]> git.saurik.com Git - bison.git/blob - src/scan-gram.l
[a-f] -> [abcdef], so that we don't assume the C locale.
[bison.git] / src / scan-gram.l
1 /* Bison Grammar Scanner -*- C -*-
2 Copyright (C) 2002 Free Software Foundation, Inc.
3
4 This file is part of Bison, the GNU Compiler Compiler.
5
6 This program is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2 of the License, or
9 (at your option) any later version.
10
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
15
16 You should have received a copy of the GNU General Public License
17 along with this program; if not, write to the Free Software
18 Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
19 02111-1307 USA
20 */
21
22 %option debug nodefault noyywrap never-interactive
23 %option prefix="gram_" outfile="lex.yy.c"
24
25 %{
26 #include "system.h"
27 #include "mbswidth.h"
28 #include "complain.h"
29 #include "quote.h"
30 #include "struniq.h"
31 #include "getargs.h"
32 #include "gram.h"
33 #include "reader.h"
34
35 #define YY_USER_INIT \
36 do { \
37 LOCATION_RESET (*loc); \
38 loc->file = current_file; \
39 } while (0)
40
41 /* Each time we match a string, move the end cursor to its end. */
42 #define STEP LOCATION_STEP (*loc)
43
44 #define YY_USER_ACTION extend_location (loc, yytext, yyleng);
45
46 #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
47
48
49 /* Read bytes from FP into buffer BUF of size SIZE. Return the
50 number of bytes read. Remove '\r' from input, treating \r\n
51 and isolated \r as \n. */
52
53 static size_t
54 no_cr_read (FILE *fp, char *buf, size_t size)
55 {
56 size_t s = fread (buf, 1, size, fp);
57 if (s)
58 {
59 char *w = memchr (buf, '\r', s);
60 if (w)
61 {
62 char const *r = ++w;
63 char const *lim = buf + s;
64
65 for (;;)
66 {
67 /* Found an '\r'. Treat it like '\n', but ignore any
68 '\n' that immediately follows. */
69 w[-1] = '\n';
70 if (r == lim)
71 {
72 int ch = getc (fp);
73 if (ch != '\n' && ungetc (ch, fp) != ch)
74 break;
75 }
76 else if (*r == '\n')
77 r++;
78
79 /* Copy until the next '\r'. */
80 do
81 {
82 if (r == lim)
83 return w - buf;
84 }
85 while ((*w++ = *r++) != '\r');
86 }
87
88 return w - buf;
89 }
90 }
91
92 return s;
93 }
94
95
96 /* Extend *LOC to account for token TOKEN of size SIZE. */
97
98 static void
99 extend_location (location_t *loc, char const *token, int size)
100 {
101 int line = loc->last_line;
102 int column = loc->last_column;
103 char const *p0 = token;
104 char const *p = token;
105 char const *lim = token + size;
106
107 for (p = token; p < lim; p++)
108 switch (*p)
109 {
110 case '\r':
111 /* \r shouldn't survive no_cr_read. */
112 abort ();
113
114 case '\n':
115 line++;
116 column = 1;
117 p0 = p + 1;
118 break;
119
120 case '\t':
121 column += mbsnwidth (p0, p - p0, 0);
122 column += 8 - ((column - 1) & 7);
123 p0 = p + 1;
124 break;
125 }
126
127 loc->last_line = line;
128 loc->last_column = column + mbsnwidth (p0, p - p0, 0);
129 }
130
131
132
133 /* STRING_OBSTACK -- Used to store all the characters that we need to
134 keep (to construct ID, STRINGS etc.). Use the following macros to
135 use it.
