Mercurial > projects > ldc
annotate dmd/lexer.c @ 1195:e961851fb8be
Merged DMD 1.042.
| author | Tomas Lindquist Olsen <tomas.l.olsen gmail.com> |
|---|---|
| date | Fri, 03 Apr 2009 17:59:34 +0200 |
| parents | 226c07c71967 |
| children | 79758fd2f48a |
| rev | line source |
|---|---|
| 159 | 1 |
| 2 // Compiler implementation of the D programming language | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
3 // Copyright (c) 1999-2009 by Digital Mars |
| 159 | 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 | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
11 #include <cmath> |
|
872
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> | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
23 #include <time.h> // for time() and ctime() |
| 159 | 24 |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
25 #include "rmem.h" |
| 159 | 26 |
| 27 #include "stringtable.h" | |
| 28 | |
| 29 #include "lexer.h" | |
| 30 #include "utf.h" | |
| 31 #include "identifier.h" | |
| 32 #include "id.h" | |
| 33 #include "module.h" | |
| 34 | |
| 35 #if _WIN32 && __DMC__ | |
| 36 // from \dm\src\include\setlocal.h | |
| 37 extern "C" char * __cdecl __locale_decpoint; | |
| 38 #endif | |
| 39 | |
| 40 extern int HtmlNamedEntity(unsigned char *p, int length); | |
| 41 | |
| 42 #define LS 0x2028 // UTF line separator | |
| 43 #define PS 0x2029 // UTF paragraph separator | |
| 44 | |
| 45 /******************************************** | |
| 46 * Do our own char maps | |
| 47 */ | |
| 48 | |
| 49 static unsigned char cmtable[256]; | |
| 50 | |
| 51 const int CMoctal = 0x1; | |
| 52 const int CMhex = 0x2; | |
| 53 const int CMidchar = 0x4; | |
| 54 | |
| 55 inline unsigned char isoctal (unsigned char c) { return cmtable[c] & CMoctal; } | |
| 56 inline unsigned char ishex (unsigned char c) { return cmtable[c] & CMhex; } | |
| 57 inline unsigned char isidchar(unsigned char c) { return cmtable[c] & CMidchar; } | |
| 58 | |
| 59 static void cmtable_init() | |
| 60 { | |
| 61 for (unsigned c = 0; c < sizeof(cmtable) / sizeof(cmtable[0]); c++) | |
| 62 { | |
| 63 if ('0' <= c && c <= '7') | |
| 64 cmtable[c] |= CMoctal; | |
| 65 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 66 cmtable[c] |= CMhex; | |
| 67 if (isalnum(c) || c == '_') | |
| 68 cmtable[c] |= CMidchar; | |
| 69 } | |
| 70 } | |
| 71 | |
| 72 | |
| 73 /************************* Token **********************************************/ | |
| 74 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
75 const char *Token::tochars[TOKMAX]; |
| 159 | 76 |
| 77 void *Token::operator new(size_t size) | |
| 78 { Token *t; | |
| 79 | |
| 80 if (Lexer::freelist) | |
| 81 { | |
| 82 t = Lexer::freelist; | |
| 83 Lexer::freelist = t->next; | |
| 84 return t; | |
| 85 } | |
| 86 | |
| 87 return ::operator new(size); | |
| 88 } | |
| 89 | |
| 90 #ifdef DEBUG | |
| 91 void Token::print() | |
| 92 { | |
| 93 fprintf(stdmsg, "%s\n", toChars()); | |
| 94 } | |
| 95 #endif | |
| 96 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
97 const char *Token::toChars() |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
98 { const char *p; |
| 159 | 99 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 100 | |
| 101 p = buffer; | |
| 102 switch (value) | |
| 103 { | |
| 104 case TOKint32v: | |
| 105 sprintf(buffer,"%d",(d_int32)int64value); | |
| 106 break; | |
| 107 | |
| 108 case TOKuns32v: | |
| 109 case TOKcharv: | |
| 110 case TOKwcharv: | |
| 111 case TOKdcharv: | |
| 112 sprintf(buffer,"%uU",(d_uns32)uns64value); | |
| 113 break; | |
| 114 | |
| 115 case TOKint64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
116 sprintf(buffer,"%jdL",int64value); |
| 159 | 117 break; |
| 118 | |
| 119 case TOKuns64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
120 sprintf(buffer,"%juUL",uns64value); |
| 159 | 121 break; |
| 122 | |
| 123 #if IN_GCC | |
| 124 case TOKfloat32v: | |
| 125 case TOKfloat64v: | |
| 126 case TOKfloat80v: | |
| 127 float80value.format(buffer, sizeof(buffer)); | |
