1 %{/* Bison Grammar Parser -*- C -*-
3 Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation,
6 This file is part of Bison, the GNU Compiler Compiler.
8 This program is free software; you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation; either version 2 of the License, or
11 (at your option) any later version.
13 This program is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with this program; if not, write to the Free Software
20 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
27 #include <strverscmp.h>
30 #include "conflicts.h"
34 #include "muscle_tab.h"
38 #include "scan-gram.h"
39 #include "scan-code.h"
41 #define YYLLOC_DEFAULT(Current, Rhs, N) (Current) = lloc_default (Rhs, N)
42 static YYLTYPE lloc_default (YYLTYPE const *, int);
44 #define YY_LOCATION_PRINT(File, Loc) \
45 location_print (File, Loc)
47 static void version_check (location const *loc, char const *version);
49 /* Request detailed syntax error messages, and pass them to GRAM_ERROR.
50 FIXME: depends on the undocumented availability of YYLLOC. */
52 #define yyerror(Msg) \
53 gram_error (&yylloc, Msg)
54 static void gram_error (location const *, char const *);
56 static char const *char_name (char);
58 /** Add a lex-param or a parse-param.
60 * \param type \a lex_param or \a parse_param
61 * \param decl the formal argument
62 * \param loc the location in the source.
64 static void add_param (char const *type, char *decl, location loc);
67 static symbol_class current_class = unknown_sym;
68 static uniqstr current_type = NULL;
69 static symbol *current_lhs;
70 static location current_lhs_location;
71 static int current_prec = 0;
73 #define YYTYPE_INT16 int_fast16_t
74 #define YYTYPE_INT8 int_fast8_t
75 #define YYTYPE_UINT16 uint_fast16_t
76 #define YYTYPE_UINT8 uint_fast8_t
90 /* Bison's grammar can initial empty locations, hence a default
91 location is needed. */
92 boundary_set (&@$.start, current_file, 1, 1);
93 boundary_set (&@$.end, current_file, 1, 1);
105 unsigned char character;
108 /* Define the tokens together with their human representation. */
109 %token GRAM_EOF 0 "end of file"
110 %token STRING "string"
113 %token PERCENT_TOKEN "%token"
114 %token PERCENT_NTERM "%nterm"
116 %token PERCENT_TYPE "%type"
117 %token PERCENT_DESTRUCTOR "%destructor"
118 %token PERCENT_PRINTER "%printer"
120 %token PERCENT_LEFT "%left"
121 %token PERCENT_RIGHT "%right"
122 %token PERCENT_NONASSOC "%nonassoc"
124 %token PERCENT_PREC "%prec"
125 %token PERCENT_DPREC "%dprec"
126 %token PERCENT_MERGE "%merge"
129 /*----------------------.
130 | Global Declarations. |
131 `----------------------*/
135 PERCENT_DEBUG "%debug"
136 PERCENT_DEFAULT_PREC "%default-prec"
137 PERCENT_DEFINE "%define"
138 PERCENT_DEFINES "%defines"
139 PERCENT_ERROR_VERBOSE "%error-verbose"
140 PERCENT_EXPECT "%expect"
141 PERCENT_EXPECT_RR "%expect-rr"
142 PERCENT_FILE_PREFIX "%file-prefix"
143 PERCENT_GLR_PARSER "%glr-parser"
144 PERCENT_INITIAL_ACTION "%initial-action"
145 PERCENT_LANGUAGE "%language"
146 PERCENT_LEX_PARAM "%lex-param"
147 PERCENT_LOCATIONS "%locations"
148 PERCENT_NAME_PREFIX "%name-prefix"
149 PERCENT_NO_DEFAULT_PREC "%no-default-prec"
150 PERCENT_NO_LINES "%no-lines"
151 PERCENT_NONDETERMINISTIC_PARSER
152 "%nondeterministic-parser"
153 PERCENT_OUTPUT "%output"
154 PERCENT_PARSE_PARAM "%parse-param"
155 PERCENT_PURE_PARSER "%pure-parser"
156 PERCENT_PUSH_PARSER "%push-parser"
157 PERCENT_PUSH_PULL_PARSER
159 PERCENT_REQUIRE "%require"
160 PERCENT_SKELETON "%skeleton"
161 PERCENT_START "%start"
162 PERCENT_TOKEN_TABLE "%token-table"
163 PERCENT_VERBOSE "%verbose"
167 %token BRACED_CODE "{...}"
169 %token EPILOGUE "epilogue"
