]> git.saurik.com Git - bison.git/blame_incremental - src/scan-gram.l
djgpp/config.site: Adjust config.sed for the use of autoconf 2.60.
[bison.git] / src / scan-gram.l
... / ...
CommitLineData
1/* Bison Grammar Scanner -*- C -*-
2
3 Copyright (C) 2002, 2003, 2004, 2005, 2006 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 2 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, write to the Free Software
19 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
20 02110-1301 USA
21*/
22
23%option debug nodefault nounput noyywrap never-interactive
24%option prefix="gram_" outfile="lex.yy.c"
25
26%{
27/* Work around a bug in flex 2.5.31. See Debian bug 333231
28 <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=333231>. */
29#undef gram_wrap
30#define gram_wrap() 1
31
32#define FLEX_PREFIX(Id) gram_ ## Id
33#include "flex-scanner.h"
34
35#include "complain.h"
36#include "files.h"
37#include "gram.h"
38#include "quotearg.h"
39#include "reader.h"
40#include "uniqstr.h"
41
42#include <mbswidth.h>
43#include <quote.h>
44
45#include "scan-gram.h"
46
47#define YY_DECL GRAM_LEX_DECL
48
49#define YY_USER_INIT \
50 code_start = scanner_cursor = loc->start; \
51
52/* Location of scanner cursor. */
53static boundary scanner_cursor;
54
55#define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng);
56
57static size_t no_cr_read (FILE *, char *, size_t);
58#define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
59
60/* A string representing the most recently saved token. */
61char *last_string;
62
63void
64gram_scanner_last_string_free (void)
65{
66 STRING_FREE;
67}
68
69static void handle_syncline (char *, location);
70static unsigned long int scan_integer (char const *p, int base, location loc);
71static int convert_ucn_to_byte (char const *hex_text);
72static void unexpected_eof (boundary, char const *);
73static void unexpected_newline (boundary, char const *);
74
75%}
76 /* A C-like comment in directives/rules. */
77%x SC_YACC_COMMENT
78 /* Strings and characters in directives/rules. */
79%x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER
80 /* A identifier was just read in directives/rules. Special state
81 to capture the sequence `identifier :'. */
82%x SC_AFTER_IDENTIFIER
83
84 /* Three types of user code:
85 - prologue (code between `%{' `%}' in the first section, before %%);
86 - actions, printers, union, etc, (between braced in the middle section);
87 - epilogue (everything after the second %%). */
88%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE
89 /* C and C++ comments in code. */
90%x SC_COMMENT SC_LINE_COMMENT
91 /* Strings and characters in code. */
92%x SC_STRING SC_CHARACTER
93
94letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
95id {letter}({letter}|[0-9])*
96directive %{letter}({letter}|[0-9]|-)*
97int [0-9]+
98
99/* POSIX says that a tag must be both an id and a C union member, but
100 historically almost any character is allowed in a tag. We disallow
101 NUL and newline, as this simplifies our implementation. */
102tag [^\0\n>]+
103
104/* Zero or more instances of backslash-newline. Following GCC, allow
105 white space between the backslash and the newline. */
106splice (\\[ \f\t\v]*\n)*
107
108%%
109%{
110 /* Nesting level of the current code in braces. */
111 int braces_level IF_LINT (= 0);
112
113 /* Parent context state, when applicable. */
114 int context_state IF_LINT (= 0);
115
116 /* Location of most recent identifier, when applicable. */
117 location id_loc IF_LINT (= empty_location);
118
119 /* Where containing code started, when applicable. Its initial
120 value is relevant only when yylex is invoked in the SC_EPILOGUE
121 start condition. */
122 boundary code_start = scanner_cursor;
123
124 /* Where containing comment or string or character literal started,
125 when applicable. */
126 boundary token_start IF_LINT (= scanner_cursor);
127%}
128
129
130 /*-----------------------.