136
137 Use STRING_GROW to append what has just been matched, and
138 STRING_FINISH to end the string (it puts the ending 0).
139 STRING_FINISH also stores this string in LAST_STRING, which can be
140 used, and which is used by STRING_FREE to free the last string. */
141
142 static struct obstack string_obstack;
143
144 /* A string representing the most recently saved token. */
145 static char *last_string;
146
147
148 #define STRING_GROW \
149 obstack_grow (&string_obstack, yytext, yyleng)
150
151 #define STRING_FINISH \
152 do { \
153 obstack_1grow (&string_obstack, '\0'); \
154 last_string = obstack_finish (&string_obstack); \
155 } while (0)
156
157 #define STRING_FREE \
158 obstack_free (&string_obstack, last_string)
159
160 void
161 scanner_last_string_free (void)
162 {
163 STRING_FREE;
164 }
165
166 /* Within well-formed rules, RULE_LENGTH is the number of values in
167 the current rule so far, which says where to find `$0' with respect
168 to the top of the stack. It is not the same as the rule->length in
169 the case of mid rule actions.
170
171 Outside of well-formed rules, RULE_LENGTH has an undefined value. */
172 static int rule_length;
173
174 static void handle_dollar (braced_code_t code_kind,
175 char *cp, location_t location);
176 static void handle_at (braced_code_t code_kind,
177 char *cp, location_t location);
178 static void handle_syncline (char *args, location_t *location);
179 static int convert_ucn_to_byte (char const *hex_text);
180 static void unexpected_end_of_file (location_t *, char const *);
181
182 %}
183 %x SC_COMMENT SC_LINE_COMMENT SC_YACC_COMMENT
184 %x SC_STRING SC_CHARACTER
185 %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER
186 %x SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE
187
188 letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
189 id {letter}({letter}|[0-9])*
190 directive %{letter}({letter}|[0-9]|-)*
191 int [0-9]+
192
193 /* POSIX says that a tag must be both an id and a C union member, but
194 historically almost any character is allowed in a tag. We disallow
195 NUL and newline, as this simplifies our implementation. */
196 tag [^\0\n>]+
197
198 /* Zero or more instances of backslash-newline. Following GCC, allow
199 white space between the backslash and the newline. */
200 splice (\\[ \f\t\v]*\n)*
201
202 %%
203 %{
204 /* Nesting level of the current code in braces. */
205 int braces_level IF_LINT (= 0);
206
207 /* Scanner context when scanning C code. */
208 int c_context IF_LINT (= 0);
209
210 /* At each yylex invocation, mark the current position as the
211 start of the next token. */
212 STEP;
213 %}
214
215
216 /*----------------------------.
217 | Scanning Bison directives. |
218 `----------------------------*/
219 <INITIAL>
220 {
221 "%binary" return PERCENT_NONASSOC;
222 "%debug" return PERCENT_DEBUG;
223 "%define" return PERCENT_DEFINE;
224 "%defines" return PERCENT_DEFINES;
225 "%destructor" return PERCENT_DESTRUCTOR;
226 "%dprec" return PERCENT_DPREC;
227 "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE;
228 "%expect" return PERCENT_EXPECT;
229 "%file-prefix" return PERCENT_FILE_PREFIX;
230 "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC;
231 "%glr-parser" return PERCENT_GLR_PARSER;
232 "%left" return PERCENT_LEFT;
233 "%locations" return PERCENT_LOCATIONS;
234 "%merge" return PERCENT_MERGE;
235 "%name"[-_]"prefix" return PERCENT_NAME_PREFIX;
236 "%no"[-_]"lines" return PERCENT_NO_LINES;
237 "%nonassoc" return PERCENT_NONASSOC;
