1 /* Output the generated parsing program for Bison.
3 Copyright (C) 1984, 1986, 1989, 1992, 2000, 2001, 2002, 2003, 2004,
4 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
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 3 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, see <http://www.gnu.org/licenses/>. */
24 #include <configmake.h>
26 #include <get-errno.h>
35 #include "muscle_tab.h"
38 #include "scan-code.h" /* max_left_semantic_context */
39 #include "scan-skel.h"
44 static struct obstack format_obstack
;
47 /*-------------------------------------------------------------------.
48 | Create a function NAME which associates to the muscle NAME the |
49 | result of formatting the FIRST and then TABLE_DATA[BEGIN..END[ (of |
50 | TYPE), and to the muscle NAME_max, the max value of the |
52 `-------------------------------------------------------------------*/
55 #define GENERATE_MUSCLE_INSERT_TABLE(Name, Type) \
58 Name (char const *name, \
71 obstack_fgrow1 (&format_obstack, "%6d", first); \
72 for (i = begin; i < end; ++i) \
74 obstack_1grow (&format_obstack, ','); \
77 obstack_sgrow (&format_obstack, "\n "); \
82 obstack_fgrow1 (&format_obstack, "%6d", table_data[i]); \
83 if (table_data[i] < min) \
84 min = table_data[i]; \
85 if (max < table_data[i]) \
86 max = table_data[i]; \
88 obstack_1grow (&format_obstack, 0); \
89 muscle_insert (name, obstack_finish (&format_obstack)); \
93 /* Build `NAME_min' and `NAME_max' in the obstack. */ \
94 obstack_fgrow1 (&format_obstack, "%s_min", name); \
95 obstack_1grow (&format_obstack, 0); \
96 MUSCLE_INSERT_LONG_INT (obstack_finish (&format_obstack), lmin); \
97 obstack_fgrow1 (&format_obstack, "%s_max", name); \
98 obstack_1grow (&format_obstack, 0); \
99 MUSCLE_INSERT_LONG_INT (obstack_finish (&format_obstack), lmax); \
102 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_unsigned_int_table
, unsigned int)
103 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_int_table
, int)
104 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_base_table
, base_number
)
105 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_rule_number_table
, rule_number
)
106 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_symbol_number_table
, symbol_number
)
107 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_item_number_table
, item_number
)
108 GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_state_number_table
, state_number
)
111 /*--------------------------------------------------------------------.
112 | Print to OUT a representation of STRING escaped both for C and M4. |
113 `--------------------------------------------------------------------*/
116 escaped_output (FILE *out
, char const *string
)
121 for (p
= quotearg_style (c_quoting_style
, string
); *p
; p
++)
124 case '$': fputs ("$][", out
); break;
125 case '@': fputs ("@@", out
); break;
126 case '[': fputs ("@{", out
); break;
127 case ']': fputs ("@}", out
); break;
128 default: fputc (*p
, out
); break;
135 /*------------------------------------------------------------------.
136 | Prepare the muscles related to the symbols: translate, tname, and |
138 `------------------------------------------------------------------*/
141 prepare_symbols (void)
143 MUSCLE_INSERT_BOOL ("token_table", token_table_flag
);
144 MUSCLE_INSERT_INT ("tokens_number", ntokens
);
145 MUSCLE_INSERT_INT ("nterms_number", nvars
);
146 MUSCLE_INSERT_INT ("undef_token_number", undeftoken
->number
);
147 MUSCLE_INSERT_INT ("user_token_number_max", max_user_token_number
);
149 muscle_insert_symbol_number_table ("translate",
151 token_translations
[0],
152 1, max_user_token_number
+ 1);
154 /* tname -- token names. */
157 /* We assume that the table will be output starting at column 2. */
159 struct quoting_options
*qo
= clone_quoting_options (0);
160 set_quoting_style (qo
, c_quoting_style
);
161 set_quoting_flags (qo
, QA_SPLIT_TRIGRAPHS
);
162 for (i
= 0; i
< nsyms
; i
++)
164 char *cp
= quotearg_alloc (symbols
[i
]->tag
, -1, qo
);
165 /* Width of the next token, including the two quotes, the
166 comma and the space. */
167 int width
= strlen (cp
) + 2;
171 obstack_sgrow (&format_obstack
, "\n ");
176 obstack_1grow (&format_obstack
, ' ');
177 MUSCLE_OBSTACK_SGROW (&format_obstack
, cp
);
179 obstack_1grow (&format_obstack
, ',');
183 obstack_sgrow (&format_obstack
, " ]b4_null[");
185 /* Finish table and store. */
186 obstack_1grow (&format_obstack
, 0);
187 muscle_insert ("tname", obstack_finish (&format_obstack
));
190 /* Output YYTOKNUM. */
193 int *values
= xnmalloc (ntokens
, sizeof *values
);
194 for (i
= 0; i
< ntokens
; ++i
)
195 values
[i
] = symbols
[i
]->user_token_number
;
196 muscle_insert_int_table ("toknum", values
,
197 values
[0], 1, ntokens
);
203 /*-------------------------------------------------------------.