131 | Scanning white space. |
132 `-----------------------*/
133
134<INITIAL,SC_AFTER_IDENTIFIER>
135{
136 /* Comments and white space. */
137 "," warn_at (*loc, _("stray `,' treated as white space"));
138 [ \f\n\t\v] |
139 "//".* ;
140 "/*" {
141 token_start = loc->start;
142 context_state = YY_START;
143 BEGIN SC_YACC_COMMENT;
144 }
145
146 /* #line directives are not documented, and may be withdrawn or
147 modified in future versions of Bison. */
148 ^"#line "{int}" \"".*"\"\n" {
149 handle_syncline (yytext + sizeof "#line " - 1, *loc);
150 }
151}
152
153
154 /*----------------------------.
155 | Scanning Bison directives. |
156 `----------------------------*/
157<INITIAL>
158{
159 "%binary" return PERCENT_NONASSOC;
160 "%code" return PERCENT_CODE;
161 "%debug" return PERCENT_DEBUG;
162 "%default"[-_]"prec" return PERCENT_DEFAULT_PREC;
163 "%define" return PERCENT_DEFINE;
164 "%defines" return PERCENT_DEFINES;
165 "%destructor" return PERCENT_DESTRUCTOR;
166 "%dprec" return PERCENT_DPREC;
167 "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE;
168 "%expect" return PERCENT_EXPECT;
169 "%expect"[-_]"rr" return PERCENT_EXPECT_RR;
170 "%file-prefix" return PERCENT_FILE_PREFIX;
171 "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC;
172 "%initial-action" return PERCENT_INITIAL_ACTION;
173 "%glr-parser" return PERCENT_GLR_PARSER;
174 "%left" return PERCENT_LEFT;
175 "%lex-param" return PERCENT_LEX_PARAM;
176 "%locations" return PERCENT_LOCATIONS;
177 "%merge" return PERCENT_MERGE;
178 "%name"[-_]"prefix" return PERCENT_NAME_PREFIX;
179 "%no"[-_]"default"[-_]"prec" return PERCENT_NO_DEFAULT_PREC;
180 "%no"[-_]"lines" return PERCENT_NO_LINES;
181 "%nonassoc" return PERCENT_NONASSOC;
182 "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER;
183 "%nterm" return PERCENT_NTERM;
184 "%output" return PERCENT_OUTPUT;
185 "%parse-param" return PERCENT_PARSE_PARAM;
186 "%prec" return PERCENT_PREC;
187 "%printer" return PERCENT_PRINTER;
188 "%provides" return PERCENT_PROVIDES;
189 "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
190 "%push"[-_]"parser" return PERCENT_PUSH_PARSER;
191 "%require" return PERCENT_REQUIRE;
192 "%requires" return PERCENT_REQUIRES;
193 "%right" return PERCENT_RIGHT;
194 "%skeleton" return PERCENT_SKELETON;
195 "%start" return PERCENT_START;
196 "%symbol-default" return PERCENT_SYMBOL_DEFAULT;
197 "%term" return PERCENT_TOKEN;
198 "%token" return PERCENT_TOKEN;
199 "%token"[-_]"table" return PERCENT_TOKEN_TABLE;
200 "%type" return PERCENT_TYPE;
201 "%union" return PERCENT_UNION;
202 "%verbose" return PERCENT_VERBOSE;
203 "%yacc" return PERCENT_YACC;
204
205 {directive} {
206 complain_at (*loc, _("invalid directive: %s"), quote (yytext));
207 }
208
209 "=" return EQUAL;
210 "|" return PIPE;
211 ";" return SEMICOLON;
212
213 {id} {
214 val->uniqstr = uniqstr_new (yytext);
215 id_loc = *loc;
216 BEGIN SC_AFTER_IDENTIFIER;
217 }
218
219 {int} {
220 val->integer = scan_integer (yytext, 10, *loc);
221 return INT;
222 }
223 0[xX][0-9abcdefABCDEF]+ {
224 val->integer = scan_integer (yytext, 16, *loc);
225 return INT;
226 }
227
228 /* Characters. We don't check there is only one. */
229 "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER;
230
231 /* Strings. */
232 "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING;
233
234 /* Prologue. */
235 "%{" code_start = loc->start; BEGIN SC_PROLOGUE;
236
237 /* Code in between braces. */
238 "{" {
239 STRING_GROW;
240 braces_level = 0;
241 code_start = loc->start;
242 BEGIN SC_BRACED_CODE;
243 }
244
245 /* A type. */
246 "<"{tag}">" {
247 obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2);
248 STRING_FINISH;
249 val->uniqstr = uniqstr_new (last_string);
250 STRING_FREE;
251 return TYPE;
252 }
253
254 "%%" {
255 static int percent_percent_count;
256 if (++percent_percent_count == 2)
257 BEGIN SC_EPILOGUE;
258 return PERCENT_PERCENT;
259 }
260
261 . {
262 complain_at (*loc, _("invalid character: %s"), quote (yytext));
263 }
264
265 <<EOF>> {
266 loc->start = loc->end = scanner_cursor;
267 yyterminate ();
268 }
269}
270
271
272 /*-----------------------------------------------------------------.
