src/reduce.c

   1 /* Grammar reduction for Bison.
   2    Copyright 1988, 1989, 2000, 2001  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 /* Reduce the grammar: Find and eliminate unreachable terminals,
  23    nonterminals, and productions.  David S. Bakin.  */
  24
  25 /* Don't eliminate unreachable terminals: They may be used by the
  26    user's parser.  */
  27
  28 #include "system.h"
  29 #include "getargs.h"
  30 #include "files.h"
  31 #include "gram.h"
  32 #include "complain.h"
  33 #include "reduce.h"
  34 #include "reader.h"
  35 #include "getargs.h"
  36
  37 typedef unsigned *BSet;
  38 typedef short *rule;
  39
  40
  41 /* Set of all nonterminals which are not useless.  */
  42 static BSet N;
  43
  44 /* Set of all rules which have no useless nonterminals in their RHS.  */
  45 static BSet P;
  46
  47 /* Set of all accessible symbols.  */
  48 static BSet V;
  49
  50 /* Set of symbols used to define rule precedence (so they are
  51    `useless', but no warning should be issued).  */
  52 static BSet V1;
  53
  54 static int nuseful_productions;
  55 static int nuseless_productions;
  56 static int nuseful_nonterminals;
  57 static int nuseless_nonterminals;
  58 \f
  59 static bool
  60 bits_equal (BSet L, BSet R, int n)
  61 {
  62   int i;
  63
  64   for (i = n - 1; i >= 0; i--)
  65     if (L[i] != R[i])
  66       return FALSE;
  67   return TRUE;
  68 }
  69
  70
  71 static int
  72 nbits (unsigned i)
  73 {
  74   int count = 0;
  75
  76   while (i != 0)
  77     {
  78       i ^= (i & ((unsigned) (-(int) i)));
  79       ++count;
  80     }
  81   return count;
  82 }
  83
  84
  85 static int
  86 bits_size (BSet S, int n)
  87 {
  88   int i, count = 0;
  89
  90   for (i = n - 1; i >= 0; i--)
  91     count += nbits (S[i]);
  92   return count;
  93 }
  94 \f
  95 /*-------------------------------------------------------------------.
  96 | Another way to do this would be with a set for each production and |
  97 | then do subset tests against N0, but even for the C grammar the    |
  98 | whole reducing process takes only 2 seconds on my 8Mhz AT.         |
  99 `-------------------------------------------------------------------*/
 100
 101 static bool
 102 useful_production (int i, BSet N0)
 103 {
 104   rule r;
 105   short n;
 106
 107   /* A production is useful if all of the nonterminals in its appear
 108      in the set of useful nonterminals.  */
 109
 110   for (r = &ritem[rule_table[i].rhs]; *r > 0; r++)
 111     if (ISVAR (n = *r))
 112       if (!BITISSET (N0, n - ntokens))
 113         return FALSE;
 114   return TRUE;
 115 }
 116
 117
 118 /*---------------------------------------------------------.
 119 | Remember that rules are 1-origin, symbols are 0-origin.  |
 120 `---------------------------------------------------------*/
 121
 122 static void
 123 useless_nonterminals (void)
 124 {
 125   BSet Np, Ns;
 126   int i;
 127
 128   /* N is set as built.  Np is set being built this iteration. P is
 129      set of all productions which have a RHS all in N.  */
 130
 131   Np = XCALLOC (unsigned, WORDSIZE (nvars));
 132
 133   /* The set being computed is a set of nonterminals which can derive
 134      the empty string or strings consisting of all terminals. At each
 135      iteration a nonterminal is added to the set if there is a
 136      production with that nonterminal as its LHS for which all the
 137      nonterminals in its RHS are already in the set.  Iterate until
 138      the set being computed remains unchanged.  Any nonterminals not
 139      in the set at that point are useless in that they will never be
 140      used in deriving a sentence of the language.