238 "%nterm" return PERCENT_NTERM;
239 "%output" return PERCENT_OUTPUT;
240 "%parse-param" return PERCENT_PARSE_PARAM;
241 "%prec" rule_length--; return PERCENT_PREC;
242 "%printer" return PERCENT_PRINTER;
243 "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
244 "%right" return PERCENT_RIGHT;
245 "%lex-param" return PERCENT_LEX_PARAM;
246 "%skeleton" return PERCENT_SKELETON;
247 "%start" return PERCENT_START;
248 "%term" return PERCENT_TOKEN;
249 "%token" return PERCENT_TOKEN;
250 "%token"[-_]"table" return PERCENT_TOKEN_TABLE;
251 "%type" return PERCENT_TYPE;
252 "%union" return PERCENT_UNION;
253 "%verbose" return PERCENT_VERBOSE;
254 "%yacc" return PERCENT_YACC;
255
256 {directive} {
257 complain_at (*loc, _("invalid directive: %s"), quote (yytext));
258 STEP;
259 }
260
261 ^"#line "{int}" \"".*"\"\n" {
262 handle_syncline (yytext + sizeof "#line " - 1, loc);
263 STEP;
264 }
265
266 "=" return EQUAL;
267 ":" rule_length = 0; return COLON;
268 "|" rule_length = 0; return PIPE;
269 ";" return SEMICOLON;
270
271 [ \f\n\t\v] STEP;
272
273 "," {
274 warn_at (*loc, _("stray `,' treated as white space"));
275 STEP;
276 }
277
278 {id} {
279 val->symbol = symbol_get (yytext, *loc);
280 rule_length++;
281 return ID;
282 }
283
284 {int} {
285 unsigned long num;
286 errno = 0;
287 num = strtoul (yytext, 0, 10);
288 if (INT_MAX < num || errno)
289 {
290 complain_at (*loc, _("integer out of range: %s"), quote (yytext));
291 num = INT_MAX;
292 }
293 val->integer = num;
294 return INT;
295 }
296
297 /* Characters. We don't check there is only one. */
298 "'" STRING_GROW; BEGIN SC_ESCAPED_CHARACTER;
299
300 /* Strings. */
301 "\"" STRING_GROW; BEGIN SC_ESCAPED_STRING;
302
303 /* Comments. */
304 "/*" BEGIN SC_YACC_COMMENT;
305 "//".* STEP;
306
307 /* Prologue. */
308 "%{" BEGIN SC_PROLOGUE;
309
310 /* Code in between braces. */
311 "{" STRING_GROW; braces_level = 0; BEGIN SC_BRACED_CODE;
312
313 /* A type. */
314 "<"{tag}">" {
315 obstack_grow (&string_obstack, yytext + 1, yyleng - 2);
316 STRING_FINISH;
317 val->struniq = struniq_new (last_string);
318 STRING_FREE;
319 return TYPE;
320 }
321
322 "%%" {
323 static int percent_percent_count;
324 if (++percent_percent_count == 2)
325 BEGIN SC_EPILOGUE;
326 return PERCENT_PERCENT;
327 }
328
329 . {
330 complain_at (*loc, _("invalid character: %s"), quote (yytext));
331 STEP;
332 }
333 }
334
335
336 /*---------------------------------------------------------------.
337 | Scanning a Yacc comment. The initial `/ *' is already eaten. |
338 `---------------------------------------------------------------*/
339
340 <SC_YACC_COMMENT>
341 {
342 "*/" {
343 STEP;
344 BEGIN INITIAL;
345 }
346
347 .|\n ;
348 <<EOF>> unexpected_end_of_file (loc, "*/");
349 }
350
351
352 /*------------------------------------------------------------.
353 | Scanning a C comment. The initial `/ *' is already eaten. |
354 `------------------------------------------------------------*/
355
356 <SC_COMMENT>
357 {
358 "*"{splice}"/" STRING_GROW; BEGIN c_context;
359 <<EOF>> unexpected_end_of_file (loc, "*/");
360 }
361
362
363 /*--------------------------------------------------------------.
364 | Scanning a line comment. The initial `//' is already eaten. |
365 `--------------------------------------------------------------*/
366
367 <SC_LINE_COMMENT>
368 {
369 "\n" STRING_GROW; BEGIN c_context;
370 {splice} STRING_GROW;
371 <<EOF>> BEGIN c_context;
372 }
373
374
375 /*----------------------------------------------------------------.