| 128 break; | |
| 129 case TOKimaginary32v: | |
| 130 case TOKimaginary64v: | |
| 131 case TOKimaginary80v: | |
| 132 float80value.format(buffer, sizeof(buffer)); | |
| 133 // %% buffer | |
| 134 strcat(buffer, "i"); | |
| 135 break; | |
| 136 #else | |
| 137 case TOKfloat32v: | |
| 138 sprintf(buffer,"%Lgf", float80value); | |
| 139 break; | |
| 140 | |
| 141 case TOKfloat64v: | |
| 142 sprintf(buffer,"%Lg", float80value); | |
| 143 break; | |
| 144 | |
| 145 case TOKfloat80v: | |
| 146 sprintf(buffer,"%LgL", float80value); | |
| 147 break; | |
| 148 | |
| 149 case TOKimaginary32v: | |
| 150 sprintf(buffer,"%Lgfi", float80value); | |
| 151 break; | |
| 152 | |
| 153 case TOKimaginary64v: | |
| 154 sprintf(buffer,"%Lgi", float80value); | |
| 155 break; | |
| 156 | |
| 157 case TOKimaginary80v: | |
| 158 sprintf(buffer,"%LgLi", float80value); | |
| 159 break; | |
| 160 #endif | |
| 161 | |
| 162 case TOKstring: | |
| 163 #if CSTRINGS | |
| 164 p = string; | |
| 165 #else | |
| 166 { OutBuffer buf; | |
| 167 | |
| 168 buf.writeByte('"'); | |
| 169 for (size_t i = 0; i < len; ) | |
| 170 { unsigned c; | |
| 171 | |
| 172 utf_decodeChar((unsigned char *)ustring, len, &i, &c); | |
| 173 switch (c) | |
| 174 { | |
| 175 case 0: | |
| 176 break; | |
| 177 | |
| 178 case '"': | |
| 179 case '\\': | |
| 180 buf.writeByte('\\'); | |
| 181 default: | |
| 182 if (isprint(c)) | |
| 183 buf.writeByte(c); | |
| 184 else if (c <= 0x7F) | |
| 185 buf.printf("\\x%02x", c); | |
| 186 else if (c <= 0xFFFF) | |
| 187 buf.printf("\\u%04x", c); | |
| 188 else | |
| 189 buf.printf("\\U%08x", c); | |
| 190 continue; | |
| 191 } | |
| 192 break; | |
| 193 } | |
| 194 buf.writeByte('"'); | |
| 195 if (postfix) | |
| 196 buf.writeByte('"'); | |
| 197 buf.writeByte(0); | |
| 198 p = (char *)buf.extractData(); | |
| 199 } | |
| 200 #endif | |
| 201 break; | |
| 202 | |
| 203 case TOKidentifier: | |
| 204 case TOKenum: | |
| 205 case TOKstruct: | |
| 206 case TOKimport: | |
| 207 CASE_BASIC_TYPES: | |
| 208 p = ident->toChars(); | |
| 209 break; | |
| 210 | |
| 211 default: | |
| 212 p = toChars(value); | |
| 213 break; | |
| 214 } | |
| 215 return p; | |
| 216 } | |
| 217 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
218 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
|
219 { const char *p; |
| 159 | 220 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 221 | |
| 222 p = tochars[value]; | |
| 223 if (!p) | |
| 224 { sprintf(buffer,"TOK%d",value); | |
| 225 p = buffer; | |
| 226 } | |
| 227 return p; | |
| 228 } | |
| 229 | |
| 230 /*************************** Lexer ********************************************/ | |
| 231 | |
| 232 Token *Lexer::freelist = NULL; | |
| 233 StringTable Lexer::stringtable; | |
| 234 OutBuffer Lexer::stringbuffer; | |
| 235 | |
| 236 Lexer::Lexer(Module *mod, | |
| 237 unsigned char *base, unsigned begoffset, unsigned endoffset, | |
| 238 int doDocComment, int commentToken) | |
| 239 : loc(mod, 1) | |
| 240 { | |
| 241 //printf("Lexer::Lexer(%p,%d)\n",base,length); | |
| 242 //printf("lexer.mod = %p, %p\n", mod, this->loc.mod); | |
| 243 memset(&token,0,sizeof(token)); | |
| 244 this->base = base; | |
| 245 this->end = base + endoffset; | |
| 246 p = base + begoffset; | |
| 247 this->mod = mod; | |
| 248 this->doDocComment = doDocComment; | |
| 249 this->anyToken = 0; | |
| 250 this->commentToken = commentToken; | |
| 251 //initKeywords(); | |
| 252 | |
| 253 /* If first line starts with '#!', ignore the line | |
| 254 */ | |
| 255 | |
| 256 if (p[0] == '#' && p[1] =='!') | |
| 257 { | |
| 258 p += 2; | |
| 259 while (1) | |
| 260 { unsigned char c = *p; | |
| 261 switch (c) | |
| 262 { | |
| 263 case '\n': | |
| 264 p++; | |
| 265 break; | |
| 266 | |
| 267 case '\r': | |
| 268 p++; | |
| 269 if (*p == '\n') | |
| 270 p++; | |
| 271 break; | |
| 272 | |
| 273 case 0: | |
| 274 case 0x1A: | |
| 275 break; | |
| 276 | |
| 277 default: | |
| 278 if (c & 0x80) | |
