]> git.saurik.com Git - bison.git/blame - src/scan-gram.l
* src/parse-gram.y (char_name): New function.
[bison.git] / src / scan-gram.l
CommitLineData
e9955c83 1/* Bison Grammar Scanner -*- C -*-
3b1e470c 2
073f9288 3 Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
e9955c83
AD
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
0fb669f9
PE
19 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
20 02110-1301 USA
e9955c83
AD
21*/
22
aa418041 23%option debug nodefault nounput noyywrap never-interactive
e9955c83
AD
24%option prefix="gram_" outfile="lex.yy.c"
25
26%{
9e668899
JD
27#include <config.h>
28#include "system.h"
29
4f6e011e
PE
30/* Work around a bug in flex 2.5.31. See Debian bug 333231
31 <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=333231>. */
32#undef gram_wrap
33#define gram_wrap() 1
34
e9071366
AD
35#define FLEX_PREFIX(Id) gram_ ## Id
36#include "flex-scanner.h"
223ff46e 37
e9955c83 38#include "complain.h"
3f2d73f1 39#include "files.h"
e9071366 40#include "getargs.h" /* yacc_flag */
e9955c83 41#include "gram.h"
ca407bdf 42#include "quotearg.h"
e9955c83 43#include "reader.h"
223ff46e 44#include "uniqstr.h"
e9955c83 45
e9071366
AD
46#include <mbswidth.h>
47#include <quote.h>
48
49#include "scan-gram.h"
50
51#define YY_DECL GRAM_LEX_DECL
2346344a 52
3f2d73f1 53#define YY_USER_INIT \
e9071366 54 code_start = scanner_cursor = loc->start; \
dc9701e8 55
3f2d73f1
PE
56/* Location of scanner cursor. */
57boundary scanner_cursor;
41141c56 58
e9071366 59#define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng);
d8d3f94a 60
6c30d641 61static size_t no_cr_read (FILE *, char *, size_t);
d8d3f94a
PE
62#define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
63
7ec2d4cd 64/* A string representing the most recently saved token. */
6b702268 65char *last_string;
7ec2d4cd 66
7ec2d4cd 67void
e9071366 68gram_scanner_last_string_free (void)
7ec2d4cd 69{
41141c56 70 STRING_FREE;
7ec2d4cd 71}
e9955c83 72
e9071366
AD
73/* The location of the most recently saved token, if it was a
74 BRACED_CODE token; otherwise, this has an unspecified value. */
75location gram_last_braced_code_loc;
4517da37 76
4517da37 77static void handle_syncline (char *, location);
1452af69 78static unsigned long int scan_integer (char const *p, int base, location loc);
d8d3f94a 79static int convert_ucn_to_byte (char const *hex_text);
aa418041 80static void unexpected_eof (boundary, char const *);
4febdd96 81static void unexpected_newline (boundary, char const *);
e9955c83
AD
82
83%}
e9071366
AD
84 /* A C-like comment in directives/rules. */
85%x SC_YACC_COMMENT
86 /* Strings and characters in directives/rules. */
e9955c83 87%x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER
e9071366
AD
88 /* A identifier was just read in directives/rules. Special state
89 to capture the sequence `identifier :'. */
90%x SC_AFTER_IDENTIFIER
e9071366
AD
91
92 /* Three types of user code:
93 - prologue (code between `%{' `%}' in the first section, before %%);
94 - actions, printers, union, etc, (between braced in the middle section);
95 - epilogue (everything after the second %%). */
96%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE
97 /* C and C++ comments in code. */
98%x SC_COMMENT SC_LINE_COMMENT
99 /* Strings and characters in code. */
100%x SC_STRING SC_CHARACTER
e9955c83 101
29c01725
AD
102letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
103id {letter}({letter}|[0-9])*
104directive %{letter}({letter}|[0-9]|-)*
624a35e2 105int [0-9]+
d8d3f94a
PE
106
107/* POSIX says that a tag must be both an id and a C union member, but
108 historically almost any character is allowed in a tag. We disallow
109 NUL and newline, as this simplifies our implementation. */
110tag [^\0\n>]+
111
112/* Zero or more instances of backslash-newline. Following GCC, allow
113 white space between the backslash and the newline. */
114splice (\\[ \f\t\v]*\n)*
e9955c83
AD
115
116%%
117%{
a706a1cc 118 /* Nesting level of the current code in braces. */
1a9e39f1
PE
119 int braces_level IF_LINT (= 0);
120
3f2d73f1
PE
121 /* Parent context state, when applicable. */
122 int context_state IF_LINT (= 0);
a706a1cc 123
3f2d73f1 124 /* Location of most recent identifier, when applicable. */
a2bc9dbc 125 location id_loc IF_LINT (= empty_location);
3f2d73f1 126
a2bc9dbc
PE
127 /* Where containing code started, when applicable. Its initial
128 value is relevant only when yylex is invoked in the SC_EPILOGUE
129 start condition. */
130 boundary code_start = scanner_cursor;
3f2d73f1 131
223ff46e
PE
132 /* Where containing comment or string or character literal started,
133 when applicable. */
a2bc9dbc 134 boundary token_start IF_LINT (= scanner_cursor);
e9955c83
AD
135%}
136
137
3f2d73f1
PE
138 /*-----------------------.
