| 1 | /* Bison Grammar Scanner -*- C -*- |
| 2 | |
| 3 | Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation, |
| 4 | Inc. |
| 5 | |
| 6 | This file is part of Bison, the GNU Compiler Compiler. |
| 7 | |
| 8 | This program is free software: you can redistribute it and/or modify |
| 9 | it under the terms of the GNU General Public License as published by |
| 10 | the Free Software Foundation, either version 3 of the License, or |
| 11 | (at your option) any later version. |
| 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 |
| 19 | along with this program. If not, see <http://www.gnu.org/licenses/>. */ |
| 20 | |
| 21 | %option debug nodefault nounput noyywrap never-interactive |
| 22 | %option prefix="gram_" outfile="lex.yy.c" |
| 23 | |
| 24 | %{ |
| 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 | |
| 30 | #define FLEX_PREFIX(Id) gram_ ## Id |
| 31 | #include "flex-scanner.h" |
| 32 | |
| 33 | #include "complain.h" |
| 34 | #include "files.h" |
| 35 | #include "gram.h" |
| 36 | #include "quotearg.h" |
| 37 | #include "reader.h" |
| 38 | #include "uniqstr.h" |
| 39 | |
| 40 | #include <mbswidth.h> |
| 41 | #include <quote.h> |
| 42 | |
| 43 | #include "scan-gram.h" |
| 44 | |
| 45 | #define YY_DECL GRAM_LEX_DECL |
| 46 | |
| 47 | #define YY_USER_INIT \ |
| 48 | code_start = scanner_cursor = loc->start; \ |
| 49 | |
| 50 | /* Location of scanner cursor. */ |
| 51 | static boundary scanner_cursor; |
| 52 | |
| 53 | #define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng); |
| 54 | |
| 55 | static size_t no_cr_read (FILE *, char *, size_t); |
| 56 | #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size)) |
| 57 | |
| 58 | /* A string representing the most recently saved token. */ |
| 59 | static char *last_string; |
| 60 | |
| 61 | void |
| 62 | gram_scanner_last_string_free (void) |
| 63 | { |
| 64 | STRING_FREE; |
| 65 | } |
| 66 | |
| 67 | static void handle_syncline (char *, location); |
| 68 | static unsigned long int scan_integer (char const *p, int base, location loc); |
| 69 | static int convert_ucn_to_byte (char const *hex_text); |
| 70 | static void unexpected_eof (boundary, char const *); |
| 71 | static void unexpected_newline (boundary, char const *); |
| 72 | |
| 73 | %} |
| 74 | /* A C-like comment in directives/rules. */ |
| 75 | %x SC_YACC_COMMENT |
| 76 | /* Strings and characters in directives/rules. */ |
| 77 | %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER |
| 78 | /* A identifier was just read in directives/rules. Special state |
| 79 | to capture the sequence `identifier :'. */ |
| 80 | %x SC_AFTER_IDENTIFIER |
| 81 | |
| 82 | /* Three types of user code: |
| 83 | - prologue (code between `%{' `%}' in the first section, before %%); |
| 84 | - actions, printers, union, etc, (between braced in the middle section); |
| 85 | - epilogue (everything after the second %%). */ |
| 86 | %x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE |
| 87 | /* C and C++ comments in code. */ |
| 88 | %x SC_COMMENT SC_LINE_COMMENT |
| 89 | /* Strings and characters in code. */ |
| 90 | %x SC_STRING SC_CHARACTER |
| 91 | |
| 92 | letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] |
| 93 | id {letter}({letter}|[0-9])* |
| 94 | directive %{letter}({letter}|[0-9]|-)* |
| 95 | int [0-9]+ |
| 96 | |
| 97 | /* POSIX says that a tag must be both an id and a C union member, but |
| 98 | historically almost any character is allowed in a tag. We disallow |
