]>
Commit | Line | Data |
---|---|---|
e9955c83 | 1 | /* Bison Grammar Scanner -*- C -*- |
3b1e470c | 2 | |
98744608 JD |
3 | Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2009 Free Software |
4 | Foundation, Inc. | |
e9955c83 AD |
5 | |
6 | This file is part of Bison, the GNU Compiler Compiler. | |
7 | ||
f16b0819 | 8 | This program is free software: you can redistribute it and/or modify |
e9955c83 | 9 | it under the terms of the GNU General Public License as published by |
f16b0819 | 10 | the Free Software Foundation, either version 3 of the License, or |
e9955c83 AD |
11 | (at your option) any later version. |
12 | ||
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. | |
17 | ||
18 | You should have received a copy of the GNU General Public License | |
f16b0819 | 19 | along with this program. If not, see <http://www.gnu.org/licenses/>. */ |
e9955c83 | 20 | |
42f8609b | 21 | %option debug nodefault noinput nounput noyywrap never-interactive |
e9955c83 AD |
22 | %option prefix="gram_" outfile="lex.yy.c" |
23 | ||
24 | %{ | |
4f6e011e PE |
25 | /* Work around a bug in flex 2.5.31. See Debian bug 333231 |
26 | <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=333231>. */ | |
27 | #undef gram_wrap | |
28 | #define gram_wrap() 1 | |
29 | ||
e9071366 AD |
30 | #define FLEX_PREFIX(Id) gram_ ## Id |
31 | #include "flex-scanner.h" | |
223ff46e | 32 | |
e9955c83 | 33 | #include "complain.h" |
3f2d73f1 | 34 | #include "files.h" |
e9955c83 | 35 | #include "gram.h" |
ca407bdf | 36 | #include "quotearg.h" |
e9955c83 | 37 | #include "reader.h" |
223ff46e | 38 | #include "uniqstr.h" |
e9955c83 | 39 | |
b1a4261e | 40 | #include <ctype.h> |
e9071366 AD |
41 | #include <mbswidth.h> |
42 | #include <quote.h> | |
43 | ||
44 | #include "scan-gram.h" | |
45 | ||
46 | #define YY_DECL GRAM_LEX_DECL | |
2346344a | 47 | |
3f2d73f1 | 48 | #define YY_USER_INIT \ |
e9071366 | 49 | code_start = scanner_cursor = loc->start; \ |
dc9701e8 | 50 | |
3f2d73f1 | 51 | /* Location of scanner cursor. */ |
4a678af8 | 52 | static boundary scanner_cursor; |
41141c56 | 53 | |
e9071366 | 54 | #define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng); |
d8d3f94a | 55 | |
6c30d641 | 56 | static size_t no_cr_read (FILE *, char *, size_t); |
d8d3f94a PE |
57 | #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size)) |
58 | ||
7ec2d4cd | 59 | /* A string representing the most recently saved token. */ |
7c0c6181 | 60 | static char *last_string; |
7ec2d4cd | 61 | |
7ec2d4cd | 62 | void |
e9071366 | 63 | gram_scanner_last_string_free (void) |
7ec2d4cd | 64 | { |
41141c56 | 65 | STRING_FREE; |
7ec2d4cd | 66 | } |
e9955c83 | 67 | |
4517da37 | 68 | static void handle_syncline (char *, location); |
1452af69 | 69 | static unsigned long int scan_integer (char const *p, int base, location loc); |
d8d3f94a | 70 | static int convert_ucn_to_byte (char const *hex_text); |
aa418041 | 71 | static void unexpected_eof (boundary, char const *); |
4febdd96 | 72 | static void unexpected_newline (boundary, char const *); |
e9955c83 AD |
73 | |
74 | %} | |
e9071366 AD |
75 | /* A C-like comment in directives/rules. */ |
76 | %x SC_YACC_COMMENT | |
