src/scan-code.l

   1 /* Bison Action Scanner                             -*- C -*-
   2
   3    Copyright (C) 2006, 2007, 2008 Free Software Foundation, Inc.
   4
   5    This file is part of Bison, the GNU Compiler Compiler.
   6
   7    This program is free software: you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation, either version 3 of the License, or
  10    (at your option) any later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  19
  20 %option debug nodefault nounput noyywrap never-interactive
  21 %option prefix="code_" outfile="lex.yy.c"
  22
  23 %{
  24 /* Work around a bug in flex 2.5.31.  See Debian bug 333231
  25    <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=333231>.  */
  26 #undef code_wrap
  27 #define code_wrap() 1
  28
  29 #define FLEX_PREFIX(Id) code_ ## Id
  30 #include "flex-scanner.h"
  31
  32 #include "complain.h"
  33 #include "reader.h"
  34 #include "getargs.h"
  35 #include <get-errno.h>
  36 #include <quote.h>
  37
  38 #include "scan-code.h"
  39 #include "symlist.h"
  40
  41 /* The current calling start condition: SC_RULE_ACTION or
  42    SC_SYMBOL_ACTION. */
  43 # define YY_DECL static char *code_lex (code_props *self, int sc_context)
  44 YY_DECL;
  45
  46 #define YY_USER_ACTION  location_compute (loc, &loc->end, yytext, yyleng);
  47
  48 static void handle_action_dollar (symbol_list *rule, char *cp,
  49                                   location dollar_loc);
  50 static void handle_action_at (symbol_list *rule, char *cp, location at_loc);
  51 static location the_location;
  52 static location *loc = &the_location;
  53
  54 /* A string representing the most recent translation.  */
  55 static char *last_string;
  56
  57 /* True if an untyped $$ or $n was seen.  */
  58 static bool untyped_var_seen;
  59 %}
  60  /* C and C++ comments in code. */
  61 %x SC_COMMENT SC_LINE_COMMENT
  62  /* Strings and characters in code. */
  63 %x SC_STRING SC_CHARACTER
  64  /* Whether in a rule or symbol action.  Specifies the translation
  65     of $ and @.  */
  66 %x SC_RULE_ACTION SC_SYMBOL_ACTION
  67
  68
  69 /* POSIX says that a tag must be both an id and a C union member, but
  70    historically almost any character is allowed in a tag.  We disallow
  71    NUL and newline, as this simplifies our implementation.  */
  72 tag      [^\0\n>]+
  73
  74 /* Zero or more instances of backslash-newline.  Following GCC, allow
  75    white space between the backslash and the newline.  */
  76 splice   (\\[ \f\t\v]*\n)*
  77
  78 %%
  79
  80 %{
  81   /* Nesting level of the current code in braces.  */
  82   int braces_level = 0;
  83
  84   /* This scanner is special: it is invoked only once, henceforth
  85      is expected to return only once.  This initialization is
  86      therefore done once per action to translate. */
  87   aver (sc_context == SC_SYMBOL_ACTION
  88         || sc_context == SC_RULE_ACTION
  89         || sc_context == INITIAL);
  90   BEGIN sc_context;
  91 %}
  92
  93   /*------------------------------------------------------------.
  94   | Scanning a C comment.  The initial `/ *' is already eaten.  |
  95   `------------------------------------------------------------*/
  96
  97 <SC_COMMENT>
  98 {
  99   "*"{splice}"/"  STRING_GROW; BEGIN sc_context;
 100 }
 101
 102
 103   /*--------------------------------------------------------------.
 104   | Scanning a line comment.  The initial `//' is already eaten.  |
 105   `--------------------------------------------------------------*/
 106
 107 <SC_LINE_COMMENT>
 108 {
 109   "\n"           STRING_GROW; BEGIN sc_context;
 110   {splice}       STRING_GROW;
 111 }
 112
 113
 114   /*--------------------------------------------.