139 | Scanning white space. |
140 `-----------------------*/
141
58d7a1a1 142<INITIAL,SC_AFTER_IDENTIFIER>
3f2d73f1 143{
4febdd96 144 /* Comments and white space. */
83adb046 145 "," warn_at (*loc, _("stray `,' treated as white space"));
4febdd96 146 [ \f\n\t\v] |
3f2d73f1 147 "//".* ;
83adb046
PE
148 "/*" {
149 token_start = loc->start;
150 context_state = YY_START;
151 BEGIN SC_YACC_COMMENT;
152 }
3f2d73f1
PE
153
154 /* #line directives are not documented, and may be withdrawn or
155 modified in future versions of Bison. */
156 ^"#line "{int}" \"".*"\"\n" {
4517da37 157 handle_syncline (yytext + sizeof "#line " - 1, *loc);
3f2d73f1
PE
158 }
159}
160
161
e9955c83
AD
162 /*----------------------------.
163 | Scanning Bison directives. |
164 `----------------------------*/
165<INITIAL>
166{
58d7a1a1
AD
167 "%binary" return PERCENT_NONASSOC;
168 "%debug" return PERCENT_DEBUG;
169 "%default"[-_]"prec" return PERCENT_DEFAULT_PREC;
170 "%define" return PERCENT_DEFINE;
171 "%defines" return PERCENT_DEFINES;
172 "%destructor" return PERCENT_DESTRUCTOR;
173 "%dprec" return PERCENT_DPREC;
174 "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE;
175 "%expect" return PERCENT_EXPECT;
176 "%expect"[-_]"rr" return PERCENT_EXPECT_RR;
177 "%file-prefix" return PERCENT_FILE_PREFIX;
e9955c83 178 "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC;
58d7a1a1
AD
179 "%initial-action" return PERCENT_INITIAL_ACTION;
180 "%glr-parser" return PERCENT_GLR_PARSER;
181 "%left" return PERCENT_LEFT;
182 "%lex-param" return PERCENT_LEX_PARAM;
183 "%locations" return PERCENT_LOCATIONS;
184 "%merge" return PERCENT_MERGE;
185 "%name"[-_]"prefix" return PERCENT_NAME_PREFIX;
186 "%no"[-_]"default"[-_]"prec" return PERCENT_NO_DEFAULT_PREC;
187 "%no"[-_]"lines" return PERCENT_NO_LINES;
188 "%nonassoc" return PERCENT_NONASSOC;
189 "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER;
190 "%nterm" return PERCENT_NTERM;
191 "%output" return PERCENT_OUTPUT;
192 "%parse-param" return PERCENT_PARSE_PARAM;
193 "%prec" return PERCENT_PREC;
194 "%printer" return PERCENT_PRINTER;
195 "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
196 "%require" return PERCENT_REQUIRE;
197 "%right" return PERCENT_RIGHT;
198 "%skeleton" return PERCENT_SKELETON;
199 "%start" return PERCENT_START;
200 "%term" return PERCENT_TOKEN;
201 "%token" return PERCENT_TOKEN;
202 "%token"[-_]"table" return PERCENT_TOKEN_TABLE;
203 "%type" return PERCENT_TYPE;
204 "%union" return PERCENT_UNION;
205 "%verbose" return PERCENT_VERBOSE;
206 "%yacc" return PERCENT_YACC;
e9955c83 207
3f2d73f1 208 {directive} {
41141c56 209 complain_at (*loc, _("invalid directive: %s"), quote (yytext));
412f8a59 210 }
900c5db5 211
e9955c83 212 "=" return EQUAL;
e9071366 213 "|" return PIPE;
e9955c83
AD
214 ";" return SEMICOLON;