77 | /* Strings and characters in directives/rules. */ | |
e9955c83 | 78 | %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER |
e9071366 AD |
79 | /* A identifier was just read in directives/rules. Special state |
80 | to capture the sequence `identifier :'. */ | |
81 | %x SC_AFTER_IDENTIFIER | |
e9071366 AD |
82 | |
83 | /* Three types of user code: | |
84 | - prologue (code between `%{' `%}' in the first section, before %%); | |
85 | - actions, printers, union, etc, (between braced in the middle section); | |
86 | - epilogue (everything after the second %%). */ | |
87 | %x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE | |
88 | /* C and C++ comments in code. */ | |
89 | %x SC_COMMENT SC_LINE_COMMENT | |
90 | /* Strings and characters in code. */ | |
91 | %x SC_STRING SC_CHARACTER | |
e9955c83 | 92 | |
29c01725 AD |
93 | letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] |
94 | id {letter}({letter}|[0-9])* | |
95 | directive %{letter}({letter}|[0-9]|-)* | |
624a35e2 | 96 | int [0-9]+ |
d8d3f94a PE |
97 | |
98 | /* POSIX says that a tag must be both an id and a C union member, but | |
99 | historically almost any character is allowed in a tag. We disallow | |
100 | NUL and newline, as this simplifies our implementation. */ | |
101 | tag [^\0\n>]+ | |
102 | ||
103 | /* Zero or more instances of backslash-newline. Following GCC, allow | |
104 | white space between the backslash and the newline. */ | |
105 | splice (\\[ \f\t\v]*\n)* | |
e9955c83 AD |
106 | |
107 | %% | |
108 | %{ | |
a706a1cc | 109 | /* Nesting level of the current code in braces. */ |
5362ed19 | 110 | int braces_level IF_LINT (= 0); |
1a9e39f1 | 111 | |
3f2d73f1 | 112 | /* Parent context state, when applicable. */ |
5362ed19 | 113 | int context_state IF_LINT (= 0); |
a706a1cc | 114 | |
3f2d73f1 | 115 | /* Location of most recent identifier, when applicable. */ |
a2bc9dbc | 116 | location id_loc IF_LINT (= empty_location); |
3f2d73f1 | 117 | |
a2bc9dbc PE |
118 | /* Where containing code started, when applicable. Its initial |
119 | value is relevant only when yylex is invoked in the SC_EPILOGUE | |
120 | start condition. */ | |
121 | boundary code_start = scanner_cursor; | |
3f2d73f1 | 122 | |
223ff46e PE |
123 | /* Where containing comment or string or character literal started, |
124 | when applicable. */ | |
a2bc9dbc | 125 | boundary token_start IF_LINT (= scanner_cursor); |
e9955c83 AD |
126 | %} |
127 | ||
128 | ||
3f2d73f1 PE |
129 | /*-----------------------. |
130 | | Scanning white space. | | |
131 | `-----------------------*/ | |
132 | ||
58d7a1a1 | 133 | <INITIAL,SC_AFTER_IDENTIFIER> |
3f2d73f1 | 134 | { |
4febdd96 | 135 | /* Comments and white space. */ |
83adb046 | 136 | "," warn_at (*loc, _("stray `,' treated as white space")); |
4febdd96 | 137 | [ \f\n\t\v] | |
3f2d73f1 | 138 | "//".* ; |
83adb046 PE |
139 | "/*" { |
140 | token_start = loc->start; | |
141 | context_state = YY_START; | |
142 | BEGIN SC_YACC_COMMENT; | |
143 | } | |
3f2d73f1 PE |
144 | |
145 | /* #line directives are not documented, and may be withdrawn or | |
146 | modified in future versions of Bison. */ | |
147 | ^"#line "{int}" \"".*"\"\n" { | |
4517da37 | 148 | handle_syncline (yytext + sizeof "#line " - 1, *loc); |