376 | Scanning a C string, including its escapes. The initial `"' is |
377 | already eaten. |
378 `----------------------------------------------------------------*/
379
380 <SC_ESCAPED_STRING>
381 {
382 "\"" {
383 STRING_GROW;
384 STRING_FINISH;
385 val->string = last_string;
386 rule_length++;
387 BEGIN INITIAL;
388 return STRING;
389 }
390
391 .|\n STRING_GROW;
392 <<EOF>> unexpected_end_of_file (loc, "\"");
393 }
394
395 /*---------------------------------------------------------------.
396 | Scanning a C character, decoding its escapes. The initial "'" |
397 | is already eaten. |
398 `---------------------------------------------------------------*/
399
400 <SC_ESCAPED_CHARACTER>
401 {
402 "'" {
403 STRING_GROW;
404 STRING_FINISH;
405 val->symbol = symbol_get (last_string, *loc);
406 symbol_class_set (val->symbol, token_sym, *loc);
407 symbol_user_token_number_set (val->symbol,
408 (unsigned char) last_string[1], *loc);
409 STRING_FREE;
410 rule_length++;
411 BEGIN INITIAL;
412 return ID;
413 }
414
415 .|\n STRING_GROW;
416 <<EOF>> unexpected_end_of_file (loc, "'");
417 }
418
419
420 /*----------------------------.
421 | Decode escaped characters. |
422 `----------------------------*/
423
424 <SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
425 {
426 \\[0-7]{1,3} {
427 unsigned long c = strtoul (yytext + 1, 0, 8);
428 if (UCHAR_MAX < c)
429 {
430 complain_at (*loc, _("invalid escape sequence: %s"),
431 quote (yytext));
432 STEP;
433 }
434 else
435 obstack_1grow (&string_obstack, c);
436 }
437
438 \\x[0-9abcdefABCDEF]+ {
439 unsigned long c;
440 errno = 0;
441 c = strtoul (yytext + 2, 0, 16);
442 if (UCHAR_MAX < c || errno)
443 {
444 complain_at (*loc, _("invalid escape sequence: %s"),
445 quote (yytext));
446 STEP;
447 }
448 else
449 obstack_1grow (&string_obstack, c);
450 }
451
452 \\a obstack_1grow (&string_obstack, '\a');
453 \\b obstack_1grow (&string_obstack, '\b');
454 \\f obstack_1grow (&string_obstack, '\f');
455 \\n obstack_1grow (&string_obstack, '\n');
456 \\r obstack_1grow (&string_obstack, '\r');
457 \\t obstack_1grow (&string_obstack, '\t');
458 \\v obstack_1grow (&string_obstack, '\v');
459
460 /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */
461 \\("\""|"'"|"?"|"\\") obstack_1grow (&string_obstack, yytext[1]);
462
463 \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
464 int c = convert_ucn_to_byte (yytext);
465 if (c < 0)
466 {
467 complain_at (*loc, _("invalid escape sequence: %s"),
468 quote (yytext));
469 STEP;
470 }
471 else
472 obstack_1grow (&string_obstack, c);
473 }
474 \\(.|\n) {
475 complain_at (*loc, _("unrecognized escape sequence: %s"),
476 quote (yytext));
477 STRING_GROW;
478 }
479 }
480
481
482 /*----------------------------------------------------------.
483 | Scanning a C character without decoding its escapes. The |
484 | initial "'" is already eaten. |
485 `----------------------------------------------------------*/
486
487 <SC_CHARACTER>
488 {
489 "'" STRING_GROW; BEGIN c_context;
490 \\{splice}[^$@\[\]] STRING_GROW;
491 <<EOF>> unexpected_end_of_file (loc, "'");
492 }
493
494
495 /*----------------------------------------------------------------.
496 | Scanning a C string, without decoding its escapes. The initial |
497 | `"' is already eaten. |
498 `----------------------------------------------------------------*/
499
500 <SC_STRING>
501 {
502 "\"" STRING_GROW; BEGIN c_context;
503 \\{splice}[^$@\[\]] STRING_GROW;
504 <<EOF>> unexpected_end_of_file (loc, "\"");
505 }
506
507
508 /*---------------------------------------------------.
