Mercurial > projects > ldc
annotate dmd/lexer.c @ 883:b52d5de7783f
GC defines and linkage changes.
| author | Christian Kamm <kamm incasoftware de> |
|---|---|
| date | Thu, 08 Jan 2009 18:20:02 +0100 |
| parents | aa953cc960b6 |
| children | b30fe7e1dbb9 |
| rev | line source |
|---|---|
| 159 | 1 |
| 2 // Compiler implementation of the D programming language | |
| 3 // Copyright (c) 1999-2008 by Digital Mars | |
| 4 // All Rights Reserved | |
| 5 // written by Walter Bright | |
| 6 // http://www.digitalmars.com | |
| 7 // License for redistribution is by either the Artistic License | |
| 8 // in artistic.txt, or the GNU General Public License in gnu.txt. | |
| 9 // See the included readme.txt for details. | |
| 10 | |
|
872
aa953cc960b6
Apply BlueZeniX's patch for OpenSolaris compatibility. Fixes #158.
Christian Kamm <kamm incasoftware de>
parents:
846
diff
changeset
|
11 #define __C99FEATURES__ 1 |
|
aa953cc960b6
Apply BlueZeniX's patch for OpenSolaris compatibility. Fixes #158.
Christian Kamm <kamm incasoftware de>
parents:
846
diff
changeset
|
12 |
| 159 | 13 /* Lexical Analyzer */ |
| 14 | |
| 15 #include <stdio.h> | |
| 16 #include <string.h> | |
| 17 #include <ctype.h> | |
| 18 #include <stdarg.h> | |
| 19 #include <errno.h> | |
| 20 #include <wchar.h> | |
| 21 #include <stdlib.h> | |
| 22 #include <assert.h> | |
| 23 #include <sys/time.h> | |
|
697
25a32766ed57
Missing include added.
Christian Kamm <kamm incasoftware de>
parents:
696
diff
changeset
|
24 #include <math.h> |
| 159 | 25 |
| 26 #ifdef IN_GCC | |
| 27 | |
| 28 #include <time.h> | |
| 29 #include "mem.h" | |
| 30 | |
| 31 #else | |
| 32 | |
| 33 #if __GNUC__ | |
| 34 #include <time.h> | |
| 35 #endif | |
| 36 | |
| 37 #if IN_LLVM | |
| 38 #include "mem.h" | |
| 39 #elif _WIN32 | |
| 40 #include "..\root\mem.h" | |
| 41 #else | |
| 42 #include "../root/mem.h" | |
| 43 #endif | |
| 44 #endif | |
| 45 | |
| 46 #include "stringtable.h" | |
| 47 | |
| 48 #include "lexer.h" | |
| 49 #include "utf.h" | |
| 50 #include "identifier.h" | |
| 51 #include "id.h" | |
| 52 #include "module.h" | |
| 53 | |
| 54 #if _WIN32 && __DMC__ | |
| 55 // from \dm\src\include\setlocal.h | |
| 56 extern "C" char * __cdecl __locale_decpoint; | |
| 57 #endif | |
| 58 | |
| 59 extern int HtmlNamedEntity(unsigned char *p, int length); | |
| 60 | |
| 61 #define LS 0x2028 // UTF line separator | |
| 62 #define PS 0x2029 // UTF paragraph separator | |
| 63 | |
| 64 /******************************************** | |
| 65 * Do our own char maps | |
| 66 */ | |
| 67 | |
| 68 static unsigned char cmtable[256]; | |
| 69 | |
| 70 const int CMoctal = 0x1; | |
| 71 const int CMhex = 0x2; | |
| 72 const int CMidchar = 0x4; | |
| 73 | |
| 74 inline unsigned char isoctal (unsigned char c) { return cmtable[c] & CMoctal; } | |
| 75 inline unsigned char ishex (unsigned char c) { return cmtable[c] & CMhex; } | |
| 76 inline unsigned char isidchar(unsigned char c) { return cmtable[c] & CMidchar; } | |
| 77 | |
| 78 static void cmtable_init() | |
| 79 { | |
| 80 for (unsigned c = 0; c < sizeof(cmtable) / sizeof(cmtable[0]); c++) | |
| 81 { | |
| 82 if ('0' <= c && c <= '7') | |
| 83 cmtable[c] |= CMoctal; | |
| 84 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 85 cmtable[c] |= CMhex; | |
| 86 if (isalnum(c) || c == '_') | |
| 87 cmtable[c] |= CMidchar; | |
| 88 } | |
| 89 } | |
| 90 | |
| 91 | |
| 92 /************************* Token **********************************************/ | |
| 93 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
94 const char *Token::tochars[TOKMAX]; |
| 159 | 95 |
| 96 void *Token::operator new(size_t size) | |
| 97 { Token *t; | |
| 98 | |
| 99 if (Lexer::freelist) | |
| 100 { | |
| 101 t = Lexer::freelist; | |
| 102 Lexer::freelist = t->next; | |
| 103 return t; | |
| 104 } | |
| 105 | |
| 106 return ::operator new(size); | |
| 107 } | |
| 108 | |
| 109 #ifdef DEBUG | |
| 110 void Token::print() | |
| 111 { | |
| 112 fprintf(stdmsg, "%s\n", toChars()); | |
| 113 } | |
| 114 #endif | |
| 115 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
116 const char *Token::toChars() |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
117 { const char *p; |
| 159 | 118 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 119 | |
| 120 p = buffer; | |
| 121 switch (value) | |
| 122 { | |
| 123 case TOKint32v: | |
| 124 #if IN_GCC | |
| 125 sprintf(buffer,"%d",(d_int32)int64value); | |
| 126 #else | |
| 127 sprintf(buffer,"%d",int32value); | |
| 128 #endif | |
| 129 break; | |
| 130 | |
| 131 case TOKuns32v: | |
| 132 case TOKcharv: | |
| 133 case TOKwcharv: | |
| 134 case TOKdcharv: | |
| 135 #if IN_GCC | |
| 136 sprintf(buffer,"%uU",(d_uns32)uns64value); | |
| 137 #else | |
| 138 sprintf(buffer,"%uU",uns32value); | |
| 139 #endif | |
| 140 break; | |
| 141 | |
| 142 case TOKint64v: | |
|
794
661384d6a936
Fix warnings on x86-64. By fvbommel.
Christian Kamm <kamm incasoftware de>
parents:
717
diff
changeset
|
143 sprintf(buffer,"%lldL",(long long)int64value); |
| 159 | 144 break; |
| 145 | |
| 146 case TOKuns64v: | |
|
794
661384d6a936
Fix warnings on x86-64. By fvbommel.
Christian Kamm <kamm incasoftware de>
parents:
717
diff
changeset
|
147 sprintf(buffer,"%lluUL",(unsigned long long)uns64value); |
| 159 | 148 break; |
| 149 | |
| 150 #if IN_GCC | |
| 151 case TOKfloat32v: | |
| 152 case TOKfloat64v: | |
| 153 case TOKfloat80v: | |
| 154 float80value.format(buffer, sizeof(buffer)); | |
| 155 break; | |
| 156 case TOKimaginary32v: | |
| 157 case TOKimaginary64v: | |
| 158 case TOKimaginary80v: | |
| 159 float80value.format(buffer, sizeof(buffer)); | |
| 160 // %% buffer | |
| 161 strcat(buffer, "i"); | |
| 162 break; | |
| 163 #else | |
| 164 case TOKfloat32v: | |
| 165 sprintf(buffer,"%Lgf", float80value); | |
| 166 break; | |
| 167 | |
| 168 case TOKfloat64v: | |
| 169 sprintf(buffer,"%Lg", float80value); | |
| 170 break; | |
| 171 | |
| 172 case TOKfloat80v: | |
| 173 sprintf(buffer,"%LgL", float80value); | |
| 174 break; | |
| 175 | |
| 176 case TOKimaginary32v: | |
| 177 sprintf(buffer,"%Lgfi", float80value); | |
| 178 break; | |
| 179 | |
| 180 case TOKimaginary64v: | |
| 181 sprintf(buffer,"%Lgi", float80value); | |
| 182 break; | |
| 183 | |
| 184 case TOKimaginary80v: | |
| 185 sprintf(buffer,"%LgLi", float80value); | |
| 186 break; | |
| 187 #endif | |
| 188 | |
| 189 case TOKstring: | |
| 190 #if CSTRINGS | |
| 191 p = string; | |
| 192 #else | |
| 193 { OutBuffer buf; | |
| 194 | |
| 195 buf.writeByte('"'); | |
| 196 for (size_t i = 0; i < len; ) | |
| 197 { unsigned c; | |
| 198 | |
| 199 utf_decodeChar((unsigned char *)ustring, len, &i, &c); | |
| 200 switch (c) | |
| 201 { | |
| 202 case 0: | |
| 203 break; | |
| 204 | |
| 205 case '"': | |
| 206 case '\\': | |
| 207 buf.writeByte('\\'); | |
| 208 default: | |
| 209 if (isprint(c)) | |
| 210 buf.writeByte(c); | |
| 211 else if (c <= 0x7F) | |
| 212 buf.printf("\\x%02x", c); | |
| 213 else if (c <= 0xFFFF) | |
| 214 buf.printf("\\u%04x", c); | |
| 215 else | |
| 216 buf.printf("\\U%08x", c); | |
| 217 continue; | |
| 218 } | |
| 219 break; | |
| 220 } | |
| 221 buf.writeByte('"'); | |
| 222 if (postfix) | |
| 223 buf.writeByte('"'); | |
| 224 buf.writeByte(0); | |
| 225 p = (char *)buf.extractData(); | |
| 226 } | |
| 227 #endif | |
| 228 break; | |
| 229 | |
| 230 case TOKidentifier: | |
| 231 case TOKenum: | |
| 232 case TOKstruct: | |
| 233 case TOKimport: | |
| 234 CASE_BASIC_TYPES: | |
| 235 p = ident->toChars(); | |
| 236 break; | |
| 237 | |
| 238 default: | |
| 239 p = toChars(value); | |
| 240 break; | |
| 241 } | |
| 242 return p; | |
| 243 } | |
| 244 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
245 const char *Token::toChars(enum TOK value) |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
246 { const char *p; |
| 159 | 247 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 248 | |
| 249 p = tochars[value]; | |
| 250 if (!p) | |
| 251 { sprintf(buffer,"TOK%d",value); | |
| 252 p = buffer; | |
| 253 } | |
| 254 return p; | |
| 255 } | |
| 256 | |
| 257 /*************************** Lexer ********************************************/ | |
| 258 | |
| 259 Token *Lexer::freelist = NULL; | |
| 260 StringTable Lexer::stringtable; | |
| 261 OutBuffer Lexer::stringbuffer; | |
| 262 | |
| 263 Lexer::Lexer(Module *mod, | |
| 264 unsigned char *base, unsigned begoffset, unsigned endoffset, | |
| 265 int doDocComment, int commentToken) | |
| 266 : loc(mod, 1) | |
| 267 { | |
| 268 //printf("Lexer::Lexer(%p,%d)\n",base,length); | |
| 269 //printf("lexer.mod = %p, %p\n", mod, this->loc.mod); | |
| 270 memset(&token,0,sizeof(token)); | |
| 271 this->base = base; | |
| 272 this->end = base + endoffset; | |
| 273 p = base + begoffset; | |
| 274 this->mod = mod; | |
| 275 this->doDocComment = doDocComment; | |
| 276 this->anyToken = 0; | |
| 277 this->commentToken = commentToken; | |
| 278 //initKeywords(); | |
| 279 | |
| 280 /* If first line starts with '#!', ignore the line | |
| 281 */ | |
| 282 | |
| 283 if (p[0] == '#' && p[1] =='!') | |
| 284 { | |
| 285 p += 2; | |
| 286 while (1) | |
| 287 { unsigned char c = *p; | |
| 288 switch (c) | |
| 289 { | |
| 290 case '\n': | |
| 291 p++; | |
| 292 break; | |
| 293 | |
| 294 case '\r': | |
| 295 p++; | |
| 296 if (*p == '\n') | |
| 297 p++; | |
| 298 break; | |
| 299 | |
| 300 case 0: | |
| 301 case 0x1A: | |
| 302 break; | |
| 303 | |
| 304 default: | |
| 305 if (c & 0x80) | |
| 306 { unsigned u = decodeUTF(); | |
| 307 if (u == PS || u == LS) | |
| 308 break; | |
| 309 } | |
| 310 p++; | |
| 311 continue; | |
| 312 } | |
| 313 break; | |
| 314 } | |
| 315 loc.linnum = 2; | |
| 316 } | |
| 317 } | |
| 318 | |
| 319 | |
| 320 void Lexer::error(const char *format, ...) | |
| 321 { | |
| 322 if (mod && !global.gag) | |
| 323 { | |
| 324 char *p = loc.toChars(); | |
| 325 if (*p) | |
| 326 fprintf(stdmsg, "%s: ", p); | |
| 327 mem.free(p); | |
| 328 | |
| 329 va_list ap; | |
| 330 va_start(ap, format); | |
| 331 vfprintf(stdmsg, format, ap); | |
| 332 va_end(ap); | |
| 333 | |
| 334 fprintf(stdmsg, "\n"); | |
| 335 fflush(stdmsg); | |
| 336 | |