171 %token ID "identifier"
172 %token ID_COLON "identifier:"
173 %token PERCENT_PERCENT "%%"
175 %token PROLOGUE "%{...%}"
178 %token TYPE_TAG_ANY "<*>"
179 %token TYPE_TAG_NONE "<>"
181 %type <character> CHAR
182 %printer { fputs (char_name ($$), stderr); } CHAR
184 /* braceless is not to be used for rule or symbol actions, as it
185 calls code_props_plain_init. */
186 %type <chars> STRING "%{...%}" EPILOGUE braceless content.opt
188 %printer { fputs (quotearg_style (c_quoting_style, $$), stderr); }
190 %printer { fprintf (stderr, "{\n%s\n}", $$); }
191 braceless content.opt "{...}" "%{...%}" EPILOGUE
193 %type <uniqstr> TYPE ID ID_COLON variable
194 %printer { fprintf (stderr, "<%s>", $$); } TYPE
195 %printer { fputs ($$, stderr); } ID variable
196 %printer { fprintf (stderr, "%s:", $$); } ID_COLON
199 %printer { fprintf (stderr, "%d", $$); } INT
201 %type <symbol> id id_colon symbol string_as_id
202 %printer { fprintf (stderr, "%s", $$->tag); } id symbol string_as_id
203 %printer { fprintf (stderr, "%s:", $$->tag); } id_colon
205 %type <assoc> precedence_declarator
206 %type <list> symbols.1 generic_symlist generic_symlist_item
210 prologue_declarations "%%" grammar epilogue.opt
214 /*------------------------------------.
215 | Declarations: before the first %%. |
216 `------------------------------------*/
218 prologue_declarations:
220 | prologue_declarations prologue_declaration
223 prologue_declaration:
227 code_props plain_code;
228 code_props_plain_init (&plain_code, $1, @1);
229 code_props_translate_code (&plain_code);
230 gram_scanner_last_string_free ();
231 muscle_code_grow (union_seen ? "post_prologue" : "pre_prologue",
232 plain_code.code, @1);
233 code_scanner_last_string_free ();
235 | "%debug" { debug_flag = true; }
236 | "%define" variable content.opt
238 char const name_prefix[] = "percent_define(";
239 size_t length = strlen ($2);
240 char *name = xmalloc (sizeof name_prefix + length + 1);
241 strcpy (name, name_prefix);
242 strcpy (name + sizeof name_prefix - 1, $2);
243 strcpy (name + sizeof name_prefix - 1 + length, ")");
244 if (muscle_find_const (name))
245 warn_at (@2, _("%s `%s' redefined"), "%define variable", $2);
246 MUSCLE_INSERT_STRING (uniqstr_new (name), $3);
248 muscle_grow_user_name_list ("percent_define_user_variables", $2, @2);
250 | "%defines" { defines_flag = true; }
254 spec_defines_file = xstrdup ($2);
256 | "%error-verbose" { error_verbose = true; }
257 | "%expect" INT { expected_sr_conflicts = $2; }
258 | "%expect-rr" INT { expected_rr_conflicts = $2; }
259 | "%file-prefix" STRING { spec_file_prefix = $2; }
260 | "%file-prefix" "=" STRING { spec_file_prefix = $3; } /* deprecated */
263 nondeterministic_parser = true;
266 | "%initial-action" "{...}"
269 code_props_symbol_action_init (&action, $2, @2);
270 code_props_translate_code (&action);
271 gram_scanner_last_string_free ();
272 muscle_code_grow ("initial_action", action.code, @2);
273 code_scanner_last_string_free ();
275 | "%language" STRING { language_argmatch ($2, 1, &@1); }
276 | "%lex-param" "{...}" { add_param ("lex_param", $2, @2); }
277 | "%locations" { locations_flag = true; }
278 | "%name-prefix" STRING { spec_name_prefix = $2; }
279 | "%name-prefix" "=" STRING { spec_name_prefix = $3; } /* deprecated */
280 | "%no-lines" { no_lines_flag = true; }
281 | "%nondeterministic-parser" { nondeterministic_parser = true; }
282 | "%output" STRING { spec_outfile = $2; }
283 | "%output" "=" STRING { spec_outfile = $3; } /* deprecated */
284 | "%parse-param" "{...}" { add_param ("parse_param", $2, @2); }
285 | "%pure-parser" { pure_parser = true; }
286 | "%push-parser" { push_parser = true; pull_parser = false; }