3f2d73f1 PE |
149 | } |
150 | } | |
151 | ||
152 | ||
e9955c83 AD |
153 | /*----------------------------. |
154 | | Scanning Bison directives. | | |
155 | `----------------------------*/ | |
156 | <INITIAL> | |
157 | { | |
58d7a1a1 | 158 | "%binary" return PERCENT_NONASSOC; |
136a0f76 | 159 | "%code" return PERCENT_CODE; |
58d7a1a1 AD |
160 | "%debug" return PERCENT_DEBUG; |
161 | "%default"[-_]"prec" return PERCENT_DEFAULT_PREC; | |
162 | "%define" return PERCENT_DEFINE; | |
163 | "%defines" return PERCENT_DEFINES; | |
164 | "%destructor" return PERCENT_DESTRUCTOR; | |
165 | "%dprec" return PERCENT_DPREC; | |
166 | "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE; | |
167 | "%expect" return PERCENT_EXPECT; | |
168 | "%expect"[-_]"rr" return PERCENT_EXPECT_RR; | |
169 | "%file-prefix" return PERCENT_FILE_PREFIX; | |
e9955c83 | 170 | "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC; |
58d7a1a1 AD |
171 | "%initial-action" return PERCENT_INITIAL_ACTION; |
172 | "%glr-parser" return PERCENT_GLR_PARSER; | |
0e021770 | 173 | "%language" return PERCENT_LANGUAGE; |
58d7a1a1 AD |
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 | "%pure"[-_]"parser" return PERCENT_PURE_PARSER; | |
189 | "%require" return PERCENT_REQUIRE; | |
190 | "%right" return PERCENT_RIGHT; | |
191 | "%skeleton" return PERCENT_SKELETON; | |
192 | "%start" return PERCENT_START; | |
193 | "%term" return PERCENT_TOKEN; | |
194 | "%token" return PERCENT_TOKEN; | |
195 | "%token"[-_]"table" return PERCENT_TOKEN_TABLE; | |
196 | "%type" return PERCENT_TYPE; | |
197 | "%union" return PERCENT_UNION; | |
198 | "%verbose" return PERCENT_VERBOSE; | |
199 | "%yacc" return PERCENT_YACC; | |
e9955c83 | 200 | |
3f2d73f1 | 201 | {directive} { |
41141c56 | 202 | complain_at (*loc, _("invalid directive: %s"), quote (yytext)); |
412f8a59 | 203 | } |
900c5db5 | 204 | |
e9955c83 | 205 | "=" return EQUAL; |
e9071366 | 206 | "|" return PIPE; |
e9955c83 | 207 | ";" return SEMICOLON; |
12e35840 | 208 | "<*>" return TYPE_TAG_ANY; |
3ebecc24 | 209 | "<>" return TYPE_TAG_NONE; |
e9955c83 | 210 | |
3f2d73f1 | 211 | {id} { |
58d7a1a1 | 212 | val->uniqstr = uniqstr_new (yytext); |
3f2d73f1 | 213 | id_loc = *loc; |
3f2d73f1 | 214 | BEGIN SC_AFTER_IDENTIFIER; |
e9955c83 AD |
215 | } |
216 | ||
d8d3f94a | 217 | {int} { |
1452af69 PE |
218 | val->integer = scan_integer (yytext, 10, *loc); |
219 | return INT; | |
220 | } | |
221 | 0[xX][0-9abcdefABCDEF]+ { | |
222 | val->integer = scan_integer (yytext, 16, *loc); | |
d8d3f94a PE |
223 | return INT; |
224 | } | |
e9955c83 AD |
225 | |
226 | /* Characters. We don't check there is only one. */ | |
3f2d73f1 | 227 | "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; |
e9955c83 AD |
228 | |
229 | /* Strings. */ | |
ca407bdf | 230 | "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING; |
e9955c83 AD |
231 | |
232 | /* Prologue. */ | |
3f2d73f1 | 233 | "%{" code_start = loc->start; BEGIN SC_PROLOGUE; |
e9955c83 AD |
234 | |
235 | /* Code in between braces. */ | |
3f2d73f1 PE |
236 | "{" { |
237 | STRING_GROW; | |
238 | braces_level = 0; | |
239 | code_start = loc->start; | |
240 | BEGIN SC_BRACED_CODE; | |
