src/LR0.c

   1 /* Generate the nondeterministic finite state machine for bison,
   2    Copyright 1984, 1986, 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 /* See comments in state.h for the data structures that represent it.
  23    The entry point is generate_states.  */
  24
  25 #include "system.h"
  26 #include "bitset.h"
  27 #include "quotearg.h"
  28 #include "symtab.h"
  29 #include "gram.h"
  30 #include "getargs.h"
  31 #include "reader.h"
  32 #include "gram.h"
  33 #include "state.h"
  34 #include "complain.h"
  35 #include "closure.h"
  36 #include "LR0.h"
  37 #include "lalr.h"
  38 #include "reduce.h"
  39
  40 typedef struct state_list_s
  41 {
  42   struct state_list_s *next;
  43   state_t *state;
  44 } state_list_t;
  45
  46 static state_list_t *first_state = NULL;
  47 static state_list_t *last_state = NULL;
  48
  49 static void
  50 state_list_append (state_t *state)
  51 {
  52   state_list_t *node = XMALLOC (state_list_t, 1);
  53   node->next = NULL;
  54   node->state = state;
  55
  56   if (!first_state)
  57     first_state = node;
  58   if (last_state)
  59     last_state->next = node;
  60   last_state = node;
  61 }
  62
  63 static int nshifts;
  64 static symbol_number_t *shift_symbol = NULL;
  65
  66 static short *redset = NULL;
  67 static state_number_t *shiftset = NULL;
  68
  69 static item_number_t **kernel_base = NULL;
  70 static int *kernel_size = NULL;
  71 static item_number_t *kernel_items = NULL;
  72
  73 \f
  74 static void
  75 allocate_itemsets (void)
  76 {
  77   symbol_number_t i;
  78   rule_number_t r;
  79   item_number_t *rhsp;
  80
  81   /* Count the number of occurrences of all the symbols in RITEMS.
  82      Note that useless productions (hence useless nonterminals) are
  83      browsed too, hence we need to allocate room for _all_ the
  84      symbols.  */
  85   int count = 0;
  86   short *symbol_count = XCALLOC (short, nsyms + nuseless_nonterminals);
  87
  88   for (r = 1; r < nrules + 1; ++r)
  89     for (rhsp = rules[r].rhs; *rhsp >= 0; ++rhsp)
  90       {
  91         count++;
  92         symbol_count[*rhsp]++;
  93       }
  94
  95   /* See comments before new_itemsets.  All the vectors of items
  96      live inside KERNEL_ITEMS.  The number of active items after
  97      some symbol cannot be more than the number of times that symbol
  98      appears as an item, which is SYMBOL_COUNT[SYMBOL].
  99      We allocate that much space for each symbol.  */
 100
 101   kernel_base = XCALLOC (item_number_t *, nsyms);
 102   if (count)
 103     kernel_items = XCALLOC (item_number_t, count);
 104
 105   count = 0;
 106   for (i = 0; i < nsyms; i++)
 107     {
 108       kernel_base[i] = kernel_items + count;
 109       count += symbol_count[i];
 110     }
 111
 112   free (symbol_count);
 113   kernel_size = XCALLOC (int, nsyms);
 114 }
 115
 116
 117 static void
 118 allocate_storage (void)
 119 {
 120   allocate_itemsets ();
 121
 122   shiftset = XCALLOC (state_number_t, nsyms);
 123   redset = XCALLOC (short, nrules + 1);
 124   state_hash_new ();
 125   shift_symbol = XCALLOC (symbol_number_t, nsyms);
 126 }
 127
 128
 129 static void
 130 free_storage (void)
 131 {
 132   free (shift_symbol);
 133   free (redset);
 134   free (shiftset);
 135   free (kernel_base);
 136   free (kernel_size);
 137   XFREE (kernel_items);
 138   state_hash_free ();
 139 }
 140
 141
 142
 143
 144 /*---------------------------------------------------------------.
