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 ("symbols_number", nsyms
);
147 MUSCLE_INSERT_INT ("undef_token_number", undeftoken
->number
);
148 MUSCLE_INSERT_INT ("user_token_number_max", max_user_token_number
);
150 muscle_insert_symbol_number_table ("translate",
152 token_translations
[0],
153 1, max_user_token_number
+ 1);
155 /* tname -- token names. */
158 /* We assume that the table will be output starting at column 2. */
160 struct quoting_options
*qo
= clone_quoting_options (0);
161 set_quoting_style (qo
, c_quoting_style
);
162 set_quoting_flags (qo
, QA_SPLIT_TRIGRAPHS
);
163 for (i
= 0; i
< nsyms
; i
++)
165 char *cp
= quotearg_alloc (symbols
[i
]->tag
, -1, qo
);
166 /* Width of the next token, including the two quotes, the
167 comma and the space. */
168 int width
= strlen (cp
) + 2;
172 obstack_sgrow (&format_obstack
, "\n ");
177 obstack_1grow (&format_obstack
, ' ');
178 MUSCLE_OBSTACK_SGROW (&format_obstack
, cp
);
180 obstack_1grow (&format_obstack
, ',');
184 obstack_sgrow (&format_obstack
, " ]b4_null[");
186 /* Finish table and store. */
187 obstack_1grow (&format_obstack
, 0);
188 muscle_insert ("tname", obstack_finish (&format_obstack
));
191 /* Output YYTOKNUM. */
194 int *values
= xnmalloc (ntokens
, sizeof *values
);
195 for (i
= 0; i
< ntokens
; ++i
)
196 values
[i
] = symbols
[i
]->user_token_number
;
197 muscle_insert_int_table ("toknum", values
,
198 values
[0], 1, ntokens
);
204 /*-------------------------------------------------------------.
205 | Prepare the muscles related to the rules: rhs, prhs, r1, r2, |
206 | rline, dprec, merger. |
207 `-------------------------------------------------------------*/
214 item_number
*rhs
= xnmalloc (nritems
, sizeof *rhs
);
215 unsigned int *prhs
= xnmalloc (nrules
, sizeof *prhs
);
216 unsigned int *rline
= xnmalloc (nrules
, sizeof *rline
);
217 symbol_number
*r1
= xnmalloc (nrules
, sizeof *r1
);
218 unsigned int *r2
= xnmalloc (nrules
, sizeof *r2
);
219 int *dprec
= xnmalloc (nrules
, sizeof *dprec
);
220 int *merger
= xnmalloc (nrules
, sizeof *merger
);
222 for (r
= 0; r
< nrules
; ++r
)
224 item_number
*rhsp
= NULL
;
225 /* Index of rule R in RHS. */
227 /* RHS of the rule R. */
228 for (rhsp
= rules
[r
].rhs
; *rhsp
>= 0; ++rhsp
)
230 /* LHS of the rule R. */
231 r1
[r
] = rules
[r
].lhs
->number
;
232 /* Length of rule R's RHS. */
234 /* Separator in RHS. */
236 /* Line where rule was defined. */
237 rline
[r
] = rules
[r
].location
.start
.line
;
238 /* Dynamic precedence (GLR). */
239 dprec
[r
] = rules
[r
].dprec
;
240 /* Merger-function index (GLR). */
241 merger
[r
] = rules
[r
].merger
;
245 muscle_insert_item_number_table ("rhs", rhs
, ritem
[0], 1, nritems
);
246 muscle_insert_unsigned_int_table ("prhs", prhs
, 0, 0, nrules
);
247 muscle_insert_unsigned_int_table ("rline", rline
, 0, 0, nrules
);
248 muscle_insert_symbol_number_table ("r1", r1
, 0, 0, nrules
);
249 muscle_insert_unsigned_int_table ("r2", r2
, 0, 0, nrules
);
250 muscle_insert_int_table ("dprec", dprec
, 0, 0, nrules
);
251 muscle_insert_int_table ("merger", merger
, 0, 0, nrules
);
253 MUSCLE_INSERT_INT ("rules_number", nrules
);
254 MUSCLE_INSERT_INT ("max_left_semantic_context", max_left_semantic_context
);
265 /*--------------------------------------------.
