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
91 /* Bison's grammar can initial empty locations, hence a default
92 location is needed. */
93 boundary_set (&@$.start, current_file, 1, 1);
94 boundary_set (&@$.end, current_file, 1, 1);
106 unsigned char character;
109 /* Define the tokens together with their human representation. */
110 %token GRAM_EOF 0 "end of file"
111 %token STRING "string"
114 %token PERCENT_TOKEN "%token"
115 %token PERCENT_NTERM "%nterm"
117 %token PERCENT_TYPE "%type"
118 %token PERCENT_DESTRUCTOR "%destructor"
119 %token PERCENT_PRINTER "%printer"
121 %token PERCENT_LEFT "%left"
122 %token PERCENT_RIGHT "%right"
123 %token PERCENT_NONASSOC "%nonassoc"
125 %token PERCENT_PREC "%prec"
126 %token PERCENT_DPREC "%dprec"
127 %token PERCENT_MERGE "%merge"
130 /*----------------------.
131 | Global Declarations. |
132 `----------------------*/
136 PERCENT_CODE_TOP "%code-top"
137 PERCENT_DEBUG "%debug"
138 PERCENT_DEFAULT_PREC "%default-prec"
139 PERCENT_DEFINE "%define"
140 PERCENT_DEFINES "%defines"
141 PERCENT_ERROR_VERBOSE "%error-verbose"
142 PERCENT_EXPECT "%expect"
143 PERCENT_EXPECT_RR "%expect-rr"
144 PERCENT_FILE_PREFIX "%file-prefix"
145 PERCENT_GLR_PARSER "%glr-parser"
146 PERCENT_INITIAL_ACTION "%initial-action"
147 PERCENT_LANGUAGE "%language"
148 PERCENT_LEX_PARAM "%lex-param"
149 PERCENT_LOCATIONS "%locations"
150 PERCENT_NAME_PREFIX "%name-prefix"
151 PERCENT_NO_DEFAULT_PREC "%no-default-prec"
152 PERCENT_NO_LINES "%no-lines"
153 PERCENT_NONDETERMINISTIC_PARSER
154 "%nondeterministic-parser"
155 PERCENT_OUTPUT "%output"
156 PERCENT_PARSE_PARAM "%parse-param"
157 PERCENT_PROVIDES "%provides"
158 PERCENT_PURE_PARSER "%pure-parser"
159 PERCENT_PUSH_PARSER "%push-parser"
160 PERCENT_PUSH_PULL_PARSER
162 PERCENT_REQUIRE "%require"
163 PERCENT_REQUIRES "%requires"
164 PERCENT_SKELETON "%skeleton"
165 PERCENT_START "%start"
166 PERCENT_TOKEN_TABLE "%token-table"
167 PERCENT_VERBOSE "%verbose"
171 %token BRACED_CODE "{...}"
173 %token EPILOGUE "epilogue"
175 %token ID "identifier"
176 %token ID_COLON "identifier:"
177 %token PERCENT_PERCENT "%%"
179 %token PROLOGUE "%{...%}"
182 %token TYPE_TAG_ANY "<*>"
183 %token TYPE_TAG_NONE "<>"
185 %type <character> CHAR
186 %printer { fputs (char_name ($$), stderr); } CHAR
188 /* braceless is not to be used for rule or symbol actions, as it
189 calls code_props_plain_init. */
190 %type <chars> STRING "%{...%}" EPILOGUE braceless content content.opt
192 %printer { fputs (quotearg_style (c_quoting_style, $$), stderr); }
194 %printer { fprintf (stderr, "{\n%s\n}", $$); }
195 braceless content content.opt "{...}" "%{...%}" EPILOGUE
197 %type <uniqstr> TYPE ID ID_COLON
198 %printer { fprintf (stderr, "<%s>", $$); } TYPE
199 %printer { fputs ($$, stderr); } ID
200 %printer { fprintf (stderr, "%s:", $$); } ID_COLON
203 %printer { fprintf (stderr, "%d", $$); } INT
205 %type <symbol> id id_colon symbol string_as_id
206 %printer { fprintf (stderr, "%s", $$->tag); } id symbol string_as_id
207 %printer { fprintf (stderr, "%s:", $$->tag); } id_colon
209 %type <assoc> precedence_declarator
210 %type <list> symbols.1 generic_symlist generic_symlist_item
214 prologue_declarations "%%" grammar epilogue.opt
218 /*------------------------------------.