273 | Scanning after an identifier, checking whether a colon is next. |
274 `-----------------------------------------------------------------*/
275
276<SC_AFTER_IDENTIFIER>
277{
278 ":" {
279 *loc = id_loc;
280 BEGIN INITIAL;
281 return ID_COLON;
282 }
283 . {
284 scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
285 yyless (0);
286 *loc = id_loc;
287 BEGIN INITIAL;
288 return ID;
289 }
290 <<EOF>> {
291 *loc = id_loc;
292 BEGIN INITIAL;
293 return ID;
294 }
295}
296
297
298 /*---------------------------------------------------------------.
299 | Scanning a Yacc comment. The initial `/ *' is already eaten. |
300 `---------------------------------------------------------------*/
301
302<SC_YACC_COMMENT>
303{
304 "*/" BEGIN context_state;
305 .|\n ;
306 <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
307}
308
309
310 /*------------------------------------------------------------.
311 | Scanning a C comment. The initial `/ *' is already eaten. |
312 `------------------------------------------------------------*/
313
314<SC_COMMENT>
315{
316 "*"{splice}"/" STRING_GROW; BEGIN context_state;
317 <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
318}
319
320
321 /*--------------------------------------------------------------.
322 | Scanning a line comment. The initial `//' is already eaten. |
323 `--------------------------------------------------------------*/
324
325<SC_LINE_COMMENT>
326{
327 "\n" STRING_GROW; BEGIN context_state;
328 {splice} STRING_GROW;
329 <<EOF>> BEGIN context_state;
330}
331
332
333 /*------------------------------------------------.
334 | Scanning a Bison string, including its escapes. |
335 | The initial quote is already eaten. |
336 `------------------------------------------------*/
337
338<SC_ESCAPED_STRING>
339{
340 "\""|"\n" {
341 if (yytext[0] == '\n')
342 unexpected_newline (token_start, "\"");
343 STRING_FINISH;
344 loc->start = token_start;
345 val->chars = last_string;
346 BEGIN INITIAL;
347 return STRING;
348 }
349 <<EOF>> {
350 unexpected_eof (token_start, "\"");
351 STRING_FINISH;
352 loc->start = token_start;
353 val->chars = last_string;
354 BEGIN INITIAL;
355 return STRING;
356 }
357}
358
359 /*----------------------------------------------------------.
360 | Scanning a Bison character literal, decoding its escapes. |
361 | The initial quote is already eaten. |
362 `----------------------------------------------------------*/
363
364<SC_ESCAPED_CHARACTER>
365{
366 "'"|"\n" {
367 if (yytext[0] == '\n')
368 unexpected_newline (token_start, "'");
369 STRING_GROW;
370 STRING_FINISH;
371 loc->start = token_start;
372 val->character = last_string[1];
373 STRING_FREE;
374 BEGIN INITIAL;
375 return CHAR;
376 }
377 <<EOF>> {
378 unexpected_eof (token_start, "'");
379 STRING_FINISH;
380 loc->start = token_start;
381 if (strlen(last_string) > 1)
382 val->character = last_string[1];
383 else
384 val->character = last_string[0];
385 STRING_FREE;
386 BEGIN INITIAL;
387 return CHAR;
388 }
389}
390
391<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING>
392{
393 \0 complain_at (*loc, _("invalid null character"));
394}
395
396
397 /*----------------------------.
