src/scan-code.l

   1 /* Bison Action Scanner                             -*- C -*-
   2
   3    Copyright (C) 2006, 2007 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.
 179
 180        FIXME: Bison should warn if a semicolon seems to be necessary
 181        here, and should omit the semicolon if it seems unnecessary
 182        (e.g., after ';', '{', or '}', each followed by comments or
 183        white space).  Such a warning shouldn't depend on --yacc; it
 184        should depend on a new --pedantic option, which would cause
 185        Bison to warn if it detects an extension to POSIX.  --pedantic
 186        should also diagnose other Bison extensions like %yacc.
 187        Perhaps there should also be a GCC-style --pedantic-errors
 188        option, so that such warnings are diagnosed as errors.  */
 189     if (outer_brace && ! yacc_flag)
 190       obstack_1grow (&obstack_for_string, ';');
 191
 192     STRING_GROW;
 193   }
 194 }
 195
 196 <SC_SYMBOL_ACTION>
 197 {
 198   "$$" {
 199     obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar[");
 200     self->is_value_used = true;
 201   }
 202   "@$" {
 203     obstack_sgrow (&obstack_for_string, "]b4_at_dollar[");
 204     locations_flag = true;
 205   }
 206 }
 207
 208
 209   /*-----------------------------------------.
 210   | Escape M4 quoting characters in C code.  |
 211   `-----------------------------------------*/
 212
 213 <*>
 214 {
 215   \$    obstack_sgrow (&obstack_for_string, "$][");
 216   \@    obstack_sgrow (&obstack_for_string, "@@");
 217   \[    obstack_sgrow (&obstack_for_string, "@{");
 218   \]    obstack_sgrow (&obstack_for_string, "@}");
 219 }
 220
 221   /*-----------------------------------------------------.
 222   | By default, grow the string obstack with the input.  |
 223   `-----------------------------------------------------*/
 224
 225 <*>.|\n STRING_GROW;
 226
 227  /* End of processing. */
 228 <*><<EOF>>       {
 229                    STRING_FINISH;
 230                    return last_string;
 231                  }
 232
 233 %%
 234
 235 /* Keeps track of the maximum number of semantic values to the left of
 236    a handle (those referenced by $0, $-1, etc.) are required by the
 237    semantic actions of this grammar. */
 238 int max_left_semantic_context = 0;
 239
 240
 241 /*------------------------------------------------------------------.
 242 | TEXT is pointing to a wannabee semantic value (i.e., a `$').      |
 243 |                                                                   |
 244 | Possible inputs: $[<TYPENAME>]($|integer)                         |
 245 |                                                                   |
 246 | Output to OBSTACK_FOR_STRING a reference to this semantic value.  |
 247 `------------------------------------------------------------------*/
 248
 249 static void
 250 handle_action_dollar (symbol_list *rule, char *text, location dollar_loc)
 251 {
 252   char const *type_name = NULL;
 253   char *cp = text + 1;
 254   symbol_list *effective_rule;
 255   int effective_rule_length;
 256
 257   if (rule->midrule_parent_rule)
 258     {
 259       effective_rule = rule->midrule_parent_rule;
 260       effective_rule_length = rule->midrule_parent_rhs_index - 1;
 261     }
 262   else
 263     {
 264       effective_rule = rule;
 265       effective_rule_length = symbol_list_length (rule->next);
 266     }
 267
 268   /* Get the type name if explicit. */
 269   if (*cp == '<')
 270     {
 271       type_name = ++cp;
 272       while (*cp != '>')
 273         ++cp;
 274       *cp = '\0';
 275       ++cp;
 276       if (untyped_var_seen)
 277         complain_at (dollar_loc, _("explicit type given in untyped grammar"));
 278       tag_seen = true;
 279     }
 280
 281   if (*cp == '$')
 282     {
 283       if (!type_name)
 284         type_name = symbol_list_n_type_name_get (rule, dollar_loc, 0);
 285
 286       if (!type_name)
 287         {
 288           if (union_seen | tag_seen)
 289             {
 290               if (rule->midrule_parent_rule)
 291                 complain_at (dollar_loc,
 292                              _("$$ for the midrule at $%d of `%s'"
 293                                " has no declared type"),
 294                              rule->midrule_parent_rhs_index,
 295                              effective_rule->content.sym->tag);
 296               else
 297                 complain_at (dollar_loc, _("$$ of `%s' has no declared type"),
 298                              rule->content.sym->tag);
 299             }
 300           else
 301             untyped_var_seen = true;
 302           type_name = "";
 303         }
 304
 305       obstack_fgrow1 (&obstack_for_string,
 306                       "]b4_lhs_value([%s])[", type_name);
 307       rule->action_props.is_value_used = true;
 308     }
 309   else
 310     {
 311       long int num = strtol (cp, NULL, 10);
 312
 313       if (1 - INT_MAX + effective_rule_length <= num
 314           && num <= effective_rule_length)
 315         {
 316           int n = num;
 317           if (max_left_semantic_context < 1 - n)
 318             max_left_semantic_context = 1 - n;
 319           if (!type_name && 0 < n)
 320             type_name =
 321               symbol_list_n_type_name_get (effective_rule, dollar_loc, n);
 322           if (!type_name)
 323             {
 324               if (union_seen | tag_seen)
 325                 complain_at (dollar_loc, _("$%d of `%s' has no declared type"),
 326                              n, effective_rule->content.sym->tag);
 327               else
 328                 untyped_var_seen = true;
 329               type_name = "";
 330             }
 331
 332           obstack_fgrow3 (&obstack_for_string,
 333                           "]b4_rhs_value(%d, %d, [%s])[",
 334                           effective_rule_length, n, type_name);
 335           if (n > 0)
 336             symbol_list_n_get (effective_rule, n)->action_props.is_value_used =
 337               true;
 338         }
 339       else
 340         complain_at (dollar_loc, _("integer out of range: %s"), quote (text));
 341     }
 342 }
 343
 344
 345 /*------------------------------------------------------.