 115   | Scanning user-code characters and strings.  |
 116   `--------------------------------------------*/
 117
 118 <SC_CHARACTER,SC_STRING>
 119 {
 120   {splice}|\\{splice}.  STRING_GROW;
 121 }
 122
 123 <SC_CHARACTER>
 124 {
 125   "'"           STRING_GROW; BEGIN sc_context;
 126 }
 127
 128 <SC_STRING>
 129 {
 130   "\""          STRING_GROW; BEGIN sc_context;
 131 }
 132
 133
 134 <SC_RULE_ACTION,SC_SYMBOL_ACTION>{
 135   "'" {
 136     STRING_GROW;
 137     BEGIN SC_CHARACTER;
 138   }
 139   "\"" {
 140     STRING_GROW;
 141     BEGIN SC_STRING;
 142   }
 143   "/"{splice}"*" {
 144     STRING_GROW;
 145     BEGIN SC_COMMENT;
 146   }
 147   "/"{splice}"/" {
 148     STRING_GROW;
 149     BEGIN SC_LINE_COMMENT;
 150   }
 151 }
 152
 153 <SC_RULE_ACTION>
 154 {
 155   "$"("<"{tag}">")?(-?[0-9]+|"$")  {
 156     handle_action_dollar (self->rule, yytext, *loc);
 157   }
 158   "@"(-?[0-9]+|"$") {
 159     handle_action_at (self->rule, yytext, *loc);
 160   }
 161
 162   "$"  {
 163     warn_at (*loc, _("stray `$'"));
 164     obstack_sgrow (&obstack_for_string, "$][");
 165   }
 166   "@"  {
 167     warn_at (*loc, _("stray `@'"));
 168     obstack_sgrow (&obstack_for_string, "@@");
 169   }
 170
 171   "{"  STRING_GROW; ++braces_level;
 172   "}"  {
 173     bool outer_brace = --braces_level == 0;
 174
 175     /* As an undocumented Bison extension, append `;' before the last
 176        brace in braced code, so that the user code can omit trailing
 177        `;'.  But do not append `;' if emulating Yacc, since Yacc does
 178        not append one.  Also, some output languages (like Java) do not
 179        accept an extra semicolon, so don't append if the user specified
 180        a skeleton or language.
 181
 182        FIXME: Bison should warn if a semicolon seems to be necessary
 183        here, and should omit the semicolon if it seems unnecessary
 184        (e.g., after ';', '{', or '}', each followed by comments or
 185        white space).  Such a warning shouldn't depend on --yacc; it
 186        should depend on a new --pedantic option, which would cause
 187        Bison to warn if it detects an extension to POSIX.  --pedantic
 188        should also diagnose other Bison extensions like %yacc.
 189        Perhaps there should also be a GCC-style --pedantic-errors
 190        option, so that such warnings are diagnosed as errors.  */
 191     if (outer_brace && !yacc_flag && language_prio == default_prio
 192         && skeleton_prio == default_prio)
 193       obstack_1grow (&obstack_for_string, ';');
 194
 195     STRING_GROW;
 196   }
 197 }
 198
 199 <SC_SYMBOL_ACTION>
 200 {
 201   "$$" {
 202     obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar[");
 203     self->is_value_used = true;
 204   }
 205   "@$" {
 206     obstack_sgrow (&obstack_for_string, "]b4_at_dollar[");
 207     locations_flag = true;
 208   }
 209 }
 210
 211
 212   /*-----------------------------------------.
 213   | Escape M4 quoting characters in C code.  |
 214   `-----------------------------------------*/
 215
 216 <*>
 217 {
 218   \$    obstack_sgrow (&obstack_for_string, "$][");
 219   \@    obstack_sgrow (&obstack_for_string, "@@");
 220   \[    obstack_sgrow (&obstack_for_string, "@{");
 221   \]    obstack_sgrow (&obstack_for_string, "@}");
 222 }
 223
 224   /*-----------------------------------------------------.
 225   | By default, grow the string obstack with the input.  |
 226   `-----------------------------------------------------*/
 227
 228 <*>.|\n STRING_GROW;
 229
 230  /* End of processing. */
 231 <*><<EOF>>       {
 232                    STRING_FINISH;
 233                    return last_string;
 234                  }
 235
 236 %%
 237
 238 /* Keeps track of the maximum number of semantic values to the left of
 239    a handle (those referenced by $0, $-1, etc.) are required by the
 240    semantic actions of this grammar. */
 241 int max_left_semantic_context = 0;
 242
 243
 244 /*------------------------------------------------------------------.
