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