| 337 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 338 fatal(); | |
| 339 } | |
| 340 global.errors++; | |
| 341 } | |
| 342 | |
| 343 void Lexer::error(Loc loc, const char *format, ...) | |
| 344 { | |
| 345 if (mod && !global.gag) | |
| 346 { | |
| 347 char *p = loc.toChars(); | |
| 348 if (*p) | |
| 349 fprintf(stdmsg, "%s: ", p); | |
| 350 mem.free(p); | |
| 351 | |
| 352 va_list ap; | |
| 353 va_start(ap, format); | |
| 354 vfprintf(stdmsg, format, ap); | |
| 355 va_end(ap); | |
| 356 | |
| 357 fprintf(stdmsg, "\n"); | |
| 358 fflush(stdmsg); | |
| 359 | |
| 360 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 361 fatal(); | |
| 362 } | |
| 363 global.errors++; | |
| 364 } | |
| 365 | |
| 366 TOK Lexer::nextToken() | |
| 367 { Token *t; | |
| 368 | |
| 369 if (token.next) | |
| 370 { | |
| 371 t = token.next; | |
| 372 memcpy(&token,t,sizeof(Token)); | |
| 373 t->next = freelist; | |
| 374 freelist = t; | |
| 375 } | |
| 376 else | |
| 377 { | |
| 378 scan(&token); | |
| 379 } | |
| 380 //token.print(); | |
| 381 return token.value; | |
| 382 } | |
| 383 | |
| 384 Token *Lexer::peek(Token *ct) | |
| 385 { Token *t; | |
| 386 | |
| 387 if (ct->next) | |
| 388 t = ct->next; | |
| 389 else | |
| 390 { | |
| 391 t = new Token(); | |
| 392 scan(t); | |
| 393 t->next = NULL; | |
| 394 ct->next = t; | |
| 395 } | |
| 396 return t; | |
| 397 } | |
| 398 | |
|
717
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
399 /*********************** |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
400 * Look ahead at next token's value. |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
401 */ |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
402 |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
403 TOK Lexer::peekNext() |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
404 { |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
405 return peek(&token)->value; |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
406 } |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
407 |
| 159 | 408 /********************************* |
| 409 * tk is on the opening (. | |
| 410 * Look ahead and return token that is past the closing ). | |
| 411 */ | |
| 412 | |
| 413 Token *Lexer::peekPastParen(Token *tk) | |
| 414 { | |
| 415 //printf("peekPastParen()\n"); | |
| 416 int parens = 1; | |
| 417 int curlynest = 0; | |
| 418 while (1) | |
| 419 { | |
| 420 tk = peek(tk); | |
| 421 //tk->print(); | |
| 422 switch (tk->value) | |
| 423 { | |
| 424 case TOKlparen: | |
| 425 parens++; | |
| 426 continue; | |
| 427 | |
| 428 case TOKrparen: | |
| 429 --parens; | |
| 430 if (parens) | |
| 431 continue; | |
| 432 tk = peek(tk); | |
| 433 break; | |
| 434 | |
| 435 case TOKlcurly: | |
| 436 curlynest++; | |
| 437 continue; | |
| 438 | |
| 439 case TOKrcurly: | |
| 440 if (--curlynest >= 0) | |
| 441 continue; | |
| 442 break; | |
| 443 | |
| 444 case TOKsemicolon: | |
| 445 if (curlynest) | |
| 446 continue; | |
| 447 break; | |
| 448 | |
| 449 case TOKeof: | |
| 450 break; | |
| 451 | |
| 452 default: | |
| 453 continue; | |
| 454 } | |
| 455 return tk; | |
| 456 } | |
| 457 } | |
| 458 | |
| 459 /********************************** | |
| 460 * Determine if string is a valid Identifier. | |
| 461 * Placed here because of commonality with Lexer functionality. | |
| 462 * Returns: | |
| 463 * 0 invalid | |
| 464 */ | |
| 465 | |
| 466 int Lexer::isValidIdentifier(char *p) | |
| 467 { | |
| 468 size_t len; | |
| 469 size_t idx; | |
| 470 | |
| 471 if (!p || !*p) | |
| 472 goto Linvalid; | |
| 473 | |
| 474 if (*p >= '0' && *p <= '9') // beware of isdigit() on signed chars | |
| 475 goto Linvalid; | |
| 476 | |
| 477 len = strlen(p); | |
| 478 idx = 0; | |
| 479 while (p[idx]) | |
| 480 { dchar_t dc; | |
| 481 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
482 const char *q = utf_decodeChar((unsigned char *)p, len, &idx, &dc); |
| 159 | 483 if (q) |
| 484 goto Linvalid; | |
| 485 | |
| 486 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) | |
| 487 goto Linvalid; | |
| 488 } | |
| 489 return 1; | |
| 490 | |
| 491 Linvalid: | |
| 492 return 0; | |
| 493 } | |
| 494 | |
| 495 /**************************** | |
| 496 * Turn next token in buffer into a token. | |
| 497 */ | |
| 498 | |
| 499 void Lexer::scan(Token *t) | |
| 500 { | |
| 501 unsigned lastLine = loc.linnum; | |
| 502 unsigned linnum; | |
| 503 | |
| 504 t->blockComment = NULL; | |
| 505 t->lineComment = NULL; | |
| 506 while (1) | |
| 507 { | |
| 508 t->ptr = p; | |
| 509 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 510 switch (*p) | |
| 511 { | |
| 512 case 0: | |
| 513 case 0x1A: | |
| 514 t->value = TOKeof; // end of file | |
| 515 return; | |
| 516 | |
| 517 case ' ': | |
| 518 case '\t': | |
| 519 case '\v': | |
| 520 case '\f': | |
| 521 p++; | |
| 522 continue; // skip white space | |
| 523 | |
| 524 case '\r': | |
| 525 p++; | |
| 526 if (*p != '\n') // if CR stands by itself | |
| 527 loc.linnum++; | |
| 528 continue; // skip white space | |
| 529 | |
| 530 case '\n': | |
| 531 p++; | |
| 532 loc.linnum++; | |
| 533 continue; // skip white space | |
| 534 | |
| 535 case '0': case '1': case '2': case '3': case '4': | |
| 536 case '5': case '6': case '7': case '8': case '9': | |
| 537 t->value = number(t); | |
| 538 return; | |
| 539 | |
| 540 #if CSTRINGS | |
| 541 case '\'': | |
| 542 t->value = charConstant(t, 0); | |
| 543 return; | |
| 544 | |
| 545 case '"': | |
| 546 t->value = stringConstant(t,0); | |
| 547 return; | |
| 548 | |
| 549 case 'l': | |
| 550 case 'L': | |
| 551 if (p[1] == '\'') | |
| 552 { | |
| 553 p++; | |
| 554 t->value = charConstant(t, 1); | |
| 555 return; | |
| 556 } | |
| 557 else if (p[1] == '"') | |
| 558 { | |
| 559 p++; | |
| 560 t->value = stringConstant(t, 1); | |
| 561 return; | |
| 562 } | |
| 563 #else | |
| 564 case '\'': | |
| 565 t->value = charConstant(t,0); | |
| 566 return; | |
| 567 | |
| 568 case 'r': | |
| 569 if (p[1] != '"') | |
| 570 goto case_ident; | |
| 571 p++; | |
| 572 case '`': | |
| 573 t->value = wysiwygStringConstant(t, *p); | |
| 574 return; | |
| 575 | |
| 576 case 'x': | |
| 577 if (p[1] != '"') | |
| 578 goto case_ident; | |
| 579 p++; | |
| 580 t->value = hexStringConstant(t); | |
| 581 return; | |
| 582 | |
| 336 | 583 #if DMDV2 |
| 159 | 584 case 'q': |
| 585 if (p[1] == '"') | |
| 586 { | |
| 587 p++; | |
| 588 t->value = delimitedStringConstant(t); | |
| 589 return; | |
| 590 } | |
| 591 else if (p[1] == '{') | |
| 592 { | |
| 593 p++; | |
| 594 t->value = tokenStringConstant(t); | |
| 595 return; | |
| 596 } | |
| 597 else | |
| 598 goto case_ident; | |
| 599 #endif | |
| 600 | |
| 601 case '"': | |
| 602 t->value = escapeStringConstant(t,0); | |
| 603 return; | |
| 604 | |
| 605 case '\\': // escaped string literal | |
| 606 { unsigned c; | |
| 607 | |
| 608 stringbuffer.reset(); | |
| 609 do | |
| 610 { | |
| 611 p++; | |
| 612 switch (*p) | |
| 613 { | |
| 614 case 'u': | |
| 615 case 'U': | |
| 616 case '&': | |
| 617 c = escapeSequence(); | |
| 618 stringbuffer.writeUTF8(c); | |
| 619 break; | |
| 620 | |
| 621 default: | |
| 622 c = escapeSequence(); | |
| 623 stringbuffer.writeByte(c); | |
| 624 break; | |
| 625 } | |
| 626 } while (*p == '\\'); | |
| 627 t->len = stringbuffer.offset; | |
| 628 stringbuffer.writeByte(0); | |
| 629 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 630 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 631 t->postfix = 0; | |
| 632 t->value = TOKstring; | |
| 633 return; | |
| 634 } | |
| 635 | |
| 636 case 'l': | |
| 637 case 'L': | |
| 638 #endif | |
| 639 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 640 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 641 case 'k': case 'm': case 'n': case 'o': | |
| 336 | 642 #if DMDV2 |
| 159 | 643 case 'p': /*case 'q': case 'r':*/ case 's': case 't': |
| 644 #else | |
| 645 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 646 #endif | |
| 647 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 648 case 'z': | |
| 649 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 650 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 651 case 'K': case 'M': case 'N': case 'O': | |
| 652 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 653 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 654 case 'Z': | |
| 655 case '_': | |
| 656 case_ident: | |
| 657 { unsigned char c; | |
| 658 StringValue *sv; | |
| 659 Identifier *id; | |
| 660 | |
| 661 do | |
| 662 { | |
| 663 c = *++p; | |
| 664 } while (isidchar(c) || (c & 0x80 && isUniAlpha(decodeUTF()))); | |
| 665 sv = stringtable.update((char *)t->ptr, p - t->ptr); | |
| 666 id = (Identifier *) sv->ptrvalue; | |
| 667 if (!id) | |
| 668 { id = new Identifier(sv->lstring.string,TOKidentifier); | |
| 669 sv->ptrvalue = id; | |
| 670 } | |
| 671 t->ident = id; | |
| 672 t->value = (enum TOK) id->value; | |
| 673 anyToken = 1; | |
| 674 if (*t->ptr == '_') // if special identifier token | |
| 675 { | |
| 676 static char date[11+1]; | |
| 677 static char time[8+1]; | |
| 678 static char timestamp[24+1]; | |
| 679 | |
| 680 if (!date[0]) // lazy evaluation | |
| 681 { time_t t; | |
| 682 char *p; | |
| 683 | |
| 684 ::time(&t); | |
| 685 p = ctime(&t); | |
| 686 assert(p); | |
| 687 sprintf(date, "%.6s %.4s", p + 4, p + 20); | |
| 688 sprintf(time, "%.8s", p + 11); | |
| 689 sprintf(timestamp, "%.24s", p); | |
| 690 } | |
| 691 | |
| 336 | 692 #if DMDV1 |
| 159 | 693 if (mod && id == Id::FILE) |
| 694 { | |
| 695 t->ustring = (unsigned char *)(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 696 goto Lstring; | |
| 697 } | |
| 698 else if (mod && id == Id::LINE) | |
| 699 { | |
| 700 t->value = TOKint64v; | |
| 701 t->uns64value = loc.linnum; | |
| 702 } | |
| 336 | 703 else |
| 704 #endif | |
| 705 if (id == Id::DATE) | |
| 159 | 706 { |
| 707 t->ustring = (unsigned char *)date; | |
| 708 goto Lstring; | |
| 709 } | |
| 710 else if (id == Id::TIME) | |
| 711 { | |
| 712 t->ustring = (unsigned char *)time; | |
| 713 goto Lstring; | |
| 714 } | |
| 715 else if (id == Id::VENDOR) | |
| 716 { | |
|
664
eef8ac26c66c
Some missed LLVMDC -> LDC.