241 | } | |
e9955c83 AD |
242 | |
243 | /* A type. */ | |
d8d3f94a | 244 | "<"{tag}">" { |
223ff46e | 245 | obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2); |
41141c56 | 246 | STRING_FINISH; |
223ff46e | 247 | val->uniqstr = uniqstr_new (last_string); |
41141c56 | 248 | STRING_FREE; |
4cdb01db AD |
249 | return TYPE; |
250 | } | |
251 | ||
a706a1cc PE |
252 | "%%" { |
253 | static int percent_percent_count; | |
e9955c83 | 254 | if (++percent_percent_count == 2) |
a2bc9dbc | 255 | BEGIN SC_EPILOGUE; |
e9955c83 AD |
256 | return PERCENT_PERCENT; |
257 | } | |
258 | ||
a706a1cc | 259 | . { |
41141c56 | 260 | complain_at (*loc, _("invalid character: %s"), quote (yytext)); |
3f2d73f1 | 261 | } |
379f0ac8 PE |
262 | |
263 | <<EOF>> { | |
264 | loc->start = loc->end = scanner_cursor; | |
265 | yyterminate (); | |
266 | } | |
3f2d73f1 PE |
267 | } |
268 | ||
269 | ||
270 | /*-----------------------------------------------------------------. | |
271 | | Scanning after an identifier, checking whether a colon is next. | | |
272 | `-----------------------------------------------------------------*/ | |
273 | ||
274 | <SC_AFTER_IDENTIFIER> | |
275 | { | |
276 | ":" { | |
3f2d73f1 PE |
277 | *loc = id_loc; |
278 | BEGIN INITIAL; | |
279 | return ID_COLON; | |
280 | } | |
281 | . { | |
282 | scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); | |
283 | yyless (0); | |
284 | *loc = id_loc; | |
285 | BEGIN INITIAL; | |
286 | return ID; | |
287 | } | |
288 | <<EOF>> { | |
289 | *loc = id_loc; | |
290 | BEGIN INITIAL; | |
291 | return ID; | |
e9955c83 AD |
292 | } |
293 | } | |
294 | ||
295 | ||
d8d3f94a PE |
296 | /*---------------------------------------------------------------. |
297 | | Scanning a Yacc comment. The initial `/ *' is already eaten. | | |
298 | `---------------------------------------------------------------*/ | |
e9955c83 | 299 | |
d8d3f94a | 300 | <SC_YACC_COMMENT> |
e9955c83 | 301 | { |
3f2d73f1 | 302 | "*/" BEGIN context_state; |
a706a1cc | 303 | .|\n ; |
aa418041 | 304 | <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state; |
d8d3f94a PE |
305 | } |
306 | ||
307 | ||
308 | /*------------------------------------------------------------. | |
309 | | Scanning a C comment. The initial `/ *' is already eaten. | | |
310 | `------------------------------------------------------------*/ | |
311 | ||
312 | <SC_COMMENT> | |
313 | { | |
3f2d73f1 | 314 | "*"{splice}"/" STRING_GROW; BEGIN context_state; |
aa418041 | 315 | <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state; |
e9955c83 AD |
316 | } |
317 | ||
318 | ||
d8d3f94a PE |
319 | /*--------------------------------------------------------------. |
320 | | Scanning a line comment. The initial `//' is already eaten. | | |
321 | `--------------------------------------------------------------*/ | |
322 | ||
323 | <SC_LINE_COMMENT> | |
324 | { | |
3f2d73f1 | 325 | "\n" STRING_GROW; BEGIN context_state; |
41141c56 | 326 | {splice} STRING_GROW; |
3f2d73f1 | 327 | <<EOF>> BEGIN context_state; |
d8d3f94a PE |
328 | } |
329 | ||
330 | ||
4febdd96 PE |
331 | /*------------------------------------------------. |
332 | | Scanning a Bison string, including its escapes. | | |
