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 "hash.h"
  24 #include "complain.h"
  25 #include "symtab.h"
  26 #include "gram.h"
  27
  28 /*------------------------.
  29 | Distinguished symbols.  |
  30 `------------------------*/
  31
  32 symbol_t *errtoken = NULL;
  33 symbol_t *undeftoken = NULL;
  34 symbol_t *eoftoken = NULL;
  35 symbol_t *axiom = NULL;
  36 symbol_t *startsymbol = NULL;
  37
  38 /*---------------------------------.
  39 | Create a new symbol, named TAG.  |
  40 `---------------------------------*/
  41
  42 static symbol_t *
  43 symbol_new (const char *tag)
  44 {
  45   symbol_t *res = XMALLOC (symbol_t, 1);
  46
  47   res->tag = xstrdup (tag);
  48   res->type_name = NULL;
  49   res->number = NUMBER_UNDEFINED;
  50   res->prec = 0;
  51   res->assoc = right_assoc;
  52   res->user_token_number = USER_NUMBER_UNDEFINED;
  53   res->alias = NULL;
  54   res->class = unknown_sym;
  55
  56   nsyms++;
  57
  58   return res;
  59 }
  60
  61
  62 /*------------.
  63 | Free THIS.  |
  64 `------------*/
  65
  66 static void
  67 symbol_free (symbol_t *this)
  68 {
  69 #if 0
  70   /* This causes crashes because one string can appear more
  71      than once.  */
  72   XFREE (this->type_name);
  73 #endif
  74   XFREE (this->tag);
  75   XFREE (this);
  76 }
  77
  78
  79 /*-----------------------------------------------------------.
  80 | If THIS is not defined, report an error, and consider it a |
  81 | nonterminal.                                               |
  82 `-----------------------------------------------------------*/
  83
  84 static bool
  85 symbol_check_defined (symbol_t *this)
  86 {
  87   if (this->class == unknown_sym)
  88     {
  89       complain
  90         (_("symbol %s is used, but is not defined as a token and has no rules"),
  91          this->tag);
  92       this->class = nterm_sym;
  93       this->number = nvars++;
  94     }
  95
  96   return TRUE;
  97 }
  98
  99
 100 /*-------------------------------------------------------------------.
 101 | Declare the new SYMBOL.  Make it an alias of SYMVAL, and type them |
 102 | with TYPENAME.                                                     |
 103 `-------------------------------------------------------------------*/
 104
 105 void
 106 symbol_make_alias (symbol_t *symbol, symbol_t *symval, char *typename)
 107 {
 108   if (symval->alias)
 109     warn (_("symbol `%s' used more than once as a literal string"),
 110           symval->tag);
 111   else if (symbol->alias)
 112     warn (_("symbol `%s' given more than one literal string"),
 113           symbol->tag);
 114   else
 115     {
 116       symval->class = token_sym;
 117       symval->type_name = typename;
 118       symval->user_token_number = symbol->user_token_number;
 119       symbol->user_token_number = USER_NUMBER_ALIAS;
 120       symval->alias = symbol;
 121       symbol->alias = symval;
 122       /* symbol and symval combined are only one symbol */
 123       nsyms--;
 124       ntokens--;
 125       assert (ntokens == symbol->number || ntokens == symval->number);
 126       symbol->number = symval->number =
 127         (symval->number < symbol->number) ? symval->number : symbol->number;
 128     }
 129 }
 130
 131
 132 /*---------------------------------------------------------.
 133 | Check that THIS, and its alias, have same precedence and |
 134 | associativity.                                           |
 135 `---------------------------------------------------------*/
 136
 137 static bool
 138 symbol_check_alias_consistence (symbol_t *this)
 139 {
 140   /* Check only those who _are_ the aliases. */
 141   if (this->alias && this->user_token_number == USER_NUMBER_ALIAS)
 142     {
 143       if (this->prec != this->alias->prec)
 144         {
 145           if (this->prec != 0 && this->alias->prec != 0)
 146             complain (_("conflicting precedences for %s and %s"),
 147                       this->tag, this->alias->tag);
 148           if (this->prec != 0)
 149             this->alias->prec = this->prec;
 150           else
 151             this->prec = this->alias->prec;
 152         }
 153
 154       if (this->assoc != this->alias->assoc)
 155         {
 156           if (this->assoc != 0 && this->alias->assoc != 0)
 157             complain (_("conflicting assoc values for %s and %s"),
 158                       this->tag, this->alias->tag);
 159           if (this->assoc != 0)
 160             this->alias->assoc = this->assoc;
 161           else
 162             this->assoc = this->alias->assoc;
 163         }
 164     }
 165   return TRUE;
 166 }
 167
 168
 169 /*-------------------------------------------------------------------.