266 | Prepare the muscles related to the states. |
267 `--------------------------------------------*/
270 prepare_states (void)
273 symbol_number
*values
= xnmalloc (nstates
, sizeof *values
);
274 for (i
= 0; i
< nstates
; ++i
)
275 values
[i
] = states
[i
]->accessing_symbol
;
276 muscle_insert_symbol_number_table ("stos", values
,
280 MUSCLE_INSERT_INT ("last", high
);
281 MUSCLE_INSERT_INT ("final_state_number", final_state
->number
);
282 MUSCLE_INSERT_INT ("states_number", nstates
);
287 /*-------------------------------------.
288 | The list of all the symbol numbers. |
289 `-------------------------------------*/
292 symbol_numbers_output (FILE *out
)
295 fputs ("m4_define([b4_symbol_numbers],\n[", out
);
296 for (i
= 0; i
< nsyms
; ++i
)
297 fprintf (out
, "%s[%d]", i
? ", " : "", i
);
298 fputs ("])\n\n", out
);
302 /*---------------------------------.
303 | Output the user actions to OUT. |
304 `---------------------------------*/
307 user_actions_output (FILE *out
)
311 fputs ("m4_define([b4_actions], \n[", out
);
312 for (r
= 0; r
< nrules
; ++r
)
315 fprintf (out
, "b4_case(%d, [b4_syncline(%d, ", r
+ 1,
316 rules
[r
].action_location
.start
.line
);
317 escaped_output (out
, rules
[r
].action_location
.start
.file
);
318 fprintf (out
, ")\n[ %s]])\n\n", rules
[r
].action
);
320 fputs ("])\n\n", out
);
323 /*------------------------------------.
324 | Output the merge functions to OUT. |
325 `------------------------------------*/
328 merger_output (FILE *out
)
333 fputs ("m4_define([b4_mergers], \n[[", out
);
334 for (n
= 1, p
= merge_functions
; p
!= NULL
; n
+= 1, p
= p
->next
)
336 if (p
->type
[0] == '\0')
337 fprintf (out
, " case %d: *yy0 = %s (*yy0, *yy1); break;\n",
340 fprintf (out
, " case %d: yy0->%s = %s (*yy0, *yy1); break;\n",
341 n
, p
->type
, p
->name
);
343 fputs ("]])\n\n", out
);
347 /*---------------------------------------.
348 | Output the symbol definitions to OUT. |
349 `---------------------------------------*/
352 symbol_definitions_output (FILE *out
)
355 for (i
= 0; i
< nsyms
; ++i
)
357 symbol
*sym
= symbols
[i
];
360 #define SET_KEY(Entry) \
361 obstack_fgrow2 (&format_obstack, "symbol(%d, %s)", i, Entry); \
362 obstack_1grow (&format_obstack, 0); \
363 key = obstack_finish (&format_obstack);
366 MUSCLE_INSERT_STRING (key
, sym
->tag
);
368 SET_KEY("user_number");
369 MUSCLE_INSERT_INT (key
, sym
->user_token_number
);
372 MUSCLE_INSERT_INT (key
, sym
->number
);
374 SET_KEY("type_name");
375 MUSCLE_INSERT_STRING (key
, sym
->type_name
? sym
->type_name
: "");
382 /*--------------------------------------.