Christian Kamm <kamm incasoftware de>
parents:
658
diff
changeset
|
717 t->ustring = (unsigned char *)"LDC"; |
| 159 | 718 goto Lstring; |
| 719 } | |
| 720 else if (id == Id::TIMESTAMP) | |
| 721 { | |
| 722 t->ustring = (unsigned char *)timestamp; | |
| 723 Lstring: | |
| 724 t->value = TOKstring; | |
| 725 Llen: | |
| 726 t->postfix = 0; | |
| 727 t->len = strlen((char *)t->ustring); | |
| 728 } | |
| 729 else if (id == Id::VERSIONX) | |
| 730 { unsigned major = 0; | |
| 731 unsigned minor = 0; | |
| 732 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
733 for (const char *p = global.version + 1; 1; p++) |
| 159 | 734 { |
| 735 char c = *p; | |
| 736 if (isdigit(c)) | |
| 737 minor = minor * 10 + c - '0'; | |
| 738 else if (c == '.') | |
| 739 { major = minor; | |
| 740 minor = 0; | |
| 741 } | |
| 742 else | |
| 743 break; | |
| 744 } | |
| 745 t->value = TOKint64v; | |
| 746 t->uns64value = major * 1000 + minor; | |
| 747 } | |
| 336 | 748 #if DMDV2 |
| 159 | 749 else if (id == Id::EOFX) |
| 750 { | |
| 751 t->value = TOKeof; | |
| 752 // Advance scanner to end of file | |
| 753 while (!(*p == 0 || *p == 0x1A)) | |
| 754 p++; | |
| 755 } | |
| 756 #endif | |
| 757 } | |
| 758 //printf("t->value = %d\n",t->value); | |
| 759 return; | |
| 760 } | |
| 761 | |
| 762 case '/': | |
| 763 p++; | |
| 764 switch (*p) | |
| 765 { | |
| 766 case '=': | |
| 767 p++; | |
| 768 t->value = TOKdivass; | |
| 769 return; | |
| 770 | |
| 771 case '*': | |
| 772 p++; | |
| 773 linnum = loc.linnum; | |
| 774 while (1) | |
| 775 { | |
| 776 while (1) | |
| 777 { unsigned char c = *p; | |
| 778 switch (c) | |
| 779 { | |
| 780 case '/': | |
| 781 break; | |
| 782 | |
| 783 case '\n': | |
| 784 loc.linnum++; | |
| 785 p++; | |
| 786 continue; | |
| 787 | |
| 788 case '\r': | |
| 789 p++; | |
| 790 if (*p != '\n') | |
| 791 loc.linnum++; | |
| 792 continue; | |
| 793 | |
| 794 case 0: | |
| 795 case 0x1A: | |
| 796 error("unterminated /* */ comment"); | |
| 797 p = end; | |
| 798 t->value = TOKeof; | |
| 799 return; | |
| 800 | |
| 801 default: | |
| 802 if (c & 0x80) | |
| 803 { unsigned u = decodeUTF(); | |
| 804 if (u == PS || u == LS) | |
| 805 loc.linnum++; | |
| 806 } | |
| 807 p++; | |
| 808 continue; | |
| 809 } | |
| 810 break; | |
| 811 } | |
| 812 p++; | |
| 813 if (p[-2] == '*' && p - 3 != t->ptr) | |
| 814 break; | |
| 815 } | |
| 816 if (commentToken) | |
| 817 { | |
| 818 t->value = TOKcomment; | |
| 819 return; | |
| 820 } | |
| 821 else if (doDocComment && t->ptr[2] == '*' && p - 4 != t->ptr) | |
| 822 { // if /** but not /**/ | |
| 823 getDocComment(t, lastLine == linnum); | |
| 824 } | |
| 825 continue; | |
| 826 | |
| 827 case '/': // do // style comments | |
| 828 linnum = loc.linnum; | |
| 829 while (1) | |
| 830 { unsigned char c = *++p; | |
| 831 switch (c) | |
| 832 { | |
| 833 case '\n': | |
| 834 break; | |
| 835 | |
| 836 case '\r': | |
| 837 if (p[1] == '\n') | |
| 838 p++; | |
| 839 break; | |
| 840 | |
| 841 case 0: | |
| 842 case 0x1A: | |
| 843 if (commentToken) | |
| 844 { | |
| 845 p = end; | |
| 846 t->value = TOKcomment; | |
| 847 return; | |
| 848 } | |
| 849 if (doDocComment && t->ptr[2] == '/') | |
| 850 getDocComment(t, lastLine == linnum); | |
| 851 p = end; | |
| 852 t->value = TOKeof; | |
| 853 return; | |
| 854 | |
| 855 default: | |
| 856 if (c & 0x80) | |
| 857 { unsigned u = decodeUTF(); | |
| 858 if (u == PS || u == LS) | |
| 859 break; | |
| 860 } | |
| 861 continue; | |
| 862 } | |
| 863 break; | |
| 864 } | |
| 865 | |
| 866 if (commentToken) | |
| 867 { | |
| 868 p++; | |
| 869 loc.linnum++; | |
| 870 t->value = TOKcomment; | |
| 871 return; | |
| 872 } | |
| 873 if (doDocComment && t->ptr[2] == '/') | |
| 874 getDocComment(t, lastLine == linnum); | |
| 875 | |
| 876 p++; | |
| 877 loc.linnum++; | |
| 878 continue; | |
| 879 | |
| 880 case '+': | |
| 881 { int nest; | |
| 882 | |
| 883 linnum = loc.linnum; | |
| 884 p++; | |
| 885 nest = 1; | |
| 886 while (1) | |
| 887 { unsigned char c = *p; | |
| 888 switch (c) | |
| 889 { | |
| 890 case '/': | |
| 891 p++; | |
| 892 if (*p == '+') | |
| 893 { | |
| 894 p++; | |
| 895 nest++; | |
| 896 } | |
| 897 continue; | |
| 898 | |
| 899 case '+': | |
| 900 p++; | |
| 901 if (*p == '/') | |
| 902 { | |
| 903 p++; | |
| 904 if (--nest == 0) | |
| 905 break; | |
| 906 } | |
| 907 continue; | |
| 908 | |
| 909 case '\r': | |
| 910 p++; | |
| 911 if (*p != '\n') | |
| 912 loc.linnum++; | |
| 913 continue; | |
| 914 | |
| 915 case '\n': | |
| 916 loc.linnum++; | |
| 917 p++; | |
| 918 continue; | |
| 919 | |
| 920 case 0: | |
| 921 case 0x1A: | |
| 922 error("unterminated /+ +/ comment"); | |
| 923 p = end; | |
| 924 t->value = TOKeof; | |
| 925 return; | |
| 926 | |
| 927 default: | |
| 928 if (c & 0x80) | |
| 929 { unsigned u = decodeUTF(); | |
| 930 if (u == PS || u == LS) | |
| 931 loc.linnum++; | |
| 932 } | |
| 933 p++; | |
| 934 continue; | |
| 935 } | |
| 936 break; | |
| 937 } | |
| 938 if (commentToken) | |
| 939 { | |
| 940 t->value = TOKcomment; | |
| 941 return; | |
| 942 } | |
| 943 if (doDocComment && t->ptr[2] == '+' && p - 4 != t->ptr) | |
| 944 { // if /++ but not /++/ | |
| 945 getDocComment(t, lastLine == linnum); | |
| 946 } | |
| 947 continue; | |
| 948 } | |
| 949 } | |
| 950 t->value = TOKdiv; | |
| 951 return; | |
| 952 | |
| 953 case '.': | |
| 954 p++; | |
| 955 if (isdigit(*p)) | |
| 956 { /* Note that we don't allow ._1 and ._ as being | |
| 957 * valid floating point numbers. | |
| 958 */ | |
| 959 p--; | |
| 960 t->value = inreal(t); | |
| 961 } | |
| 962 else if (p[0] == '.') | |
| 963 { | |
| 964 if (p[1] == '.') | |
| 965 { p += 2; | |
| 966 t->value = TOKdotdotdot; | |
| 967 } | |
| 968 else | |
| 969 { p++; | |
| 970 t->value = TOKslice; | |
| 971 } | |
| 972 } | |
| 973 else | |
| 974 t->value = TOKdot; | |
| 975 return; | |
| 976 | |
| 977 case '&': | |
| 978 p++; | |
| 979 if (*p == '=') | |
| 980 { p++; | |
| 981 t->value = TOKandass; | |
| 982 } | |
| 983 else if (*p == '&') | |
| 984 { p++; | |
| 985 t->value = TOKandand; | |
| 986 } | |
| 987 else | |
| 988 t->value = TOKand; | |
| 989 return; | |
| 990 | |
| 991 case '|': | |
| 992 p++; | |
| 993 if (*p == '=') | |
| 994 { p++; | |
| 995 t->value = TOKorass; | |
| 996 } | |
| 997 else if (*p == '|') | |
| 998 { p++; | |
| 999 t->value = TOKoror; | |
| 1000 } | |
| 1001 else | |
| 1002 t->value = TOKor; | |
| 1003 return; | |
| 1004 | |
| 1005 case '-': | |
| 1006 p++; | |
| 1007 if (*p == '=') | |
| 1008 { p++; | |
| 1009 t->value = TOKminass; | |
| 1010 } | |
| 1011 #if 0 | |
| 1012 else if (*p == '>') | |
| 1013 { p++; | |
| 1014 t->value = TOKarrow; | |
| 1015 } | |
| 1016 #endif | |
| 1017 else if (*p == '-') | |
| 1018 { p++; | |
| 1019 t->value = TOKminusminus; | |
| 1020 } | |
| 1021 else | |
| 1022 t->value = TOKmin; | |
| 1023 return; | |
| 1024 | |
| 1025 case '+': | |
| 1026 p++; | |
| 1027 if (*p == '=') | |
| 1028 { p++; | |
| 1029 t->value = TOKaddass; | |
| 1030 } | |
| 1031 else if (*p == '+') | |
| 1032 { p++; | |
| 1033 t->value = TOKplusplus; | |
| 1034 } | |
| 1035 else | |
| 1036 t->value = TOKadd; | |
| 1037 return; | |
| 1038 | |
| 1039 case '<': | |
| 1040 p++; | |
| 1041 if (*p == '=') | |
| 1042 { p++; | |
| 1043 t->value = TOKle; // <= | |
| 1044 } | |
| 1045 else if (*p == '<') | |
| 1046 { p++; | |
| 1047 if (*p == '=') | |
| 1048 { p++; | |
| 1049 t->value = TOKshlass; // <<= | |
| 1050 } | |
| 1051 else | |
| 1052 t->value = TOKshl; // << | |
| 1053 } | |
| 1054 else if (*p == '>') | |
| 1055 { p++; | |
| 1056 if (*p == '=') | |
| 1057 { p++; | |
| 1058 t->value = TOKleg; // <>= | |
| 1059 } | |
| 1060 else | |
| 1061 t->value = TOKlg; // <> | |
| 1062 } | |
| 1063 else | |
| 1064 t->value = TOKlt; // < | |
| 1065 return; | |
| 1066 | |
| 1067 case '>': | |
| 1068 p++; | |
| 1069 if (*p == '=') | |
| 1070 { p++; | |
| 1071 t->value = TOKge; // >= | |
| 1072 } | |
| 1073 else if (*p == '>') | |
| 1074 { p++; | |
| 1075 if (*p == '=') | |
| 1076 { p++; | |
| 1077 t->value = TOKshrass; // >>= | |
| 1078 } | |
| 1079 else if (*p == '>') | |
| 1080 { p++; | |
| 1081 if (*p == '=') | |
| 1082 { p++; | |
| 1083 t->value = TOKushrass; // >>>= | |
| 1084 } | |
| 1085 else | |
| 1086 t->value = TOKushr; // >>> | |
| 1087 } | |
| 1088 else | |
| 1089 t->value = TOKshr; // >> | |
| 1090 } | |
| 1091 else | |
| 1092 t->value = TOKgt; // > | |
| 1093 return; | |
| 1094 | |
| 1095 case '!': | |
| 1096 p++; | |
| 1097 if (*p == '=') | |
| 1098 { p++; | |
| 1099 if (*p == '=' && global.params.Dversion == 1) | |
| 1100 { p++; | |
| 1101 t->value = TOKnotidentity; // !== | |
| 1102 } | |
| 1103 else | |
| 1104 t->value = TOKnotequal; // != | |
| 1105 } | |
| 1106 else if (*p == '<') | |
| 1107 { p++; | |
| 1108 if (*p == '>') | |
| 1109 { p++; | |
| 1110 if (*p == '=') | |
| 1111 { p++; | |
| 1112 t->value = TOKunord; // !<>= | |
| 1113 } | |
| 1114 else | |
| 1115 t->value = TOKue; // !<> | |
| 1116 } | |
| 1117 else if (*p == '=') | |
| 1118 { p++; | |
| 1119 t->value = TOKug; // !<= | |
| 1120 } | |
| 1121 else | |
| 1122 t->value = TOKuge; // !< | |
| 1123 } | |
| 1124 else if (*p == '>') | |
| 1125 { p++; | |
| 1126 if (*p == '=') | |
| 1127 { p++; | |
| 1128 t->value = TOKul; // !>= | |
| 1129 } | |
| 1130 else | |
| 1131 t->value = TOKule; // !> | |
| 1132 } | |
| 1133 else | |
| 1134 t->value = TOKnot; // ! | |
| 1135 return; | |
| 1136 | |
| 1137 case '=': | |
| 1138 p++; | |
| 1139 if (*p == '=') | |
| 1140 { p++; | |
| 1141 if (*p == '=' && global.params.Dversion == 1) | |
| 1142 { p++; | |
| 1143 t->value = TOKidentity; // === | |
| 1144 } | |
| 1145 else | |
| 1146 t->value = TOKequal; // == | |
| 1147 } | |
| 1148 else | |
| 1149 t->value = TOKassign; // = | |
| 1150 return; | |
| 1151 | |
| 1152 case '~': | |
| 1153 p++; | |
| 1154 if (*p == '=') | |
| 1155 { p++; | |
| 1156 t->value = TOKcatass; // ~= | |
| 1157 } | |
| 1158 else | |
| 1159 t->value = TOKtilde; // ~ | |
| 1160 return; | |
| 1161 | |
| 1162 #define SINGLE(c,tok) case c: p++; t->value = tok; return; | |
| 1163 | |
| 1164 SINGLE('(', TOKlparen) | |
| 1165 SINGLE(')', TOKrparen) | |
| 1166 SINGLE('[', TOKlbracket) | |
| 1167 SINGLE(']', TOKrbracket) | |
| 1168 SINGLE('{', TOKlcurly) | |
| 1169 SINGLE('}', TOKrcurly) | |
| 1170 SINGLE('?', TOKquestion) | |
| 1171 SINGLE(',', TOKcomma) | |
| 1172 SINGLE(';', TOKsemicolon) | |
| 1173 SINGLE(':', TOKcolon) | |
| 1174 SINGLE('$', TOKdollar) | |
| 1175 | |
| 1176 #undef SINGLE | |
| 1177 | |
| 1178 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1179 case c1: \ | |
| 1180 p++; \ | |
| 1181 if (*p == c2) \ | |
| 1182 { p++; \ | |
| 1183 t->value = tok2; \ | |
| 1184 } \ | |
| 1185 else \ | |
| 1186 t->value = tok1; \ | |
| 1187 return; | |
| 1188 | |
| 1189 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1190 DOUBLE('%', TOKmod, '=', TOKmodass) | |
| 1191 DOUBLE('^', TOKxor, '=', TOKxorass) | |
| 1192 | |
| 1193 #undef DOUBLE | |
| 1194 | |
| 1195 case '#': | |
| 1196 p++; | |
| 1197 pragma(); | |
| 1198 continue; | |
| 1199 | |
| 1200 default: | |