| 99 | NUL and newline, as this simplifies our implementation. */ |
| 100 | tag [^\0\n>]+ |
| 101 | |
| 102 | /* Zero or more instances of backslash-newline. Following GCC, allow |
| 103 | white space between the backslash and the newline. */ |
| 104 | splice (\\[ \f\t\v]*\n)* |
| 105 | |
| 106 | %% |
| 107 | %{ |
| 108 | /* Nesting level of the current code in braces. */ |
| 109 | int braces_level IF_LINT (= 0); |
| 110 | |
| 111 | /* Parent context state, when applicable. */ |
| 112 | int context_state IF_LINT (= 0); |
| 113 | |
| 114 | /* Location of most recent identifier, when applicable. */ |
| 115 | location id_loc IF_LINT (= empty_location); |
| 116 | |
| 117 | /* Where containing code started, when applicable. Its initial |
| 118 | value is relevant only when yylex is invoked in the SC_EPILOGUE |
| 119 | start condition. */ |
| 120 | boundary code_start = scanner_cursor; |
| 121 | |
| 122 | /* Where containing comment or string or character literal started, |
| 123 | when applicable. */ |
| 124 | boundary token_start IF_LINT (= scanner_cursor); |
| 125 | %} |
| 126 | |
| 127 | |
| 128 | /*-----------------------. |
| 129 | | Scanning white space. | |
| 130 | `-----------------------*/ |
| 131 | |
| 132 | <INITIAL,SC_AFTER_IDENTIFIER> |
| 133 | { |
| 134 | /* Comments and white space. */ |
| 135 | "," warn_at (*loc, _("stray `,' treated as white space")); |
| 136 | [ \f\n\t\v] | |
| 137 | "//".* ; |
| 138 | "/*" { |
| 139 | token_start = loc->start; |
| 140 | context_state = YY_START; |
| 141 | BEGIN SC_YACC_COMMENT; |
| 142 | } |
| 143 | |
| 144 | /* #line directives are not documented, and may be withdrawn or |
| 145 | modified in future versions of Bison. */ |
| 146 | ^"#line "{int}" \"".*"\"\n" { |
| 147 | handle_syncline (yytext + sizeof "#line " - 1, *loc); |
| 148 | } |
| 149 | } |
| 150 | |
| 151 | |
| 152 | /*----------------------------. |
| 153 | | Scanning Bison directives. | |
| 154 | `----------------------------*/ |
| 155 | <INITIAL> |
| 156 | { |
| 157 | "%binary" return PERCENT_NONASSOC; |
| 158 | "%code" return PERCENT_CODE; |
| 159 | "%debug" return PERCENT_DEBUG; |
| 160 | "%default"[-_]"prec" return PERCENT_DEFAULT_PREC; |
| 161 | "%define" return PERCENT_DEFINE; |
| 162 | "%defines" return PERCENT_DEFINES; |
| 163 | "%destructor" return PERCENT_DESTRUCTOR; |
| 164 | "%dprec" return PERCENT_DPREC; |
| 165 | "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE; |
| 166 | "%expect" return PERCENT_EXPECT; |
| 167 | "%expect"[-_]"rr" return PERCENT_EXPECT_RR; |
| 168 | "%file-prefix" return PERCENT_FILE_PREFIX; |
| 169 | "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC; |
| 170 | "%initial-action" return PERCENT_INITIAL_ACTION; |
| 171 | "%glr-parser" return PERCENT_GLR_PARSER; |
| 172 | "%language" return PERCENT_LANGUAGE; |
| 173 | "%left" return PERCENT_LEFT; |
| 174 | "%lex-param" return PERCENT_LEX_PARAM; |
| 175 | "%locations" return PERCENT_LOCATIONS; |
| 176 | "%merge" return PERCENT_MERGE; |
| 177 | "%name"[-_]"prefix" return PERCENT_NAME_PREFIX; |
| 178 | "%no"[-_]"default"[-_]"prec" return PERCENT_NO_DEFAULT_PREC; |
| 179 | "%no"[-_]"lines" return PERCENT_NO_LINES; |
| 180 | "%nonassoc" return PERCENT_NONASSOC; |
| 181 | "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER; |
| 182 | "%nterm" return PERCENT_NTERM; |
| 183 | "%output" return PERCENT_OUTPUT; |
| 184 | "%parse-param" return PERCENT_PARSE_PARAM; |