383 | Output the tokens definition to OUT. |
384 `--------------------------------------*/
387 token_definitions_output (FILE *out
)
390 char const *sep
= "";
392 fputs ("m4_define([b4_tokens], \n[", out
);
393 for (i
= 0; i
< ntokens
; ++i
)
395 symbol
*sym
= symbols
[i
];
396 int number
= sym
->user_token_number
;
398 /* At this stage, if there are literal aliases, they are part of
399 SYMBOLS, so we should not find symbols which are the aliases
401 aver (number
!= USER_NUMBER_ALIAS
);
403 /* Skip error token. */
407 /* If this string has an alias, then it is necessarily the alias
408 which is to be output. */
412 /* Don't output literal chars or strings (when defined only as a
413 string). Note that must be done after the alias resolution:
414 think about `%token 'f' "f"'. */
415 if (sym
->tag
[0] == '\'' || sym
->tag
[0] == '\"')
418 /* Don't #define nonliteral tokens whose names contain periods
419 or '$' (as does the default value of the EOF token). */
420 if (strchr (sym
->tag
, '.') || strchr (sym
->tag
, '$'))
423 fprintf (out
, "%s[[[%s]], %d]",
424 sep
, sym
->tag
, number
);
427 fputs ("])\n\n", out
);
431 /*---------------------------------------------------.
432 | Output the symbol destructors or printers to OUT. |
433 `---------------------------------------------------*/
436 symbol_code_props_output (FILE *out
, char const *what
,
437 code_props
const *(*get
)(symbol
const *))
440 char const *sep
= "";
442 fputs ("m4_define([b4_symbol_", out
);
444 fputs ("], \n[", out
);
445 for (i
= 0; i
< nsyms
; ++i
)
447 symbol
*sym
= symbols
[i
];
448 char const *code
= (*get
) (sym
)->code
;
451 location loc
= (*get
) (sym
)->location
;
453 Symbol-name, Symbol-number,
454 code, optional typename. */
455 fprintf (out
, "%s[", sep
);
457 escaped_output (out
, loc
.start
.file
);
458 fprintf (out
, ", %d, ", loc
.start
.line
);
459 escaped_output (out
, sym
->tag
);
460 fprintf (out
, ", %d, [[%s]]", sym
->number
, code
);
462 fprintf (out
, ", [[%s]]", sym
->type_name
);
466 fputs ("])\n\n", out
);
471 prepare_actions (void)
473 /* Figure out the actions for the specified state, indexed by
474 lookahead token type. */
476 muscle_insert_rule_number_table ("defact", yydefact
,
477 yydefact
[0], 1, nstates
);
479 /* Figure out what to do after reducing with each rule, depending on
480 the saved state from before the beginning of parsing the data
481 that matched this rule. */
482 muscle_insert_state_number_table ("defgoto", yydefgoto
,
483 yydefgoto
[0], 1, nsyms
- ntokens
);
487 muscle_insert_base_table ("pact", base
,
488 base
[0], 1, nstates
);
489 MUSCLE_INSERT_INT ("pact_ninf", base_ninf
);
492 muscle_insert_base_table ("pgoto", base
,
493 base
[nstates
], nstates
+ 1, nvectors
);
495 muscle_insert_base_table ("table", table
,
496 table
[0], 1, high
+ 1);
497 MUSCLE_INSERT_INT ("table_ninf", table_ninf
);
499 muscle_insert_base_table ("check", check
,
500 check
[0], 1, high
+ 1);
502 /* GLR parsing slightly modifies YYTABLE and YYCHECK (and thus
503 YYPACT) so that in states with unresolved conflicts, the default
504 reduction is not used in the conflicted entries, so that there is
505 a place to put a conflict pointer.
507 This means that YYCONFLP and YYCONFL are nonsense for a non-GLR
508 parser, so we could avoid accidents by not writing them out in
509 that case. Nevertheless, it seems even better to be able to use
510 the GLR skeletons even without the non-deterministic tables. */
511 muscle_insert_unsigned_int_table ("conflict_list_heads", conflict_table
,
512 conflict_table
[0], 1, high
+ 1);
513 muscle_insert_unsigned_int_table ("conflicting_rules", conflict_list
,
514 0, 1, conflict_list_cnt
);
518 /*--------------------------------------------.