 145 | Find which symbols can be shifted in STATE, and for each one   |
 146 | record which items would be active after that shift.  Uses the |
 147 | contents of itemset.                                           |
 148 |                                                                |
 149 | shift_symbol is set to a vector of the symbols that can be     |
 150 | shifted.  For each symbol in the grammar, kernel_base[symbol]  |
 151 | points to a vector of item numbers activated if that symbol is |
 152 | shifted, and kernel_size[symbol] is their numbers.             |
 153 `---------------------------------------------------------------*/
 154
 155 static void
 156 new_itemsets (state_t *state)
 157 {
 158   int i;
 159
 160   if (trace_flag)
 161     fprintf (stderr, "Entering new_itemsets, state = %d\n",
 162              state->number);
 163
 164   for (i = 0; i < nsyms; i++)
 165     kernel_size[i] = 0;
 166
 167   nshifts = 0;
 168
 169   for (i = 0; i < nritemset; ++i)
 170     if (ritem[itemset[i]] >= 0)
 171       {
 172         symbol_number_t symbol
 173           = item_number_as_symbol_number (ritem[itemset[i]]);
 174         if (!kernel_size[symbol])
 175           {
 176             shift_symbol[nshifts] = symbol;
 177             nshifts++;
 178           }
 179
 180         kernel_base[symbol][kernel_size[symbol]] = itemset[i] + 1;
 181         kernel_size[symbol]++;
 182       }
 183 }
 184
 185
 186
 187 /*-----------------------------------------------------------------.
 188 | Subroutine of get_state.  Create a new state for those items, if |
 189 | necessary.                                                       |
 190 `-----------------------------------------------------------------*/
 191
 192 static state_t *
 193 new_state (symbol_number_t symbol, size_t core_size, item_number_t *core)
 194 {
 195   state_t *res;
 196
 197   if (trace_flag)
 198     fprintf (stderr, "Entering new_state, state = %d, symbol = %d (%s)\n",
 199              nstates, symbol, symbol_tag_get (symbols[symbol]));
 200
 201   res = state_new (symbol, core_size, core);
 202   state_hash_insert (res);
 203
 204   /* If this is the eoftoken, and this is not the initial state, then
 205      this is the final state.  */
 206   if (symbol == 0 && first_state)
 207     final_state = res;
 208
 209   state_list_append (res);
 210   return res;
 211 }
 212
 213
 214 /*--------------------------------------------------------------.
 215 | Find the state number for the state we would get to (from the |
 216 | current state) by shifting symbol.  Create a new state if no  |
 217 | equivalent one exists already.  Used by append_states.        |
 218 `--------------------------------------------------------------*/
 219
 220 static state_number_t
 221 get_state (symbol_number_t symbol, size_t core_size, item_number_t *core)
 222 {
 223   state_t *sp;
 224
 225   if (trace_flag)
 226     fprintf (stderr, "Entering get_state, symbol = %d (%s)\n",
 227              symbol, symbol_tag_get (symbols[symbol]));
 228
 229   sp = state_hash_lookup (core_size, core);
 230   if (!sp)
 231     sp = new_state (symbol, core_size, core);
 232
 233   if (trace_flag)
 234     fprintf (stderr, "Exiting get_state => %d\n", sp->number);
 235
 236   return sp->number;
 237 }
 238
 239 /*------------------------------------------------------------------.
 240 | Use the information computed by new_itemsets to find the state    |
 241 | numbers reached by each shift transition from STATE.              |
 242 |                                                                   |
 243 | SHIFTSET is set up as a vector of state numbers of those states.  |
 244 `------------------------------------------------------------------*/
 245
 246 static void
 247 append_states (state_t *state)
 248 {
 249   int i;
 250   int j;
 251   symbol_number_t symbol;
 252
 253   if (trace_flag)
 254     fprintf (stderr, "Entering append_states, state = %d\n",
 255              state->number);
 256
 257   /* first sort shift_symbol into increasing order */
 258
 259   for (i = 1; i < nshifts; i++)
 260     {
 261       symbol = shift_symbol[i];
 262       j = i;
 263       while (j > 0 && shift_symbol[j - 1] > symbol)
 264         {
 265           shift_symbol[j] = shift_symbol[j - 1];
 266           j--;
 267         }
 268       shift_symbol[j] = symbol;
 269     }
 270
 271   for (i = 0; i < nshifts; i++)
 272     {
 273       symbol = shift_symbol[i];
 274       shiftset[i] = get_state (symbol,
 275                                kernel_size[symbol], kernel_base[symbol]);
 276     }
 277 }
 278
 279
 280 static void
 281 new_states (void)
 282 {
 283   /* The 0 at the lhs is the index of the item of this initial rule.  */
 284   kernel_base[0][0] = 0;
 285   kernel_size[0] = 1;
 286   state_list_append (new_state (0, kernel_size[0], kernel_base[0]));
 287 }
 288
 289
 290
 291 /*----------------------------------------------------------------.