| 185 | "%prec" return PERCENT_PREC; |
| 186 | "%printer" return PERCENT_PRINTER; |
| 187 | "%pure"[-_]"parser" return PERCENT_PURE_PARSER; |
| 188 | "%require" return PERCENT_REQUIRE; |
| 189 | "%right" return PERCENT_RIGHT; |
| 190 | "%skeleton" return PERCENT_SKELETON; |
| 191 | "%start" return PERCENT_START; |
| 192 | "%term" return PERCENT_TOKEN; |
| 193 | "%token" return PERCENT_TOKEN; |
| 194 | "%token"[-_]"table" return PERCENT_TOKEN_TABLE; |
| 195 | "%type" return PERCENT_TYPE; |
| 196 | "%union" return PERCENT_UNION; |
| 197 | "%verbose" return PERCENT_VERBOSE; |
| 198 | "%yacc" return PERCENT_YACC; |
| 199 | |
| 200 | {directive} { |
| 201 | complain_at (*loc, _("invalid directive: %s"), quote (yytext)); |
| 202 | } |
| 203 | |
| 204 | "=" return EQUAL; |
| 205 | "|" return PIPE; |
| 206 | ";" return SEMICOLON; |
| 207 | "<*>" return TYPE_TAG_ANY; |
| 208 | "<>" return TYPE_TAG_NONE; |
| 209 | |
| 210 | {id} { |
| 211 | val->uniqstr = uniqstr_new (yytext); |
| 212 | id_loc = *loc; |
| 213 | BEGIN SC_AFTER_IDENTIFIER; |
| 214 | } |
| 215 | |
| 216 | {int} { |
| 217 | val->integer = scan_integer (yytext, 10, *loc); |
| 218 | return INT; |
| 219 | } |
| 220 | 0[xX][0-9abcdefABCDEF]+ { |
| 221 | val->integer = scan_integer (yytext, 16, *loc); |
| 222 | return INT; |
| 223 | } |
| 224 | |
| 225 | /* Characters. We don't check there is only one. */ |
| 226 | "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; |
| 227 | |
| 228 | /* Strings. */ |
| 229 | "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING; |
| 230 | |
| 231 | /* Prologue. */ |
| 232 | "%{" code_start = loc->start; BEGIN SC_PROLOGUE; |
| 233 | |
| 234 | /* Code in between braces. */ |
| 235 | "{" { |
| 236 | STRING_GROW; |
| 237 | braces_level = 0; |
| 238 | code_start = loc->start; |
| 239 | BEGIN SC_BRACED_CODE; |
| 240 | } |
| 241 | |
| 242 | /* A type. */ |
| 243 | "<"{tag}">" { |
| 244 | obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2); |
| 245 | STRING_FINISH; |
| 246 | val->uniqstr = uniqstr_new (last_string); |
| 247 | STRING_FREE; |
| 248 | return TYPE; |
| 249 | } |
| 250 | |
| 251 | "%%" { |
| 252 | static int percent_percent_count; |
| 253 | if (++percent_percent_count == 2) |
| 254 | BEGIN SC_EPILOGUE; |
| 255 | return PERCENT_PERCENT; |
| 256 | } |
| 257 | |
| 258 | . { |
| 259 | complain_at (*loc, _("invalid character: %s"), quote (yytext)); |
| 260 | } |
| 261 | |
| 262 | <<EOF>> { |
| 263 | loc->start = loc->end = scanner_cursor; |
| 264 | yyterminate (); |
| 265 | } |
| 266 | } |
| 267 | |
| 268 | |
| 269 | /*-----------------------------------------------------------------. |
| 270 | | Scanning after an identifier, checking whether a colon is next. | |
| 271 | `-----------------------------------------------------------------*/ |
| 272 | |
| 273 | <SC_AFTER_IDENTIFIER> |
| 274 | { |
| 275 | ":" { |
| 276 | *loc = id_loc; |
| 277 | BEGIN INITIAL; |
| 278 | return ID_COLON; |
| 279 | } |
| 280 | . { |
| 281 | scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); |
| 282 | yyless (0); |
| 283 | *loc = id_loc; |
| 284 | BEGIN INITIAL; |
| 285 | return ID; |
| 286 | } |
| 287 | <<EOF>> { |
| 288 | *loc = id_loc; |
| 289 | BEGIN INITIAL; |
| 290 | return ID; |
| 291 | } |
| 292 | } |
| 293 | |
| 294 | |
| 295 | /*---------------------------------------------------------------. |
| 296 | | Scanning a Yacc comment. The initial `/ *' is already eaten. | |