333 | | The initial quote is already eaten. | | |
334 | `------------------------------------------------*/ | |
e9955c83 AD |
335 | |
336 | <SC_ESCAPED_STRING> | |
337 | { | |
47aee066 JD |
338 | "\""|"\n" { |
339 | if (yytext[0] == '\n') | |
340 | unexpected_newline (token_start, "\""); | |
341 | STRING_FINISH; | |
342 | loc->start = token_start; | |
343 | val->chars = last_string; | |
344 | BEGIN INITIAL; | |
345 | return STRING; | |
346 | } | |
347 | <<EOF>> { | |
348 | unexpected_eof (token_start, "\""); | |
41141c56 | 349 | STRING_FINISH; |
3f2d73f1 | 350 | loc->start = token_start; |
223ff46e | 351 | val->chars = last_string; |
a706a1cc | 352 | BEGIN INITIAL; |
e9955c83 AD |
353 | return STRING; |
354 | } | |
e9955c83 AD |
355 | } |
356 | ||
4febdd96 PE |
357 | /*----------------------------------------------------------. |
358 | | Scanning a Bison character literal, decoding its escapes. | | |
359 | | The initial quote is already eaten. | | |
360 | `----------------------------------------------------------*/ | |
e9955c83 AD |
361 | |
362 | <SC_ESCAPED_CHARACTER> | |
363 | { | |
47aee066 JD |
364 | "'"|"\n" { |
365 | if (yytext[0] == '\n') | |
366 | unexpected_newline (token_start, "'"); | |
41141c56 PE |
367 | STRING_GROW; |
368 | STRING_FINISH; | |
3f2d73f1 | 369 | loc->start = token_start; |
58d7a1a1 | 370 | val->character = last_string[1]; |
41141c56 | 371 | STRING_FREE; |
a706a1cc | 372 | BEGIN INITIAL; |
58d7a1a1 | 373 | return CHAR; |
e9955c83 | 374 | } |
47aee066 JD |
375 | <<EOF>> { |
376 | unexpected_eof (token_start, "'"); | |
377 | STRING_FINISH; | |
378 | loc->start = token_start; | |
379 | if (strlen(last_string) > 1) | |
380 | val->character = last_string[1]; | |
381 | else | |
382 | val->character = last_string[0]; | |
383 | STRING_FREE; | |
384 | BEGIN INITIAL; | |
385 | return CHAR; | |
386 | } | |
4febdd96 | 387 | } |
a706a1cc | 388 | |
4febdd96 PE |
389 | <SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING> |
390 | { | |
92ac3705 | 391 | \0 complain_at (*loc, _("invalid null character")); |
e9955c83 AD |
392 | } |
393 | ||
394 | ||
395 | /*----------------------------. | |
396 | | Decode escaped characters. | | |
397 | `----------------------------*/ | |
398 | ||
399 | <SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER> | |
400 | { | |
d8d3f94a | 401 | \\[0-7]{1,3} { |
4517da37 | 402 | unsigned long int c = strtoul (yytext + 1, NULL, 8); |
b1a4261e JD |
403 | if (!c || UCHAR_MAX < c) |
404 | complain_at (*loc, _("invalid number after \\-escape: %s"), | |
405 | yytext+1); | |
e9955c83 | 406 | else |
223ff46e | 407 | obstack_1grow (&obstack_for_string, c); |
e9955c83 AD |
408 | } |
409 | ||
6b0d38ab | 410 | \\x[0-9abcdefABCDEF]+ { |
4517da37 PE |
411 | verify (UCHAR_MAX < ULONG_MAX); |
412 | unsigned long int c = strtoul (yytext + 2, NULL, 16); | |
b1a4261e JD |
413 | if (!c || UCHAR_MAX < c) |
414 | complain_at (*loc, _("invalid number after \\-escape: %s"), | |
415 | yytext+1); | |
d8d3f94a | 416 | else |
223ff46e | 417 | obstack_1grow (&obstack_for_string, c); |
e9955c83 AD |
418 | } |
419 | ||
223ff46e PE |
420 | \\a obstack_1grow (&obstack_for_string, '\a'); |