215
3f2d73f1 216 {id} {
58d7a1a1 217 val->uniqstr = uniqstr_new (yytext);
3f2d73f1 218 id_loc = *loc;
3f2d73f1 219 BEGIN SC_AFTER_IDENTIFIER;
e9955c83
AD
220 }
221
d8d3f94a 222 {int} {
1452af69
PE
223 val->integer = scan_integer (yytext, 10, *loc);
224 return INT;
225 }
226 0[xX][0-9abcdefABCDEF]+ {
227 val->integer = scan_integer (yytext, 16, *loc);
d8d3f94a
PE
228 return INT;
229 }
e9955c83
AD
230
231 /* Characters. We don't check there is only one. */
3f2d73f1 232 "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER;
e9955c83
AD
233
234 /* Strings. */
ca407bdf 235 "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING;
e9955c83
AD
236
237 /* Prologue. */
3f2d73f1 238 "%{" code_start = loc->start; BEGIN SC_PROLOGUE;
e9955c83
AD
239
240 /* Code in between braces. */
3f2d73f1 241 "{" {
b2ddc3f3 242 if (current_rule && current_rule->action)
6b702268 243 grammar_midrule_action ();
3f2d73f1
PE
244 STRING_GROW;
245 braces_level = 0;
246 code_start = loc->start;
247 BEGIN SC_BRACED_CODE;
248 }
e9955c83
AD
249
250 /* A type. */
d8d3f94a 251 "<"{tag}">" {
223ff46e 252 obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2);
41141c56 253 STRING_FINISH;
223ff46e 254 val->uniqstr = uniqstr_new (last_string);
41141c56 255 STRING_FREE;
4cdb01db
AD
256 return TYPE;
257 }
258
a706a1cc
PE
259 "%%" {
260 static int percent_percent_count;
e9955c83 261 if (++percent_percent_count == 2)
a2bc9dbc 262 BEGIN SC_EPILOGUE;
e9955c83
AD
263 return PERCENT_PERCENT;
264 }
265
a706a1cc 266 . {
41141c56 267 complain_at (*loc, _("invalid character: %s"), quote (yytext));
3f2d73f1 268 }
379f0ac8
PE
269
270 <<EOF>> {
271 loc->start = loc->end = scanner_cursor;
272 yyterminate ();
273 }
3f2d73f1
PE
274}
275
276
277 /*-----------------------------------------------------------------.
278 | Scanning after an identifier, checking whether a colon is next. |
279 `-----------------------------------------------------------------*/
280
281<SC_AFTER_IDENTIFIER>
282{
283 ":" {
3f2d73f1
PE
284 *loc = id_loc;
285 BEGIN INITIAL;
286 return ID_COLON;
287 }
288 . {
289 scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
290 yyless (0);
291 *loc = id_loc;
292 BEGIN INITIAL;
293 return ID;
294 }
295 <<EOF>> {
296 *loc = id_loc;
297 BEGIN INITIAL;
298 return ID;
e9955c83
AD
299 }
300}
301
302
d8d3f94a
PE
303 /*---------------------------------------------------------------.
304 | Scanning a Yacc comment. The initial `/ *' is already eaten. |
305 `---------------------------------------------------------------*/
e9955c83 306
d8d3f94a 307<SC_YACC_COMMENT>
e9955c83 308{
3f2d73f1 309 "*/" BEGIN context_state;
a706a1cc 310 .|\n ;
aa418041 311 <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
d8d3f94a
PE
312}
313
314
315 /*------------------------------------------------------------.