 170 | Assign a symbol number, and write the definition of the token name |
 171 | into FDEFINES.  Put in SYMBOLS.                                    |
 172 `-------------------------------------------------------------------*/
 173
 174 static bool
 175 symbol_pack (symbol_t *this)
 176 {
 177   if (this->class == nterm_sym)
 178     {
 179       this->number += ntokens;
 180     }
 181   else if (this->alias)
 182     {
 183       /* This symbol and its alias are a single token defn.
 184          Allocate a tokno, and assign to both check agreement of
 185          prec and assoc fields and make both the same */
 186       if (this->number == NUMBER_UNDEFINED)
 187         {
 188           if (this == eoftoken || this->alias == eoftoken)
 189             this->number = this->alias->number = 0;
 190           else
 191             {
 192               assert (this->alias->number != NUMBER_UNDEFINED);
 193               this->number = this->alias->number;
 194             }
 195         }
 196       /* Do not do processing below for USER_NUMBER_ALIASs.  */
 197       if (this->user_token_number == USER_NUMBER_ALIAS)
 198         return TRUE;
 199     }
 200   else /* this->class == token_sym */
 201     {
 202       assert (this->number != NUMBER_UNDEFINED);
 203     }
 204
 205   symbols[this->number] = this;
 206   return TRUE;
 207 }
 208
 209
 210
 211
 212 /*--------------------------------------------------.
 213 | Put THIS in TOKEN_TRANSLATIONS if it is a token.  |
 214 `--------------------------------------------------*/
 215
 216 static bool
 217 symbol_translation (symbol_t *this)
 218 {
 219   /* Non-terminal? */
 220   if (this->class == token_sym
 221       && this->user_token_number != USER_NUMBER_ALIAS)
 222     {
 223       /* A token which translation has already been set? */
 224       if (token_translations[this->user_token_number] != undeftoken->number)
 225         complain (_("tokens %s and %s both assigned number %d"),
 226                   symbols[token_translations[this->user_token_number]]->tag,
 227                   this->tag, this->user_token_number);
 228
 229       token_translations[this->user_token_number] = this->number;
 230     }
 231
 232   return TRUE;
 233 }
 234
 235
 236 /*----------------------.
 237 | A symbol hash table.  |
 238 `----------------------*/
 239
 240 /* Initial capacity of symbols hash table.  */
 241 #define HT_INITIAL_CAPACITY 257
 242
 243 static struct hash_table *symbol_table = NULL;
 244
 245 static bool
 246 hash_compare_symbol_t (const symbol_t *m1, const symbol_t *m2)
 247 {
 248   return strcmp (m1->tag, m2->tag) ? FALSE : TRUE;
 249 }
 250
 251 static unsigned int
 252 hash_symbol_t (const symbol_t *m, unsigned int tablesize)
 253 {
 254   return hash_string (m->tag, tablesize);
 255 }
 256
 257
 258 /*-------------------------------.
 259 | Create the symbol hash table.  |
 260 `-------------------------------*/
 261
 262 void
 263 symbols_new (void)
 264 {
 265   symbol_table = hash_initialize (HT_INITIAL_CAPACITY,
 266                                   NULL,
 267                                   (Hash_hasher) hash_symbol_t,
 268                                   (Hash_comparator) hash_compare_symbol_t,
 269                                   (Hash_data_freer) symbol_free);
 270 }
 271
 272
 273 /*----------------------------------------------------------------.
 274 | Find the symbol named KEY, and return it.  If it does not exist |
 275 | yet, create it.                                                 |
 276 `----------------------------------------------------------------*/
 277
 278 symbol_t *
 279 getsym (const char *key)
 280 {
 281   symbol_t probe;
 282   symbol_t *entry;
 283
 284   (const char *) probe.tag = key;
 285   entry = hash_lookup (symbol_table, &probe);
 286
 287   if (!entry)
 288     {
 289       /* First insertion in the hash. */
 290       entry = symbol_new (key);
 291       hash_insert (symbol_table, entry);
 292     }
 293   return entry;
 294 }
 295
 296
 297 /*-------------------.