287 | "%push-pull-parser" { push_parser = true; pull_parser = true; }
288 | "%require" STRING { version_check (&@2, $2); }
289 | "%skeleton" STRING { skeleton_arg ($2, 1, &@1); }
290 | "%token-table" { token_table_flag = true; }
291 | "%verbose" { report_flag = report_states; }
292 | "%yacc" { yacc_flag = true; }
293 | /*FIXME: Err? What is this horror doing here? */ ";"
297 precedence_declaration
301 grammar_start_symbol_set ($2, @2);
303 | "%destructor" "{...}" generic_symlist
306 for (list = $3; list; list = list->next)
307 symbol_list_destructor_set (list, $2, @2);
308 symbol_list_free ($3);
310 | "%printer" "{...}" generic_symlist
313 for (list = $3; list; list = list->next)
314 symbol_list_printer_set (list, $2, @2);
315 symbol_list_free ($3);
323 default_prec = false;
327 muscle_code_grow ("percent_code_unqualified", $2, @2);
328 code_scanner_last_string_free ();
330 | "%code" ID braceless
332 char const name_prefix[] = "percent_code(";
333 size_t length = strlen ($2);
334 char *name = xmalloc (sizeof name_prefix + length + 1);
335 strcpy (name, name_prefix);
336 strcpy (name + sizeof name_prefix - 1, $2);
337 strcpy (name + sizeof name_prefix - 1 + length, ")");
338 muscle_code_grow (uniqstr_new (name), $3, @3);
340 code_scanner_last_string_free ();
341 muscle_grow_user_name_list ("percent_code_user_qualifiers", $2, @2);
350 %token PERCENT_UNION "%union";
354 | ID { muscle_code_grow ("union_name", $1, @1); }
358 "%union" union_name braceless
361 muscle_code_grow ("stype", $3, @3);
362 code_scanner_last_string_free ();
370 "%nterm" { current_class = nterm_sym; } symbol_defs.1
372 current_class = unknown_sym;
375 | "%token" { current_class = token_sym; } symbol_defs.1
377 current_class = unknown_sym;
380 | "%type" TYPE symbols.1
384 for (list = $3; list; list = list->next)
385 symbol_type_set (list->content.sym, $2, @2);
386 symbol_list_free ($3);
390 precedence_declaration:
391 precedence_declarator type.opt symbols.1
395 for (list = $3; list; list = list->next)
397 symbol_type_set (list->content.sym, current_type, @2);
398 symbol_precedence_set (list->content.sym, current_prec, $1, @1);
400 symbol_list_free ($3);
405 precedence_declarator:
406 "%left" { $$ = left_assoc; }
407 | "%right" { $$ = right_assoc; }
408 | "%nonassoc" { $$ = non_assoc; }
412 /* Nothing. */ { current_type = NULL; }
413 | TYPE { current_type = $1; tag_seen = true; }
416 /* One or more symbols to be %typed. */
419 { $$ = symbol_list_sym_new ($1, @1); }
421 { $$ = symbol_list_prepend ($1, symbol_list_sym_new ($2, @2)); }
425 generic_symlist_item { $$ = $1; }
426 | generic_symlist generic_symlist_item { $$ = symbol_list_prepend ($1, $2); }
429 generic_symlist_item:
430 symbol { $$ = symbol_list_sym_new ($1, @1); }
431 | TYPE { $$ = symbol_list_type_new ($1, @1); }
432 | "<*>" { $$ = symbol_list_default_tagged_new (@1); }
433 | "<>" { $$ = symbol_list_default_tagless_new (@1); }
436 /* One token definition. */
445 symbol_class_set ($1, current_class, @1, true);
446 symbol_type_set ($1, current_type, @1);
450 symbol_class_set ($1, current_class, @1, true);
451 symbol_type_set ($1, current_type, @1);
452 symbol_user_token_number_set ($1, $2, @2);
456 symbol_class_set ($1, current_class, @1, true);
457 symbol_type_set ($1, current_type, @1);
458 symbol_make_alias ($1, $2, @$);
460 | id INT string_as_id
462 symbol_class_set ($1, current_class, @1, true);
463 symbol_type_set ($1, current_type, @1);
464 symbol_user_token_number_set ($1, $2, @2);
465 symbol_make_alias ($1, $3, @$);
469 /* One or more symbol definitions. */
472 | symbol_defs.1 symbol_def
476 /*------------------------------------------.
