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, location_t location, char *type_name)
 106 {
 107   if (type_name)
 108     {
 109       if (symbol->type_name)
 110         complain_at (location,
 111                      _("type redeclaration for %s"), symbol->tag);
 112       symbol->type_name = type_name;
 113     }
 114 }
 115
 116
 117 /*------------------------------------------------------------------.
 118 | Set the PRECEDENCE associated to SYMBOL.  Does nothing if invoked |
 119 | with UNDEF_ASSOC as ASSOC.                                        |
 120 `------------------------------------------------------------------*/
 121
 122 void
 123 symbol_precedence_set (symbol_t *symbol, location_t location,
 124                        int prec, associativity assoc)
 125 {
 126   if (assoc != undef_assoc)
 127     {
 128       if (symbol->prec != 0)
 129         complain_at (location,
 130                      _("redefining precedence of %s"), symbol->tag);
 131       symbol->prec = prec;
 132       symbol->assoc = assoc;
 133     }
 134
 135   /* Only terminals have a precedence. */
 136   symbol_class_set (symbol, token_sym);
 137 }
 138
 139
 140 /*-------------------------------------.
 141 | Set the CLASS associated to SYMBOL.  |
 142 `-------------------------------------*/
 143
 144 void
 145 symbol_class_set (symbol_t *symbol, symbol_class class)
 146 {
 147   if (symbol->class != unknown_sym && symbol->class != class)
 148     complain (_("symbol %s redefined"), symbol->tag);
 149
 150   if (class == nterm_sym && symbol->class != nterm_sym)
 151     symbol->number = nvars++;
 152   else if (class == token_sym && symbol->number == NUMBER_UNDEFINED)
 153     symbol->number = ntokens++;
 154
 155   symbol->class = class;
 156 }
 157
 158
 159 /*-------------------------------------------------.
 160 | Set the USER_TOKEN_NUMBER associated to SYMBOL.  |
 161 `-------------------------------------------------*/
 162
 163 void
 164 symbol_user_token_number_set (symbol_t *symbol, int user_token_number)
 165 {
 166   assert (symbol->class == token_sym);
 167
 168   if (symbol->user_token_number != USER_NUMBER_UNDEFINED
 169       && symbol->user_token_number != user_token_number)
 170     complain (_("redefining user token number of %s"), symbol->tag);
 171
 172   symbol->user_token_number = user_token_number;
 173   /* User defined EOF token? */
 174   if (user_token_number == 0)
 175     {
 176       eoftoken = symbol;
 177       eoftoken->number = 0;
 178       /* It is always mapped to 0, so it was already counted in
 179          NTOKENS.  */
 180       --ntokens;
 181     }
 182 }
 183
 184
 185 /*------------.
 186 | Free THIS.  |
 187 `------------*/
 188
 189 static void
 190 symbol_free (symbol_t *this)
 191 {
 192 #if 0
 193   /* This causes crashes because one string can appear more
 194      than once.  */
 195   XFREE (this->type_name);
 196 #endif
 197   XFREE (this->tag);
 198   XFREE (this);
 199 }
 200
 201
 202 /*-----------------------------------------------------------.
 203 | If THIS is not defined, report an error, and consider it a |
 204 | nonterminal.                                               |
 205 `-----------------------------------------------------------*/
 206
 207 static bool
 208 symbol_check_defined (symbol_t *this)
 209 {
 210   if (this->class == unknown_sym)
 211     {
 212       complain_at
 213         (this->location,
 214          _("symbol %s is used, but is not defined as a token and has no rules"),
 215          this->tag);
 216       this->class = nterm_sym;
 217       this->number = nvars++;
 218     }
 219
 220   return TRUE;
 221 }
 222
 223
 224 /*-------------------------------------------------------------------.
 225 | Declare the new SYMBOL.  Make it an alias of SYMVAL, and type them |
 226 | with TYPENAME.                                                     |
 227 `-------------------------------------------------------------------*/
 228
 229 void
 230 symbol_make_alias (symbol_t *symbol, symbol_t *symval)
 231 {
 232   if (symval->alias)
 233     warn (_("symbol `%s' used more than once as a literal string"),
 234           symval->tag);
 235   else if (symbol->alias)
 236     warn (_("symbol `%s' given more than one literal string"),
 237           symbol->tag);
 238   else
 239     {
 240       symval->class = token_sym;
 241       symval->user_token_number = symbol->user_token_number;
 242       symbol->user_token_number = USER_NUMBER_ALIAS;
 243       symval->alias = symbol;
 244       symbol->alias = symval;
 245       /* symbol and symval combined are only one symbol */
 246       nsyms--;
 247       ntokens--;
 248       assert (ntokens == symbol->number || ntokens == symval->number);
 249       symbol->number = symval->number =
 250         (symval->number < symbol->number) ? symval->number : symbol->number;
 251     }
 252 }
 253
 254
 255 /*---------------------------------------------------------.