| 297 | `---------------------------------------------------------------*/ |
| 298 | |
| 299 | <SC_YACC_COMMENT> |
| 300 | { |
| 301 | "*/" BEGIN context_state; |
| 302 | .|\n ; |
| 303 | <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state; |
| 304 | } |
| 305 | |
| 306 | |
| 307 | /*------------------------------------------------------------. |
| 308 | | Scanning a C comment. The initial `/ *' is already eaten. | |
| 309 | `------------------------------------------------------------*/ |
| 310 | |
| 311 | <SC_COMMENT> |
| 312 | { |
| 313 | "*"{splice}"/" STRING_GROW; BEGIN context_state; |
| 314 | <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state; |
| 315 | } |
| 316 | |
| 317 | |
| 318 | /*--------------------------------------------------------------. |
| 319 | | Scanning a line comment. The initial `//' is already eaten. | |
| 320 | `--------------------------------------------------------------*/ |
| 321 | |
| 322 | <SC_LINE_COMMENT> |
| 323 | { |
| 324 | "\n" STRING_GROW; BEGIN context_state; |
| 325 | {splice} STRING_GROW; |
| 326 | <<EOF>> BEGIN context_state; |
| 327 | } |
| 328 | |
| 329 | |
| 330 | /*------------------------------------------------. |
| 331 | | Scanning a Bison string, including its escapes. | |
| 332 | | The initial quote is already eaten. | |
| 333 | `------------------------------------------------*/ |
| 334 | |
| 335 | <SC_ESCAPED_STRING> |
| 336 | { |
| 337 | "\""|"\n" { |
| 338 | if (yytext[0] == '\n') |
| 339 | unexpected_newline (token_start, "\""); |
| 340 | STRING_FINISH; |
| 341 | loc->start = token_start; |
| 342 | val->chars = last_string; |
| 343 | BEGIN INITIAL; |
| 344 | return STRING; |
| 345 | } |
| 346 | <<EOF>> { |
| 347 | unexpected_eof (token_start, "\""); |
| 348 | STRING_FINISH; |
| 349 | loc->start = token_start; |
| 350 | val->chars = last_string; |
| 351 | BEGIN INITIAL; |
| 352 | return STRING; |
| 353 | } |
| 354 | } |
| 355 | |
| 356 | /*----------------------------------------------------------. |
| 357 | | Scanning a Bison character literal, decoding its escapes. | |
| 358 | | The initial quote is already eaten. | |
| 359 | `----------------------------------------------------------*/ |
| 360 | |
| 361 | <SC_ESCAPED_CHARACTER> |
| 362 | { |
| 363 | "'"|"\n" { |
| 364 | if (yytext[0] == '\n') |
| 365 | unexpected_newline (token_start, "'"); |
| 366 | STRING_GROW; |
| 367 | STRING_FINISH; |
| 368 | loc->start = token_start; |
| 369 | val->character = last_string[1]; |
| 370 | STRING_FREE; |
| 371 | BEGIN INITIAL; |
| 372 | return CHAR; |
| 373 | } |
| 374 | <<EOF>> { |
| 375 | unexpected_eof (token_start, "'"); |
| 376 | STRING_FINISH; |
| 377 | loc->start = token_start; |
| 378 | if (strlen(last_string) > 1) |
| 379 | val->character = last_string[1]; |
| 380 | else |
| 381 | val->character = last_string[0]; |
| 382 | STRING_FREE; |
| 383 | BEGIN INITIAL; |
| 384 | return CHAR; |
| 385 | } |
| 386 | } |
| 387 | |
| 388 | <SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING> |
| 389 | { |
| 390 | \0 complain_at (*loc, _("invalid null character")); |
| 391 | } |
| 392 | |
| 393 | |
| 394 | /*----------------------------. |
| 395 | | Decode escaped characters. | |
| 396 | `----------------------------*/ |
| 397 | |
| 398 | <SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER> |
| 399 | { |
| 400 | \\[0-7]{1,3} { |
| 401 | unsigned long int c = strtoul (yytext + 1, NULL, 8); |
| 402 | if (UCHAR_MAX < c) |