| 1201 { unsigned char c = *p; | |
| 1202 | |
| 1203 if (c & 0x80) | |
| 1204 { unsigned u = decodeUTF(); | |
| 1205 | |
| 1206 // Check for start of unicode identifier | |
| 1207 if (isUniAlpha(u)) | |
| 1208 goto case_ident; | |
| 1209 | |
| 1210 if (u == PS || u == LS) | |
| 1211 { | |
| 1212 loc.linnum++; | |
| 1213 p++; | |
| 1214 continue; | |
| 1215 } | |
| 1216 } | |
| 1217 if (isprint(c)) | |
| 1218 error("unsupported char '%c'", c); | |
| 1219 else | |
| 1220 error("unsupported char 0x%02x", c); | |
| 1221 p++; | |
| 1222 continue; | |
| 1223 } | |
| 1224 } | |
| 1225 } | |
| 1226 } | |
| 1227 | |
| 1228 /******************************************* | |
| 1229 * Parse escape sequence. | |
| 1230 */ | |
| 1231 | |
| 1232 unsigned Lexer::escapeSequence() | |
| 1233 { unsigned c; | |
| 1234 int n; | |
| 1235 int ndigits; | |
| 1236 | |
| 1237 c = *p; | |
| 1238 switch (c) | |
| 1239 { | |
| 1240 case '\'': | |
| 1241 case '"': | |
| 1242 case '?': | |
| 1243 case '\\': | |
| 1244 Lconsume: | |
| 1245 p++; | |
| 1246 break; | |
| 1247 | |
| 1248 case 'a': c = 7; goto Lconsume; | |
| 1249 case 'b': c = 8; goto Lconsume; | |
| 1250 case 'f': c = 12; goto Lconsume; | |
| 1251 case 'n': c = 10; goto Lconsume; | |
| 1252 case 'r': c = 13; goto Lconsume; | |
| 1253 case 't': c = 9; goto Lconsume; | |
| 1254 case 'v': c = 11; goto Lconsume; | |
| 1255 | |
| 1256 case 'u': | |
| 1257 ndigits = 4; | |
| 1258 goto Lhex; | |
| 1259 case 'U': | |
| 1260 ndigits = 8; | |
| 1261 goto Lhex; | |
| 1262 case 'x': | |
| 1263 ndigits = 2; | |
| 1264 Lhex: | |
| 1265 p++; | |
| 1266 c = *p; | |
| 1267 if (ishex(c)) | |
| 1268 { unsigned v; | |
| 1269 | |
| 1270 n = 0; | |
| 1271 v = 0; | |
| 1272 while (1) | |
| 1273 { | |
| 1274 if (isdigit(c)) | |
| 1275 c -= '0'; | |
| 1276 else if (islower(c)) | |
| 1277 c -= 'a' - 10; | |
| 1278 else | |
| 1279 c -= 'A' - 10; | |
| 1280 v = v * 16 + c; | |
| 1281 c = *++p; | |
| 1282 if (++n == ndigits) | |
| 1283 break; | |
| 1284 if (!ishex(c)) | |
| 1285 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1286 break; | |
| 1287 } | |
| 1288 } | |
| 1289 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1290 error("invalid UTF character \\U%08x", v); | |
| 1291 c = v; | |
| 1292 } | |
| 1293 else | |
| 1294 error("undefined escape hex sequence \\%c\n",c); | |
| 1295 break; | |
| 1296 | |
| 1297 case '&': // named character entity | |
| 1298 for (unsigned char *idstart = ++p; 1; p++) | |
| 1299 { | |
| 1300 switch (*p) | |
| 1301 { | |
| 1302 case ';': | |
| 1303 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1304 if (c == ~0) | |
| 1305 { error("unnamed character entity &%.*s;", (int)(p - idstart), idstart); | |
| 1306 c = ' '; | |
| 1307 } | |
| 1308 p++; | |
| 1309 break; | |
| 1310 | |
| 1311 default: | |
| 1312 if (isalpha(*p) || | |
| 1313 (p != idstart + 1 && isdigit(*p))) | |
| 1314 continue; | |
| 1315 error("unterminated named entity"); | |
| 1316 break; | |
| 1317 } | |
| 1318 break; | |
| 1319 } | |
| 1320 break; | |
| 1321 | |
| 1322 case 0: | |
| 1323 case 0x1A: // end of file | |
| 1324 c = '\\'; | |
| 1325 break; | |
| 1326 | |
| 1327 default: | |
| 1328 if (isoctal(c)) | |
| 1329 { unsigned v; | |
| 1330 | |
| 1331 n = 0; | |
| 1332 v = 0; | |
| 1333 do | |
| 1334 { | |
| 1335 v = v * 8 + (c - '0'); | |
| 1336 c = *++p; | |
| 1337 } while (++n < 3 && isoctal(c)); | |
| 1338 c = v; | |
| 1339 if (c > 0xFF) | |
| 1340 error("0%03o is larger than a byte", c); | |
| 1341 } | |
| 1342 else | |
| 1343 error("undefined escape sequence \\%c\n",c); | |
| 1344 break; | |
| 1345 } | |
| 1346 return c; | |
| 1347 } | |
| 1348 | |
| 1349 /************************************** | |
| 1350 */ | |
| 1351 | |
| 1352 TOK Lexer::wysiwygStringConstant(Token *t, int tc) | |
| 1353 { unsigned c; | |
| 1354 Loc start = loc; | |
| 1355 | |
| 1356 p++; | |
| 1357 stringbuffer.reset(); | |
| 1358 while (1) | |
| 1359 { | |
| 1360 c = *p++; | |
| 1361 switch (c) | |
| 1362 { | |
| 1363 case '\n': | |
| 1364 loc.linnum++; | |
| 1365 break; | |
| 1366 | |
| 1367 case '\r': | |
| 1368 if (*p == '\n') | |
| 1369 continue; // ignore | |
| 1370 c = '\n'; // treat EndOfLine as \n character | |
| 1371 loc.linnum++; | |
| 1372 break; | |
| 1373 | |
| 1374 case 0: | |
| 1375 case 0x1A: | |
| 1376 error("unterminated string constant starting at %s", start.toChars()); | |
| 1377 t->ustring = (unsigned char *)""; | |
| 1378 t->len = 0; | |
| 1379 t->postfix = 0; | |
| 1380 return TOKstring; | |
| 1381 | |
| 1382 case '"': | |
| 1383 case '`': | |
| 1384 if (c == tc) | |
| 1385 { | |
| 1386 t->len = stringbuffer.offset; | |
| 1387 stringbuffer.writeByte(0); | |
| 1388 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1389 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1390 stringPostfix(t); | |
| 1391 return TOKstring; | |
| 1392 } | |
| 1393 break; | |
| 1394 | |
| 1395 default: | |
| 1396 if (c & 0x80) | |
| 1397 { p--; | |
| 1398 unsigned u = decodeUTF(); | |
| 1399 p++; | |
| 1400 if (u == PS || u == LS) | |
| 1401 loc.linnum++; | |
| 1402 stringbuffer.writeUTF8(u); | |
| 1403 continue; | |
| 1404 } | |
| 1405 break; | |
| 1406 } | |
| 1407 stringbuffer.writeByte(c); | |
| 1408 } | |
| 1409 } | |
| 1410 | |
| 1411 /************************************** | |
| 1412 * Lex hex strings: | |
| 1413 * x"0A ae 34FE BD" | |
| 1414 */ | |
| 1415 | |
| 1416 TOK Lexer::hexStringConstant(Token *t) | |
| 1417 { unsigned c; | |
| 1418 Loc start = loc; | |
| 1419 unsigned n = 0; | |
| 1420 unsigned v; | |
| 1421 | |
| 1422 p++; | |
| 1423 stringbuffer.reset(); | |
| 1424 while (1) | |
| 1425 { | |
| 1426 c = *p++; | |
| 1427 switch (c) | |
| 1428 { | |
| 1429 case ' ': | |
| 1430 case '\t': | |
| 1431 case '\v': | |
| 1432 case '\f': | |
| 1433 continue; // skip white space | |
| 1434 | |
| 1435 case '\r': | |
| 1436 if (*p == '\n') | |
| 1437 continue; // ignore | |
| 1438 // Treat isolated '\r' as if it were a '\n' | |
| 1439 case '\n': | |
| 1440 loc.linnum++; | |
| 1441 continue; | |
| 1442 | |
| 1443 case 0: | |
| 1444 case 0x1A: | |
| 1445 error("unterminated string constant starting at %s", start.toChars()); | |
| 1446 t->ustring = (unsigned char *)""; | |
| 1447 t->len = 0; | |
| 1448 t->postfix = 0; | |
| 1449 return TOKstring; | |
| 1450 | |
| 1451 case '"': | |
| 1452 if (n & 1) | |
| 1453 { error("odd number (%d) of hex characters in hex string", n); | |
| 1454 stringbuffer.writeByte(v); | |
| 1455 } | |
| 1456 t->len = stringbuffer.offset; | |
| 1457 stringbuffer.writeByte(0); | |
| 1458 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1459 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1460 stringPostfix(t); | |
| 1461 return TOKstring; | |
| 1462 | |
| 1463 default: | |
| 1464 if (c >= '0' && c <= '9') | |
| 1465 c -= '0'; | |
| 1466 else if (c >= 'a' && c <= 'f') | |
| 1467 c -= 'a' - 10; | |
| 1468 else if (c >= 'A' && c <= 'F') | |
| 1469 c -= 'A' - 10; | |
| 1470 else if (c & 0x80) | |
| 1471 { p--; | |
| 1472 unsigned u = decodeUTF(); | |
| 1473 p++; | |
| 1474 if (u == PS || u == LS) | |
| 1475 loc.linnum++; | |
| 1476 else | |
| 1477 error("non-hex character \\u%x", u); | |
| 1478 } | |
| 1479 else | |
| 1480 error("non-hex character '%c'", c); | |
| 1481 if (n & 1) | |
| 1482 { v = (v << 4) | c; | |
| 1483 stringbuffer.writeByte(v); | |
| 1484 } | |
| 1485 else | |
| 1486 v = c; | |
| 1487 n++; | |
| 1488 break; | |
| 1489 } | |
| 1490 } | |
| 1491 } | |
| 1492 | |
| 1493 | |
| 336 | 1494 #if DMDV2 |
| 159 | 1495 /************************************** |
| 1496 * Lex delimited strings: | |
| 1497 * q"(foo(xxx))" // "foo(xxx)" | |
| 1498 * q"[foo(]" // "foo(" | |
| 1499 * q"/foo]/" // "foo]" | |
| 1500 * q"HERE | |
| 1501 * foo | |
| 1502 * HERE" // "foo\n" | |
| 1503 * Input: | |
| 1504 * p is on the " | |
| 1505 */ | |
| 1506 | |
| 1507 TOK Lexer::delimitedStringConstant(Token *t) | |
| 1508 { unsigned c; | |
| 1509 Loc start = loc; | |
| 1510 unsigned delimleft = 0; | |
| 1511 unsigned delimright = 0; | |
| 1512 unsigned nest = 1; | |
| 1513 unsigned nestcount; | |
| 1514 Identifier *hereid = NULL; | |
| 1515 unsigned blankrol = 0; | |
| 1516 unsigned startline = 0; | |
| 1517 | |
| 1518 p++; | |
| 1519 stringbuffer.reset(); | |
| 1520 while (1) | |
| 1521 { | |
| 1522 c = *p++; | |
| 1523 //printf("c = '%c'\n", c); | |
| 1524 switch (c) | |
| 1525 { | |
| 1526 case '\n': | |
| 1527 Lnextline: | |
| 1528 loc.linnum++; | |
| 1529 startline = 1; | |
| 1530 if (blankrol) | |
| 1531 { blankrol = 0; | |
| 1532 continue; | |
| 1533 } | |
| 1534 if (hereid) | |
| 1535 { | |
| 1536 stringbuffer.writeUTF8(c); | |
| 1537 continue; | |
| 1538 } | |
| 1539 break; | |
| 1540 | |
| 1541 case '\r': | |
| 1542 if (*p == '\n') | |
| 1543 continue; // ignore | |
| 1544 c = '\n'; // treat EndOfLine as \n character | |
| 1545 goto Lnextline; | |
| 1546 | |
| 1547 case 0: | |
| 1548 case 0x1A: | |
| 1549 goto Lerror; | |
| 1550 | |
| 1551 default: | |
| 1552 if (c & 0x80) | |
| 1553 { p--; | |
| 1554 c = decodeUTF(); | |
| 1555 p++; | |
| 1556 if (c == PS || c == LS) | |
| 1557 goto Lnextline; | |
| 1558 } | |
| 1559 break; | |
| 1560 } | |
| 1561 if (delimleft == 0) | |
| 1562 { delimleft = c; | |
| 1563 nest = 1; | |
| 1564 nestcount = 1; | |
| 1565 if (c == '(') | |
| 1566 delimright = ')'; | |
| 1567 else if (c == '{') | |
| 1568 delimright = '}'; | |
| 1569 else if (c == '[') | |
| 1570 delimright = ']'; | |
| 1571 else if (c == '<') | |
| 1572 delimright = '>'; | |
| 1573 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1574 { // Start of identifier; must be a heredoc | |
| 1575 Token t; | |
| 1576 p--; | |
| 1577 scan(&t); // read in heredoc identifier | |
| 1578 if (t.value != TOKidentifier) | |
| 1579 { error("identifier expected for heredoc, not %s", t.toChars()); | |
| 1580 delimright = c; | |
| 1581 } | |
| 1582 else | |
| 1583 { hereid = t.ident; | |
| 1584 //printf("hereid = '%s'\n", hereid->toChars()); | |
| 1585 blankrol = 1; | |
| 1586 } | |
| 1587 nest = 0; | |
| 1588 } | |
| 1589 else | |
| 1590 { delimright = c; | |
| 1591 nest = 0; | |
| 1592 } | |
| 1593 } | |
| 1594 else | |
| 1595 { | |
| 1596 if (blankrol) | |
| 1597 { error("heredoc rest of line should be blank"); | |
| 1598 blankrol = 0; | |
| 1599 continue; | |
| 1600 } | |
| 1601 if (nest == 1) | |
| 1602 { | |
| 1603 if (c == delimleft) | |
| 1604 nestcount++; | |
| 1605 else if (c == delimright) | |
| 1606 { nestcount--; | |
| 1607 if (nestcount == 0) | |
| 1608 goto Ldone; | |
| 1609 } | |
| 1610 } | |
| 1611 else if (c == delimright) | |
| 1612 goto Ldone; | |
| 1613 if (startline && isalpha(c)) | |
| 1614 { Token t; | |
| 1615 unsigned char *psave = p; | |
| 1616 p--; | |
| 1617 scan(&t); // read in possible heredoc identifier | |
| 1618 //printf("endid = '%s'\n", t.ident->toChars()); | |
| 1619 if (t.value == TOKidentifier && t.ident->equals(hereid)) | |
| 1620 { /* should check that rest of line is blank | |
| 1621 */ | |
| 1622 goto Ldone; | |
| 1623 } | |
| 1624 p = psave; | |
| 1625 } | |
| 1626 stringbuffer.writeUTF8(c); | |
| 1627 startline = 0; | |
| 1628 } | |