 245 | TEXT is pointing to a wannabee semantic value (i.e., a `$').      |
 246 |                                                                   |
 247 | Possible inputs: $[<TYPENAME>]($|integer)                         |
 248 |                                                                   |
 249 | Output to OBSTACK_FOR_STRING a reference to this semantic value.  |
 250 `------------------------------------------------------------------*/
 251
 252 static void
 253 handle_action_dollar (symbol_list *rule, char *text, location dollar_loc)
 254 {
 255   char const *type_name = NULL;
 256   char *cp = text + 1;
 257   symbol_list *effective_rule;
 258   int effective_rule_length;
 259
 260   if (rule->midrule_parent_rule)
 261     {
 262       effective_rule = rule->midrule_parent_rule;
 263       effective_rule_length = rule->midrule_parent_rhs_index - 1;
 264     }
 265   else
 266     {
 267       effective_rule = rule;
 268       effective_rule_length = symbol_list_length (rule->next);
 269     }
 270
 271   /* Get the type name if explicit. */
 272   if (*cp == '<')
 273     {
 274       type_name = ++cp;
 275       while (*cp != '>')
 276         ++cp;
 277       *cp = '\0';
 278       ++cp;
 279       if (untyped_var_seen)
 280         complain_at (dollar_loc, _("explicit type given in untyped grammar"));
 281       tag_seen = true;
 282     }
 283
 284   if (*cp == '$')
 285     {
 286       if (!type_name)
 287         type_name = symbol_list_n_type_name_get (rule, dollar_loc, 0);
 288
 289       if (!type_name)
 290         {
 291           if (union_seen | tag_seen)
 292             {
 293               if (rule->midrule_parent_rule)
 294                 complain_at (dollar_loc,
 295                              _("$$ for the midrule at $%d of `%s'"
 296                                " has no declared type"),
 297                              rule->midrule_parent_rhs_index,
 298                              effective_rule->content.sym->tag);
 299               else
 300                 complain_at (dollar_loc, _("$$ of `%s' has no declared type"),
 301                              rule->content.sym->tag);
 302             }
 303           else
 304             untyped_var_seen = true;
 305           type_name = "";
 306         }
 307
 308       obstack_fgrow1 (&obstack_for_string,
 309                       "]b4_lhs_value([%s])[", type_name);
 310       rule->action_props.is_value_used = true;
 311     }
 312   else
 313     {
 314       long int num = strtol (cp, NULL, 10);
 315
 316       if (1 - INT_MAX + effective_rule_length <= num
 317           && num <= effective_rule_length)
 318         {
 319           int n = num;
 320           if (max_left_semantic_context < 1 - n)
 321             max_left_semantic_context = 1 - n;
 322           if (!type_name && 0 < n)
 323             type_name =
 324               symbol_list_n_type_name_get (effective_rule, dollar_loc, n);
 325           if (!type_name)
 326             {
 327               if (union_seen | tag_seen)
 328                 complain_at (dollar_loc, _("$%d of `%s' has no declared type"),
 329                              n, effective_rule->content.sym->tag);
 330               else
 331                 untyped_var_seen = true;
 332               type_name = "";
 333             }
 334
 335           obstack_fgrow3 (&obstack_for_string,
 336                           "]b4_rhs_value(%d, %d, [%s])[",
 337                           effective_rule_length, n, type_name);
 338           if (n > 0)
 339             symbol_list_n_get (effective_rule, n)->action_props.is_value_used =
 340               true;
 341         }
 342       else
 343         complain_at (dollar_loc, _("integer out of range: %s"), quote (text));
 344     }
 345 }
 346
 347
 348 /*------------------------------------------------------.