477 | The grammar section: between the two %%. |
478 `------------------------------------------*/
481 rules_or_grammar_declaration
482 | grammar rules_or_grammar_declaration
485 /* As a Bison extension, one can use the grammar declarations in the
486 body of the grammar. */
487 rules_or_grammar_declaration:
489 | grammar_declaration ";"
497 id_colon { current_lhs = $1; current_lhs_location = @1; } rhses.1
501 rhs { grammar_current_rule_end (@1); }
502 | rhses.1 "|" rhs { grammar_current_rule_end (@3); }
508 { grammar_current_rule_begin (current_lhs, current_lhs_location); }
510 { grammar_current_rule_symbol_append ($2, @2); }
512 { grammar_current_rule_action_append ($2, @2); }
514 { grammar_current_rule_prec_set ($3, @3); }
516 { grammar_current_rule_dprec_set ($3, @3); }
518 { grammar_current_rule_merge_set ($3, @3); }
522 /*----------------------------*
523 | variable and content.opt. |
524 *---------------------------*/
528 | STRING { $$ = uniqstr_new ($1); } /* deprecated and not M4-friendly */
531 /* Some content or "1" by default. */
535 static char one[] = "1";
549 code_props plain_code;
550 $1[strlen ($1) - 1] = '\n';
551 code_props_plain_init (&plain_code, $1+1, @1);
552 code_props_translate_code (&plain_code);
553 gram_scanner_last_string_free ();
554 $$ = plain_code.code;
563 /* Identifiers are returned as uniqstr values by the scanner.
564 Depending on their use, we may need to make them genuine symbols. */
568 { $$ = symbol_from_uniqstr ($1, @1); }
571 $$ = symbol_get (char_name ($1), @1);
572 symbol_class_set ($$, token_sym, @1, false);
573 symbol_user_token_number_set ($$, $1, @1);
578 ID_COLON { $$ = symbol_from_uniqstr ($1, @1); }
587 /* A string used as an ID: quote it. */
591 $$ = symbol_get (quotearg_style (c_quoting_style, $1), @1);
592 symbol_class_set ($$, token_sym, @1, false);
600 code_props plain_code;
601 code_props_plain_init (&plain_code, $2, @2);
602 code_props_translate_code (&plain_code);
603 gram_scanner_last_string_free ();
604 muscle_code_grow ("epilogue", plain_code.code, @2);
605 code_scanner_last_string_free ();
612 /* Return the location of the left-hand side of a rule whose
613 right-hand side is RHS[1] ... RHS[N]. Ignore empty nonterminals in
614 the right-hand side, and return an empty location equal to the end
615 boundary of RHS[0] if the right-hand side is empty. */
618 lloc_default (YYLTYPE const *rhs, int n)
623 /* SGI MIPSpro 7.4.1m miscompiles "loc.start = loc.end = rhs[n].end;".
624 The bug is fixed in 7.4.2m, but play it safe for now. */
625 loc.start = rhs[n].end;
626 loc.end = rhs[n].end;
628 /* Ignore empty nonterminals the start of the the right-hand side.
629 Do not bother to ignore them at the end of the right-hand side,
630 since empty nonterminals have the same end as their predecessors. */
631 for (i = 1; i <= n; i++)
632 if (! equal_boundaries (rhs[i].start, rhs[i].end))
634 loc.start = rhs[i].start;
642 /* Add a lex-param or a parse-param (depending on TYPE) with
643 declaration DECL and location LOC. */
646 add_param (char const *type, char *decl, location loc)
648 static char const alphanum[26 + 26 + 1 + 10] =
649 "abcdefghijklmnopqrstuvwxyz"
650 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
653 char const *name_start = NULL;
656 /* Stop on last actual character. */
657 for (p = decl; p[1]; p++)
659 || ! memchr (alphanum, p[-1], sizeof alphanum))
660 && memchr (alphanum, p[0], sizeof alphanum - 10))
663 /* Strip the surrounding '{' and '}', and any blanks just inside
665 while (*--p == ' ' || *p == '\t')
668 while (*++decl == ' ' || *decl == '\t')
672 complain_at (loc, _("missing identifier in parameter declaration"));
679 memchr (alphanum, name_start[name_len], sizeof alphanum);
683 name = xmalloc (name_len + 1);
684 memcpy (name, name_start, name_len);
685 name[name_len] = '\0';
686 muscle_pair_list_grow (type, decl, name);
690 gram_scanner_last_string_free ();
695 version_check (location const *loc, char const *version)
697 if (strverscmp (version, PACKAGE_VERSION) > 0)
699 complain_at (*loc, "require bison %s, but have %s",
700 version, PACKAGE_VERSION);
706 gram_error (location const *loc, char const *msg)
708 complain_at (*loc, "%s", msg);
712 token_name (int type)
714 return yytname[YYTRANSLATE (type)];
725 buf[0] = '\''; buf[1] = c; buf[2] = '\''; buf[3] = '\0';
726 return quotearg_style (escape_quoting_style, buf);