| 1629 } | |
| 1630 | |
| 1631 Ldone: | |
| 1632 if (*p == '"') | |
| 1633 p++; | |
| 1634 else | |
| 1635 error("delimited string must end in %c\"", delimright); | |
| 1636 t->len = stringbuffer.offset; | |
| 1637 stringbuffer.writeByte(0); | |
| 1638 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1639 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1640 stringPostfix(t); | |
| 1641 return TOKstring; | |
| 1642 | |
| 1643 Lerror: | |
| 1644 error("unterminated string constant starting at %s", start.toChars()); | |
| 1645 t->ustring = (unsigned char *)""; | |
| 1646 t->len = 0; | |
| 1647 t->postfix = 0; | |
| 1648 return TOKstring; | |
| 1649 } | |
| 1650 | |
| 1651 /************************************** | |
| 1652 * Lex delimited strings: | |
| 1653 * q{ foo(xxx) } // " foo(xxx) " | |
| 1654 * q{foo(} // "foo(" | |
| 1655 * q{{foo}"}"} // "{foo}"}"" | |
| 1656 * Input: | |
| 1657 * p is on the q | |
| 1658 */ | |
| 1659 | |
| 1660 TOK Lexer::tokenStringConstant(Token *t) | |
| 1661 { | |
| 1662 unsigned nest = 1; | |
| 1663 Loc start = loc; | |
| 1664 unsigned char *pstart = ++p; | |
| 1665 | |
| 1666 while (1) | |
| 1667 { Token tok; | |
| 1668 | |
| 1669 scan(&tok); | |
| 1670 switch (tok.value) | |
| 1671 { | |
| 1672 case TOKlcurly: | |
| 1673 nest++; | |
| 1674 continue; | |
| 1675 | |
| 1676 case TOKrcurly: | |
| 1677 if (--nest == 0) | |
| 1678 goto Ldone; | |
| 1679 continue; | |
| 1680 | |
| 1681 case TOKeof: | |
| 1682 goto Lerror; | |
| 1683 | |
| 1684 default: | |
| 1685 continue; | |
| 1686 } | |
| 1687 } | |
| 1688 | |
| 1689 Ldone: | |
| 1690 t->len = p - 1 - pstart; | |
| 1691 t->ustring = (unsigned char *)mem.malloc(t->len + 1); | |
| 1692 memcpy(t->ustring, pstart, t->len); | |
| 1693 t->ustring[t->len] = 0; | |
| 1694 stringPostfix(t); | |
| 1695 return TOKstring; | |
| 1696 | |
| 1697 Lerror: | |
| 1698 error("unterminated token string constant starting at %s", start.toChars()); | |
| 1699 t->ustring = (unsigned char *)""; | |
| 1700 t->len = 0; | |
| 1701 t->postfix = 0; | |
| 1702 return TOKstring; | |
| 1703 } | |
| 1704 | |
| 1705 #endif | |
| 1706 | |
| 1707 | |
| 1708 /************************************** | |
| 1709 */ | |
| 1710 | |
| 1711 TOK Lexer::escapeStringConstant(Token *t, int wide) | |
| 1712 { unsigned c; | |
| 1713 Loc start = loc; | |
| 1714 | |
| 1715 p++; | |
| 1716 stringbuffer.reset(); | |
| 1717 while (1) | |
| 1718 { | |
| 1719 c = *p++; | |
| 1720 switch (c) | |
| 1721 { | |
| 1722 case '\\': | |
| 1723 switch (*p) | |
| 1724 { | |
| 1725 case 'u': | |
| 1726 case 'U': | |
| 1727 case '&': | |
| 1728 c = escapeSequence(); | |
| 1729 stringbuffer.writeUTF8(c); | |
| 1730 continue; | |
| 1731 | |
| 1732 default: | |
| 1733 c = escapeSequence(); | |
| 1734 break; | |
| 1735 } | |
| 1736 break; | |
| 1737 | |
| 1738 case '\n': | |
| 1739 loc.linnum++; | |
| 1740 break; | |
| 1741 | |
| 1742 case '\r': | |
| 1743 if (*p == '\n') | |
| 1744 continue; // ignore | |
| 1745 c = '\n'; // treat EndOfLine as \n character | |
| 1746 loc.linnum++; | |
| 1747 break; | |
| 1748 | |
| 1749 case '"': | |
| 1750 t->len = stringbuffer.offset; | |
| 1751 stringbuffer.writeByte(0); | |
| 1752 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1753 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1754 stringPostfix(t); | |
| 1755 return TOKstring; | |
| 1756 | |
| 1757 case 0: | |
| 1758 case 0x1A: | |
| 1759 p--; | |
| 1760 error("unterminated string constant starting at %s", start.toChars()); | |
| 1761 t->ustring = (unsigned char *)""; | |
| 1762 t->len = 0; | |
| 1763 t->postfix = 0; | |
| 1764 return TOKstring; | |
| 1765 | |
| 1766 default: | |
| 1767 if (c & 0x80) | |
| 1768 { | |
| 1769 p--; | |
| 1770 c = decodeUTF(); | |
| 1771 if (c == LS || c == PS) | |
| 1772 { c = '\n'; | |
| 1773 loc.linnum++; | |
| 1774 } | |
| 1775 p++; | |
| 1776 stringbuffer.writeUTF8(c); | |
| 1777 continue; | |
| 1778 } | |
| 1779 break; | |
| 1780 } | |
| 1781 stringbuffer.writeByte(c); | |
| 1782 } | |
| 1783 } | |
| 1784 | |
| 1785 /************************************** | |
| 1786 */ | |
| 1787 | |
| 1788 TOK Lexer::charConstant(Token *t, int wide) | |
| 1789 { | |
| 1790 unsigned c; | |
| 1791 TOK tk = TOKcharv; | |
| 1792 | |
| 1793 //printf("Lexer::charConstant\n"); | |
| 1794 p++; | |
| 1795 c = *p++; | |
| 1796 switch (c) | |
| 1797 { | |
| 1798 case '\\': | |
| 1799 switch (*p) | |
| 1800 { | |
| 1801 case 'u': | |
| 1802 t->uns64value = escapeSequence(); | |
| 1803 tk = TOKwcharv; | |
| 1804 break; | |
| 1805 | |
| 1806 case 'U': | |
| 1807 case '&': | |
| 1808 t->uns64value = escapeSequence(); | |
| 1809 tk = TOKdcharv; | |
| 1810 break; | |
| 1811 | |
| 1812 default: | |
| 1813 t->uns64value = escapeSequence(); | |
| 1814 break; | |
| 1815 } | |
| 1816 break; | |
| 1817 | |
| 1818 case '\n': | |
| 1819 L1: | |
| 1820 loc.linnum++; | |
| 1821 case '\r': | |
| 1822 case 0: | |
| 1823 case 0x1A: | |
| 1824 case '\'': | |
| 1825 error("unterminated character constant"); | |
| 1826 return tk; | |
| 1827 | |
| 1828 default: | |
| 1829 if (c & 0x80) | |
| 1830 { | |
| 1831 p--; | |
| 1832 c = decodeUTF(); | |
| 1833 p++; | |
| 1834 if (c == LS || c == PS) | |
| 1835 goto L1; | |
| 1836 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 1837 tk = TOKwcharv; | |
| 1838 else | |
| 1839 tk = TOKdcharv; | |
| 1840 } | |
| 1841 t->uns64value = c; | |
| 1842 break; | |
| 1843 } | |
| 1844 | |
| 1845 if (*p != '\'') | |
| 1846 { error("unterminated character constant"); | |
| 1847 return tk; | |
| 1848 } | |
| 1849 p++; | |
| 1850 return tk; | |
| 1851 } | |
| 1852 | |
| 1853 /*************************************** | |
| 1854 * Get postfix of string literal. | |
| 1855 */ | |
| 1856 | |
| 1857 void Lexer::stringPostfix(Token *t) | |
| 1858 { | |
| 1859 switch (*p) | |
| 1860 { | |
| 1861 case 'c': | |
| 1862 case 'w': | |
| 1863 case 'd': | |
| 1864 t->postfix = *p; | |
| 1865 p++; | |
| 1866 break; | |
| 1867 | |
| 1868 default: | |
| 1869 t->postfix = 0; | |
| 1870 break; | |
| 1871 } | |
| 1872 } | |
| 1873 | |
| 1874 /*************************************** | |
| 1875 * Read \u or \U unicode sequence | |
| 1876 * Input: | |
| 1877 * u 'u' or 'U' | |
| 1878 */ | |
| 1879 | |
| 1880 #if 0 | |
| 1881 unsigned Lexer::wchar(unsigned u) | |
| 1882 { | |
| 1883 unsigned value; | |
| 1884 unsigned n; | |
| 1885 unsigned char c; | |
| 1886 unsigned nchars; | |
| 1887 | |
| 1888 nchars = (u == 'U') ? 8 : 4; | |
| 1889 value = 0; | |
| 1890 for (n = 0; 1; n++) | |
| 1891 { | |
| 1892 ++p; | |
| 1893 if (n == nchars) | |
| 1894 break; | |
| 1895 c = *p; | |
| 1896 if (!ishex(c)) | |
| 1897 { error("\\%c sequence must be followed by %d hex characters", u, nchars); | |
| 1898 break; | |
| 1899 } | |
| 1900 if (isdigit(c)) | |
| 1901 c -= '0'; | |
| 1902 else if (islower(c)) | |
| 1903 c -= 'a' - 10; | |
| 1904 else | |
| 1905 c -= 'A' - 10; | |
| 1906 value <<= 4; | |
| 1907 value |= c; | |
| 1908 } | |
| 1909 return value; | |
| 1910 } | |
| 1911 #endif | |
| 1912 | |
| 1913 /************************************** | |
| 1914 * Read in a number. | |
| 1915 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 1916 * integers can be decimal, octal or hex | |
| 1917 * Handle the suffixes U, UL, LU, L, etc. | |
| 1918 * If it's double, store it in tok.TKutok.Vdouble. | |
| 1919 * Returns: | |
| 1920 * TKnum | |
| 1921 * TKdouble,... | |
| 1922 */ | |
| 1923 | |
| 1924 TOK Lexer::number(Token *t) | |
| 1925 { | |
| 1926 // We use a state machine to collect numbers | |
| 1927 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 1928 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 1929 STATE_hexh, STATE_error }; | |
| 1930 enum STATE state; | |
| 1931 | |
| 1932 enum FLAGS | |
| 1933 { FLAGS_decimal = 1, // decimal | |
| 1934 FLAGS_unsigned = 2, // u or U suffix | |
| 1935 FLAGS_long = 4, // l or L suffix | |
| 1936 }; | |
| 1937 enum FLAGS flags = FLAGS_decimal; | |
| 1938 | |
| 1939 int i; | |
| 1940 int base; | |
| 1941 unsigned c; | |
| 1942 unsigned char *start; | |
| 1943 TOK result; | |
| 1944 | |
| 1945 //printf("Lexer::number()\n"); | |
| 1946 state = STATE_initial; | |
| 1947 base = 0; | |
| 1948 stringbuffer.reset(); | |
| 1949 start = p; | |
| 1950 while (1) | |
| 1951 { | |
| 1952 c = *p; | |
| 1953 switch (state) | |
| 1954 { | |
| 1955 case STATE_initial: // opening state | |
| 1956 if (c == '0') | |
| 1957 state = STATE_0; | |
| 1958 else | |
| 1959 state = STATE_decimal; | |
| 1960 break; | |
| 1961 | |
| 1962 case STATE_0: | |
| 1963 flags = (FLAGS) (flags & ~FLAGS_decimal); | |
| 1964 switch (c) | |
| 1965 { | |
| 1966 #if ZEROH | |
| 1967 case 'H': // 0h | |
| 1968 case 'h': | |
| 1969 goto hexh; | |
| 1970 #endif | |
| 1971 case 'X': | |
| 1972 case 'x': | |
| 1973 state = STATE_hex0; | |
| 1974 break; | |
| 1975 | |
| 1976 case '.': | |
| 1977 if (p[1] == '.') // .. is a separate token | |
| 1978 goto done; | |
| 1979 case 'i': | |
| 1980 case 'f': | |
| 1981 case 'F': | |
| 1982 goto real; | |
| 1983 #if ZEROH | |
| 1984 case 'E': | |
| 1985 case 'e': | |
| 1986 goto case_hex; | |
| 1987 #endif | |
| 1988 case 'B': | |
| 1989 case 'b': | |
| 1990 state = STATE_binary0; | |
| 1991 break; | |
| 1992 | |
| 1993 case '0': case '1': case '2': case '3': | |
| 1994 case '4': case '5': case '6': case '7': | |
| 1995 state = STATE_octal; | |
| 1996 break; | |
| 1997 | |
| 1998 #if ZEROH | |
| 1999 case '8': case '9': case 'A': | |
| 2000 case 'C': case 'D': case 'F': | |
| 2001 case 'a': case 'c': case 'd': case 'f': | |
| 2002 case_hex: | |
| 2003 state = STATE_hexh; | |
| 2004 break; | |
| 2005 #endif | |
| 2006 case '_': | |
| 2007 state = STATE_octal; | |
| 2008 p++; | |
| 2009 continue; | |
| 2010 | |
| 2011 case 'L': | |
| 2012 if (p[1] == 'i') | |
| 2013 goto real; | |
| 2014 goto done; | |
| 2015 | |
| 2016 default: | |
| 2017 goto done; | |
| 2018 } | |
| 2019 break; | |
| 2020 | |
| 2021 case STATE_decimal: // reading decimal number | |
| 2022 if (!isdigit(c)) | |
| 2023 { | |
| 2024 #if ZEROH | |
| 2025 if (ishex(c) | |
| 2026 || c == 'H' || c == 'h' | |
| 2027 ) | |
| 2028 goto hexh; | |
| 2029 #endif | |
| 2030 if (c == '_') // ignore embedded _ | |
| 2031 { p++; | |
| 2032 continue; | |
| 2033 } | |
| 2034 if (c == '.' && p[1] != '.') | |
| 2035 goto real; | |
| 2036 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2037 c == 'e' || c == 'E') | |
| 2038 { | |
| 2039 real: // It's a real number. Back up and rescan as a real | |
| 2040 p = start; | |
| 2041 return inreal(t); | |
| 2042 } | |
| 2043 else if (c == 'L' && p[1] == 'i') | |
| 2044 goto real; | |
| 2045 goto done; | |
| 2046 } | |
| 2047 break; | |
| 2048 | |
| 2049 case STATE_hex0: // reading hex number | |
| 2050 case STATE_hex: | |
| 2051 if (!ishex(c)) | |
| 2052 { | |
| 2053 if (c == '_') // ignore embedded _ | |
| 2054 { p++; | |
| 2055 continue; | |
| 2056 } | |
| 2057 if (c == '.' && p[1] != '.') | |
| 2058 goto real; | |
| 2059 if (c == 'P' || c == 'p' || c == 'i') | |
| 2060 goto real; | |
| 2061 if (state == STATE_hex0) | |