398 | Decode escaped characters. |
399 `----------------------------*/
400
401<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
402{
403 \\[0-7]{1,3} {
404 unsigned long int c = strtoul (yytext + 1, NULL, 8);
405 if (UCHAR_MAX < c)
406 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
407 else if (! c)
408 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
409 else
410 obstack_1grow (&obstack_for_string, c);
411 }
412
413 \\x[0-9abcdefABCDEF]+ {
414 verify (UCHAR_MAX < ULONG_MAX);
415 unsigned long int c = strtoul (yytext + 2, NULL, 16);
416 if (UCHAR_MAX < c)
417 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
418 else if (! c)
419 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
420 else
421 obstack_1grow (&obstack_for_string, c);
422 }
423
424 \\a obstack_1grow (&obstack_for_string, '\a');
425 \\b obstack_1grow (&obstack_for_string, '\b');
426 \\f obstack_1grow (&obstack_for_string, '\f');
427 \\n obstack_1grow (&obstack_for_string, '\n');
428 \\r obstack_1grow (&obstack_for_string, '\r');
429 \\t obstack_1grow (&obstack_for_string, '\t');
430 \\v obstack_1grow (&obstack_for_string, '\v');
431
432 /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */
433 \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]);
434
435 \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
436 int c = convert_ucn_to_byte (yytext);
437 if (c < 0)
438 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
439 else if (! c)
440 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
441 else
442 obstack_1grow (&obstack_for_string, c);
443 }
444 \\(.|\n) {
445 complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext));
446 STRING_GROW;
447 }
448}
449
450 /*--------------------------------------------.
451 | Scanning user-code characters and strings. |
452 `--------------------------------------------*/
453
454<SC_CHARACTER,SC_STRING>
455{
456 {splice}|\\{splice}[^\n\[\]] STRING_GROW;
457}
458
459<SC_CHARACTER>
460{
461 "'" STRING_GROW; BEGIN context_state;
462 \n unexpected_newline (token_start, "'"); BEGIN context_state;
463 <<EOF>> unexpected_eof (token_start, "'"); BEGIN context_state;
464}
465
466<SC_STRING>
467{
468 "\"" STRING_GROW; BEGIN context_state;
469 \n unexpected_newline (token_start, "\""); BEGIN context_state;
470 <<EOF>> unexpected_eof (token_start, "\""); BEGIN context_state;
471}
472
473
474 /*---------------------------------------------------.
475 | Strings, comments etc. can be found in user code. |
476 `---------------------------------------------------*/
477
478<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
479{
480 "'" {
481 STRING_GROW;
482 context_state = YY_START;
483 token_start = loc->start;
484 BEGIN SC_CHARACTER;
485 }
486 "\"" {
487 STRING_GROW;
488 context_state = YY_START;
489 token_start = loc->start;
490 BEGIN SC_STRING;
491 }
492 "/"{splice}"*" {
493 STRING_GROW;
494 context_state = YY_START;
495 token_start = loc->start;
496 BEGIN SC_COMMENT;
497 }
498 "/"{splice}"/" {
499 STRING_GROW;
500 context_state = YY_START;
501 BEGIN SC_LINE_COMMENT;
502 }
503}
504
505
506
507 /*-----------------------------------------------------------.