316 | Scanning a C comment. The initial `/ *' is already eaten. |
317 `------------------------------------------------------------*/
318
319<SC_COMMENT>
320{
3f2d73f1 321 "*"{splice}"/" STRING_GROW; BEGIN context_state;
aa418041 322 <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
e9955c83
AD
323}
324
325
d8d3f94a
PE
326 /*--------------------------------------------------------------.
327 | Scanning a line comment. The initial `//' is already eaten. |
328 `--------------------------------------------------------------*/
329
330<SC_LINE_COMMENT>
331{
3f2d73f1 332 "\n" STRING_GROW; BEGIN context_state;
41141c56 333 {splice} STRING_GROW;
3f2d73f1 334 <<EOF>> BEGIN context_state;
d8d3f94a
PE
335}
336
337
4febdd96
PE
338 /*------------------------------------------------.
339 | Scanning a Bison string, including its escapes. |
340 | The initial quote is already eaten. |
341 `------------------------------------------------*/
e9955c83
AD
342
343<SC_ESCAPED_STRING>
344{
db2cc12f 345 "\"" {
41141c56 346 STRING_FINISH;
3f2d73f1 347 loc->start = token_start;
223ff46e 348 val->chars = last_string;
a706a1cc 349 BEGIN INITIAL;
e9955c83
AD
350 return STRING;
351 }
4febdd96
PE
352 \n unexpected_newline (token_start, "\""); BEGIN INITIAL;
353 <<EOF>> unexpected_eof (token_start, "\""); BEGIN INITIAL;
e9955c83
AD
354}
355
4febdd96
PE
356 /*----------------------------------------------------------.
357 | Scanning a Bison character literal, decoding its escapes. |
358 | The initial quote is already eaten. |
359 `----------------------------------------------------------*/
e9955c83
AD
360
361<SC_ESCAPED_CHARACTER>
362{
db2cc12f 363 "'" {
3b1e470c 364 unsigned char last_string_1;
41141c56
PE
365 STRING_GROW;
366 STRING_FINISH;
3f2d73f1 367 loc->start = token_start;
58d7a1a1 368 val->character = last_string[1];
41141c56 369 STRING_FREE;
a706a1cc 370 BEGIN INITIAL;
58d7a1a1 371 return CHAR;
e9955c83 372 }
4febdd96
PE
373 \n unexpected_newline (token_start, "'"); BEGIN INITIAL;
374 <<EOF>> unexpected_eof (token_start, "'"); BEGIN INITIAL;
375}
a706a1cc 376
4febdd96
PE
377<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING>
378{
92ac3705 379 \0 complain_at (*loc, _("invalid null character"));
e9955c83
AD
380}
381
382
383 /*----------------------------.
384 | Decode escaped characters. |
385 `----------------------------*/
386
387<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
388{
d8d3f94a 389 \\[0-7]{1,3} {
4517da37 390 unsigned long int c = strtoul (yytext + 1, NULL, 8);
d8d3f94a 391 if (UCHAR_MAX < c)
3f2d73f1 392 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
05ac60f3 393 else if (! c)
92ac3705 394 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
e9955c83 395 else
223ff46e 396 obstack_1grow (&obstack_for_string, c);
e9955c83
AD
397 }
398
6b0d38ab 399 \\x[0-9abcdefABCDEF]+ {
4517da37
PE
400 verify (UCHAR_MAX < ULONG_MAX);
401 unsigned long int c = strtoul (yytext + 2, NULL, 16);
402 if (UCHAR_MAX < c)
3f2d73f1 403 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
92ac3705
PE
404 else if (! c)
405 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
d8d3f94a 406 else
223ff46e 407 obstack_1grow (&obstack_for_string, c);
e9955c83
AD
408 }
409
223ff46e
PE
410 \\a obstack_1grow (&obstack_for_string, '\a');
411 \\b obstack_1grow (&obstack_for_string, '\b');
412 \\f obstack_1grow (&obstack_for_string, '\f');
413 \\n obstack_1grow (&obstack_for_string, '\n');
414 \\r obstack_1grow (&obstack_for_string, '\r');
415 \\t obstack_1grow (&obstack_for_string, '\t');
416 \\v obstack_1grow (&obstack_for_string, '\v');
412f8a59
PE
417
418 /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */
223ff46e 419 \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]);
412f8a59 420
6b0d38ab 421 \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
d8d3f94a
PE
422 int c = convert_ucn_to_byte (yytext);
423 if (c < 0)
3f2d73f1 424 complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
92ac3705
PE
425 else if (! c)
426 complain_at (*loc, _("invalid null character: %s"), quote (yytext));
d8d3f94a 427 else
223ff46e 428 obstack_1grow (&obstack_for_string, c);
d8d3f94a 429 }
4f25ebb0 430 \\(.|\n) {
3f2d73f1 431 complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext));
41141c56 432 STRING_GROW;
e9955c83
AD
433 }
434}
435
4febdd96
PE
436 /*--------------------------------------------.