 256 | Check that THIS, and its alias, have same precedence and |
 257 | associativity.                                           |
 258 `---------------------------------------------------------*/
 259
 260 static bool
 261 symbol_check_alias_consistence (symbol_t *this)
 262 {
 263   /* Check only those who _are_ the aliases. */
 264   if (this->alias && this->user_token_number == USER_NUMBER_ALIAS)
 265     {
 266       if (this->prec != this->alias->prec)
 267         {
 268           if (this->prec != 0 && this->alias->prec != 0)
 269             complain (_("conflicting precedences for %s and %s"),
 270                       this->tag, this->alias->tag);
 271           if (this->prec != 0)
 272             this->alias->prec = this->prec;
 273           else
 274             this->prec = this->alias->prec;
 275         }
 276
 277       if (this->assoc != this->alias->assoc)
 278         {
 279           /* FIXME: For some reason (probably the S/R => keep the S),
 280              the right assoc is chosen has the ``not set''.  This is
 281              not nice, fix this!  */
 282           if (this->assoc != right_assoc
 283               && this->alias->assoc != right_assoc)
 284             complain (_("conflicting associativities for %s and %s"),
 285                       this->tag, this->alias->tag);
 286           if (this->assoc != 0)
 287             this->alias->assoc = this->assoc;
 288           else
 289             this->assoc = this->alias->assoc;
 290         }
 291     }
 292   return TRUE;
 293 }
 294
 295
 296 /*-------------------------------------------------------------------.
 297 | Assign a symbol number, and write the definition of the token name |
 298 | into FDEFINES.  Put in SYMBOLS.                                    |
 299 `-------------------------------------------------------------------*/
 300
 301 static bool
 302 symbol_pack (symbol_t *this)
 303 {
 304   if (this->class == nterm_sym)
 305     {
 306       this->number += ntokens;
 307     }
 308   else if (this->alias)
 309     {
 310       /* This symbol and its alias are a single token defn.
 311          Allocate a tokno, and assign to both check agreement of
 312          prec and assoc fields and make both the same */
 313       if (this->number == NUMBER_UNDEFINED)
 314         {
 315           if (this == eoftoken || this->alias == eoftoken)
 316             this->number = this->alias->number = 0;
 317           else
 318             {
 319               assert (this->alias->number != NUMBER_UNDEFINED);
 320               this->number = this->alias->number;
 321             }
 322         }
 323       /* Do not do processing below for USER_NUMBER_ALIASs.  */
 324       if (this->user_token_number == USER_NUMBER_ALIAS)
 325         return TRUE;
 326     }
 327   else /* this->class == token_sym */
 328     {
 329       assert (this->number != NUMBER_UNDEFINED);
 330     }
 331
 332   symbols[this->number] = this;
 333   return TRUE;
 334 }
 335
 336
 337
 338
 339 /*--------------------------------------------------.
 340 | Put THIS in TOKEN_TRANSLATIONS if it is a token.  |
 341 `--------------------------------------------------*/
 342
 343 static bool
 344 symbol_translation (symbol_t *this)
 345 {
 346   /* Non-terminal? */
 347   if (this->class == token_sym
 348       && this->user_token_number != USER_NUMBER_ALIAS)
 349     {
 350       /* A token which translation has already been set? */
 351       if (token_translations[this->user_token_number] != undeftoken->number)
 352         complain (_("tokens %s and %s both assigned number %d"),
 353                   symbols[token_translations[this->user_token_number]]->tag,
 354                   this->tag, this->user_token_number);
 355
 356       token_translations[this->user_token_number] = this->number;
 357     }
 358
 359   return TRUE;
 360 }
 361
 362
 363 /*----------------------.
 364 | A symbol hash table.  |
 365 `----------------------*/
 366
 367 /* Initial capacity of symbols hash table.  */
 368 #define HT_INITIAL_CAPACITY 257
 369
 370 static struct hash_table *symbol_table = NULL;
 371
 372 static bool
 373 hash_compare_symbol_t (const symbol_t *m1, const symbol_t *m2)
 374 {
 375   return strcmp (m1->tag, m2->tag) ? FALSE : TRUE;
 376 }
 377
 378 static unsigned int
 379 hash_symbol_t (const symbol_t *m, unsigned int tablesize)
 380 {
 381   return hash_string (m->tag, tablesize);
 382 }
 383
 384
 385 /*-------------------------------.
 386 | Create the symbol hash table.  |
 387 `-------------------------------*/
 388
 389 void
 390 symbols_new (void)
 391 {
 392   symbol_table = hash_initialize (HT_INITIAL_CAPACITY,
 393                                   NULL,
 394                                   (Hash_hasher) hash_symbol_t,
 395                                   (Hash_comparator) hash_compare_symbol_t,
 396                                   (Hash_data_freer) symbol_free);
 397 }
 398
 399
 400 /*----------------------------------------------------------------.