 346 | TEXT is a location token (i.e., a `@...').  Output to |
 347 | OBSTACK_FOR_STRING a reference to this location.      |
 348 `------------------------------------------------------*/
 349
 350 static void
 351 handle_action_at (symbol_list *rule, char *text, location at_loc)
 352 {
 353   char *cp = text + 1;
 354   int effective_rule_length =
 355     (rule->midrule_parent_rule
 356      ? rule->midrule_parent_rhs_index - 1
 357      : symbol_list_length (rule->next));
 358
 359   locations_flag = true;
 360
 361   if (*cp == '$')
 362     obstack_sgrow (&obstack_for_string, "]b4_lhs_location[");
 363   else
 364     {
 365       long int num = strtol (cp, NULL, 10);
 366
 367       if (1 - INT_MAX + effective_rule_length <= num
 368           && num <= effective_rule_length)
 369         {
 370           int n = num;
 371           obstack_fgrow2 (&obstack_for_string, "]b4_rhs_location(%d, %d)[",
 372                           effective_rule_length, n);
 373         }
 374       else
 375         complain_at (at_loc, _("integer out of range: %s"), quote (text));
 376     }
 377 }
 378
 379
 380 /*-------------------------.
 381 | Initialize the scanner.  |
 382 `-------------------------*/
 383
 384 /* Translate the dollars and ats in \a self, in the context \a sc_context
 385    (SC_RULE_ACTION, SC_SYMBOL_ACTION, INITIAL).  */
 386
 387 static char const *
 388 translate_action (code_props *self, int sc_context)
 389 {
 390   char *res;
 391   static bool initialized = false;
 392   if (!initialized)
 393     {
 394       obstack_init (&obstack_for_string);
 395       yy_flex_debug = 0;
 396       initialized = true;
 397     }
 398
 399   loc->start = loc->end = self->location.start;
 400   yy_switch_to_buffer (yy_scan_string (self->code));
 401   res = code_lex (self, sc_context);
 402   yy_delete_buffer (YY_CURRENT_BUFFER);
 403
 404   return res;
 405 }
 406
 407 /*------------------------------------------------------------------------.
 408 | Implementation of the public interface as documented in "scan-code.h".  |
 409 `------------------------------------------------------------------------*/
 410
 411 void
 412 code_props_none_init (code_props *self)
 413 {
 414   *self = code_props_none;
 415 }
 416
 417 code_props const code_props_none = CODE_PROPS_NONE_INIT;
 418
 419 void
 420 code_props_plain_init (code_props *self, char const *code, location code_loc)
 421 {
 422   self->kind = CODE_PROPS_PLAIN;
 423   self->code = code;
 424   self->location = code_loc;
 425   self->is_value_used = false;
 426   self->rule = NULL;
 427 }
 428
 429 void
 430 code_props_symbol_action_init (code_props *self, char const *code,
 431                                location code_loc)
 432 {
 433   self->kind = CODE_PROPS_SYMBOL_ACTION;
 434   self->code = code;
 435   self->location = code_loc;
 436   self->is_value_used = false;
 437   self->rule = NULL;
 438 }
 439
 440 void
 441 code_props_rule_action_init (code_props *self, char const *code,
 442                              location code_loc, symbol_list *rule)
 443 {
 444   self->kind = CODE_PROPS_RULE_ACTION;
 445   self->code = code;
 446   self->location = code_loc;
 447   self->is_value_used = false;
 448   self->rule = rule;
 449 }
 450
 451 void
 452 code_props_translate_code (code_props *self)
 453 {
 454   switch (self->kind)
 455     {
 456       case CODE_PROPS_NONE:
 457         break;
 458       case CODE_PROPS_PLAIN:
 459         self->code = translate_action (self, INITIAL);
 460         break;
 461       case CODE_PROPS_SYMBOL_ACTION:
 462         self->code = translate_action (self, SC_SYMBOL_ACTION);
 463         break;
 464       case CODE_PROPS_RULE_ACTION:
 465         self->code = translate_action (self, SC_RULE_ACTION);
 466         break;
 467     }
 468 }
 469
 470 void
 471 code_scanner_last_string_free (void)
 472 {
 473   STRING_FREE;
 474 }
 475
 476 void
 477 code_scanner_free (void)
 478 {
 479   obstack_free (&obstack_for_string, 0);
 480   /* Reclaim Flex's buffers.  */
 481   yylex_destroy ();
 482 }