509 | Strings, comments etc. can be found in user code. |
510 `---------------------------------------------------*/
511
512 <SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
513 {
514 "'" STRING_GROW; c_context = YY_START; BEGIN SC_CHARACTER;
515 "\"" STRING_GROW; c_context = YY_START; BEGIN SC_STRING;
516 "/"{splice}"*" STRING_GROW; c_context = YY_START; BEGIN SC_COMMENT;
517 "/"{splice}"/" STRING_GROW; c_context = YY_START; BEGIN SC_LINE_COMMENT;
518 }
519
520
521 /*---------------------------------------------------------------.
522 | Scanning some code in braces (%union and actions). The initial |
523 | "{" is already eaten. |
524 `---------------------------------------------------------------*/
525
526 <SC_BRACED_CODE>
527 {
528 "{"|"<"{splice}"%" STRING_GROW; braces_level++;
529 "%"{splice}">" STRING_GROW; braces_level--;
530 "}" {
531 STRING_GROW;
532 braces_level--;
533 if (braces_level < 0)
534 {
535 STRING_FINISH;
536 val->string = last_string;
537 rule_length++;
538 BEGIN INITIAL;
539 return BRACED_CODE;
540 }
541 }
542
543 /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
544 (as `<' `<%'). */
545 "<"{splice}"<" STRING_GROW;
546
547 "$"("<"{tag}">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code,
548 yytext, *loc); }
549 "@"(-?[0-9]+|"$") { handle_at (current_braced_code,
550 yytext, *loc); }
551
552 <<EOF>> unexpected_end_of_file (loc, "}");
553 }
554
555
556 /*--------------------------------------------------------------.
557 | Scanning some prologue: from "%{" (already scanned) to "%}". |
558 `--------------------------------------------------------------*/
559
560 <SC_PROLOGUE>
561 {
562 "%}" {
563 STRING_FINISH;
564 val->string = last_string;
565 BEGIN INITIAL;
566 return PROLOGUE;
567 }
568
569 <<EOF>> unexpected_end_of_file (loc, "%}");
570 }
571
572
573 /*---------------------------------------------------------------.
574 | Scanning the epilogue (everything after the second "%%", which |
575 | has already been eaten). |
576 `---------------------------------------------------------------*/
577
578 <SC_EPILOGUE>
579 {
580 <<EOF>> {
581 STRING_FINISH;
582 val->string = last_string;
583 BEGIN INITIAL;
584 return EPILOGUE;
585 }
586 }
587
588
589 /*----------------------------------------------------------------.
590 | By default, grow the string obstack with the input, escaping M4 |
591 | quoting characters. |
592 `----------------------------------------------------------------*/
593
594 <SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
595 {
596 \$ obstack_sgrow (&string_obstack, "$][");
597 \@ obstack_sgrow (&string_obstack, "@@");
598 \[ obstack_sgrow (&string_obstack, "@{");
599 \] obstack_sgrow (&string_obstack, "@}");
600 .|\n STRING_GROW;
601 }
602
603
604 %%
605
606 /*------------------------------------------------------------------.