204 | Prepare the muscles related to the rules: rhs, prhs, r1, r2, |
205 | rline, dprec, merger. |
206 `-------------------------------------------------------------*/
213 item_number
*rhs
= xnmalloc (nritems
, sizeof *rhs
);
214 unsigned int *prhs
= xnmalloc (nrules
, sizeof *prhs
);
215 unsigned int *rline
= xnmalloc (nrules
, sizeof *rline
);
216 symbol_number
*r1
= xnmalloc (nrules
, sizeof *r1
);
217 unsigned int *r2
= xnmalloc (nrules
, sizeof *r2
);
218 int *dprec
= xnmalloc (nrules
, sizeof *dprec
);
219 int *merger
= xnmalloc (nrules
, sizeof *merger
);
221 for (r
= 0; r
< nrules
; ++r
)
223 item_number
*rhsp
= NULL
;
224 /* Index of rule R in RHS. */
226 /* RHS of the rule R. */
227 for (rhsp
= rules
[r
].rhs
; *rhsp
>= 0; ++rhsp
)
229 /* LHS of the rule R. */
230 r1
[r
] = rules
[r
].lhs
->number
;
231 /* Length of rule R's RHS. */
233 /* Separator in RHS. */
235 /* Line where rule was defined. */
236 rline
[r
] = rules
[r
].location
.start
.line
;
237 /* Dynamic precedence (GLR). */
238 dprec
[r
] = rules
[r
].dprec
;
239 /* Merger-function index (GLR). */
240 merger
[r
] = rules
[r
].merger
;
244 muscle_insert_item_number_table ("rhs", rhs
, ritem
[0], 1, nritems
);
245 muscle_insert_unsigned_int_table ("prhs", prhs
, 0, 0, nrules
);
246 muscle_insert_unsigned_int_table ("rline", rline
, 0, 0, nrules
);
247 muscle_insert_symbol_number_table ("r1", r1
, 0, 0, nrules
);
248 muscle_insert_unsigned_int_table ("r2", r2
, 0, 0, nrules
);
249 muscle_insert_int_table ("dprec", dprec
, 0, 0, nrules
);
250 muscle_insert_int_table ("merger", merger
, 0, 0, nrules
);
252 MUSCLE_INSERT_INT ("rules_number", nrules
);
253 MUSCLE_INSERT_INT ("max_left_semantic_context", max_left_semantic_context
);
264 /*--------------------------------------------.
265 | Prepare the muscles related to the states. |
266 `--------------------------------------------*/
269 prepare_states (void)
272 symbol_number
*values
= xnmalloc (nstates
, sizeof *values
);
273 for (i
= 0; i
< nstates
; ++i
)
274 values
[i
] = states
[i
]->accessing_symbol
;
275 muscle_insert_symbol_number_table ("stos", values
,
279 MUSCLE_INSERT_INT ("last", high
);
280 MUSCLE_INSERT_INT ("final_state_number", final_state
->number
);
281 MUSCLE_INSERT_INT ("states_number", nstates
);
286 /*---------------------------------.
287 | Output the user actions to OUT. |
288 `---------------------------------*/
291 user_actions_output (FILE *out
)
295 fputs ("m4_define([b4_actions], \n[", out
);
296 for (r
= 0; r
< nrules
; ++r
)
299 fprintf (out
, "b4_case(%d, [b4_syncline(%d, ", r
+ 1,
300 rules
[r
].action_location
.start
.line
);
301 escaped_output (out
, rules
[r
].action_location
.start
.file
);
302 fprintf (out
, ")\n[ %s]])\n\n", rules
[r
].action
);
304 fputs ("])\n\n", out
);
307 /*--------------------------------------.