519 | Output the definitions of all the muscles. |
520 `--------------------------------------------*/
523 muscles_output (FILE *out
)
525 fputs ("m4_init()\n", out
);
528 symbol_code_props_output (out
, "destructors", &symbol_destructor_get
);
529 symbol_code_props_output (out
, "printers", &symbol_printer_get
);
530 symbol_definitions_output (out
);
531 symbol_numbers_output (out
);
532 token_definitions_output (out
);
533 user_actions_output (out
);
535 muscles_m4_output (out
);
538 /*---------------------------.
539 | Call the skeleton parser. |
540 `---------------------------*/
543 output_skeleton (void)
550 /* Compute the names of the package data dir and skeleton files. */
551 char const m4sugar
[] = "m4sugar/m4sugar.m4";
552 char const m4bison
[] = "bison.m4";
557 char const *m4
= (p
= getenv ("M4")) ? p
: M4
;
558 char const *pkgdatadir
= compute_pkgdatadir ();
559 size_t skeleton_size
= strlen (skeleton
) + 1;
560 size_t pkgdatadirlen
= strlen (pkgdatadir
);
561 while (pkgdatadirlen
&& pkgdatadir
[pkgdatadirlen
- 1] == '/')
563 full_skeleton
= xmalloc (pkgdatadirlen
+ 1
564 + (skeleton_size
< sizeof m4sugar
565 ? sizeof m4sugar
: skeleton_size
));
566 strncpy (full_skeleton
, pkgdatadir
, pkgdatadirlen
);
567 full_skeleton
[pkgdatadirlen
] = '/';
568 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, m4sugar
);
569 full_m4sugar
= xstrdup (full_skeleton
);
570 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, m4bison
);
571 full_m4bison
= xstrdup (full_skeleton
);
572 if (strchr (skeleton
, '/'))
573 strcpy (full_skeleton
, skeleton
);
575 strcpy (full_skeleton
+ pkgdatadirlen
+ 1, skeleton
);
577 /* Test whether m4sugar.m4 is readable, to check for proper
578 installation. A faulty installation can cause deadlock, so a
579 cheap sanity check is worthwhile. */
580 xfclose (xfopen (full_m4sugar
, "r"));
582 /* Create an m4 subprocess connected to us via two pipes. */
584 if (trace_flag
& trace_tools
)
585 fprintf (stderr
, "running: %s %s - %s %s\n",
586 m4
, full_m4sugar
, full_m4bison
, full_skeleton
);
588 /* Some future version of GNU M4 (most likely 1.6) may treat the -dV in a
589 position-dependent manner. Keep it as the first argument so that all
592 See the thread starting at
593 <http://lists.gnu.org/archive/html/bug-bison/2008-07/msg00000.html>
599 argv
[i
++] = pkgdatadir
;
600 if (trace_flag
& trace_m4
)
602 argv
[i
++] = full_m4sugar
;
604 argv
[i
++] = full_m4bison
;
605 argv
[i
++] = full_skeleton
;
608 /* When POSIXLY_CORRECT is set, some future versions of GNU M4 (most likely
609 2.0) may drop some of the GNU extensions that Bison's skeletons depend
610 upon. So that the next release of Bison is forward compatible with those
611 future versions of GNU M4, we unset POSIXLY_CORRECT here.