421 | \\b obstack_1grow (&obstack_for_string, '\b'); | |
422 | \\f obstack_1grow (&obstack_for_string, '\f'); | |
423 | \\n obstack_1grow (&obstack_for_string, '\n'); | |
424 | \\r obstack_1grow (&obstack_for_string, '\r'); | |
425 | \\t obstack_1grow (&obstack_for_string, '\t'); | |
426 | \\v obstack_1grow (&obstack_for_string, '\v'); | |
412f8a59 PE |
427 | |
428 | /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */ | |
223ff46e | 429 | \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]); |
412f8a59 | 430 | |
6b0d38ab | 431 | \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} { |
d8d3f94a | 432 | int c = convert_ucn_to_byte (yytext); |
b1a4261e JD |
433 | if (c <= 0) |
434 | complain_at (*loc, _("invalid number after \\-escape: %s"), | |
435 | yytext+1); | |
d8d3f94a | 436 | else |
223ff46e | 437 | obstack_1grow (&obstack_for_string, c); |
d8d3f94a | 438 | } |
4f25ebb0 | 439 | \\(.|\n) { |
b1a4261e | 440 | char const *p = yytext + 1; |
321fe707 | 441 | /* Quote only if escaping won't make the character visible. */ |
bbbbe221 | 442 | if (isspace ((unsigned char) *p) && isprint ((unsigned char) *p)) |
321fe707 | 443 | p = quote (p); |
b1a4261e JD |
444 | else |
445 | p = quotearg_style_mem (escape_quoting_style, p, 1); | |
446 | complain_at (*loc, _("invalid character after \\-escape: %s"), p); | |
e9955c83 AD |
447 | } |
448 | } | |
449 | ||
4febdd96 PE |
450 | /*--------------------------------------------. |
451 | | Scanning user-code characters and strings. | | |
452 | `--------------------------------------------*/ | |
e9955c83 | 453 | |
4febdd96 PE |
454 | <SC_CHARACTER,SC_STRING> |
455 | { | |
e9071366 | 456 | {splice}|\\{splice}[^\n\[\]] STRING_GROW; |
4febdd96 | 457 | } |
e9955c83 AD |
458 | |
459 | <SC_CHARACTER> | |
460 | { | |
4febdd96 PE |
461 | "'" STRING_GROW; BEGIN context_state; |
462 | \n unexpected_newline (token_start, "'"); BEGIN context_state; | |
463 | <<EOF>> unexpected_eof (token_start, "'"); BEGIN context_state; | |
e9955c83 AD |
464 | } |
465 | ||
e9955c83 AD |
466 | <SC_STRING> |
467 | { | |
4febdd96 PE |
468 | "\"" STRING_GROW; BEGIN context_state; |
469 | \n unexpected_newline (token_start, "\""); BEGIN context_state; | |
470 | <<EOF>> unexpected_eof (token_start, "\""); BEGIN context_state; | |
e9955c83 AD |
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 | { | |
3f2d73f1 PE |
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 | } | |
e9955c83 AD |
503 | } |
504 | ||
505 | ||
624a35e2 | 506 | |
58d7a1a1 AD |
507 | /*-----------------------------------------------------------. |
508 | | Scanning some code in braces (actions). The initial "{" is | | |
509 | | already eaten. | | |
510 | `-----------------------------------------------------------*/ | |
e9955c83 AD |
511 | |
512 | <SC_BRACED_CODE> | |
513 | { | |
41141c56 PE |
514 | "{"|"<"{splice}"%" STRING_GROW; braces_level++; |
515 | "%"{splice}">" STRING_GROW; braces_level--; | |
e9955c83 | 516 | "}" { |
25522739 PE |
517 | obstack_1grow (&obstack_for_string, '}'); |
518 | ||
2346344a AD |
519 | --braces_level; |
520 | if (braces_level < 0) | |
e9955c83 | 521 | { |
41141c56 | 522 | STRING_FINISH; |
3f2d73f1 | 523 | loc->start = code_start; |