308 | Output the merge functions to OUT. |
309 `--------------------------------------*/
312 merger_output (FILE *out
)
317 fputs ("m4_define([b4_mergers], \n[[", out
);
318 for (n
= 1, p
= merge_functions
; p
!= NULL
; n
+= 1, p
= p
->next
)
320 if (p
->type
[0] == '\0')
321 fprintf (out
, " case %d: *yy0 = %s (*yy0, *yy1); break;\n",
324 fprintf (out
, " case %d: yy0->%s = %s (*yy0, *yy1); break;\n",
325 n
, p
->type
, p
->name
);
327 fputs ("]])\n\n", out
);
330 /*--------------------------------------.
331 | Output the tokens definition to OUT. |
332 `--------------------------------------*/
335 token_definitions_output (FILE *out
)
338 char const *sep
= "";
340 fputs ("m4_define([b4_tokens], \n[", out
);
341 for (i
= 0; i
< ntokens
; ++i
)
343 symbol
*sym
= symbols
[i
];
344 int number
= sym
->user_token_number
;
346 /* At this stage, if there are literal aliases, they are part of
347 SYMBOLS, so we should not find symbols which are the aliases
349 aver (number
!= USER_NUMBER_ALIAS
);
351 /* Skip error token. */
355 /* If this string has an alias, then it is necessarily the alias
356 which is to be output. */
360 /* Don't output literal chars or strings (when defined only as a
361 string). Note that must be done after the alias resolution:
362 think about `%token 'f' "f"'. */
363 if (sym
->tag
[0] == '\'' || sym
->tag
[0] == '\"')
366 /* Don't #define nonliteral tokens whose names contain periods
367 or '$' (as does the default value of the EOF token). */
368 if (strchr (sym
->tag
, '.') || strchr (sym
->tag
, '$'))
371 fprintf (out
, "%s[[[%s]], %d]",
372 sep
, sym
->tag
, number
);
375 fputs ("])\n\n", out
);
379 /*---------------------------------------------------.
380 | Output the symbol destructors or printers to OUT. |
381 `---------------------------------------------------*/
384 symbol_code_props_output (FILE *out
, char const *what
,
385 code_props
const *(*get
)(symbol
const *))
388 char const *sep
= "";
390 fputs ("m4_define([b4_symbol_", out
);
392 fputs ("], \n[", out
);
393 for (i
= 0; i
< nsyms
; ++i
)
395 symbol
*sym
= symbols
[i
];
396 char const *code
= (*get
) (sym
)->code
;
399 location loc
= (*get
) (sym
)->location
;
401 Symbol-name, Symbol-number,
402 code, optional typename. */
403 fprintf (out
, "%s[", sep
);
405 escaped_output (out
, loc
.start
.file
);
406 fprintf (out
, ", %d, ", loc
.start
.line
);
407 escaped_output (out
, sym
->tag
);
408 fprintf (out
, ", %d, [[%s]]", sym
->number
, code
);
410 fprintf (out
, ", [[%s]]", sym
->type_name
);
414 fputs ("])\n\n", out
);
419 prepare_actions (void)
421 /* Figure out the actions for the specified state, indexed by
422 lookahead token type. */
424 muscle_insert_rule_number_table ("defact", yydefact
,
425 yydefact
[0], 1, nstates
);
427 /* Figure out what to do after reducing with each rule, depending on
428 the saved state from before the beginning of parsing the data
429 that matched this rule. */
430 muscle_insert_state_number_table ("defgoto", yydefgoto
,
431 yydefgoto
[0], 1, nsyms
- ntokens
);
435 muscle_insert_base_table ("pact", base
,
436 base
[0], 1, nstates
);
437 MUSCLE_INSERT_INT ("pact_ninf", base_ninf
);
440 muscle_insert_base_table ("pgoto", base
,
441 base
[nstates
], nstates
+ 1, nvectors
);
443 muscle_insert_base_table ("table", table
,
444 table
[0], 1, high
+ 1);
445 MUSCLE_INSERT_INT ("table_ninf", table_ninf
);
447 muscle_insert_base_table ("check", check
,
448 check
[0], 1, high
+ 1);
450 /* GLR parsing slightly modifies YYTABLE and YYCHECK (and thus
451 YYPACT) so that in states with unresolved conflicts, the default
452 reduction is not used in the conflicted entries, so that there is
453 a place to put a conflict pointer.
