
 350         complain (_("tokens %s and %s both assigned number %d"),
 351                   symbols[token_translations[this->user_token_number]]->tag,
 352                   this->tag, this->user_token_number);
 353
 354       token_translations[this->user_token_number] = this->number;
 355     }
 356
 357   return TRUE;
 358 }
 359
 360
 361 /*----------------------.
 362 | A symbol hash table.  |
 363 `----------------------*/
 364
 365 /* Initial capacity of symbols hash table.  */
 366 #define HT_INITIAL_CAPACITY 257
 367
 368 static struct hash_table *symbol_table = NULL;
 369
 370 static bool
 371 hash_compare_symbol_t (const symbol_t *m1, const symbol_t *m2)
 372 {
 373   return strcmp (m1->tag, m2->tag) ? FALSE : TRUE;
 374 }
 375
 376 static unsigned int
 377 hash_symbol_t (const symbol_t *m, unsigned int tablesize)
 378 {
 379   return hash_string (m->tag, tablesize);
 380 }
 381
 382
 383 /*-------------------------------.
 384 | Create the symbol hash table.  |
 385 `-------------------------------*/
 386
 387 void
 388 symbols_new (void)
 389 {
 390   symbol_table = hash_initialize (HT_INITIAL_CAPACITY,
 391                                   NULL,
 392                                   (Hash_hasher) hash_symbol_t,
 393                                   (Hash_comparator) hash_compare_symbol_t,
 394                                   (Hash_data_freer) symbol_free);
 395 }
 396
 397
 398 /*----------------------------------------------------------------.
 399 | Find the symbol named KEY, and return it.  If it does not exist |
 400 | yet, create it.                                                 |
 401 `----------------------------------------------------------------*/
 402
 403 symbol_t *
 404 getsym (const char *key, location_t location)
 405 {
 406   symbol_t probe;
 407   symbol_t *entry;
 408
 409   (const char *) probe.tag = key;
 410   entry = hash_lookup (symbol_table, &probe);
 411
 412   if (!entry)
 413     {
 414       /* First insertion in the hash. */
 415       entry = symbol_new (key, location);
 416       hash_insert (symbol_table, entry);
 417     }
 418   return entry;
 419 }
 420
 421
 422 /*-------------------.
 423 | Free the symbols.  |
 424 `-------------------*/
 425
 426 void
 427 symbols_free (void)
 428 {
 429   hash_free (symbol_table);
 430 }
 431
 432
 433 /*---------------------------------------------------------------.
 434 | Look for undefined symbols, report an error, and consider them |
 435 | terminals.                                                     |
 436 `---------------------------------------------------------------*/
 437
 438 void
 439 symbols_do (symbol_processor processor, void *processor_data)
 440 {
 441   hash_do_for_each (symbol_table,
 442                     (Hash_processor) processor,
 443                     processor_data);
 444 }
 445
 446
 447 /*--------------------------------------------------------------.
 448 | Check that all the symbols are defined.  Report any undefined |
 449 | symbols and consider them nonterminals.                       |
 450 `--------------------------------------------------------------*/
 451
 452 void
 453 symbols_check_defined (void)
 454 {
 455   symbols_do (symbol_check_defined, NULL);
 456 }
 457
 458 /*------------------------------------------------------------------.
 459 | Set TOKEN_TRANSLATIONS.  Check that no two symbols share the same |
 460 | number.                                                           |
 461 `------------------------------------------------------------------*/
 462
 463 static void
 464 symbols_token_translations_init (void)
 465 {
 466   int num_256_available_p = TRUE;
 467   int i;
 468
 469   /* Find the highest user token number, and whether 256, the POSIX
 470      preferred user token number for the error token, is used.  */
 471   max_user_token_number = 0;
 472   for (i = 0; i < ntokens; ++i)
 473     {
 474       symbol_t *this = symbols[i];
 475       if (this->user_token_number != USER_NUMBER_UNDEFINED)
 476         {
 477           if (this->user_token_number > max_user_token_number)
 478             max_user_token_number = this->user_token_number;
 479           if (this->user_token_number == 256)
 480             num_256_available_p = FALSE;
 481         }
 482     }
 483
 484   /* If 256 is not used, assign it to error, to follow POSIX.  */
 485   if (num_256_available_p
 486       && errtoken->user_token_number == USER_NUMBER_UNDEFINED)
 487     errtoken->user_token_number = 256;
 488
 489   /* Set the missing user numbers. */
 490   if (max_user_token_number < 256)
 491     max_user_token_number = 256;
 492
 493   for (i = 0; i < ntokens; ++i)
 494     {
 495       symbol_t *this = symbols[i];
 496       if (this->user_token_number == USER_NUMBER_UNDEFINED)
 497         this->user_token_number = ++max_user_token_number;
 498       if (this->user_token_number > max_user_token_number)
 499         max_user_token_number = this->user_token_number;
 500     }
 501
 502   token_translations = XCALLOC (symbol_number_t, max_user_token_number + 1);
 503
 504   /* Initialize all entries for literal tokens to 2, the internal
 505      token number for $undefined., which represents all invalid
 506      inputs.  */
 507   for (i = 0; i < max_user_token_number + 1; i++)
 508     token_translations[i] = undeftoken->number;
 509   symbols_do (symbol_translation, NULL);
 510 }
 511
 512
 513 /*----------------------------------------------------------------.
 514 | Assign symbol numbers, and write definition of token names into |
 515 | FDEFINES.  Set up vectors SYMBOL_TABLE, TAGS of symbols.        |
 516 `----------------------------------------------------------------*/
 517
 518 void
 519 symbols_pack (void)
 520 {
 521   symbols = XCALLOC (symbol_t *, nsyms);
 522
 523   symbols_do (symbol_check_alias_consistence, NULL);
 524   symbols_do (symbol_pack, NULL);
 525
 526   symbols_token_translations_init ();
 527
 528   if (startsymbol->class == unknown_sym)
 529     fatal_at (startsymbol_location,
 530               _("the start symbol %s is undefined"), startsymbol->tag);
 531   else if (startsymbol->class == token_sym)
 532     fatal_at (startsymbol_location,
 533               _("the start symbol %s is a token"), startsymbol->tag);
 534 }