437 | Scanning user-code characters and strings. |
438 `--------------------------------------------*/
e9955c83 439
4febdd96
PE
440<SC_CHARACTER,SC_STRING>
441{
e9071366 442 {splice}|\\{splice}[^\n\[\]] STRING_GROW;
4febdd96 443}
e9955c83
AD
444
445<SC_CHARACTER>
446{
4febdd96
PE
447 "'" STRING_GROW; BEGIN context_state;
448 \n unexpected_newline (token_start, "'"); BEGIN context_state;
449 <<EOF>> unexpected_eof (token_start, "'"); BEGIN context_state;
e9955c83
AD
450}
451
e9955c83
AD
452<SC_STRING>
453{
4febdd96
PE
454 "\"" STRING_GROW; BEGIN context_state;
455 \n unexpected_newline (token_start, "\""); BEGIN context_state;
456 <<EOF>> unexpected_eof (token_start, "\""); BEGIN context_state;
e9955c83
AD
457}
458
459
460 /*---------------------------------------------------.
461 | Strings, comments etc. can be found in user code. |
462 `---------------------------------------------------*/
463
464<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
465{
3f2d73f1
PE
466 "'" {
467 STRING_GROW;
468 context_state = YY_START;
469 token_start = loc->start;
470 BEGIN SC_CHARACTER;
471 }
472 "\"" {
473 STRING_GROW;
474 context_state = YY_START;
475 token_start = loc->start;
476 BEGIN SC_STRING;
477 }
478 "/"{splice}"*" {
479 STRING_GROW;
480 context_state = YY_START;
481 token_start = loc->start;
482 BEGIN SC_COMMENT;
483 }
484 "/"{splice}"/" {
485 STRING_GROW;
486 context_state = YY_START;
487 BEGIN SC_LINE_COMMENT;
488 }
e9955c83
AD
489}
490
491
624a35e2 492
58d7a1a1
AD
493 /*-----------------------------------------------------------.
494 | Scanning some code in braces (actions). The initial "{" is |
495 | already eaten. |
496 `-----------------------------------------------------------*/
e9955c83
AD
497
498<SC_BRACED_CODE>
499{
41141c56
PE
500 "{"|"<"{splice}"%" STRING_GROW; braces_level++;
501 "%"{splice}">" STRING_GROW; braces_level--;
e9955c83 502 "}" {
25522739
PE
503 obstack_1grow (&obstack_for_string, '}');
504
2346344a
AD
505 --braces_level;
506 if (braces_level < 0)
e9955c83 507 {
41141c56 508 STRING_FINISH;
3f2d73f1 509 loc->start = code_start;
223ff46e 510 val->chars = last_string;
e9071366 511 gram_last_braced_code_loc = *loc;
a706a1cc 512 BEGIN INITIAL;
58d7a1a1 513 return BRACED_CODE;
e9955c83
AD
514 }
515 }
516
a706a1cc
PE
517 /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
518 (as `<' `<%'). */
41141c56 519 "<"{splice}"<" STRING_GROW;
a706a1cc 520
aa418041 521 <<EOF>> unexpected_eof (code_start, "}"); BEGIN INITIAL;
e9955c83
AD
522}
523
524
525 /*--------------------------------------------------------------.