 141
 142      This iteration doesn't use any special traversal over the
 143      productions.  A set is kept of all productions for which all the
 144      nonterminals in the RHS are in useful.  Only productions not in
 145      this set are scanned on each iteration.  At the end, this set is
 146      saved to be used when finding useful productions: only
 147      productions in this set will appear in the final grammar.  */
 148
 149   while (1)
 150     {
 151       for (i = WORDSIZE (nvars) - 1; i >= 0; i--)
 152         Np[i] = N[i];
 153       for (i = 1; i <= nrules; i++)
 154         {
 155           if (!BITISSET (P, i))
 156             {
 157               if (useful_production (i, N))
 158                 {
 159                   SETBIT (Np, rule_table[i].lhs - ntokens);
 160                   SETBIT (P, i);
 161                 }
 162             }
 163         }
 164       if (bits_equal (N, Np, WORDSIZE (nvars)))
 165         break;
 166       Ns = Np;
 167       Np = N;
 168       N = Ns;
 169     }
 170   XFREE (N);
 171   N = Np;
 172 }
 173
 174
 175 static void
 176 inaccessable_symbols (void)
 177 {
 178   BSet Vp, Vs, Pp;
 179   int i;
 180   short t;
 181   rule r;
 182
 183   /* Find out which productions are reachable and which symbols are
 184      used.  Starting with an empty set of productions and a set of
 185      symbols which only has the start symbol in it, iterate over all
 186      productions until the set of productions remains unchanged for an
 187      iteration.  For each production which has a LHS in the set of
 188      reachable symbols, add the production to the set of reachable
 189      productions, and add all of the nonterminals in the RHS of the
 190      production to the set of reachable symbols.
 191
 192      Consider only the (partially) reduced grammar which has only
 193      nonterminals in N and productions in P.
 194
 195      The result is the set P of productions in the reduced grammar,
 196      and the set V of symbols in the reduced grammar.
 197
 198      Although this algorithm also computes the set of terminals which
 199      are reachable, no terminal will be deleted from the grammar. Some
 200      terminals might not be in the grammar but might be generated by
 201      semantic routines, and so the user might want them available with
 202      specified numbers.  (Is this true?)  However, the nonreachable
 203      terminals are printed (if running in verbose mode) so that the
 204      user can know.  */
 205
 206   Vp = XCALLOC (unsigned, WORDSIZE (nsyms));
 207   Pp = XCALLOC (unsigned, WORDSIZE (nrules + 1));
 208
 209   /* If the start symbol isn't useful, then nothing will be useful. */
 210   if (BITISSET (N, start_symbol - ntokens))
 211     {
 212       SETBIT (V, start_symbol);
 213
 214       while (1)
 215         {
 216           for (i = WORDSIZE (nsyms) - 1; i >= 0; i--)
 217             Vp[i] = V[i];
 218           for (i = 1; i <= nrules; i++)
 219             {
 220               if (!BITISSET (Pp, i)
 221                   && BITISSET (P, i)
 222                   && BITISSET (V, rule_table[i].lhs))
 223                 {
 224                   for (r = &ritem[rule_table[i].rhs]; *r >= 0; r++)
 225                     if (ISTOKEN (t = *r) || BITISSET (N, t - ntokens))
 226                       SETBIT (Vp, t);
 227                   SETBIT (Pp, i);
 228                 }
 229             }
 230           if (bits_equal (V, Vp, WORDSIZE (nsyms)))
 231             break;
 232           Vs = Vp;
 233           Vp = V;
 234           V = Vs;
 235         }
 236     }
 237
 238   XFREE (V);
 239   V = Vp;
 240
 241   /* Tokens 0, 1, and 2 are internal to Bison.  Consider them useful. */
 242   SETBIT (V, 0);                /* end-of-input token */
 243   SETBIT (V, 1);                /* error token */
 244   SETBIT (V, 2);                /* some undefined token */
 245
 246   XFREE (P);
 247   P = Pp;
 248
 249   nuseful_productions = bits_size (P, WORDSIZE (nrules + 1));