eb095650 | 524 | val->code = last_string; |
a706a1cc | 525 | BEGIN INITIAL; |
58d7a1a1 | 526 | return BRACED_CODE; |
e9955c83 AD |
527 | } |
528 | } | |
529 | ||
a706a1cc PE |
530 | /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly |
531 | (as `<' `<%'). */ | |
41141c56 | 532 | "<"{splice}"<" STRING_GROW; |
a706a1cc | 533 | |
47aee066 JD |
534 | <<EOF>> { |
535 | unexpected_eof (code_start, "}"); | |
536 | STRING_FINISH; | |
537 | loc->start = code_start; | |
eb095650 | 538 | val->code = last_string; |
47aee066 JD |
539 | BEGIN INITIAL; |
540 | return BRACED_CODE; | |
541 | } | |
e9955c83 AD |
542 | } |
543 | ||
544 | ||
545 | /*--------------------------------------------------------------. | |
546 | | Scanning some prologue: from "%{" (already scanned) to "%}". | | |
547 | `--------------------------------------------------------------*/ | |
548 | ||
549 | <SC_PROLOGUE> | |
550 | { | |
551 | "%}" { | |
41141c56 | 552 | STRING_FINISH; |
3f2d73f1 | 553 | loc->start = code_start; |
223ff46e | 554 | val->chars = last_string; |
a706a1cc | 555 | BEGIN INITIAL; |
e9955c83 AD |
556 | return PROLOGUE; |
557 | } | |
558 | ||
47aee066 JD |
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 | } | |
e9955c83 AD |
567 | } |
568 | ||
569 | ||
570 | /*---------------------------------------------------------------. | |
571 | | Scanning the epilogue (everything after the second "%%", which | | |
d8d3f94a | 572 | | has already been eaten). | |
e9955c83 AD |
573 | `---------------------------------------------------------------*/ |
574 | ||
575 | <SC_EPILOGUE> | |
576 | { | |
e9955c83 | 577 | <<EOF>> { |
41141c56 | 578 | STRING_FINISH; |
3f2d73f1 | 579 | loc->start = code_start; |
223ff46e | 580 | val->chars = last_string; |
a706a1cc | 581 | BEGIN INITIAL; |
e9955c83 AD |
582 | return EPILOGUE; |
583 | } | |
584 | } | |
585 | ||
586 | ||
4febdd96 PE |
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 | ||
e9955c83 AD |
594 | %% |
595 | ||
6c30d641 PE |
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 | ||
600 | static size_t | |
601 | no_cr_read (FILE *fp, char *buf, size_t size) | |
602 | { | |
a737b216 PE |
603 | size_t bytes_read = fread (buf, 1, size, fp); |
604 | if (bytes_read) | |
6c30d641 | 605 | { |
a737b216 | 606 | char *w = memchr (buf, '\r', bytes_read); |
6c30d641 PE |
607 | if (w) |
608 | { | |
609 | char const *r = ++w; | |
a737b216 | 610 | char const *lim = buf + bytes_read; |
6c30d641 PE |
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 | ||
a737b216 | 639 | return bytes_read; |
6c30d641 PE |
640 | } |
641 | ||
642 | ||
f25bfb75 | 643 | |
1452af69 PE |
644 | /*------------------------------------------------------. |
645 | | Scan NUMBER for a base-BASE integer at location LOC. | | |
646 | `------------------------------------------------------*/ | |
647 | ||
648 | static unsigned long int | |
649 | scan_integer (char const *number, int base, location loc) | |
650 | { | |
4517da37 PE |
651 | verify (INT_MAX < ULONG_MAX); |
652 | unsigned long int num = strtoul (number, NULL, base); | |
653 | ||
654 | if (INT_MAX < num) | |
1452af69 PE |
655 | { |
656 | complain_at (loc, _("integer out of range: %s"), quote (number)); | |