219 | Declarations: before the first %%. |
220 `------------------------------------*/
222 prologue_declarations:
224 | prologue_declarations prologue_declaration
227 prologue_declaration:
231 code_props plain_code;
232 code_props_plain_init (&plain_code, $1, @1);
233 code_props_translate_code (&plain_code);
234 gram_scanner_last_string_free ();
235 prologue_augment (plain_code.code, @1, union_seen);
236 code_scanner_last_string_free ();
238 | "%debug" { debug_flag = true; }
239 | "%define" STRING content.opt { muscle_insert ($2, $3); }
240 | "%defines" { defines_flag = true; }
244 spec_defines_file = xstrdup ($2);
246 | "%error-verbose" { error_verbose = true; }
247 | "%expect" INT { expected_sr_conflicts = $2; }
248 | "%expect-rr" INT { expected_rr_conflicts = $2; }
249 | "%file-prefix" STRING { spec_file_prefix = $2; }
250 | "%file-prefix" "=" STRING { spec_file_prefix = $3; } /* deprecated */
253 nondeterministic_parser = true;
256 | "%initial-action" "{...}"
259 code_props_symbol_action_init (&action, $2, @2);
260 code_props_translate_code (&action);
261 gram_scanner_last_string_free ();
262 muscle_code_grow ("initial_action", action.code, @2);
263 code_scanner_last_string_free ();
265 | "%language" STRING { language_argmatch ($2, 1, &@1); }
266 | "%lex-param" "{...}" { add_param ("lex_param", $2, @2); }
267 | "%locations" { locations_flag = true; }
268 | "%name-prefix" STRING { spec_name_prefix = $2; }
269 | "%name-prefix" "=" STRING { spec_name_prefix = $3; } /* deprecated */
270 | "%no-lines" { no_lines_flag = true; }
271 | "%nondeterministic-parser" { nondeterministic_parser = true; }
272 | "%output" STRING { spec_outfile = $2; }
273 | "%output" "=" STRING { spec_outfile = $3; } /* deprecated */
274 | "%parse-param" "{...}" { add_param ("parse_param", $2, @2); }
275 | "%pure-parser" { pure_parser = true; }
276 | "%push-parser" { push_parser = true; pull_parser = false; }
277 | "%push-pull-parser" { push_parser = true; pull_parser = true; }
278 | "%require" STRING { version_check (&@2, $2); }
279 | "%skeleton" STRING { skeleton_arg ($2, 1, &@1); }
280 | "%token-table" { token_table_flag = true; }
281 | "%verbose" { report_flag = report_states; }
282 | "%yacc" { yacc_flag = true; }
283 | /*FIXME: Err? What is this horror doing here? */ ";"
287 precedence_declaration
291 grammar_start_symbol_set ($2, @2);
293 | "%destructor" "{...}" generic_symlist
296 for (list = $3; list; list = list->next)
297 symbol_list_destructor_set (list, $2, @2);
298 symbol_list_free ($3);
300 | "%printer" "{...}" generic_symlist
303 for (list = $3; list; list = list->next)
304 symbol_list_printer_set (list, $2, @2);
305 symbol_list_free ($3);
313 default_prec = false;
315 | "%code" braceless { prologue_augment ($2, @2, true); }
316 | "%code-top" braceless { prologue_augment ($2, @2, false); }
317 | "%provides" braceless { muscle_code_grow ("provides", $2, @2); }
318 | "%requires" braceless { muscle_code_grow ("requires", $2, @2); }
326 %token PERCENT_UNION "%union";
330 | ID { muscle_code_grow ("union_name", $1, @1); }
334 "%union" union_name "{...}"
336 char const *body = $3;
338 /* Concatenate the %union bodies. If this is the first %union, make sure
339 the synchronization line appears after the opening '{' so as not to
340 confuse Doxygen. Otherwise, turn the previous %union's trailing '}'
341 into '\n', and omit the new %union's leading '{'. */
344 muscle_grow ("stype", "{", "");