607 | TEXT is pointing to a wannabee semantic value (i.e., a `$'). |
608 | |
609 | Possible inputs: $[<TYPENAME>]($|integer) |
610 | |
611 | Output to the STRING_OBSTACK a reference to this semantic value. |
612 `------------------------------------------------------------------*/
613
614 static inline void
615 handle_action_dollar (char *text, location_t location)
616 {
617 const char *type_name = NULL;
618 char *cp = text + 1;
619
620 /* Get the type name if explicit. */
621 if (*cp == '<')
622 {
623 type_name = ++cp;
624 while (*cp != '>')
625 ++cp;
626 *cp = '\0';
627 ++cp;
628 }
629
630 if (*cp == '$')
631 {
632 if (!type_name)
633 type_name = symbol_list_n_type_name_get (current_rule, location, 0);
634 if (!type_name && typed)
635 complain_at (location, _("$$ of `%s' has no declared type"),
636 current_rule->sym->tag);
637 if (!type_name)
638 type_name = "";
639 obstack_fgrow1 (&string_obstack,
640 "]b4_lhs_value([%s])[", type_name);
641 }
642 else
643 {
644 long num;
645 errno = 0;
646 num = strtol (cp, 0, 10);
647
648 if (INT_MIN <= num && num <= rule_length && ! errno)
649 {
650 int n = num;
651 if (!type_name && n > 0)
652 type_name = symbol_list_n_type_name_get (current_rule, location,
653 n);
654 if (!type_name && typed)
655 complain_at (location, _("$%d of `%s' has no declared type"),
656 n, current_rule->sym->tag);
657 if (!type_name)
658 type_name = "";
659 obstack_fgrow3 (&string_obstack,
660 "]b4_rhs_value([%d], [%d], [%s])[",
661 rule_length, n, type_name);
662 }
663 else
664 complain_at (location, _("integer out of range: %s"), quote (text));
665 }
666 }
667
668
669 /*---------------------------------------------------------------.
670 | TEXT is expected to be $$ in some code associated to a symbol: |
671 | destructor or printer. |
672 `---------------------------------------------------------------*/
673
674 static inline void
675 handle_symbol_code_dollar (char *text, location_t location)
676 {
677 char *cp = text + 1;
678 if (*cp == '$')
679 obstack_sgrow (&string_obstack, "]b4_dollar_dollar[");
680 else
681 complain_at (location, _("invalid value: %s"), quote (text));
682 }
683
684
685 /*-----------------------------------------------------------------.
686 | Dispatch onto handle_action_dollar, or handle_destructor_dollar, |
687 | depending upon CODE_KIND. |
688 `-----------------------------------------------------------------*/
689
690 static void
691 handle_dollar (braced_code_t braced_code_kind,
692 char *text, location_t location)
693 {
694 switch (braced_code_kind)
695 {
696 case action_braced_code:
697 handle_action_dollar (text, location);
698 break;
699
700 case destructor_braced_code:
701 case printer_braced_code:
702 handle_symbol_code_dollar (text, location);
703 break;
704 }
705 }
706
707
708 /*------------------------------------------------------.
709 | TEXT is a location token (i.e., a `@...'). Output to |
710 | STRING_OBSTACK a reference to this location. |
711 `------------------------------------------------------*/
712
713 static inline void
714 handle_action_at (char *text, location_t location)
715 {
716 char *cp = text + 1;
717 locations_flag = 1;
718
719 if (*cp == '$')
720 {
721 obstack_sgrow (&string_obstack, "]b4_lhs_location[");
722 }
723 else
724 {
725 long num;
726 errno = 0;
727 num = strtol (cp, 0, 10);
728
729 if (INT_MIN <= num && num <= rule_length && ! errno)
730 {
731 int n = num;
732 obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
733 rule_length, n);
734 }
735 else
736 complain_at (location, _("integer out of range: %s"), quote (text));
737 }
738 }
739
740
741 /*---------------------------------------------------------------.
742 | TEXT is expected to be @$ in some code associated to a symbol: |
743 | destructor or printer. |
744 `---------------------------------------------------------------*/
745
746 static inline void
747 handle_symbol_code_at (char *text, location_t location)
748 {
749 char *cp = text + 1;
750 if (*cp == '$')
751 obstack_sgrow (&string_obstack, "]b4_at_dollar[");
752 else
753 complain_at (location, _("invalid value: %s"), quote (text));
754 }
755
756
757 /*-------------------------------------------------------------------.