 250   nuseless_productions = nrules - nuseful_productions;
 251
 252   nuseful_nonterminals = 0;
 253   for (i = ntokens; i < nsyms; i++)
 254     if (BITISSET (V, i))
 255       nuseful_nonterminals++;
 256   nuseless_nonterminals = nvars - nuseful_nonterminals;
 257
 258   /* A token that was used in %prec should not be warned about.  */
 259   for (i = 1; i < nrules; i++)
 260     if (rule_table[i].precsym != 0)
 261       SETBIT (V1, rule_table[i].precsym);
 262 }
 263
 264 static void
 265 reduce_grammar_tables (void)
 266 {
 267 /* This is turned off because we would need to change the numbers
 268    in the case statements in the actions file.  */
 269 #if 0
 270   /* remove useless productions */
 271   if (nuseless_productions > 0)
 272     {
 273       short np, pn, ni, pi;
 274
 275       np = 0;
 276       ni = 0;
 277       for (pn = 1; pn <= nrules; pn++)
 278         {
 279           if (BITISSET (P, pn))
 280             {
 281               np++;
 282               if (pn != np)
 283                 {
 284                   rule_table[np].lhs = rule_table[pn].lhs;
 285                   rline[np] = rline[pn];
 286                   rule_table[np].prec = rule_table[pn].prec;
 287                   rule_table[np].assoc = rule_table[pn].assoc;
 288                   rule_table[np].rhs = rule_table[pn].rhs;
 289                   if (rule_table[np].rhs != ni)
 290                     {
 291                       pi = rule_table[np].rhs;
 292                       rule_table[np].rhs = ni;
 293                       while (ritem[pi] >= 0)
 294                         ritem[ni++] = ritem[pi++];
 295                       ritem[ni++] = -np;
 296                     }
 297                 }
 298               else
 299                 {
 300                   while (ritem[ni++] >= 0);
 301                 }
 302             }
 303         }
 304       ritem[ni] = 0;
 305       nrules -= nuseless_productions;
 306       nitems = ni;
 307
 308       /* Is it worth it to reduce the amount of memory for the
 309          grammar? Probably not.  */
 310
 311     }
 312 #endif /* 0 */
 313   /* Disable useless productions,
 314      since they may contain useless nonterms
 315      that would get mapped below to -1 and confuse everyone.  */
 316   if (nuseless_productions > 0)
 317     {
 318       int pn;
 319
 320       for (pn = 1; pn <= nrules; pn++)
 321         {
 322           if (!BITISSET (P, pn))
 323             {
 324               rule_table[pn].lhs = -1;
 325             }
 326         }
 327     }
 328 }
 329
 330
 331 /*------------------------------.
 332 | Remove useless nonterminals.  |
 333 `------------------------------*/
 334
 335 static void
 336 nonterminals_reduce (void)
 337 {
 338   int i, n;
 339   rule r;
 340
 341   /* Create a map of nonterminal number to new nonterminal number. -1
 342      in the map means it was useless and is being eliminated.  */
 343
 344   short *nontermmap = XCALLOC (short, nvars) - ntokens;
 345   n = ntokens;
 346   for (i = ntokens; i < nsyms; i++)
 347     if (BITISSET (V, i))
 348       nontermmap[i] = n++;
 349   for (i = ntokens; i < nsyms; i++)
 350     if (!BITISSET (V, i))
 351       nontermmap[i] = n++;
 352
 353
 354   /* Shuffle elements of tables indexed by symbol number.  */
 355   {
 356     short *sassoc_sorted = XMALLOC (short, nvars) - ntokens;
 357     short *sprec_sorted  = XMALLOC (short, nvars) - ntokens;
 358     char **tags_sorted   = XMALLOC (char *, nvars) - ntokens;
 359
 360     for (i = ntokens; i < nsyms; i++)
 361       {
 362         n = nontermmap[i];
 363         sassoc_sorted[n] = sassoc[i];
 364         sprec_sorted[n]  = sprec[i];
 365         tags_sorted[n]   = tags[i];
 366       }
 367     for (i = ntokens; i < nsyms; i++)
 368       {
 369         sassoc[i] = sassoc_sorted[i];
 370         sprec[i]  = sprec_sorted[i];
 371         tags[i]   = tags_sorted[i];
 372       }
 373     free (sassoc_sorted + ntokens);
 374     free (sprec_sorted + ntokens);