348 char *code = muscle_find ("stype");
349 code[strlen (code) - 1] = '\n';
354 muscle_code_grow ("stype", body, @3);
355 gram_scanner_last_string_free ();
363 "%nterm" { current_class = nterm_sym; } symbol_defs.1
365 current_class = unknown_sym;
368 | "%token" { current_class = token_sym; } symbol_defs.1
370 current_class = unknown_sym;
373 | "%type" TYPE symbols.1
377 for (list = $3; list; list = list->next)
378 symbol_type_set (list->content.sym, $2, @2);
379 symbol_list_free ($3);
383 precedence_declaration:
384 precedence_declarator type.opt symbols.1
388 for (list = $3; list; list = list->next)
390 symbol_type_set (list->content.sym, current_type, @2);
391 symbol_precedence_set (list->content.sym, current_prec, $1, @1);
393 symbol_list_free ($3);
398 precedence_declarator:
399 "%left" { $$ = left_assoc; }
400 | "%right" { $$ = right_assoc; }
401 | "%nonassoc" { $$ = non_assoc; }
405 /* Nothing. */ { current_type = NULL; }
406 | TYPE { current_type = $1; tag_seen = true; }
409 /* One or more symbols to be %typed. */
412 { $$ = symbol_list_sym_new ($1, @1); }
414 { $$ = symbol_list_prepend ($1, symbol_list_sym_new ($2, @2)); }
418 generic_symlist_item { $$ = $1; }
419 | generic_symlist generic_symlist_item { $$ = symbol_list_prepend ($1, $2); }
422 generic_symlist_item:
423 symbol { $$ = symbol_list_sym_new ($1, @1); }
424 | TYPE { $$ = symbol_list_type_new ($1, @1); }
425 | "<*>" { $$ = symbol_list_default_tagged_new (@1); }
426 | "<>" { $$ = symbol_list_default_tagless_new (@1); }
429 /* One token definition. */
438 symbol_class_set ($1, current_class, @1, true);
439 symbol_type_set ($1, current_type, @1);
443 symbol_class_set ($1, current_class, @1, true);
444 symbol_type_set ($1, current_type, @1);
445 symbol_user_token_number_set ($1, $2, @2);
449 symbol_class_set ($1, current_class, @1, true);
450 symbol_type_set ($1, current_type, @1);
451 symbol_make_alias ($1, $2, @$);
453 | id INT string_as_id
455 symbol_class_set ($1, current_class, @1, true);
456 symbol_type_set ($1, current_type, @1);
457 symbol_user_token_number_set ($1, $2, @2);
458 symbol_make_alias ($1, $3, @$);
462 /* One or more symbol definitions. */
465 | symbol_defs.1 symbol_def
469 /*------------------------------------------.
470 | The grammar section: between the two %%. |
471 `------------------------------------------*/
474 rules_or_grammar_declaration
475 | grammar rules_or_grammar_declaration
478 /* As a Bison extension, one can use the grammar declarations in the
479 body of the grammar. */
480 rules_or_grammar_declaration:
482 | grammar_declaration ";"
490 id_colon { current_lhs = $1; current_lhs_location = @1; } rhses.1
494 rhs { grammar_current_rule_end (@1); }
495 | rhses.1 "|" rhs { grammar_current_rule_end (@3); }
501 { grammar_current_rule_begin (current_lhs, current_lhs_location); }
503 { grammar_current_rule_symbol_append ($2, @2); }
505 { grammar_current_rule_action_append ($2, @2); }
507 { grammar_current_rule_prec_set ($3, @3); }
509 { grammar_current_rule_dprec_set ($3, @3); }
511 { grammar_current_rule_merge_set ($3, @3); }
524 /* Some content or "1" by default. */
528 static char one[] = "1";
538 code_props plain_code;
539 $1[strlen ($1) - 1] = '\n';
540 code_props_plain_init (&plain_code, $1+1, @1);
541 code_props_translate_code (&plain_code);
542 gram_scanner_last_string_free ();
543 $$ = plain_code.code;
552 /* Identifiers are returned as uniqstr values by the scanner.