657 | num = INT_MAX; | |
658 | } | |
4517da37 | 659 | |
1452af69 PE |
660 | return num; |
661 | } | |
662 | ||
663 | ||
d8d3f94a PE |
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 | ||
670 | static int | |
671 | convert_ucn_to_byte (char const *ucn) | |
672 | { | |
4517da37 PE |
673 | verify (UCHAR_MAX <= INT_MAX); |
674 | unsigned long int code = strtoul (ucn + 2, NULL, 16); | |
d8d3f94a PE |
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 ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e) | |
686 | { | |
687 | /* A non-ASCII host. Use CODE to index into a table of the C | |
688 | basic execution character set, which is guaranteed to exist on | |
689 | all Standard C platforms. This table also includes '$', '@', | |
8e6ef483 | 690 | and '`', which are not in the basic execution character set but |
d8d3f94a PE |
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 | |
c4d720cd | 716 | |
d8d3f94a PE |
717 | return code; |
718 | } | |
719 | ||
720 | ||
900c5db5 AD |
721 | /*----------------------------------------------------------------. |
722 | | Handle `#line INT "FILE"'. ARGS has already skipped `#line '. | | |
723 | `----------------------------------------------------------------*/ | |
724 | ||
725 | static void | |
4517da37 | 726 | handle_syncline (char *args, location loc) |
900c5db5 | 727 | { |
4517da37 PE |
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 | } | |
e9071366 | 737 | current_file = uniqstr_new (file); |
0c8e079f | 738 | boundary_set (&scanner_cursor, current_file, lineno, 1); |
4517da37 PE |
739 | } |
740 | ||
741 | ||
4febdd96 PE |
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 | ||
748 | static void | |
749 | unexpected_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 | ||
3f2d73f1 PE |
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. | | |
3f2d73f1 | 761 | `------------------------------------------------------------------------*/ |
a706a1cc PE |
762 | |
763 | static void | |
aa418041 | 764 | unexpected_eof (boundary start, char const *token_end) |
a706a1cc | 765 | { |
4febdd96 PE |
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 | ||
774 | static void | |
775 | unexpected_newline (boundary start, char const *token_end) | |
776 | { | |
777 | unexpected_end (start, N_("missing `%s' at end of line"), token_end); | |
a706a1cc PE |
778 | } |
779 | ||
780 | ||
f25bfb75 AD |
781 | /*-------------------------. |
782 | | Initialize the scanner. | | |
783 | `-------------------------*/ | |
784 | ||
1d6412ad | 785 | void |
e9071366 | 786 | gram_scanner_initialize (void) |
1d6412ad | 787 | { |
223ff46e | 788 | obstack_init (&obstack_for_string); |
1d6412ad AD |
789 | } |
790 | ||
791 | ||
f25bfb75 AD |
792 | /*-----------------------------------------------. |
793 | | Free all the memory allocated to the scanner. | | |
794 | `-----------------------------------------------*/ | |
795 | ||
4cdb01db | 796 | void |
e9071366 | 797 | gram_scanner_free (void) |
4cdb01db | 798 | { |
223ff46e | 799 | obstack_free (&obstack_for_string, 0); |
536545f3 | 800 | /* Reclaim Flex's buffers. */ |
580b8926 | 801 | yylex_destroy (); |
4cdb01db | 802 | } |