| 2062 error("Hex digit expected, not '%c'", c); | |
| 2063 goto done; | |
| 2064 } | |
| 2065 state = STATE_hex; | |
| 2066 break; | |
| 2067 | |
| 2068 #if ZEROH | |
| 2069 hexh: | |
| 2070 state = STATE_hexh; | |
| 2071 case STATE_hexh: // parse numbers like 0FFh | |
| 2072 if (!ishex(c)) | |
| 2073 { | |
| 2074 if (c == 'H' || c == 'h') | |
| 2075 { | |
| 2076 p++; | |
| 2077 base = 16; | |
| 2078 goto done; | |
| 2079 } | |
| 2080 else | |
| 2081 { | |
| 2082 // Check for something like 1E3 or 0E24 | |
| 2083 if (memchr((char *)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2084 memchr((char *)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2085 goto real; | |
| 2086 error("Hex digit expected, not '%c'", c); | |
| 2087 goto done; | |
| 2088 } | |
| 2089 } | |
| 2090 break; | |
| 2091 #endif | |
| 2092 | |
| 2093 case STATE_octal: // reading octal number | |
| 2094 case STATE_octale: // reading octal number with non-octal digits | |
| 2095 if (!isoctal(c)) | |
| 2096 { | |
| 2097 #if ZEROH | |
| 2098 if (ishex(c) | |
| 2099 || c == 'H' || c == 'h' | |
| 2100 ) | |
| 2101 goto hexh; | |
| 2102 #endif | |
| 2103 if (c == '_') // ignore embedded _ | |
| 2104 { p++; | |
| 2105 continue; | |
| 2106 } | |
| 2107 if (c == '.' && p[1] != '.') | |
| 2108 goto real; | |
| 2109 if (c == 'i') | |
| 2110 goto real; | |
| 2111 if (isdigit(c)) | |
| 2112 { | |
| 2113 state = STATE_octale; | |
| 2114 } | |
| 2115 else | |
| 2116 goto done; | |
| 2117 } | |
| 2118 break; | |
| 2119 | |
| 2120 case STATE_binary0: // starting binary number | |
| 2121 case STATE_binary: // reading binary number | |
| 2122 if (c != '0' && c != '1') | |
| 2123 { | |
| 2124 #if ZEROH | |
| 2125 if (ishex(c) | |
| 2126 || c == 'H' || c == 'h' | |
| 2127 ) | |
| 2128 goto hexh; | |
| 2129 #endif | |
| 2130 if (c == '_') // ignore embedded _ | |
| 2131 { p++; | |
| 2132 continue; | |
| 2133 } | |
| 2134 if (state == STATE_binary0) | |
| 2135 { error("binary digit expected"); | |
| 2136 state = STATE_error; | |
| 2137 break; | |
| 2138 } | |
| 2139 else | |
| 2140 goto done; | |
| 2141 } | |
| 2142 state = STATE_binary; | |
| 2143 break; | |
| 2144 | |
| 2145 case STATE_error: // for error recovery | |
| 2146 if (!isdigit(c)) // scan until non-digit | |
| 2147 goto done; | |
| 2148 break; | |
| 2149 | |
| 2150 default: | |
| 2151 assert(0); | |
| 2152 } | |
| 2153 stringbuffer.writeByte(c); | |
| 2154 p++; | |
| 2155 } | |
| 2156 done: | |
| 2157 stringbuffer.writeByte(0); // terminate string | |
| 2158 if (state == STATE_octale) | |
| 2159 error("Octal digit expected"); | |
| 2160 | |
| 2161 uinteger_t n; // unsigned >=64 bit integer type | |
| 2162 | |
| 2163 if (stringbuffer.offset == 2 && (state == STATE_decimal || state == STATE_0)) | |
| 2164 n = stringbuffer.data[0] - '0'; | |
| 2165 else | |
| 2166 { | |
| 2167 // Convert string to integer | |
| 2168 #if __DMC__ | |
| 2169 errno = 0; | |
| 2170 n = strtoull((char *)stringbuffer.data,NULL,base); | |
| 2171 if (errno == ERANGE) | |
| 2172 error("integer overflow"); | |
| 2173 #else | |
| 2174 // Not everybody implements strtoull() | |
| 2175 char *p = (char *)stringbuffer.data; | |
| 2176 int r = 10, d; | |
| 2177 | |
| 2178 if (*p == '0') | |
| 2179 { | |
| 2180 if (p[1] == 'x' || p[1] == 'X') | |
| 2181 p += 2, r = 16; | |
| 2182 else if (p[1] == 'b' || p[1] == 'B') | |
| 2183 p += 2, r = 2; | |
| 2184 else if (isdigit(p[1])) | |
| 2185 p += 1, r = 8; | |
| 2186 } | |
| 2187 | |
| 2188 n = 0; | |
| 2189 while (1) | |
| 2190 { | |
| 2191 if (*p >= '0' && *p <= '9') | |
| 2192 d = *p - '0'; | |
| 2193 else if (*p >= 'a' && *p <= 'z') | |
| 2194 d = *p - 'a' + 10; | |
| 2195 else if (*p >= 'A' && *p <= 'Z') | |
| 2196 d = *p - 'A' + 10; | |
| 2197 else | |
| 2198 break; | |
| 2199 if (d >= r) | |
| 2200 break; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2201 uinteger_t n2 = n * r; |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2202 //printf("n2 / r = %llx, n = %llx\n", n2/r, n); |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2203 if (n2 / r != n || n2 + d < n) |
| 159 | 2204 { |
| 2205 error ("integer overflow"); | |
| 2206 break; | |
| 2207 } | |
| 2208 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2209 n = n2 + d; |
| 159 | 2210 p++; |
| 2211 } | |
| 2212 #endif | |
| 2213 if (sizeof(n) > 8 && | |
| 2214 n > 0xFFFFFFFFFFFFFFFFULL) // if n needs more than 64 bits | |
| 2215 error("integer overflow"); | |
| 2216 } | |
| 2217 | |
| 2218 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2219 while (1) | |
| 2220 { unsigned char f; | |
| 2221 | |
| 2222 switch (*p) | |
| 2223 { case 'U': | |
| 2224 case 'u': | |
| 2225 f = FLAGS_unsigned; | |
| 2226 goto L1; | |
| 2227 | |
| 2228 case 'l': | |
| 2229 if (1 || !global.params.useDeprecated) | |
| 2230 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2231 case 'L': | |
| 2232 f = FLAGS_long; | |
| 2233 L1: | |
| 2234 p++; | |
| 2235 if (flags & f) | |
| 2236 error("unrecognized token"); | |
| 2237 flags = (FLAGS) (flags | f); | |
| 2238 continue; | |
| 2239 default: | |
| 2240 break; | |
| 2241 } | |
| 2242 break; | |
| 2243 } | |
| 2244 | |
| 2245 switch (flags) | |
| 2246 { | |
| 2247 case 0: | |
| 2248 /* Octal or Hexadecimal constant. | |
| 2249 * First that fits: int, uint, long, ulong | |
| 2250 */ | |
| 2251 if (n & 0x8000000000000000LL) | |
| 2252 result = TOKuns64v; | |
| 2253 else if (n & 0xFFFFFFFF00000000LL) | |
| 2254 result = TOKint64v; | |
| 2255 else if (n & 0x80000000) | |
| 2256 result = TOKuns32v; | |
| 2257 else | |
| 2258 result = TOKint32v; | |
| 2259 break; | |
| 2260 | |
| 2261 case FLAGS_decimal: | |
| 2262 /* First that fits: int, long, long long | |
| 2263 */ | |
| 2264 if (n & 0x8000000000000000LL) | |
| 2265 { error("signed integer overflow"); | |
| 2266 result = TOKuns64v; | |
| 2267 } | |
| 2268 else if (n & 0xFFFFFFFF80000000LL) | |
| 2269 result = TOKint64v; | |
| 2270 else | |
| 2271 result = TOKint32v; | |
| 2272 break; | |
| 2273 | |
| 2274 case FLAGS_unsigned: | |
| 2275 case FLAGS_decimal | FLAGS_unsigned: | |
| 2276 /* First that fits: uint, ulong | |
| 2277 */ | |
| 2278 if (n & 0xFFFFFFFF00000000LL) | |
| 2279 result = TOKuns64v; | |
| 2280 else | |
| 2281 result = TOKuns32v; | |
| 2282 break; | |
| 2283 | |
| 2284 case FLAGS_decimal | FLAGS_long: | |
| 2285 if (n & 0x8000000000000000LL) | |
| 2286 { error("signed integer overflow"); | |
| 2287 result = TOKuns64v; | |
| 2288 } | |
| 2289 else | |
| 2290 result = TOKint64v; | |
| 2291 break; | |
| 2292 | |
| 2293 case FLAGS_long: | |
| 2294 if (n & 0x8000000000000000LL) | |
| 2295 result = TOKuns64v; | |
| 2296 else | |
| 2297 result = TOKint64v; | |
| 2298 break; | |
| 2299 | |
| 2300 case FLAGS_unsigned | FLAGS_long: | |
| 2301 case FLAGS_decimal | FLAGS_unsigned | FLAGS_long: | |
| 2302 result = TOKuns64v; | |
| 2303 break; | |
| 2304 | |
| 2305 default: | |
| 2306 #ifdef DEBUG | |
| 2307 printf("%x\n",flags); | |
| 2308 #endif | |
| 2309 assert(0); | |
| 2310 } | |
| 2311 t->uns64value = n; | |
| 2312 return result; | |
| 2313 } | |
| 2314 | |
| 2315 /************************************** | |
| 2316 * Read in characters, converting them to real. | |
| 2317 * Bugs: | |
| 2318 * Exponent overflow not detected. | |
| 2319 * Too much requested precision is not detected. | |
| 2320 */ | |
| 2321 | |
| 2322 TOK Lexer::inreal(Token *t) | |
| 2323 #ifdef __DMC__ | |
| 2324 __in | |
| 2325 { | |
| 2326 assert(*p == '.' || isdigit(*p)); | |
| 2327 } | |
| 2328 __out (result) | |
| 2329 { | |
| 2330 switch (result) | |
| 2331 { | |
| 2332 case TOKfloat32v: | |
| 2333 case TOKfloat64v: | |
| 2334 case TOKfloat80v: | |
| 2335 case TOKimaginary32v: | |
| 2336 case TOKimaginary64v: | |
| 2337 case TOKimaginary80v: | |
| 2338 break; | |
| 2339 | |
| 2340 default: | |
| 2341 assert(0); | |
| 2342 } | |
| 2343 } | |
| 2344 __body | |
| 2345 #endif /* __DMC__ */ | |
| 2346 { int dblstate; | |
| 2347 unsigned c; | |
| 2348 char hex; // is this a hexadecimal-floating-constant? | |
| 2349 TOK result; | |
| 2350 | |
| 2351 //printf("Lexer::inreal()\n"); | |
| 2352 stringbuffer.reset(); | |
| 2353 dblstate = 0; | |
| 2354 hex = 0; | |
| 2355 Lnext: | |
| 2356 while (1) | |
| 2357 { | |
| 2358 // Get next char from input | |
| 2359 c = *p++; | |
| 2360 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2361 while (1) | |
| 2362 { | |
| 2363 switch (dblstate) | |
| 2364 { | |
| 2365 case 0: // opening state | |
| 2366 if (c == '0') | |
| 2367 dblstate = 9; | |
| 2368 else if (c == '.') | |
| 2369 dblstate = 3; | |
| 2370 else | |
| 2371 dblstate = 1; | |
| 2372 break; | |
| 2373 | |
| 2374 case 9: | |
| 2375 dblstate = 1; | |
| 2376 if (c == 'X' || c == 'x') | |
| 2377 { hex++; | |
| 2378 break; | |
| 2379 } | |
| 2380 case 1: // digits to left of . | |
| 2381 case 3: // digits to right of . | |
| 2382 case 7: // continuing exponent digits | |
| 2383 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2384 { | |
| 2385 if (c == '_') | |
| 2386 goto Lnext; // ignore embedded '_' | |
| 2387 dblstate++; | |
| 2388 continue; | |
| 2389 } | |
| 2390 break; | |
| 2391 | |
| 2392 case 2: // no more digits to left of . | |
| 2393 if (c == '.') | |
| 2394 { dblstate++; | |
| 2395 break; | |
| 2396 } | |
| 2397 case 4: // no more digits to right of . | |
| 2398 if ((c == 'E' || c == 'e') || | |
| 2399 hex && (c == 'P' || c == 'p')) | |
| 2400 { dblstate = 5; | |
| 2401 hex = 0; // exponent is always decimal | |
| 2402 break; | |
| 2403 } | |
| 2404 if (hex) | |
| 2405 error("binary-exponent-part required"); | |
| 2406 goto done; | |
| 2407 | |
| 2408 case 5: // looking immediately to right of E | |
| 2409 dblstate++; | |
| 2410 if (c == '-' || c == '+') | |
| 2411 break; | |
| 2412 case 6: // 1st exponent digit expected | |
| 2413 if (!isdigit(c)) | |
| 2414 error("exponent expected"); | |
| 2415 dblstate++; | |
| 2416 break; | |
| 2417 | |
| 2418 case 8: // past end of exponent digits | |
| 2419 goto done; | |
| 2420 } | |
| 2421 break; | |
| 2422 } | |
| 2423 stringbuffer.writeByte(c); | |
| 2424 } | |
| 2425 done: | |
| 2426 p--; | |
| 2427 | |
| 2428 stringbuffer.writeByte(0); | |
| 2429 | |
| 2430 #if _WIN32 && __DMC__ | |
| 2431 char *save = __locale_decpoint; | |
| 2432 __locale_decpoint = "."; | |
| 2433 #endif | |
| 2434 #ifdef IN_GCC | |
| 2435 t->float80value = real_t::parse((char *)stringbuffer.data, real_t::LongDouble); | |
| 2436 #else | |
| 2437 t->float80value = strtold((char *)stringbuffer.data, NULL); | |
| 2438 #endif | |
| 2439 errno = 0; | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2440 float strtofres; |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2441 double strtodres; |
| 159 | 2442 switch (*p) |
| 2443 { | |
| 2444 case 'F': | |
| 2445 case 'f': | |
| 2446 #ifdef IN_GCC | |
| 2447 real_t::parse((char *)stringbuffer.data, real_t::Float); | |
| 2448 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2449 strtofres = strtof((char *)stringbuffer.data, NULL); |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2450 // LDC change: don't error on gradual underflow |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2451 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2452 strtofres != 0 && strtofres != HUGE_VALF && strtofres != -HUGE_VALF) |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2453 errno = 0; |