553 Depending on their use, we may need to make them genuine symbols. */
557 { $$ = symbol_from_uniqstr ($1, @1); }
560 $$ = symbol_get (char_name ($1), @1);
561 symbol_class_set ($$, token_sym, @1, false);
562 symbol_user_token_number_set ($$, $1, @1);
567 ID_COLON { $$ = symbol_from_uniqstr ($1, @1); }
576 /* A string used as an ID: quote it. */
580 $$ = symbol_get (quotearg_style (c_quoting_style, $1), @1);
581 symbol_class_set ($$, token_sym, @1, false);
589 code_props plain_code;
590 code_props_plain_init (&plain_code, $2, @2);
591 code_props_translate_code (&plain_code);
592 gram_scanner_last_string_free ();
593 muscle_code_grow ("epilogue", plain_code.code, @2);
594 code_scanner_last_string_free ();
601 /* Return the location of the left-hand side of a rule whose
602 right-hand side is RHS[1] ... RHS[N]. Ignore empty nonterminals in
603 the right-hand side, and return an empty location equal to the end
604 boundary of RHS[0] if the right-hand side is empty. */
607 lloc_default (YYLTYPE const *rhs, int n)
612 /* SGI MIPSpro 7.4.1m miscompiles "loc.start = loc.end = rhs[n].end;".
613 The bug is fixed in 7.4.2m, but play it safe for now. */
614 loc.start = rhs[n].end;
615 loc.end = rhs[n].end;
617 /* Ignore empty nonterminals the start of the the right-hand side.
618 Do not bother to ignore them at the end of the right-hand side,
619 since empty nonterminals have the same end as their predecessors. */
620 for (i = 1; i <= n; i++)
621 if (! equal_boundaries (rhs[i].start, rhs[i].end))
623 loc.start = rhs[i].start;
631 /* Add a lex-param or a parse-param (depending on TYPE) with
632 declaration DECL and location LOC. */
635 add_param (char const *type, char *decl, location loc)
637 static char const alphanum[26 + 26 + 1 + 10] =
638 "abcdefghijklmnopqrstuvwxyz"
639 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
642 char const *name_start = NULL;
645 /* Stop on last actual character. */
646 for (p = decl; p[1]; p++)
648 || ! memchr (alphanum, p[-1], sizeof alphanum))
649 && memchr (alphanum, p[0], sizeof alphanum - 10))
652 /* Strip the surrounding '{' and '}', and any blanks just inside
654 while (*--p == ' ' || *p == '\t')
657 while (*++decl == ' ' || *decl == '\t')
661 complain_at (loc, _("missing identifier in parameter declaration"));
668 memchr (alphanum, name_start[name_len], sizeof alphanum);
672 name = xmalloc (name_len + 1);
673 memcpy (name, name_start, name_len);
674 name[name_len] = '\0';
675 muscle_pair_list_grow (type, decl, name);
679 gram_scanner_last_string_free ();
684 version_check (location const *loc, char const *version)
686 if (strverscmp (version, PACKAGE_VERSION) > 0)
688 complain_at (*loc, "require bison %s, but have %s",
689 version, PACKAGE_VERSION);
695 gram_error (location const *loc, char const *msg)
697 complain_at (*loc, "%s", msg);
701 token_name (int type)
703 return yytname[YYTRANSLATE (type)];
714 buf[0] = '\''; buf[1] = c; buf[2] = '\''; buf[3] = '\0';
715 return quotearg_style (escape_quoting_style, buf);