 375     free (tags_sorted + ntokens);
 376   }
 377
 378   /* Replace all symbol numbers in valid data structures.  */
 379
 380   for (i = 1; i <= nrules; i++)
 381     {
 382       /* Ignore the rules disabled above.  */
 383       if (rule_table[i].lhs >= 0)
 384         rule_table[i].lhs = nontermmap[rule_table[i].lhs];
 385       if (ISVAR (rule_table[i].precsym))
 386         /* Can this happen?  */
 387         rule_table[i].precsym = nontermmap[rule_table[i].precsym];
 388     }
 389
 390   for (r = ritem; *r; r++)
 391     if (ISVAR (*r))
 392       *r = nontermmap[*r];
 393
 394   start_symbol = nontermmap[start_symbol];
 395
 396   nsyms -= nuseless_nonterminals;
 397   nvars -= nuseless_nonterminals;
 398
 399   free (&nontermmap[ntokens]);
 400 }
 401
 402
 403 /*------------------------------------------------------------------.
 404 | Output the detailed results of the reductions.  For FILE.output.  |
 405 `------------------------------------------------------------------*/
 406
 407 void
 408 reduce_output (FILE *out)
 409 {
 410   int i;
 411   rule r;
 412   bool b;
 413
 414   if (nuseless_nonterminals > 0)
 415     {
 416       fprintf (out, _("Useless nonterminals:"));
 417       fprintf (out, "\n\n");
 418       for (i = 0; i < nuseless_nonterminals; ++i)
 419         fprintf (out, "   %s\n", tags[nsyms + i]);
 420     }
 421   b = FALSE;
 422   for (i = 0; i < ntokens; i++)
 423     {
 424       if (!BITISSET (V, i) && !BITISSET (V1, i))
 425         {
 426           if (!b)
 427             {
 428               fprintf (out, "\n\n");
 429               fprintf (out, _("Terminals which are not used:"));
 430               fprintf (out, "\n\n");
 431               b = TRUE;
 432             }
 433           fprintf (out, "   %s\n", tags[i]);
 434         }
 435     }
 436
 437   if (nuseless_productions > 0)
 438     {
 439       fprintf (out, "\n\n");
 440       fprintf (out, _("Useless rules:"));
 441       fprintf (out, "\n\n");
 442       for (i = 1; i <= nrules; i++)
 443         if (!BITISSET (P, i))
 444           {
 445             fprintf (out, "#%-4d  ", i);
 446             fprintf (out, "%s :\t", tags[rule_table[i].lhs]);
 447             for (r = &ritem[rule_table[i].rhs]; *r >= 0; r++)
 448               fprintf (out, " %s", tags[*r]);
 449             fprintf (out, ";\n");
 450           }
 451     }
 452   if (nuseless_nonterminals > 0 || nuseless_productions > 0 || b)
 453     fprintf (out, "\n\n");
 454 }
 455 \f
 456 static void
 457 dump_grammar (FILE *out)
 458 {
 459   int i;
 460   rule r;
 461
 462   fprintf (out, "REDUCED GRAMMAR\n\n");
 463   fprintf (out,
 464            "ntokens = %d, nvars = %d, nsyms = %d, nrules = %d, nitems = %d\n\n",
 465            ntokens, nvars, nsyms, nrules, nitems);
 466   fprintf (out, "Variables\n---------\n\n");
 467   fprintf (out, "Value  Sprec  Sassoc  Tag\n");
 468   for (i = ntokens; i < nsyms; i++)
 469     fprintf (out, "%5d  %5d   %5d  %s\n", i, sprec[i], sassoc[i], tags[i]);
 470   fprintf (out, "\n\n");
 471   fprintf (out, "Rules\n-----\n\n");
 472   fprintf (out, "Num (Prec, Assoc) Lhs : (@Rhs) Ritems [Num?]\n");
 473   for (i = 1; i <= nrules; i++)
 474     {
 475       fprintf (out, "%-5d(%5d%5d)%5d : (@%-5d)",
 476                i,
 477                rule_table[i].prec,
 478                rule_table[i].assoc,
 479                rule_table[i].lhs,
 480                rule_table[i].rhs);
 481       for (r = &ritem[rule_table[i].rhs]; *r > 0; r++)
 482         fprintf (out, "%5d", *r);
 483       fprintf (out, " [%d]\n", -(*r));
 484     }
 485   fprintf (out, "\n\n");
 486   fprintf (out, "Rules interpreted\n-----------------\n\n");
 487   for (i = 1; i <= nrules; i++)
 488     {
 489       fprintf (out, "%-5d  %s :", i, tags[rule_table[i].lhs]);
 490       for (r = &ritem[rule_table[i].rhs]; *r > 0; r++)
 491         fprintf (out, " %s", tags[*r]);
 492       fputc ('\n', out);
 493     }
 494   fprintf (out, "\n\n");
 495 }
 496
 497
 498
 499 /*-------------------------------.