| 403 | complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext)); |
| 404 | else if (! c) |
| 405 | complain_at (*loc, _("invalid null character: %s"), quote (yytext)); |
| 406 | else |
| 407 | obstack_1grow (&obstack_for_string, c); |
| 408 | } |
| 409 | |
| 410 | \\x[0-9abcdefABCDEF]+ { |
| 411 | verify (UCHAR_MAX < ULONG_MAX); |
| 412 | unsigned long int c = strtoul (yytext + 2, NULL, 16); |
| 413 | if (UCHAR_MAX < c) |
| 414 | complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext)); |
| 415 | else if (! c) |
| 416 | complain_at (*loc, _("invalid null character: %s"), quote (yytext)); |
| 417 | else |
| 418 | obstack_1grow (&obstack_for_string, c); |
| 419 | } |
| 420 | |
| 421 | \\a obstack_1grow (&obstack_for_string, '\a'); |
| 422 | \\b obstack_1grow (&obstack_for_string, '\b'); |
| 423 | \\f obstack_1grow (&obstack_for_string, '\f'); |
| 424 | \\n obstack_1grow (&obstack_for_string, '\n'); |
| 425 | \\r obstack_1grow (&obstack_for_string, '\r'); |
| 426 | \\t obstack_1grow (&obstack_for_string, '\t'); |
| 427 | \\v obstack_1grow (&obstack_for_string, '\v'); |
| 428 | |
| 429 | /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */ |
| 430 | \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]); |
| 431 | |
| 432 | \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} { |
| 433 | int c = convert_ucn_to_byte (yytext); |
| 434 | if (c < 0) |
| 435 | complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext)); |
| 436 | else if (! c) |
| 437 | complain_at (*loc, _("invalid null character: %s"), quote (yytext)); |
| 438 | else |
| 439 | obstack_1grow (&obstack_for_string, c); |
| 440 | } |
| 441 | \\(.|\n) { |
| 442 | complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext)); |
| 443 | STRING_GROW; |
| 444 | } |
| 445 | } |
| 446 | |
| 447 | /*--------------------------------------------. |
| 448 | | Scanning user-code characters and strings. | |
| 449 | `--------------------------------------------*/ |
| 450 | |
| 451 | <SC_CHARACTER,SC_STRING> |
| 452 | { |
| 453 | {splice}|\\{splice}[^\n\[\]] STRING_GROW; |
| 454 | } |
| 455 | |
| 456 | <SC_CHARACTER> |
| 457 | { |
| 458 | "'" STRING_GROW; BEGIN context_state; |
| 459 | \n unexpected_newline (token_start, "'"); BEGIN context_state; |
| 460 | <<EOF>> unexpected_eof (token_start, "'"); BEGIN context_state; |
| 461 | } |
| 462 | |
| 463 | <SC_STRING> |
| 464 | { |
| 465 | "\"" STRING_GROW; BEGIN context_state; |
| 466 | \n unexpected_newline (token_start, "\""); BEGIN context_state; |
| 467 | <<EOF>> unexpected_eof (token_start, "\""); BEGIN context_state; |
| 468 | } |
| 469 | |
| 470 | |
| 471 | /*---------------------------------------------------. |
| 472 | | Strings, comments etc. can be found in user code. | |
| 473 | `---------------------------------------------------*/ |
| 474 | |
| 475 | <SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE> |
| 476 | { |
| 477 | "'" { |
| 478 | STRING_GROW; |
| 479 | context_state = YY_START; |
| 480 | token_start = loc->start; |
| 481 | BEGIN SC_CHARACTER; |
| 482 | } |
| 483 | "\"" { |
| 484 | STRING_GROW; |
| 485 | context_state = YY_START; |
| 486 | token_start = loc->start; |
| 487 | BEGIN SC_STRING; |
| 488 | } |
| 489 | "/"{splice}"*" { |
| 490 | STRING_GROW; |
| 491 | context_state = YY_START; |
| 492 | token_start = loc->start; |
| 493 | BEGIN SC_COMMENT; |
| 494 | } |
| 495 | "/"{splice}"/" { |
| 496 | STRING_GROW; |
| 497 | context_state = YY_START; |