508 | Scanning some code in braces (actions). The initial "{" is |
509 | already eaten. |
510 `-----------------------------------------------------------*/
511
512<SC_BRACED_CODE>
513{
514 "{"|"<"{splice}"%" STRING_GROW; braces_level++;
515 "%"{splice}">" STRING_GROW; braces_level--;
516 "}" {
517 obstack_1grow (&obstack_for_string, '}');
518
519 --braces_level;
520 if (braces_level < 0)
521 {
522 STRING_FINISH;
523 loc->start = code_start;
524 val->chars = last_string;
525 BEGIN INITIAL;
526 return BRACED_CODE;
527 }
528 }
529
530 /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
531 (as `<' `<%'). */
532 "<"{splice}"<" STRING_GROW;
533
534 <<EOF>> {
535 unexpected_eof (code_start, "}");
536 STRING_FINISH;
537 loc->start = code_start;
538 val->chars = last_string;
539 BEGIN INITIAL;
540 return BRACED_CODE;
541 }
542}
543
544
545 /*--------------------------------------------------------------.
546 | Scanning some prologue: from "%{" (already scanned) to "%}". |
547 `--------------------------------------------------------------*/
548
549<SC_PROLOGUE>
550{
551 "%}" {
552 STRING_FINISH;
553 loc->start = code_start;
554 val->chars = last_string;
555 BEGIN INITIAL;
556 return PROLOGUE;
557 }
558
559 <<EOF>> {
560 unexpected_eof (code_start, "%}");
561 STRING_FINISH;
562 loc->start = code_start;
563 val->chars = last_string;
564 BEGIN INITIAL;
565 return PROLOGUE;
566 }
567}
568
569
570 /*---------------------------------------------------------------.
571 | Scanning the epilogue (everything after the second "%%", which |
572 | has already been eaten). |
573 `---------------------------------------------------------------*/
574
575<SC_EPILOGUE>
576{
577 <<EOF>> {
578 STRING_FINISH;
579 loc->start = code_start;
580 val->chars = last_string;
581 BEGIN INITIAL;
582 return EPILOGUE;
583 }
584}
585
586
587 /*-----------------------------------------------------.
588 | By default, grow the string obstack with the input. |
589 `-----------------------------------------------------*/
590
591<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>. |
592<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n STRING_GROW;
593
594%%
595
596/* Read bytes from FP into buffer BUF of size SIZE. Return the
597 number of bytes read. Remove '\r' from input, treating \r\n
598 and isolated \r as \n. */
599
600static size_t
601no_cr_read (FILE *fp, char *buf, size_t size)
602{
603 size_t bytes_read = fread (buf, 1, size, fp);
604 if (bytes_read)
605 {
606 char *w = memchr (buf, '\r', bytes_read);
607 if (w)
608 {
609 char const *r = ++w;
610 char const *lim = buf + bytes_read;
611
612 for (;;)
613 {
614 /* Found an '\r'. Treat it like '\n', but ignore any
615 '\n' that immediately follows. */
616 w[-1] = '\n';
617 if (r == lim)
618 {
619 int ch = getc (fp);
620 if (ch != '\n' && ungetc (ch, fp) != ch)
621 break;
622 }
623 else if (*r == '\n')
624 r++;
625
626 /* Copy until the next '\r'. */
627 do
628 {
629 if (r == lim)
630 return w - buf;
631 }
632 while ((*w++ = *r++) != '\r');
633 }
634
635 return w - buf;
636 }
637 }
638
639 return bytes_read;
640}
641
642
643
644/*------------------------------------------------------.
645| Scan NUMBER for a base-BASE integer at location LOC. |
646`------------------------------------------------------*/
647
648static unsigned long int
649scan_integer (char const *number, int base, location loc)
650{
651 verify (INT_MAX < ULONG_MAX);
652 unsigned long int num = strtoul (number, NULL, base);
653
654 if (INT_MAX < num)
655 {
656 complain_at (loc, _("integer out of range: %s"), quote (number));
657 num = INT_MAX;
658 }
659
660 return num;
661}
662
663
664/*------------------------------------------------------------------.