455 This means that YYCONFLP and YYCONFL are nonsense for a non-GLR
456 parser, so we could avoid accidents by not writing them out in
457 that case. Nevertheless, it seems even better to be able to use
458 the GLR skeletons even without the non-deterministic tables. */
459 muscle_insert_unsigned_int_table ("conflict_list_heads", conflict_table
,
460 conflict_table
[0], 1, high
+ 1);
461 muscle_insert_unsigned_int_table ("conflicting_rules", conflict_list
,
462 0, 1, conflict_list_cnt
);
465 /*--------------------------------------------.
466 | Output the definitions of all the muscles. |
467 `--------------------------------------------*/
470 muscles_output (FILE *out
)
472 fputs ("m4_init()\n", out
);
474 user_actions_output (out
);
476 token_definitions_output (out
);
477 symbol_code_props_output (out
, "destructors", &symbol_destructor_get
);
478 symbol_code_props_output (out
, "printers", &symbol_printer_get
);
480 muscles_m4_output (out
);
483 /*---------------------------.
484 | Call the skeleton parser. |
485 `---------------------------*/
488 output_skeleton (void)
495 /* Compute the names of the package data dir and skeleton files. */
496 char const m4sugar
[] = "m4sugar/m4sugar.m4";
497 char const m4bison
[] = "bison.m4";
502 char const *m4
= (p
= getenv ("M4")) ? p
: M4
;
503 char const *pkgdatadir
= compute_pkgdatadir ();
504 size_t skeleton_size
= strlen (skeleton
) + 1;
505 size_t pkgdatadirlen
= strlen (pkgdatadir
);
506 while (pkgdatadirlen
&& pkgdatadir
[pkgdatadirlen
- 1] == '/')
508 full_skeleton
= xmalloc (pkgdatadirlen
+ 1
509 + (skeleton_size
< sizeof m4sugar
510 ? sizeof m4sugar
: skeleton_size
));
511 strncpy (full_skeleton
, pkgdatadir
, pkgdatadirlen
);
512 full_skeleton
[pkgdatadirlen
] = '/';
513 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, m4sugar
);
514 full_m4sugar
= xstrdup (full_skeleton
);
515 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, m4bison
);
516 full_m4bison
= xstrdup (full_skeleton
);
517 if (strchr (skeleton
, '/'))
518 strcpy (full_skeleton
, skeleton
);
520 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, skeleton
);
522 /* Test whether m4sugar.m4 is readable, to check for proper
523 installation. A faulty installation can cause deadlock, so a
524 cheap sanity check is worthwhile. */
525 xfclose (xfopen (full_m4sugar
, "r"));
527 /* Create an m4 subprocess connected to us via two pipes. */
529 if (trace_flag
& trace_tools
)
530 fprintf (stderr
, "running: %s %s - %s %s\n",
531 m4
, full_m4sugar
, full_m4bison
, full_skeleton
);
533 /* Some future version of GNU M4 (most likely 1.6) may treat the -dV in a
534 position-dependent manner. Keep it as the first argument so that all
537 See the thread starting at
538 <http://lists.gnu.org/archive/html/bug-bison/2008-07/msg00000.html>
544 argv
[i
++] = pkgdatadir
;
545 if (trace_flag
& trace_m4
)
547 argv
[i
++] = full_m4sugar
;
549 argv
[i
++] = full_m4bison
;
550 argv
[i
++] = full_skeleton
;
553 /* When POSIXLY_CORRECT is set, some future versions of GNU M4 (most likely
554 2.0) may drop some of the GNU extensions that Bison's skeletons depend
555 upon. So that the next release of Bison is forward compatible with those
556 future versions of GNU M4, we unset POSIXLY_CORRECT here.