| 498 | BEGIN SC_LINE_COMMENT; |
| 499 | } |
| 500 | } |
| 501 | |
| 502 | |
| 503 | |
| 504 | /*-----------------------------------------------------------. |
| 505 | | Scanning some code in braces (actions). The initial "{" is | |
| 506 | | already eaten. | |
| 507 | `-----------------------------------------------------------*/ |
| 508 | |
| 509 | <SC_BRACED_CODE> |
| 510 | { |
| 511 | "{"|"<"{splice}"%" STRING_GROW; braces_level++; |
| 512 | "%"{splice}">" STRING_GROW; braces_level--; |
| 513 | "}" { |
| 514 | obstack_1grow (&obstack_for_string, '}'); |
| 515 | |
| 516 | --braces_level; |
| 517 | if (braces_level < 0) |
| 518 | { |
| 519 | STRING_FINISH; |
| 520 | loc->start = code_start; |
| 521 | val->code = last_string; |
| 522 | BEGIN INITIAL; |
| 523 | return BRACED_CODE; |
| 524 | } |
| 525 | } |
| 526 | |
| 527 | /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly |
| 528 | (as `<' `<%'). */ |
| 529 | "<"{splice}"<" STRING_GROW; |
| 530 | |
| 531 | <<EOF>> { |
| 532 | unexpected_eof (code_start, "}"); |
| 533 | STRING_FINISH; |
| 534 | loc->start = code_start; |
| 535 | val->code = last_string; |
| 536 | BEGIN INITIAL; |
| 537 | return BRACED_CODE; |
| 538 | } |
| 539 | } |
| 540 | |
| 541 | |
| 542 | /*--------------------------------------------------------------. |
| 543 | | Scanning some prologue: from "%{" (already scanned) to "%}". | |
| 544 | `--------------------------------------------------------------*/ |
| 545 | |
| 546 | <SC_PROLOGUE> |
| 547 | { |
| 548 | "%}" { |
| 549 | STRING_FINISH; |
| 550 | loc->start = code_start; |
| 551 | val->chars = last_string; |
| 552 | BEGIN INITIAL; |
| 553 | return PROLOGUE; |
| 554 | } |
| 555 | |
| 556 | <<EOF>> { |
| 557 | unexpected_eof (code_start, "%}"); |
| 558 | STRING_FINISH; |
| 559 | loc->start = code_start; |
| 560 | val->chars = last_string; |
| 561 | BEGIN INITIAL; |
| 562 | return PROLOGUE; |
| 563 | } |
| 564 | } |
| 565 | |
| 566 | |
| 567 | /*---------------------------------------------------------------. |
| 568 | | Scanning the epilogue (everything after the second "%%", which | |
| 569 | | has already been eaten). | |
| 570 | `---------------------------------------------------------------*/ |
| 571 | |
| 572 | <SC_EPILOGUE> |
| 573 | { |
| 574 | <<EOF>> { |
| 575 | STRING_FINISH; |
| 576 | loc->start = code_start; |
| 577 | val->chars = last_string; |
| 578 | BEGIN INITIAL; |
| 579 | return EPILOGUE; |
| 580 | } |
| 581 | } |
| 582 | |
| 583 | |
| 584 | /*-----------------------------------------------------. |
| 585 | | By default, grow the string obstack with the input. | |
| 586 | `-----------------------------------------------------*/ |
| 587 | |
| 588 | <SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>. | |
| 589 | <SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n STRING_GROW; |
| 590 | |
| 591 | %% |
| 592 | |
| 593 | /* Read bytes from FP into buffer BUF of size SIZE. Return the |
| 594 | number of bytes read. Remove '\r' from input, treating \r\n |
| 595 | and isolated \r as \n. */ |
| 596 | |
| 597 | static size_t |
| 598 | no_cr_read (FILE *fp, char *buf, size_t size) |
| 599 | { |
| 600 | size_t bytes_read = fread (buf, 1, size, fp); |
| 601 | if (bytes_read) |
| 602 | { |
| 603 | char *w = memchr (buf, '\r', bytes_read); |
| 604 | if (w) |
| 605 | { |
| 606 | char const *r = ++w; |
| 607 | char const *lim = buf + bytes_read; |
| 608 | |
| 609 | for (;;) |
| 610 | { |