| 159 | 2454 #endif |
| 2455 result = TOKfloat32v; | |
| 2456 p++; | |
| 2457 break; | |
| 2458 | |
| 2459 default: | |
| 2460 #ifdef IN_GCC | |
| 2461 real_t::parse((char *)stringbuffer.data, real_t::Double); | |
| 2462 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2463 strtodres = strtod((char *)stringbuffer.data, NULL); |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2464 // LDC change: don't error on gradual underflow |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2465 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2466 strtodres != 0 && strtodres != HUGE_VAL && strtodres != -HUGE_VAL) |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2467 errno = 0; |
| 159 | 2468 #endif |
| 2469 result = TOKfloat64v; | |
| 2470 break; | |
| 2471 | |
| 2472 case 'l': | |
| 2473 if (!global.params.useDeprecated) | |
| 2474 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2475 case 'L': | |
| 2476 result = TOKfloat80v; | |
| 2477 p++; | |
| 2478 break; | |
| 2479 } | |
| 2480 if (*p == 'i' || *p == 'I') | |
| 2481 { | |
| 2482 if (!global.params.useDeprecated && *p == 'I') | |
| 2483 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2484 p++; | |
| 2485 switch (result) | |
| 2486 { | |
| 2487 case TOKfloat32v: | |
| 2488 result = TOKimaginary32v; | |
| 2489 break; | |
| 2490 case TOKfloat64v: | |
| 2491 result = TOKimaginary64v; | |
| 2492 break; | |
| 2493 case TOKfloat80v: | |
| 2494 result = TOKimaginary80v; | |
| 2495 break; | |
| 2496 } | |
| 2497 } | |
| 2498 #if _WIN32 && __DMC__ | |
| 2499 __locale_decpoint = save; | |
| 2500 #endif | |
| 2501 if (errno == ERANGE) | |
| 2502 error("number is not representable"); | |
| 2503 return result; | |
| 2504 } | |
| 2505 | |
| 2506 /********************************************* | |
| 2507 * Do pragma. | |
| 2508 * Currently, the only pragma supported is: | |
| 2509 * #line linnum [filespec] | |
| 2510 */ | |
| 2511 | |
| 2512 void Lexer::pragma() | |
| 2513 { | |
| 2514 Token tok; | |
| 2515 int linnum; | |
| 2516 char *filespec = NULL; | |
| 2517 Loc loc = this->loc; | |
| 2518 | |
| 2519 scan(&tok); | |
| 2520 if (tok.value != TOKidentifier || tok.ident != Id::line) | |
| 2521 goto Lerr; | |
| 2522 | |
| 2523 scan(&tok); | |
| 2524 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 2525 linnum = tok.uns64value - 1; | |
| 2526 else | |
| 2527 goto Lerr; | |
| 2528 | |
| 2529 while (1) | |
| 2530 { | |
| 2531 switch (*p) | |
| 2532 { | |
| 2533 case 0: | |
| 2534 case 0x1A: | |
| 2535 case '\n': | |
| 2536 Lnewline: | |
| 2537 this->loc.linnum = linnum; | |
| 2538 if (filespec) | |
| 2539 this->loc.filename = filespec; | |
| 2540 return; | |
| 2541 | |
| 2542 case '\r': | |
| 2543 p++; | |
| 2544 if (*p != '\n') | |
| 2545 { p--; | |
| 2546 goto Lnewline; | |
| 2547 } | |
| 2548 continue; | |
| 2549 | |
| 2550 case ' ': | |
| 2551 case '\t': | |
| 2552 case '\v': | |
| 2553 case '\f': | |
| 2554 p++; | |
| 2555 continue; // skip white space | |
| 2556 | |
| 2557 case '_': | |
| 2558 if (mod && memcmp(p, "__FILE__", 8) == 0) | |
| 2559 { | |
| 2560 p += 8; | |
| 2561 filespec = mem.strdup(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 2562 } | |
| 2563 continue; | |
| 2564 | |
| 2565 case '"': | |
| 2566 if (filespec) | |
| 2567 goto Lerr; | |
| 2568 stringbuffer.reset(); | |
| 2569 p++; | |
| 2570 while (1) | |
| 2571 { unsigned c; | |
| 2572 | |
| 2573 c = *p; | |
| 2574 switch (c) | |
| 2575 { | |
| 2576 case '\n': | |
| 2577 case '\r': | |
| 2578 case 0: | |
| 2579 case 0x1A: | |
| 2580 goto Lerr; | |
| 2581 | |
| 2582 case '"': | |
| 2583 stringbuffer.writeByte(0); | |
| 2584 filespec = mem.strdup((char *)stringbuffer.data); | |
| 2585 p++; | |
| 2586 break; | |
| 2587 | |
| 2588 default: | |
| 2589 if (c & 0x80) | |
| 2590 { unsigned u = decodeUTF(); | |
| 2591 if (u == PS || u == LS) | |
| 2592 goto Lerr; | |
| 2593 } | |
| 2594 stringbuffer.writeByte(c); | |
| 2595 p++; | |
| 2596 continue; | |
| 2597 } | |
| 2598 break; | |
| 2599 } | |
| 2600 continue; | |
| 2601 | |
| 2602 default: | |
| 2603 if (*p & 0x80) | |
| 2604 { unsigned u = decodeUTF(); | |
| 2605 if (u == PS || u == LS) | |
| 2606 goto Lnewline; | |
| 2607 } | |
| 2608 goto Lerr; | |
| 2609 } | |
| 2610 } | |
| 2611 | |
| 2612 Lerr: | |
| 2613 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 2614 } | |
| 2615 | |
| 2616 | |
| 2617 /******************************************** | |
| 2618 * Decode UTF character. | |
| 2619 * Issue error messages for invalid sequences. | |
| 2620 * Return decoded character, advance p to last character in UTF sequence. | |
| 2621 */ | |
| 2622 | |
| 2623 unsigned Lexer::decodeUTF() | |
| 2624 { | |
| 2625 dchar_t u; | |
| 2626 unsigned char c; | |
| 2627 unsigned char *s = p; | |
| 2628 size_t len; | |
| 2629 size_t idx; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2630 const char *msg; |
| 159 | 2631 |
| 2632 c = *s; | |
| 2633 assert(c & 0x80); | |
| 2634 | |
| 2635 // Check length of remaining string up to 6 UTF-8 characters | |
| 2636 for (len = 1; len < 6 && s[len]; len++) | |
| 2637 ; | |
| 2638 | |
| 2639 idx = 0; | |
| 2640 msg = utf_decodeChar(s, len, &idx, &u); | |
| 2641 p += idx - 1; | |
| 2642 if (msg) | |
| 2643 { | |
| 2644 error("%s", msg); | |
| 2645 } | |
| 2646 return u; | |
| 2647 } | |
| 2648 | |
| 2649 | |
| 2650 /*************************************************** | |
| 2651 * Parse doc comment embedded between t->ptr and p. | |
| 2652 * Remove trailing blanks and tabs from lines. | |
| 2653 * Replace all newlines with \n. | |
| 2654 * Remove leading comment character from each line. | |
| 2655 * Decide if it's a lineComment or a blockComment. | |
| 2656 * Append to previous one for this token. | |
| 2657 */ | |
| 2658 | |
| 2659 void Lexer::getDocComment(Token *t, unsigned lineComment) | |
| 2660 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2661 /* ct tells us which kind of comment it is: '/', '*', or '+' |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2662 */ |
| 159 | 2663 unsigned char ct = t->ptr[2]; |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2664 |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2665 /* Start of comment text skips over / * *, / + +, or / / / |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2666 */ |
| 159 | 2667 unsigned char *q = t->ptr + 3; // start of comment text |
| 2668 | |
| 2669 unsigned char *qend = p; | |
| 2670 if (ct == '*' || ct == '+') | |
| 2671 qend -= 2; | |
| 2672 | |
| 2673 /* Scan over initial row of ****'s or ++++'s or ////'s | |
| 2674 */ | |
| 2675 for (; q < qend; q++) | |
| 2676 { | |
| 2677 if (*q != ct) | |
| 2678 break; | |
| 2679 } | |
| 2680 | |
| 2681 /* Remove trailing row of ****'s or ++++'s | |
| 2682 */ | |
| 2683 if (ct != '/') | |
| 2684 { | |
| 2685 for (; q < qend; qend--) | |
| 2686 { | |
| 2687 if (qend[-1] != ct) | |
| 2688 break; | |
| 2689 } | |
| 2690 } | |
| 2691 | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2692 /* Comment is now [q .. qend]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2693 * Canonicalize it into buf[]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2694 */ |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2695 OutBuffer buf; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2696 int linestart = 0; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2697 |
| 159 | 2698 for (; q < qend; q++) |
| 2699 { | |
| 2700 unsigned char c = *q; | |
| 2701 | |
| 2702 switch (c) | |
| 2703 { | |
| 2704 case '*': | |
| 2705 case '+': | |
| 2706 if (linestart && c == ct) | |
| 2707 { linestart = 0; | |
| 2708 /* Trim preceding whitespace up to preceding \n | |
| 2709 */ | |
| 2710 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2711 buf.offset--; | |
| 2712 continue; | |
| 2713 } | |
| 2714 break; | |
| 2715 | |
| 2716 case ' ': | |
| 2717 case '\t': | |
| 2718 break; | |
| 2719 | |
| 2720 case '\r': | |
| 2721 if (q[1] == '\n') | |
| 2722 continue; // skip the \r | |
| 2723 goto Lnewline; | |
| 2724 | |
| 2725 default: | |
| 2726 if (c == 226) | |
| 2727 { | |
| 2728 // If LS or PS | |
| 2729 if (q[1] == 128 && | |
| 2730 (q[2] == 168 || q[2] == 169)) | |
| 2731 { | |
| 2732 q += 2; | |
| 2733 goto Lnewline; | |
| 2734 } | |
| 2735 } | |
| 2736 linestart = 0; | |
| 2737 break; | |
| 2738 | |
| 2739 Lnewline: | |
| 2740 c = '\n'; // replace all newlines with \n | |
| 2741 case '\n': | |
| 2742 linestart = 1; | |
| 2743 | |
| 2744 /* Trim trailing whitespace | |
| 2745 */ | |
| 2746 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2747 buf.offset--; | |
| 2748 | |
| 2749 break; | |
| 2750 } | |
| 2751 buf.writeByte(c); | |
| 2752 } | |
| 2753 | |
| 2754 // Always end with a newline | |
| 2755 if (!buf.offset || buf.data[buf.offset - 1] != '\n') | |
| 2756 buf.writeByte('\n'); | |
| 2757 | |
| 2758 buf.writeByte(0); | |
| 2759 | |
| 2760 // It's a line comment if the start of the doc comment comes | |
| 2761 // after other non-whitespace on the same line. | |
| 2762 unsigned char** dc = (lineComment && anyToken) | |
| 2763 ? &t->lineComment | |
| 2764 : &t->blockComment; | |
| 2765 | |
| 2766 // Combine with previous doc comment, if any | |
| 2767 if (*dc) | |
| 2768 *dc = combineComments(*dc, (unsigned char *)buf.data); | |
| 2769 else | |
| 2770 *dc = (unsigned char *)buf.extractData(); | |
| 2771 } | |
| 2772 | |
| 2773 /******************************************** | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2774 * Combine two document comments into one, |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2775 * separated by a newline. |
| 159 | 2776 */ |
| 2777 | |
| 2778 unsigned char *Lexer::combineComments(unsigned char *c1, unsigned char *c2) | |
| 2779 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2780 //printf("Lexer::combineComments('%s', '%s')\n", c1, c2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2781 |
| 159 | 2782 unsigned char *c = c2; |
| 2783 | |
| 2784 if (c1) | |
| 2785 { c = c1; | |
| 2786 if (c2) | |
| 2787 { size_t len1 = strlen((char *)c1); | |
| 2788 size_t len2 = strlen((char *)c2); | |
| 2789 | |
| 2790 c = (unsigned char *)mem.malloc(len1 + 1 + len2 + 1); | |
| 2791 memcpy(c, c1, len1); | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2792 if (len1 && c1[len1 - 1] != '\n') |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2793 { c[len1] = '\n'; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2794 len1++; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2795 } |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2796 memcpy(c + len1, c2, len2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2797 c[len1 + len2] = 0; |
| 159 | 2798 } |
| 2799 } | |
| 2800 return c; | |
| 2801 } | |
| 2802 | |
| 2803 /******************************************** | |
| 2804 * Create an identifier in the string table. | |
| 2805 */ | |