 500 | Report the results to STDERR.  |
 501 `-------------------------------*/
 502
 503 static void
 504 reduce_print (void)
 505 {
 506   if (yacc_flag && nuseless_productions)
 507     fprintf (stderr, ngettext ("%d rule never reduced\n",
 508                                "%d rules never reduced\n",
 509                                nuseless_productions),
 510              nuseless_productions);
 511
 512   fprintf (stderr, _("%s contains "), infile);
 513
 514   if (nuseless_nonterminals > 0)
 515     fprintf (stderr, ngettext ("%d useless nonterminal",
 516                                "%d useless nonterminals",
 517                                nuseless_nonterminals),
 518              nuseless_nonterminals);
 519
 520   if (nuseless_nonterminals > 0 && nuseless_productions > 0)
 521     fprintf (stderr, _(" and "));
 522
 523   if (nuseless_productions > 0)
 524     fprintf (stderr, ngettext ("%d useless rule",
 525                                "%d useless rules",
 526                                nuseless_productions),
 527              nuseless_productions);
 528   fprintf (stderr, "\n");
 529   fflush (stderr);
 530 }
 531 \f
 532 void
 533 reduce_grammar (void)
 534 {
 535   bool reduced;
 536
 537   /* Allocate the global sets used to compute the reduced grammar */
 538
 539   N = XCALLOC (unsigned, WORDSIZE (nvars));
 540   P = XCALLOC (unsigned, WORDSIZE (nrules + 1));
 541   V = XCALLOC (unsigned, WORDSIZE (nsyms));
 542   V1 = XCALLOC (unsigned, WORDSIZE (nsyms));
 543
 544   useless_nonterminals ();
 545   inaccessable_symbols ();
 546
 547   reduced = (bool) (nuseless_nonterminals + nuseless_productions > 0);
 548
 549   if (!reduced)
 550     return;
 551
 552   reduce_print ();
 553
 554   if (!BITISSET (N, start_symbol - ntokens))
 555     fatal (_("Start symbol %s does not derive any sentence"),
 556            tags[start_symbol]);
 557
 558   reduce_grammar_tables ();
 559   if (nuseless_nonterminals > 0)
 560     nonterminals_reduce ();
 561
 562   if (trace_flag)
 563     {
 564       dump_grammar (stderr);
 565
 566       fprintf (stderr, "reduced %s defines %d terminals, %d nonterminals\
 567 , and %d productions.\n",
 568                infile, ntokens, nvars, nrules);
 569     }
 570 }
 571
 572
 573 /*-----------------------------------------------------------.
 574 | Free the global sets used to compute the reduced grammar.  |
 575 `-----------------------------------------------------------*/
 576
 577 void
 578 reduce_free (void)
 579 {
 580   XFREE (N);
 581   XFREE (V);
 582   XFREE (V1);
 583   XFREE (P);
 584 }