| 611 | /* Found an '\r'. Treat it like '\n', but ignore any |
| 612 | '\n' that immediately follows. */ |
| 613 | w[-1] = '\n'; |
| 614 | if (r == lim) |
| 615 | { |
| 616 | int ch = getc (fp); |
| 617 | if (ch != '\n' && ungetc (ch, fp) != ch) |
| 618 | break; |
| 619 | } |
| 620 | else if (*r == '\n') |
| 621 | r++; |
| 622 | |
| 623 | /* Copy until the next '\r'. */ |
| 624 | do |
| 625 | { |
| 626 | if (r == lim) |
| 627 | return w - buf; |
| 628 | } |
| 629 | while ((*w++ = *r++) != '\r'); |
| 630 | } |
| 631 | |
| 632 | return w - buf; |
| 633 | } |
| 634 | } |
| 635 | |
| 636 | return bytes_read; |
| 637 | } |
| 638 | |
| 639 | |
| 640 | |
| 641 | /*------------------------------------------------------. |
| 642 | | Scan NUMBER for a base-BASE integer at location LOC. | |
| 643 | `------------------------------------------------------*/ |
| 644 | |
| 645 | static unsigned long int |
| 646 | scan_integer (char const *number, int base, location loc) |
| 647 | { |
| 648 | verify (INT_MAX < ULONG_MAX); |
| 649 | unsigned long int num = strtoul (number, NULL, base); |
| 650 | |
| 651 | if (INT_MAX < num) |
| 652 | { |
| 653 | complain_at (loc, _("integer out of range: %s"), quote (number)); |
| 654 | num = INT_MAX; |
| 655 | } |
| 656 | |
| 657 | return num; |
| 658 | } |
| 659 | |
| 660 | |
| 661 | /*------------------------------------------------------------------. |
| 662 | | Convert universal character name UCN to a single-byte character, | |
| 663 | | and return that character. Return -1 if UCN does not correspond | |
| 664 | | to a single-byte character. | |
| 665 | `------------------------------------------------------------------*/ |
| 666 | |
| 667 | static int |
| 668 | convert_ucn_to_byte (char const *ucn) |
| 669 | { |
| 670 | verify (UCHAR_MAX <= INT_MAX); |
| 671 | unsigned long int code = strtoul (ucn + 2, NULL, 16); |
| 672 | |
| 673 | /* FIXME: Currently we assume Unicode-compatible unibyte characters |
| 674 | on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On |
| 675 | non-ASCII hosts we support only the portable C character set. |
| 676 | These limitations should be removed once we add support for |
| 677 | multibyte characters. */ |
| 678 | |
| 679 | if (UCHAR_MAX < code) |
| 680 | return -1; |
| 681 | |
| 682 | #if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e) |
| 683 | { |
| 684 | /* A non-ASCII host. Use CODE to index into a table of the C |
| 685 | basic execution character set, which is guaranteed to exist on |
| 686 | all Standard C platforms. This table also includes '$', '@', |
| 687 | and '`', which are not in the basic execution character set but |
| 688 | which are unibyte characters on all the platforms that we know |
| 689 | about. */ |
| 690 | static signed char const table[] = |
| 691 | { |
| 692 | '\0', -1, -1, -1, -1, -1, -1, '\a', |
| 693 | '\b', '\t', '\n', '\v', '\f', '\r', -1, -1, |
| 694 | -1, -1, -1, -1, -1, -1, -1, -1, |
| 695 | -1, -1, -1, -1, -1, -1, -1, -1, |
| 696 | ' ', '!', '"', '#', '$', '%', '&', '\'', |
| 697 | '(', ')', '*', '+', ',', '-', '.', '/', |
| 698 | '0', '1', '2', '3', '4', '5', '6', '7', |
| 699 | '8', '9', ':', ';', '<', '=', '>', '?', |
| 700 | '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G', |
| 701 | 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', |
| 702 | 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', |
| 703 | 'X', 'Y', 'Z', '[', '\\', ']', '^', '_', |