613 FIXME: A user might set POSIXLY_CORRECT to affect processes run from
614 macros like m4_syscmd in a custom skeleton. For now, Bison makes no
615 promises about the behavior of custom skeletons, so this scenario is not a
616 concern. However, we eventually want to eliminate this shortcoming. The
617 next release of GNU M4 (1.4.12 or 1.6) will accept the -g command-line
618 option as a no-op, and later releases will accept it to indicate that
619 POSIXLY_CORRECT should be ignored. Once the GNU M4 versions that accept
620 -g are pervasive, Bison should use -g instead of unsetting
623 See the thread starting at
624 <http://lists.gnu.org/archive/html/bug-bison/2008-07/msg00000.html>
626 unsetenv ("POSIXLY_CORRECT");
628 pid
= create_subpipe (argv
, filter_fd
);
631 free (full_skeleton
);
634 if (trace_flag
& trace_muscles
)
635 muscles_output (stderr
);
637 FILE *out
= fdopen (filter_fd
[0], "w");
639 error (EXIT_FAILURE
, get_errno (),
641 muscles_output (out
);
645 /* Read and process m4's output. */
646 timevar_push (TV_M4
);
647 end_of_output_subpipe (pid
, filter_fd
);
648 in
= fdopen (filter_fd
[1], "r");
650 error (EXIT_FAILURE
, get_errno (),
654 reap_subpipe (pid
, m4
);
661 /* BISON_USE_PUSH_FOR_PULL is for the test suite and should not be documented
663 char const *use_push_for_pull_env
= getenv ("BISON_USE_PUSH_FOR_PULL");
664 bool use_push_for_pull_flag
= false;
665 if (use_push_for_pull_env
!= NULL
666 && use_push_for_pull_env
[0] != '\0'
667 && 0 != strcmp (use_push_for_pull_env
, "0"))
668 use_push_for_pull_flag
= true;
671 MUSCLE_INSERT_BOOL ("debug_flag", debug_flag
);
672 MUSCLE_INSERT_BOOL ("defines_flag", defines_flag
);
673 MUSCLE_INSERT_BOOL ("error_verbose_flag", error_verbose
);
674 MUSCLE_INSERT_BOOL ("glr_flag", glr_parser
);
675 MUSCLE_INSERT_BOOL ("locations_flag", locations_flag
);
676 MUSCLE_INSERT_BOOL ("nondeterministic_flag", nondeterministic_parser
);
677 MUSCLE_INSERT_BOOL ("synclines_flag", !no_lines_flag
);
678 MUSCLE_INSERT_BOOL ("tag_seen_flag", tag_seen
);
679 MUSCLE_INSERT_BOOL ("use_push_for_pull_flag", use_push_for_pull_flag
);
680 MUSCLE_INSERT_BOOL ("yacc_flag", yacc_flag
);
683 if (spec_name_prefix
)
684 MUSCLE_INSERT_STRING ("prefix", spec_name_prefix
);
686 MUSCLE_INSERT_STRING ("file_name_all_but_ext", all_but_ext
);
688 #define DEFINE(Name) MUSCLE_INSERT_STRING (#Name, Name ? Name : "")
690 DEFINE (parser_file_name
);
691 DEFINE (spec_defines_file
);
692 DEFINE (spec_file_prefix
);
693 DEFINE (spec_graph_file
);
694 DEFINE (spec_name_prefix
);
695 DEFINE (spec_outfile
);
696 DEFINE (spec_verbose_file
);
699 /* Find the right skeleton file, and add muscles about the skeletons. */
701 MUSCLE_INSERT_C_STRING ("skeleton", skeleton
);
703 skeleton
= language
->skeleton
;
705 /* About the skeletons. */
707 /* b4_pkgdatadir is used inside m4_include in the skeletons, so digraphs
708 would never be expanded. Hopefully no one has M4-special characters in
709 his Bison installation path. */
710 MUSCLE_INSERT_STRING_RAW ("pkgdatadir", compute_pkgdatadir ());
715 /*----------------------------------------------------------.
716 | Output the parsing tables and the parser code to ftable. |
717 `----------------------------------------------------------*/
722 obstack_init (&format_obstack
);
731 /* Process the selected skeleton file. */
734 obstack_free (&format_obstack
, NULL
);
738 compute_pkgdatadir (void)
740 char const *pkgdatadir
= getenv ("BISON_PKGDATADIR");
741 return pkgdatadir
? pkgdatadir
: PKGDATADIR
;