| 2806 | |
| 2807 Identifier *Lexer::idPool(const char *s) | |
| 2808 { | |
| 2809 size_t len = strlen(s); | |
| 2810 StringValue *sv = stringtable.update(s, len); | |
| 2811 Identifier *id = (Identifier *) sv->ptrvalue; | |
| 2812 if (!id) | |
| 2813 { | |
| 2814 id = new Identifier(sv->lstring.string, TOKidentifier); | |
| 2815 sv->ptrvalue = id; | |
| 2816 } | |
| 2817 return id; | |
| 2818 } | |
| 2819 | |
| 2820 /********************************************* | |
| 2821 * Create a unique identifier using the prefix s. | |
| 2822 */ | |
| 2823 | |
| 2824 Identifier *Lexer::uniqueId(const char *s, int num) | |
| 2825 { char buffer[32]; | |
| 2826 size_t slen = strlen(s); | |
| 2827 | |
| 2828 assert(slen + sizeof(num) * 3 + 1 <= sizeof(buffer)); | |
| 2829 sprintf(buffer, "%s%d", s, num); | |
| 2830 return idPool(buffer); | |
| 2831 } | |
| 2832 | |
| 2833 Identifier *Lexer::uniqueId(const char *s) | |
| 2834 { | |
| 2835 static int num; | |
| 2836 return uniqueId(s, ++num); | |
| 2837 } | |
| 2838 | |
| 2839 /**************************************** | |
| 2840 */ | |
| 2841 | |
| 2842 struct Keyword | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2843 { const char *name; |
| 159 | 2844 enum TOK value; |
| 2845 }; | |
| 2846 | |
| 2847 static Keyword keywords[] = | |
| 2848 { | |
| 2849 // { "", TOK }, | |
| 2850 | |
| 2851 { "this", TOKthis }, | |
| 2852 { "super", TOKsuper }, | |
| 2853 { "assert", TOKassert }, | |
| 2854 { "null", TOKnull }, | |
| 2855 { "true", TOKtrue }, | |
| 2856 { "false", TOKfalse }, | |
| 2857 { "cast", TOKcast }, | |
| 2858 { "new", TOKnew }, | |
| 2859 { "delete", TOKdelete }, | |
| 2860 { "throw", TOKthrow }, | |
| 2861 { "module", TOKmodule }, | |
| 2862 { "pragma", TOKpragma }, | |
| 2863 { "typeof", TOKtypeof }, | |
| 2864 { "typeid", TOKtypeid }, | |
| 2865 | |
| 2866 { "template", TOKtemplate }, | |
| 2867 | |
| 2868 { "void", TOKvoid }, | |
| 2869 { "byte", TOKint8 }, | |
| 2870 { "ubyte", TOKuns8 }, | |
| 2871 { "short", TOKint16 }, | |
| 2872 { "ushort", TOKuns16 }, | |
| 2873 { "int", TOKint32 }, | |
| 2874 { "uint", TOKuns32 }, | |
| 2875 { "long", TOKint64 }, | |
| 2876 { "ulong", TOKuns64 }, | |
| 2877 { "cent", TOKcent, }, | |
| 2878 { "ucent", TOKucent, }, | |
| 2879 { "float", TOKfloat32 }, | |
| 2880 { "double", TOKfloat64 }, | |
| 2881 { "real", TOKfloat80 }, | |
| 2882 | |
| 2883 { "bool", TOKbool }, | |
| 2884 { "char", TOKchar }, | |
| 2885 { "wchar", TOKwchar }, | |
| 2886 { "dchar", TOKdchar }, | |
| 2887 | |
| 2888 { "ifloat", TOKimaginary32 }, | |
| 2889 { "idouble", TOKimaginary64 }, | |
| 2890 { "ireal", TOKimaginary80 }, | |
| 2891 | |
| 2892 { "cfloat", TOKcomplex32 }, | |
| 2893 { "cdouble", TOKcomplex64 }, | |
| 2894 { "creal", TOKcomplex80 }, | |
| 2895 | |
| 2896 { "delegate", TOKdelegate }, | |
| 2897 { "function", TOKfunction }, | |
| 2898 | |
| 2899 { "is", TOKis }, | |
| 2900 { "if", TOKif }, | |
| 2901 { "else", TOKelse }, | |
| 2902 { "while", TOKwhile }, | |
| 2903 { "for", TOKfor }, | |
| 2904 { "do", TOKdo }, | |
| 2905 { "switch", TOKswitch }, | |
| 2906 { "case", TOKcase }, | |
| 2907 { "default", TOKdefault }, | |
| 2908 { "break", TOKbreak }, | |
| 2909 { "continue", TOKcontinue }, | |
| 2910 { "synchronized", TOKsynchronized }, | |
| 2911 { "return", TOKreturn }, | |
| 2912 { "goto", TOKgoto }, | |
| 2913 { "try", TOKtry }, | |
| 2914 { "catch", TOKcatch }, | |
| 2915 { "finally", TOKfinally }, | |
| 2916 { "with", TOKwith }, | |
| 2917 { "asm", TOKasm }, | |
| 2918 { "foreach", TOKforeach }, | |
| 2919 { "foreach_reverse", TOKforeach_reverse }, | |
| 2920 { "scope", TOKscope }, | |
| 2921 | |
| 2922 { "struct", TOKstruct }, | |
| 2923 { "class", TOKclass }, | |
| 2924 { "interface", TOKinterface }, | |
| 2925 { "union", TOKunion }, | |
| 2926 { "enum", TOKenum }, | |
| 2927 { "import", TOKimport }, | |
| 2928 { "mixin", TOKmixin }, | |
| 2929 { "static", TOKstatic }, | |
| 2930 { "final", TOKfinal }, | |
| 2931 { "const", TOKconst }, | |
| 2932 { "typedef", TOKtypedef }, | |
| 2933 { "alias", TOKalias }, | |
| 2934 { "override", TOKoverride }, | |
| 2935 { "abstract", TOKabstract }, | |
| 2936 { "volatile", TOKvolatile }, | |
| 2937 { "debug", TOKdebug }, | |
| 2938 { "deprecated", TOKdeprecated }, | |
| 2939 { "in", TOKin }, | |
| 2940 { "out", TOKout }, | |
| 2941 { "inout", TOKinout }, | |
| 2942 { "lazy", TOKlazy }, | |
| 2943 { "auto", TOKauto }, | |
| 2944 | |
| 2945 { "align", TOKalign }, | |
| 2946 { "extern", TOKextern }, | |
| 2947 { "private", TOKprivate }, | |
| 2948 { "package", TOKpackage }, | |
| 2949 { "protected", TOKprotected }, | |
| 2950 { "public", TOKpublic }, | |
| 2951 { "export", TOKexport }, | |
| 2952 | |
| 2953 { "body", TOKbody }, | |
| 2954 { "invariant", TOKinvariant }, | |
| 2955 { "unittest", TOKunittest }, | |
| 2956 { "version", TOKversion }, | |
| 2957 //{ "manifest", TOKmanifest }, | |
| 2958 | |
| 2959 // Added after 1.0 | |
| 2960 { "ref", TOKref }, | |
| 2961 { "macro", TOKmacro }, | |
| 336 | 2962 #if DMDV2 |
| 159 | 2963 { "pure", TOKpure }, |
| 2964 { "nothrow", TOKnothrow }, | |
| 336 | 2965 { "__thread", TOKtls }, |
| 159 | 2966 { "__traits", TOKtraits }, |
| 2967 { "__overloadset", TOKoverloadset }, | |
| 336 | 2968 { "__FILE__", TOKfile }, |
| 2969 { "__LINE__", TOKline }, | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2970 { "shared", TOKshared }, |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2971 { "immutable", TOKimmutable }, |
| 159 | 2972 #endif |
| 2973 }; | |
| 2974 | |
| 2975 int Token::isKeyword() | |
| 2976 { | |
| 2977 for (unsigned u = 0; u < sizeof(keywords) / sizeof(keywords[0]); u++) | |
| 2978 { | |
| 2979 if (keywords[u].value == value) | |
| 2980 return 1; | |
| 2981 } | |
| 2982 return 0; | |
| 2983 } | |
| 2984 | |
| 2985 void Lexer::initKeywords() | |
| 2986 { StringValue *sv; | |
| 2987 unsigned u; | |
| 2988 enum TOK v; | |
| 2989 unsigned nkeywords = sizeof(keywords) / sizeof(keywords[0]); | |
| 2990 | |
| 2991 if (global.params.Dversion == 1) | |
| 2992 nkeywords -= 2; | |
| 2993 | |
| 2994 cmtable_init(); | |
| 2995 | |
| 2996 for (u = 0; u < nkeywords; u++) | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2997 { const char *s; |
| 159 | 2998 |
| 2999 //printf("keyword[%d] = '%s'\n",u, keywords[u].name); | |
| 3000 s = keywords[u].name; | |
| 3001 v = keywords[u].value; | |
| 3002 sv = stringtable.insert(s, strlen(s)); | |
| 3003 sv->ptrvalue = (void *) new Identifier(sv->lstring.string,v); | |
| 3004 | |
| 3005 //printf("tochars[%d] = '%s'\n",v, s); | |
| 3006 Token::tochars[v] = s; | |
| 3007 } | |
| 3008 | |
| 3009 Token::tochars[TOKeof] = "EOF"; | |
| 3010 Token::tochars[TOKlcurly] = "{"; | |
| 3011 Token::tochars[TOKrcurly] = "}"; | |
| 3012 Token::tochars[TOKlparen] = "("; | |
| 3013 Token::tochars[TOKrparen] = ")"; | |
| 3014 Token::tochars[TOKlbracket] = "["; | |
| 3015 Token::tochars[TOKrbracket] = "]"; | |
| 3016 Token::tochars[TOKsemicolon] = ";"; | |
| 3017 Token::tochars[TOKcolon] = ":"; | |
| 3018 Token::tochars[TOKcomma] = ","; | |
| 3019 Token::tochars[TOKdot] = "."; | |
| 3020 Token::tochars[TOKxor] = "^"; | |
| 3021 Token::tochars[TOKxorass] = "^="; | |
| 3022 Token::tochars[TOKassign] = "="; | |
| 3023 Token::tochars[TOKconstruct] = "="; | |
| 336 | 3024 #if DMDV2 |
| 159 | 3025 Token::tochars[TOKblit] = "="; |
| 3026 #endif | |
| 3027 Token::tochars[TOKlt] = "<"; | |
| 3028 Token::tochars[TOKgt] = ">"; | |
| 3029 Token::tochars[TOKle] = "<="; | |
| 3030 Token::tochars[TOKge] = ">="; | |
| 3031 Token::tochars[TOKequal] = "=="; | |
| 3032 Token::tochars[TOKnotequal] = "!="; | |
| 3033 Token::tochars[TOKnotidentity] = "!is"; | |
| 3034 Token::tochars[TOKtobool] = "!!"; | |
| 3035 | |
| 3036 Token::tochars[TOKunord] = "!<>="; | |
| 3037 Token::tochars[TOKue] = "!<>"; | |
| 3038 Token::tochars[TOKlg] = "<>"; | |
| 3039 Token::tochars[TOKleg] = "<>="; | |
| 3040 Token::tochars[TOKule] = "!>"; | |
| 3041 Token::tochars[TOKul] = "!>="; | |
| 3042 Token::tochars[TOKuge] = "!<"; | |
| 3043 Token::tochars[TOKug] = "!<="; | |
| 3044 | |
| 3045 Token::tochars[TOKnot] = "!"; | |
| 3046 Token::tochars[TOKtobool] = "!!"; | |
| 3047 Token::tochars[TOKshl] = "<<"; | |
| 3048 Token::tochars[TOKshr] = ">>"; | |
| 3049 Token::tochars[TOKushr] = ">>>"; | |
| 3050 Token::tochars[TOKadd] = "+"; | |
| 3051 Token::tochars[TOKmin] = "-"; | |
| 3052 Token::tochars[TOKmul] = "*"; | |
| 3053 Token::tochars[TOKdiv] = "/"; | |
| 3054 Token::tochars[TOKmod] = "%"; | |
| 3055 Token::tochars[TOKslice] = ".."; | |
| 3056 Token::tochars[TOKdotdotdot] = "..."; | |
| 3057 Token::tochars[TOKand] = "&"; | |
| 3058 Token::tochars[TOKandand] = "&&"; | |
| 3059 Token::tochars[TOKor] = "|"; | |
| 3060 Token::tochars[TOKoror] = "||"; | |
| 3061 Token::tochars[TOKarray] = "[]"; | |
| 3062 Token::tochars[TOKindex] = "[i]"; | |
| 3063 Token::tochars[TOKaddress] = "&"; | |
| 3064 Token::tochars[TOKstar] = "*"; | |
| 3065 Token::tochars[TOKtilde] = "~"; | |
| 3066 Token::tochars[TOKdollar] = "$"; | |
| 3067 Token::tochars[TOKcast] = "cast"; | |
| 3068 Token::tochars[TOKplusplus] = "++"; | |
| 3069 Token::tochars[TOKminusminus] = "--"; | |
| 3070 Token::tochars[TOKtype] = "type"; | |
| 3071 Token::tochars[TOKquestion] = "?"; | |
| 3072 Token::tochars[TOKneg] = "-"; | |
| 3073 Token::tochars[TOKuadd] = "+"; | |
| 3074 Token::tochars[TOKvar] = "var"; | |
| 3075 Token::tochars[TOKaddass] = "+="; | |
| 3076 Token::tochars[TOKminass] = "-="; | |
| 3077 Token::tochars[TOKmulass] = "*="; | |
| 3078 Token::tochars[TOKdivass] = "/="; | |
| 3079 Token::tochars[TOKmodass] = "%="; | |
| 3080 Token::tochars[TOKshlass] = "<<="; | |
| 3081 Token::tochars[TOKshrass] = ">>="; | |
| 3082 Token::tochars[TOKushrass] = ">>>="; | |
| 3083 Token::tochars[TOKandass] = "&="; | |
| 3084 Token::tochars[TOKorass] = "|="; | |
| 3085 Token::tochars[TOKcatass] = "~="; | |
| 3086 Token::tochars[TOKcat] = "~"; | |
| 3087 Token::tochars[TOKcall] = "call"; | |
| 3088 Token::tochars[TOKidentity] = "is"; | |
| 3089 Token::tochars[TOKnotidentity] = "!is"; | |
| 3090 | |
| 3091 Token::tochars[TOKorass] = "|="; | |
| 3092 Token::tochars[TOKidentifier] = "identifier"; | |
| 3093 | |
| 3094 // For debugging | |
| 3095 Token::tochars[TOKdotexp] = "dotexp"; | |
| 3096 Token::tochars[TOKdotti] = "dotti"; | |
| 3097 Token::tochars[TOKdotvar] = "dotvar"; | |
| 3098 Token::tochars[TOKdottype] = "dottype"; | |
| 3099 Token::tochars[TOKsymoff] = "symoff"; | |
| 3100 Token::tochars[TOKtypedot] = "typedot"; | |
| 3101 Token::tochars[TOKarraylength] = "arraylength"; | |
| 3102 Token::tochars[TOKarrayliteral] = "arrayliteral"; | |
| 3103 Token::tochars[TOKassocarrayliteral] = "assocarrayliteral"; | |
| 3104 Token::tochars[TOKstructliteral] = "structliteral"; | |
| 3105 Token::tochars[TOKstring] = "string"; | |
| 3106 Token::tochars[TOKdsymbol] = "symbol"; | |
| 3107 Token::tochars[TOKtuple] = "tuple"; | |
| 3108 Token::tochars[TOKdeclaration] = "declaration"; | |
| 3109 Token::tochars[TOKdottd] = "dottd"; | |
| 3110 Token::tochars[TOKon_scope_exit] = "scope(exit)"; | |
| 3111 } |