558 FIXME: A user might set POSIXLY_CORRECT to affect processes run from
559 macros like m4_syscmd in a custom skeleton. For now, Bison makes no
560 promises about the behavior of custom skeletons, so this scenario is not a
561 concern. However, we eventually want to eliminate this shortcoming. The
562 next release of GNU M4 (1.4.12 or 1.6) will accept the -g command-line
563 option as a no-op, and later releases will accept it to indicate that
564 POSIXLY_CORRECT should be ignored. Once the GNU M4 versions that accept
565 -g are pervasive, Bison should use -g instead of unsetting
568 See the thread starting at
569 <http://lists.gnu.org/archive/html/bug-bison/2008-07/msg00000.html>
571 unsetenv ("POSIXLY_CORRECT");
573 pid
= create_subpipe (argv
, filter_fd
);
576 free (full_skeleton
);
579 if (trace_flag
& trace_muscles
)
580 muscles_output (stderr
);
582 FILE *out
= fdopen (filter_fd
[0], "w");
584 error (EXIT_FAILURE
, get_errno (),
586 muscles_output (out
);
590 /* Read and process m4's output. */
591 timevar_push (TV_M4
);
592 end_of_output_subpipe (pid
, filter_fd
);
593 in
= fdopen (filter_fd
[1], "r");
595 error (EXIT_FAILURE
, get_errno (),
599 reap_subpipe (pid
, m4
);
606 /* BISON_USE_PUSH_FOR_PULL is for the test suite and should not be documented
608 char const *use_push_for_pull_env
= getenv ("BISON_USE_PUSH_FOR_PULL");
609 bool use_push_for_pull_flag
= false;
610 if (use_push_for_pull_env
!= NULL
611 && use_push_for_pull_env
[0] != '\0'
612 && 0 != strcmp (use_push_for_pull_env
, "0"))
613 use_push_for_pull_flag
= true;
616 MUSCLE_INSERT_BOOL ("debug_flag", debug_flag
);
617 MUSCLE_INSERT_BOOL ("defines_flag", defines_flag
);
618 MUSCLE_INSERT_BOOL ("error_verbose_flag", error_verbose
);
619 MUSCLE_INSERT_BOOL ("glr_flag", glr_parser
);
620 MUSCLE_INSERT_BOOL ("locations_flag", locations_flag
);
621 MUSCLE_INSERT_BOOL ("nondeterministic_flag", nondeterministic_parser
);
622 MUSCLE_INSERT_BOOL ("synclines_flag", !no_lines_flag
);
623 MUSCLE_INSERT_BOOL ("tag_seen_flag", tag_seen
);
624 MUSCLE_INSERT_BOOL ("use_push_for_pull_flag", use_push_for_pull_flag
);
625 MUSCLE_INSERT_BOOL ("yacc_flag", yacc_flag
);
628 if (spec_name_prefix
)
629 MUSCLE_INSERT_STRING ("prefix", spec_name_prefix
);
631 MUSCLE_INSERT_STRING ("file_name_all_but_ext", all_but_ext
);
633 #define DEFINE(Name) MUSCLE_INSERT_STRING (#Name, Name ? Name : "")
635 DEFINE (parser_file_name
);
636 DEFINE (spec_defines_file
);
637 DEFINE (spec_file_prefix
);
638 DEFINE (spec_graph_file
);
639 DEFINE (spec_name_prefix
);
640 DEFINE (spec_outfile
);
641 DEFINE (spec_verbose_file
);
644 /* Find the right skeleton file, and add muscles about the skeletons. */
646 MUSCLE_INSERT_C_STRING ("skeleton", skeleton
);
648 skeleton
= language
->skeleton
;
650 /* About the skeletons. */
652 /* b4_pkgdatadir is used inside m4_include in the skeletons, so digraphs
653 would never be expanded. Hopefully no one has M4-special characters in
654 his Bison installation path. */
655 MUSCLE_INSERT_STRING_RAW ("pkgdatadir", compute_pkgdatadir ());
660 /*----------------------------------------------------------.
661 | Output the parsing tables and the parser code to ftable. |
662 `----------------------------------------------------------*/
667 obstack_init (&format_obstack
);
676 /* Process the selected skeleton file. */
679 obstack_free (&format_obstack
, NULL
);
683 compute_pkgdatadir (void)
685 char const *pkgdatadir
= getenv ("BISON_PKGDATADIR");
686 return pkgdatadir
? pkgdatadir
: PKGDATADIR
;