| 704 | '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', |
| 705 | 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', |
| 706 | 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', |
| 707 | 'x', 'y', 'z', '{', '|', '}', '~' |
| 708 | }; |
| 709 | |
| 710 | code = code < sizeof table ? table[code] : -1; |
| 711 | } |
| 712 | #endif |
| 713 | |
| 714 | return code; |
| 715 | } |
| 716 | |
| 717 | |
| 718 | /*----------------------------------------------------------------. |
| 719 | | Handle `#line INT "FILE"'. ARGS has already skipped `#line '. | |
| 720 | `----------------------------------------------------------------*/ |
| 721 | |
| 722 | static void |
| 723 | handle_syncline (char *args, location loc) |
| 724 | { |
| 725 | char *after_num; |
| 726 | unsigned long int lineno = strtoul (args, &after_num, 10); |
| 727 | char *file = strchr (after_num, '"') + 1; |
| 728 | *strchr (file, '"') = '\0'; |
| 729 | if (INT_MAX <= lineno) |
| 730 | { |
| 731 | warn_at (loc, _("line number overflow")); |
| 732 | lineno = INT_MAX; |
| 733 | } |
| 734 | current_file = uniqstr_new (file); |
| 735 | boundary_set (&scanner_cursor, current_file, lineno, 1); |
| 736 | } |
| 737 | |
| 738 | |
| 739 | /*----------------------------------------------------------------. |
| 740 | | For a token or comment starting at START, report message MSGID, | |
| 741 | | which should say that an end marker was found before | |
| 742 | | the expected TOKEN_END. | |
| 743 | `----------------------------------------------------------------*/ |
| 744 | |
| 745 | static void |
| 746 | unexpected_end (boundary start, char const *msgid, char const *token_end) |
| 747 | { |
| 748 | location loc; |
| 749 | loc.start = start; |
| 750 | loc.end = scanner_cursor; |
| 751 | complain_at (loc, _(msgid), token_end); |
| 752 | } |
| 753 | |
| 754 | |
| 755 | /*------------------------------------------------------------------------. |
| 756 | | Report an unexpected EOF in a token or comment starting at START. | |
| 757 | | An end of file was encountered and the expected TOKEN_END was missing. | |
| 758 | `------------------------------------------------------------------------*/ |
| 759 | |
| 760 | static void |
| 761 | unexpected_eof (boundary start, char const *token_end) |
| 762 | { |
| 763 | unexpected_end (start, N_("missing `%s' at end of file"), token_end); |
| 764 | } |
| 765 | |
| 766 | |
| 767 | /*----------------------------------------. |
| 768 | | Likewise, but for unexpected newlines. | |
| 769 | `----------------------------------------*/ |
| 770 | |
| 771 | static void |
| 772 | unexpected_newline (boundary start, char const *token_end) |
| 773 | { |
| 774 | unexpected_end (start, N_("missing `%s' at end of line"), token_end); |
| 775 | } |
| 776 | |
| 777 | |
| 778 | /*-------------------------. |
| 779 | | Initialize the scanner. | |
| 780 | `-------------------------*/ |
| 781 | |
| 782 | void |
| 783 | gram_scanner_initialize (void) |
| 784 | { |
| 785 | obstack_init (&obstack_for_string); |
| 786 | } |
| 787 | |
| 788 | |
| 789 | /*-----------------------------------------------. |
| 790 | | Free all the memory allocated to the scanner. | |
| 791 | `-----------------------------------------------*/ |
| 792 | |
| 793 | void |
| 794 | gram_scanner_free (void) |
| 795 | { |
| 796 | obstack_free (&obstack_for_string, 0); |
| 797 | /* Reclaim Flex's buffers. */ |
| 798 | yylex_destroy (); |
| 799 | } |