665| Convert universal character name UCN to a single-byte character, |
666| and return that character. Return -1 if UCN does not correspond |
667| to a single-byte character. |
668`------------------------------------------------------------------*/
669
670static int
671convert_ucn_to_byte (char const *ucn)
672{
673 verify (UCHAR_MAX <= INT_MAX);
674 unsigned long int code = strtoul (ucn + 2, NULL, 16);
675
676 /* FIXME: Currently we assume Unicode-compatible unibyte characters
677 on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On
678 non-ASCII hosts we support only the portable C character set.
679 These limitations should be removed once we add support for
680 multibyte characters. */
681
682 if (UCHAR_MAX < code)
683 return -1;
684
685#if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e)
686 {
687 /* A non-ASCII host. Use CODE to index into a table of the C
688 basic execution character set, which is guaranteed to exist on
689 all Standard C platforms. This table also includes '$', '@',
690 and '`', which are not in the basic execution character set but
691 which are unibyte characters on all the platforms that we know
692 about. */
693 static signed char const table[] =
694 {
695 '\0', -1, -1, -1, -1, -1, -1, '\a',
696 '\b', '\t', '\n', '\v', '\f', '\r', -1, -1,
697 -1, -1, -1, -1, -1, -1, -1, -1,
698 -1, -1, -1, -1, -1, -1, -1, -1,
699 ' ', '!', '"', '#', '$', '%', '&', '\'',
700 '(', ')', '*', '+', ',', '-', '.', '/',
701 '0', '1', '2', '3', '4', '5', '6', '7',
702 '8', '9', ':', ';', '<', '=', '>', '?',
703 '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
704 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
705 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
706 'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
707 '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
708 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
709 'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
710 'x', 'y', 'z', '{', '|', '}', '~'
711 };
712
713 code = code < sizeof table ? table[code] : -1;
714 }
715#endif
716
717 return code;
718}
719
720
721/*----------------------------------------------------------------.
722| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. |
723`----------------------------------------------------------------*/
724
725static void
726handle_syncline (char *args, location loc)
727{
728 char *after_num;
729 unsigned long int lineno = strtoul (args, &after_num, 10);
730 char *file = strchr (after_num, '"') + 1;
731 *strchr (file, '"') = '\0';
732 if (INT_MAX <= lineno)
733 {
734 warn_at (loc, _("line number overflow"));
735 lineno = INT_MAX;
736 }
737 current_file = uniqstr_new (file);
738 boundary_set (&scanner_cursor, current_file, lineno, 1);
739}
740
741
742/*----------------------------------------------------------------.
743| For a token or comment starting at START, report message MSGID, |
744| which should say that an end marker was found before |
745| the expected TOKEN_END. |
746`----------------------------------------------------------------*/
747
748static void
749unexpected_end (boundary start, char const *msgid, char const *token_end)
750{
751 location loc;
752 loc.start = start;
753 loc.end = scanner_cursor;
754 complain_at (loc, _(msgid), token_end);
755}
756
757
758/*------------------------------------------------------------------------.
759| Report an unexpected EOF in a token or comment starting at START. |
760| An end of file was encountered and the expected TOKEN_END was missing. |
761`------------------------------------------------------------------------*/
762
763static void
764unexpected_eof (boundary start, char const *token_end)
765{
766 unexpected_end (start, N_("missing `%s' at end of file"), token_end);
767}
768
769
770/*----------------------------------------.
771| Likewise, but for unexpected newlines. |
772`----------------------------------------*/
773
774static void
775unexpected_newline (boundary start, char const *token_end)
776{
777 unexpected_end (start, N_("missing `%s' at end of line"), token_end);
778}
779
780
781/*-------------------------.
782| Initialize the scanner. |
783`-------------------------*/
784
785void
786gram_scanner_initialize (void)
787{
788 obstack_init (&obstack_for_string);
789}
790
791
792/*-----------------------------------------------.
793| Free all the memory allocated to the scanner. |
794`-----------------------------------------------*/
795
796void
797gram_scanner_free (void)
798{
799 obstack_free (&obstack_for_string, 0);
800 /* Reclaim Flex's buffers. */
801 yy_delete_buffer (YY_CURRENT_BUFFER);
802}