526 | Scanning some prologue: from "%{" (already scanned) to "%}". |
527 `--------------------------------------------------------------*/
528
529<SC_PROLOGUE>
530{
531 "%}" {
41141c56 532 STRING_FINISH;
3f2d73f1 533 loc->start = code_start;
223ff46e 534 val->chars = last_string;
a706a1cc 535 BEGIN INITIAL;
e9955c83
AD
536 return PROLOGUE;
537 }
538
aa418041 539 <<EOF>> unexpected_eof (code_start, "%}"); BEGIN INITIAL;
e9955c83
AD
540}
541
542
543 /*---------------------------------------------------------------.
544 | Scanning the epilogue (everything after the second "%%", which |
d8d3f94a 545 | has already been eaten). |
e9955c83
AD
546 `---------------------------------------------------------------*/
547
548<SC_EPILOGUE>
549{
e9955c83 550 <<EOF>> {
41141c56 551 STRING_FINISH;
3f2d73f1 552 loc->start = code_start;
223ff46e 553 val->chars = last_string;
a706a1cc 554 BEGIN INITIAL;
e9955c83
AD
555 return EPILOGUE;
556 }
557}
558
559
4febdd96
PE
560 /*-----------------------------------------------------.
561 | By default, grow the string obstack with the input. |
562 `-----------------------------------------------------*/
563
564<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>. |
565<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n STRING_GROW;
566
e9955c83
AD
567%%
568
6c30d641
PE
569/* Read bytes from FP into buffer BUF of size SIZE. Return the
570 number of bytes read. Remove '\r' from input, treating \r\n
571 and isolated \r as \n. */
572
573static size_t
574no_cr_read (FILE *fp, char *buf, size_t size)
575{
a737b216
PE
576 size_t bytes_read = fread (buf, 1, size, fp);
577 if (bytes_read)
6c30d641 578 {
a737b216 579 char *w = memchr (buf, '\r', bytes_read);
6c30d641
PE
580 if (w)
581 {
582 char const *r = ++w;
a737b216 583 char const *lim = buf + bytes_read;
6c30d641
PE
584
585 for (;;)
586 {
587 /* Found an '\r'. Treat it like '\n', but ignore any
588 '\n' that immediately follows. */
589 w[-1] = '\n';
590 if (r == lim)
591 {
592 int ch = getc (fp);
593 if (ch != '\n' && ungetc (ch, fp) != ch)
594 break;
595 }
596 else if (*r == '\n')
597 r++;
598
599 /* Copy until the next '\r'. */
600 do
601 {
602 if (r == lim)
603 return w - buf;
604 }
605 while ((*w++ = *r++) != '\r');
606 }
607
608 return w - buf;
609 }
610 }
611
a737b216 612 return bytes_read;
6c30d641
PE
613}
614
615
f25bfb75 616
1452af69
PE
617/*------------------------------------------------------.
618| Scan NUMBER for a base-BASE integer at location LOC. |
619`------------------------------------------------------*/
620
621static unsigned long int
622scan_integer (char const *number, int base, location loc)
623{
4517da37
PE
624 verify (INT_MAX < ULONG_MAX);
625 unsigned long int num = strtoul (number, NULL, base);
626
627 if (INT_MAX < num)
1452af69
PE
628 {
629 complain_at (loc, _("integer out of range: %s"), quote (number));
630 num = INT_MAX;
631 }
4517da37 632
1452af69
PE
633 return num;
634}
635
636
d8d3f94a
PE
637/*------------------------------------------------------------------.
638| Convert universal character name UCN to a single-byte character, |
639| and return that character. Return -1 if UCN does not correspond |
640| to a single-byte character. |
641`------------------------------------------------------------------*/
642
643static int
644convert_ucn_to_byte (char const *ucn)
645{
4517da37
PE
646 verify (UCHAR_MAX <= INT_MAX);
647 unsigned long int code = strtoul (ucn + 2, NULL, 16);
d8d3f94a
PE
648
649 /* FIXME: Currently we assume Unicode-compatible unibyte characters
650 on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On
651 non-ASCII hosts we support only the portable C character set.