 298 | Free the symbols.  |
 299 `-------------------*/
 300
 301 void
 302 symbols_free (void)
 303 {
 304   hash_free (symbol_table);
 305 }
 306
 307
 308 /*---------------------------------------------------------------.
 309 | Look for undefined symbols, report an error, and consider them |
 310 | terminals.                                                     |
 311 `---------------------------------------------------------------*/
 312
 313 void
 314 symbols_do (symbol_processor processor, void *processor_data)
 315 {
 316   hash_do_for_each (symbol_table,
 317                     (Hash_processor) processor,
 318                     processor_data);
 319 }
 320
 321
 322 /*--------------------------------------------------------------.
 323 | Check that all the symbols are defined.  Report any undefined |
 324 | symbols and consider them nonterminals.                       |
 325 `--------------------------------------------------------------*/
 326
 327 void
 328 symbols_check_defined (void)
 329 {
 330   symbols_do (symbol_check_defined, NULL);
 331 }
 332
 333 /*------------------------------------------------------------------.
 334 | Set TOKEN_TRANSLATIONS.  Check that no two symbols share the same |
 335 | number.                                                           |
 336 `------------------------------------------------------------------*/
 337
 338 static void
 339 symbols_token_translations_init (void)
 340 {
 341   int num_256_available_p = TRUE;
 342   int i;
 343
 344   /* Find the highest user token number, and whether 256, the POSIX
 345      preferred user token number for the error token, is used.  */
 346   max_user_token_number = 0;
 347   for (i = 0; i < ntokens; ++i)
 348     {
 349       symbol_t *this = symbols[i];
 350       if (this->user_token_number != USER_NUMBER_UNDEFINED)
 351         {
 352           if (this->user_token_number > max_user_token_number)
 353             max_user_token_number = this->user_token_number;
 354           if (this->user_token_number == 256)
 355             num_256_available_p = FALSE;
 356         }
 357     }
 358
 359   /* If 256 is not used, assign it to error, to follow POSIX.  */
 360   if (num_256_available_p
 361       && errtoken->user_token_number == USER_NUMBER_UNDEFINED)
 362     errtoken->user_token_number = 256;
 363
 364   /* Set the missing user numbers. */
 365   if (max_user_token_number < 256)
 366     max_user_token_number = 256;
 367
 368   for (i = 0; i < ntokens; ++i)
 369     {
 370       symbol_t *this = symbols[i];
 371       if (this->user_token_number == USER_NUMBER_UNDEFINED)
 372         this->user_token_number = ++max_user_token_number;
 373       if (this->user_token_number > max_user_token_number)
 374         max_user_token_number = this->user_token_number;
 375     }
 376
 377   token_translations = XCALLOC (symbol_number_t, max_user_token_number + 1);
 378
 379   /* Initialize all entries for literal tokens to 2, the internal
 380      token number for $undefined., which represents all invalid
 381      inputs.  */
 382   for (i = 0; i < max_user_token_number + 1; i++)
 383     token_translations[i] = undeftoken->number;
 384   symbols_do (symbol_translation, NULL);
 385 }
 386
 387
 388 /*----------------------------------------------------------------.
 389 | Assign symbol numbers, and write definition of token names into |
 390 | FDEFINES.  Set up vectors SYMBOL_TABLE, TAGS of symbols.        |
 391 `----------------------------------------------------------------*/
 392
 393 void
 394 symbols_pack (void)
 395 {
 396   symbols = XCALLOC (symbol_t *, nsyms);
 397
 398   symbols_do (symbol_check_alias_consistence, NULL);
 399   symbols_do (symbol_pack, NULL);
 400
 401   symbols_token_translations_init ();
 402
 403   if (startsymbol->class == unknown_sym)
 404     fatal (_("the start symbol %s is undefined"), startsymbol->tag);
 405   else if (startsymbol->class == token_sym)
 406     fatal (_("the start symbol %s is a token"), startsymbol->tag);
 407 }