 292 | Find which rules can be used for reduction transitions from the |
 293 | current state and make a reductions structure for the state to  |
 294 | record their rule numbers.                                      |
 295 `----------------------------------------------------------------*/
 296
 297 static void
 298 save_reductions (state_t *state)
 299 {
 300   int count = 0;
 301   int i;
 302
 303   /* If this is the final state, we want it to have no reductions at
 304      all, although it has one for `START_SYMBOL EOF .'.  */
 305   if (final_state && state->number == final_state->number)
 306     return;
 307
 308   /* Find and count the active items that represent ends of rules. */
 309   for (i = 0; i < nritemset; ++i)
 310     {
 311       int item = ritem[itemset[i]];
 312       if (item < 0)
 313         redset[count++] = -item;
 314     }
 315
 316   /* Make a reductions structure and copy the data into it.  */
 317   state_reductions_set (state, count, redset);
 318 }
 319
 320 \f
 321 /*---------------.
 322 | Build STATES.  |
 323 `---------------*/
 324
 325 static void
 326 set_states (void)
 327 {
 328   states = XCALLOC (state_t *, nstates);
 329
 330   while (first_state)
 331     {
 332       state_list_t *this = first_state;
 333
 334       /* Pessimization, but simplification of the code: make sure all
 335          the states have a shifts, errs, and reductions, even if
 336          reduced to 0.  */
 337       state_t *state = this->state;
 338       if (!state->shifts)
 339         state_shifts_set (state, 0, 0);
 340       if (!state->errs)
 341         state->errs = errs_new (0);
 342       if (!state->reductions)
 343         state_reductions_set (state, 0, 0);
 344
 345       states[state->number] = state;
 346
 347       first_state = this->next;
 348       free (this);
 349     }
 350   first_state = NULL;
 351   last_state = NULL;
 352 }
 353
 354
 355 /*-------------------------------------------------------------------.
 356 | Compute the nondeterministic finite state machine (see state.h for |
 357 | details) from the grammar.                                         |
 358 `-------------------------------------------------------------------*/
 359
 360 void
 361 generate_states (void)
 362 {
 363   state_list_t *list = NULL;
 364   allocate_storage ();
 365   new_closure (nritems);
 366   new_states ();
 367   list = first_state;
 368
 369   while (list)
 370     {
 371       state_t *state = list->state;
 372       if (trace_flag)
 373         fprintf (stderr, "Processing state %d (reached by %s)\n",
 374                  state->number,
 375                  symbol_tag_get (symbols[state->accessing_symbol]));
 376       /* Set up ruleset and itemset for the transitions out of this
 377          state.  ruleset gets a 1 bit for each rule that could reduce
 378          now.  itemset gets a vector of all the items that could be
 379          accepted next.  */
 380       closure (state->items, state->nitems);
 381       /* Record the reductions allowed out of this state.  */
 382       save_reductions (state);
 383       /* Find the itemsets of the states that shifts can reach.  */
 384       new_itemsets (state);
 385       /* Find or create the core structures for those states.  */
 386       append_states (state);
 387
 388       /* Create the shifts structures for the shifts to those states,
 389          now that the state numbers transitioning to are known.  */
 390       state_shifts_set (state, nshifts, shiftset);
 391
 392       /* States are queued when they are created; process them all.
 393          */
 394       list = list->next;
 395     }
 396
 397   /* discard various storage */
 398   free_closure ();
 399   free_storage ();
 400
 401   /* Set up STATES. */
 402   set_states ();
 403 }