652 These limitations should be removed once we add support for
653 multibyte characters. */
654
655 if (UCHAR_MAX < code)
656 return -1;
657
658#if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e)
659 {
660 /* A non-ASCII host. Use CODE to index into a table of the C
661 basic execution character set, which is guaranteed to exist on
662 all Standard C platforms. This table also includes '$', '@',
8e6ef483 663 and '`', which are not in the basic execution character set but
d8d3f94a
PE
664 which are unibyte characters on all the platforms that we know
665 about. */
666 static signed char const table[] =
667 {
668 '\0', -1, -1, -1, -1, -1, -1, '\a',
669 '\b', '\t', '\n', '\v', '\f', '\r', -1, -1,
670 -1, -1, -1, -1, -1, -1, -1, -1,
671 -1, -1, -1, -1, -1, -1, -1, -1,
672 ' ', '!', '"', '#', '$', '%', '&', '\'',
673 '(', ')', '*', '+', ',', '-', '.', '/',
674 '0', '1', '2', '3', '4', '5', '6', '7',
675 '8', '9', ':', ';', '<', '=', '>', '?',
676 '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
677 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
678 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
679 'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
680 '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
681 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
682 'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
683 'x', 'y', 'z', '{', '|', '}', '~'
684 };
685
686 code = code < sizeof table ? table[code] : -1;
687 }
688#endif
c4d720cd 689
d8d3f94a
PE
690 return code;
691}
692
693
900c5db5
AD
694/*----------------------------------------------------------------.
695| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. |
696`----------------------------------------------------------------*/
697
698static void
4517da37 699handle_syncline (char *args, location loc)
900c5db5 700{
4517da37
PE
701 char *after_num;
702 unsigned long int lineno = strtoul (args, &after_num, 10);
703 char *file = strchr (after_num, '"') + 1;
704 *strchr (file, '"') = '\0';
705 if (INT_MAX <= lineno)
706 {
707 warn_at (loc, _("line number overflow"));
708 lineno = INT_MAX;
709 }
e9071366 710 current_file = uniqstr_new (file);
0c8e079f 711 boundary_set (&scanner_cursor, current_file, lineno, 1);
4517da37
PE
712}
713
714
4febdd96
PE
715/*----------------------------------------------------------------.
716| For a token or comment starting at START, report message MSGID, |
717| which should say that an end marker was found before |
718| the expected TOKEN_END. |
719`----------------------------------------------------------------*/
720
721static void
722unexpected_end (boundary start, char const *msgid, char const *token_end)
723{
724 location loc;
725 loc.start = start;
726 loc.end = scanner_cursor;
727 complain_at (loc, _(msgid), token_end);
728}
729
730
3f2d73f1
PE
731/*------------------------------------------------------------------------.
732| Report an unexpected EOF in a token or comment starting at START. |
733| An end of file was encountered and the expected TOKEN_END was missing. |
3f2d73f1 734`------------------------------------------------------------------------*/
a706a1cc
PE
735
736static void
aa418041 737unexpected_eof (boundary start, char const *token_end)
a706a1cc 738{
4febdd96
PE
739 unexpected_end (start, N_("missing `%s' at end of file"), token_end);
740}
741
742
743/*----------------------------------------.
744| Likewise, but for unexpected newlines. |
745`----------------------------------------*/
746
747static void
748unexpected_newline (boundary start, char const *token_end)
749{
750 unexpected_end (start, N_("missing `%s' at end of line"), token_end);
a706a1cc
PE
751}
752
753
f25bfb75
AD
754/*-------------------------.
755| Initialize the scanner. |
756`-------------------------*/
757
1d6412ad 758void
e9071366 759gram_scanner_initialize (void)
1d6412ad 760{
223ff46e 761 obstack_init (&obstack_for_string);
1d6412ad
AD
762}
763
764
f25bfb75
AD
765/*-----------------------------------------------.
766| Free all the memory allocated to the scanner. |
767`-----------------------------------------------*/
768
4cdb01db 769void
e9071366 770gram_scanner_free (void)
4cdb01db 771{
223ff46e 772 obstack_free (&obstack_for_string, 0);
536545f3
AD
773 /* Reclaim Flex's buffers. */
774 yy_delete_buffer (YY_CURRENT_BUFFER);
4cdb01db 775}