 349 | TEXT is a location token (i.e., a `@...').  Output to |
 350 | OBSTACK_FOR_STRING a reference to this location.      |
 351 `------------------------------------------------------*/
 352
 353 static void
 354 handle_action_at (symbol_list *rule, char *text, location at_loc)
 355 {
 356   char *cp = text + 1;
 357   int effective_rule_length =
 358     (rule->midrule_parent_rule
 359      ? rule->midrule_parent_rhs_index - 1
 360      : symbol_list_length (rule->next));
 361
 362   locations_flag = true;
 363
 364   if (*cp == '$')
 365     obstack_sgrow (&obstack_for_string, "]b4_lhs_location[");
 366   else
 367     {
 368       long int num = strtol (cp, NULL, 10);
 369
 370       if (1 - INT_MAX + effective_rule_length <= num
 371           && num <= effective_rule_length)
 372         {
 373           int n = num;
 374           obstack_fgrow2 (&obstack_for_string, "]b4_rhs_location(%d, %d)[",
 375                           effective_rule_length, n);
 376         }
 377       else
 378         complain_at (at_loc, _("integer out of range: %s"), quote (text));
 379     }
 380 }
 381
 382
 383 /*-------------------------.
 384 | Initialize the scanner.  |
 385 `-------------------------*/
 386
 387 /* Translate the dollars and ats in \a self, in the context \a sc_context
 388    (SC_RULE_ACTION, SC_SYMBOL_ACTION, INITIAL).  */
 389
 390 static char const *
 391 translate_action (code_props *self, int sc_context)
 392 {
 393   char *res;
 394   static bool initialized = false;
 395   if (!initialized)
 396     {
 397       obstack_init (&obstack_for_string);
 398       yy_flex_debug = 0;
 399       initialized = true;
 400     }
 401
 402   loc->start = loc->end = self->location.start;
 403   yy_switch_to_buffer (yy_scan_string (self->code));
 404   res = code_lex (self, sc_context);
 405   yy_delete_buffer (YY_CURRENT_BUFFER);
 406
 407   return res;
 408 }
 409
 410 /*------------------------------------------------------------------------.
 411 | Implementation of the public interface as documented in "scan-code.h".  |
 412 `------------------------------------------------------------------------*/
 413
 414 void
 415 code_props_none_init (code_props *self)
 416 {
 417   *self = code_props_none;
 418 }
 419
 420 code_props const code_props_none = CODE_PROPS_NONE_INIT;
 421
 422 void
 423 code_props_plain_init (code_props *self, char const *code, location code_loc)
 424 {
 425   self->kind = CODE_PROPS_PLAIN;
 426   self->code = code;
 427   self->location = code_loc;
 428   self->is_value_used = false;
 429   self->rule = NULL;
 430 }
 431
 432 void
 433 code_props_symbol_action_init (code_props *self, char const *code,
 434                                location code_loc)
 435 {
 436   self->kind = CODE_PROPS_SYMBOL_ACTION;
 437   self->code = code;
 438   self->location = code_loc;
 439   self->is_value_used = false;
 440   self->rule = NULL;
 441 }
 442
 443 void
 444 code_props_rule_action_init (code_props *self, char const *code,
 445                              location code_loc, symbol_list *rule)
 446 {
 447   self->kind = CODE_PROPS_RULE_ACTION;
 448   self->code = code;
 449   self->location = code_loc;
 450   self->is_value_used = false;
 451   self->rule = rule;
 452 }
 453
 454 void
 455 code_props_translate_code (code_props *self)
 456 {
 457   switch (self->kind)
 458     {
 459       case CODE_PROPS_NONE:
 460         break;
 461       case CODE_PROPS_PLAIN:
 462         self->code = translate_action (self, INITIAL);
 463         break;
 464       case CODE_PROPS_SYMBOL_ACTION:
 465         self->code = translate_action (self, SC_SYMBOL_ACTION);
 466         break;
 467       case CODE_PROPS_RULE_ACTION:
 468         self->code = translate_action (self, SC_RULE_ACTION);
 469         break;
 470     }
 471 }
 472
 473 void
 474 code_scanner_last_string_free (void)
 475 {
 476   STRING_FREE;
 477 }
 478
 479 void
 480 code_scanner_free (void)
 481 {
 482   obstack_free (&obstack_for_string, 0);
 483   /* Reclaim Flex's buffers.  */
 484   yylex_destroy ();
 485 }