 401 | Find the symbol named KEY, and return it.  If it does not exist |
 402 | yet, create it.                                                 |
 403 `----------------------------------------------------------------*/
 404
 405 symbol_t *
 406 getsym (const char *key, location_t location)
 407 {
 408   symbol_t probe;
 409   symbol_t *entry;
 410
 411   (const char *) probe.tag = key;
 412   entry = hash_lookup (symbol_table, &probe);
 413
 414   if (!entry)
 415     {
 416       /* First insertion in the hash. */
 417       entry = symbol_new (key, location);
 418       hash_insert (symbol_table, entry);
 419     }
 420   return entry;
 421 }
 422
 423
 424 /*-------------------.
 425 | Free the symbols.  |
 426 `-------------------*/
 427
 428 void
 429 symbols_free (void)
 430 {
 431   hash_free (symbol_table);
 432 }
 433
 434
 435 /*---------------------------------------------------------------.
 436 | Look for undefined symbols, report an error, and consider them |
 437 | terminals.                                                     |
 438 `---------------------------------------------------------------*/
 439
 440 void
 441 symbols_do (symbol_processor processor, void *processor_data)
 442 {
 443   hash_do_for_each (symbol_table,
 444                     (Hash_processor) processor,
 445                     processor_data);
 446 }
 447
 448
 449 /*--------------------------------------------------------------.
 450 | Check that all the symbols are defined.  Report any undefined |
 451 | symbols and consider them nonterminals.                       |
 452 `--------------------------------------------------------------*/
 453
 454 void
 455 symbols_check_defined (void)
 456 {
 457   symbols_do (symbol_check_defined, NULL);
 458 }
 459
 460 /*------------------------------------------------------------------.
 461 | Set TOKEN_TRANSLATIONS.  Check that no two symbols share the same |
 462 | number.                                                           |
 463 `------------------------------------------------------------------*/
 464
 465 static void
 466 symbols_token_translations_init (void)
 467 {
 468   int num_256_available_p = TRUE;
 469   int i;
 470
 471   /* Find the highest user token number, and whether 256, the POSIX
 472      preferred user token number for the error token, is used.  */
 473   max_user_token_number = 0;
 474   for (i = 0; i < ntokens; ++i)
 475     {
 476       symbol_t *this = symbols[i];
 477       if (this->user_token_number != USER_NUMBER_UNDEFINED)
 478         {
 479           if (this->user_token_number > max_user_token_number)
 480             max_user_token_number = this->user_token_number;
 481           if (this->user_token_number == 256)
 482             num_256_available_p = FALSE;
 483         }
 484     }
 485
 486   /* If 256 is not used, assign it to error, to follow POSIX.  */
 487   if (num_256_available_p
 488       && errtoken->user_token_number == USER_NUMBER_UNDEFINED)
 489     errtoken->user_token_number = 256;
 490
 491   /* Set the missing user numbers. */
 492   if (max_user_token_number < 256)
 493     max_user_token_number = 256;
 494
 495   for (i = 0; i < ntokens; ++i)
 496     {
 497       symbol_t *this = symbols[i];
 498       if (this->user_token_number == USER_NUMBER_UNDEFINED)
 499         this->user_token_number = ++max_user_token_number;
 500       if (this->user_token_number > max_user_token_number)
 501         max_user_token_number = this->user_token_number;
 502     }
 503
 504   token_translations = XCALLOC (symbol_number_t, max_user_token_number + 1);
 505
 506   /* Initialize all entries for literal tokens to 2, the internal
 507      token number for $undefined., which represents all invalid
 508      inputs.  */
 509   for (i = 0; i < max_user_token_number + 1; i++)
 510     token_translations[i] = undeftoken->number;
 511   symbols_do (symbol_translation, NULL);
 512 }
 513
 514
 515 /*----------------------------------------------------------------.
 516 | Assign symbol numbers, and write definition of token names into |
 517 | FDEFINES.  Set up vectors SYMBOL_TABLE, TAGS of symbols.        |
 518 `----------------------------------------------------------------*/
 519
 520 void
 521 symbols_pack (void)
 522 {
 523   symbols = XCALLOC (symbol_t *, nsyms);
 524
 525   symbols_do (symbol_check_alias_consistence, NULL);
 526   symbols_do (symbol_pack, NULL);
 527
 528   symbols_token_translations_init ();
 529
 530   if (startsymbol->class == unknown_sym)
 531     fatal_at (startsymbol_location,
 532               _("the start symbol %s is undefined"), startsymbol->tag);
 533   else if (startsymbol->class == token_sym)
 534     fatal_at (startsymbol_location,
 535               _("the start symbol %s is a token"), startsymbol->tag);
 536 }