758 | Dispatch onto handle_action_at, or handle_destructor_at, depending |
759 | upon CODE_KIND. |
760 `-------------------------------------------------------------------*/
761
762 static void
763 handle_at (braced_code_t braced_code_kind,
764 char *text, location_t location)
765 {
766 switch (braced_code_kind)
767 {
768 case action_braced_code:
769 handle_action_at (text, location);
770 break;
771
772 case destructor_braced_code:
773 case printer_braced_code:
774 handle_symbol_code_at (text, location);
775 break;
776 }
777 }
778
779
780 /*------------------------------------------------------------------.
781 | Convert universal character name UCN to a single-byte character, |
782 | and return that character. Return -1 if UCN does not correspond |
783 | to a single-byte character. |
784 `------------------------------------------------------------------*/
785
786 static int
787 convert_ucn_to_byte (char const *ucn)
788 {
789 unsigned long code = strtoul (ucn + 2, 0, 16);
790
791 /* FIXME: Currently we assume Unicode-compatible unibyte characters
792 on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On
793 non-ASCII hosts we support only the portable C character set.
794 These limitations should be removed once we add support for
795 multibyte characters. */
796
797 if (UCHAR_MAX < code)
798 return -1;
799
800 #if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e)
801 {
802 /* A non-ASCII host. Use CODE to index into a table of the C
803 basic execution character set, which is guaranteed to exist on
804 all Standard C platforms. This table also includes '$', '@',
805 and '`', which are not in the basic execution character set but
806 which are unibyte characters on all the platforms that we know
807 about. */
808 static signed char const table[] =
809 {
810 '\0', -1, -1, -1, -1, -1, -1, '\a',
811 '\b', '\t', '\n', '\v', '\f', '\r', -1, -1,
812 -1, -1, -1, -1, -1, -1, -1, -1,
813 -1, -1, -1, -1, -1, -1, -1, -1,
814 ' ', '!', '"', '#', '$', '%', '&', '\'',
815 '(', ')', '*', '+', ',', '-', '.', '/',
816 '0', '1', '2', '3', '4', '5', '6', '7',
817 '8', '9', ':', ';', '<', '=', '>', '?',
818 '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
819 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
820 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
821 'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
822 '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
823 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
824 'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
825 'x', 'y', 'z', '{', '|', '}', '~'
826 };
827
828 code = code < sizeof table ? table[code] : -1;
829 }
830 #endif
831
832 return code;
833 }
834
835
836 /*----------------------------------------------------------------.
837 | Handle `#line INT "FILE"'. ARGS has already skipped `#line '. |
838 `----------------------------------------------------------------*/
839
840 static void
841 handle_syncline (char *args, location_t *location)
842 {
843 int lineno = strtol (args, &args, 10);
844 const char *file = NULL;
845 file = strchr (args, '"') + 1;
846 *strchr (file, '"') = 0;
847 current_file = xstrdup (file);
848 location->file = current_file;
849 location->last_line = lineno;
850 }
851
852
853 /*-------------------------------------------------------------.
854 | Report an unexpected end of file at LOC. An end of file was |
855 | encountered and the expected TOKEN_END was missing. After |
856 | reporting the problem, pretend that TOKEN_END was found. |
857 `-------------------------------------------------------------*/
858
859 static void
860 unexpected_end_of_file (location_t *loc, char const *token_end)
861 {
862 size_t i = strlen (token_end);
863
864 complain_at (*loc, _("missing `%s' at end of file"), token_end);
865
866 /* Adjust location's last column so that any later message does not
867 mention the characters just inserted. */
868 loc->last_column -= i;
869
870 while (i != 0)
871 unput (token_end[--i]);
872 }
873
874
875 /*-------------------------.
876 | Initialize the scanner. |
877 `-------------------------*/
878
879 void
880 scanner_initialize (void)
881 {
882 obstack_init (&string_obstack);
883 }
884
885
886 /*-----------------------------------------------.
887 | Free all the memory allocated to the scanner. |
888 `-----------------------------------------------*/
889
890 void
891 scanner_free (void)
892 {
893 obstack_free (&string_obstack, 0);
894 /* Reclaim Flex's buffers. */
895 yy_delete_buffer (YY_CURRENT_BUFFER);
896 }