| 279 { unsigned u = decodeUTF(); | |
| 280 if (u == PS || u == LS) | |
| 281 break; | |
| 282 } | |
| 283 p++; | |
| 284 continue; | |
| 285 } | |
| 286 break; | |
| 287 } | |
| 288 loc.linnum = 2; | |
| 289 } | |
| 290 } | |
| 291 | |
| 292 | |
| 293 void Lexer::error(const char *format, ...) | |
| 294 { | |
| 295 if (mod && !global.gag) | |
| 296 { | |
| 297 char *p = loc.toChars(); | |
| 298 if (*p) | |
| 299 fprintf(stdmsg, "%s: ", p); | |
| 300 mem.free(p); | |
| 301 | |
| 302 va_list ap; | |
| 303 va_start(ap, format); | |
| 304 vfprintf(stdmsg, format, ap); | |
| 305 va_end(ap); | |
| 306 | |
| 307 fprintf(stdmsg, "\n"); | |
| 308 fflush(stdmsg); | |
| 309 | |
| 310 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 311 fatal(); | |
| 312 } | |
| 313 global.errors++; | |
| 314 } | |
| 315 | |
| 316 void Lexer::error(Loc loc, const char *format, ...) | |
| 317 { | |
| 318 if (mod && !global.gag) | |
| 319 { | |
| 320 char *p = loc.toChars(); | |
| 321 if (*p) | |
| 322 fprintf(stdmsg, "%s: ", p); | |
| 323 mem.free(p); | |
| 324 | |
| 325 va_list ap; | |
| 326 va_start(ap, format); | |
| 327 vfprintf(stdmsg, format, ap); | |
| 328 va_end(ap); | |
| 329 | |
| 330 fprintf(stdmsg, "\n"); | |
| 331 fflush(stdmsg); | |
| 332 | |
| 333 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 334 fatal(); | |
| 335 } | |
| 336 global.errors++; | |
| 337 } | |
| 338 | |
| 339 TOK Lexer::nextToken() | |
| 340 { Token *t; | |
| 341 | |
| 342 if (token.next) | |
| 343 { | |
| 344 t = token.next; | |
| 345 memcpy(&token,t,sizeof(Token)); | |
| 346 t->next = freelist; | |
| 347 freelist = t; | |
| 348 } | |
| 349 else | |
| 350 { | |
| 351 scan(&token); | |
| 352 } | |
| 353 //token.print(); | |
| 354 return token.value; | |
| 355 } | |
| 356 | |
| 357 Token *Lexer::peek(Token *ct) | |
| 358 { Token *t; | |
| 359 | |
| 360 if (ct->next) | |
| 361 t = ct->next; | |
| 362 else | |
| 363 { | |
| 364 t = new Token(); | |
| 365 scan(t); | |
| 366 t->next = NULL; | |
| 367 ct->next = t; | |
| 368 } | |
| 369 return t; | |
| 370 } | |
| 371 | |
|
717
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
372 /*********************** |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
373 * Look ahead at next token's value. |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
374 */ |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
375 |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
376 TOK Lexer::peekNext() |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
377 { |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
378 return peek(&token)->value; |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
379 } |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
380 |
| 159 | 381 /********************************* |
| 382 * tk is on the opening (. | |
| 383 * Look ahead and return token that is past the closing ). | |
| 384 */ | |
| 385 | |
| 386 Token *Lexer::peekPastParen(Token *tk) | |
| 387 { | |
| 388 //printf("peekPastParen()\n"); | |
| 389 int parens = 1; | |
| 390 int curlynest = 0; | |
| 391 while (1) | |
| 392 { | |
| 393 tk = peek(tk); | |
| 394 //tk->print(); | |
| 395 switch (tk->value) | |
| 396 { | |
| 397 case TOKlparen: | |
| 398 parens++; | |
| 399 continue; | |
| 400 | |
| 401 case TOKrparen: | |
| 402 --parens; | |
| 403 if (parens) | |
| 404 continue; | |
| 405 tk = peek(tk); | |
| 406 break; | |
| 407 | |
| 408 case TOKlcurly: | |
| 409 curlynest++; | |
| 410 continue; | |
| 411 | |
| 412 case TOKrcurly: | |
| 413 if (--curlynest >= 0) | |
| 414 continue; | |
| 415 break; | |
| 416 | |
| 417 case TOKsemicolon: | |
| 418 if (curlynest) | |
| 419 continue; | |
| 420 break; | |
| 421 | |
| 422 case TOKeof: | |
| 423 break; | |
| 424 | |
| 425 default: | |
| 426 continue; | |
| 427 } | |
| 428 return tk; | |
| 429 } | |
| 430 } | |
| 431 | |
| 432 /********************************** | |
| 433 * Determine if string is a valid Identifier. | |
| 434 * Placed here because of commonality with Lexer functionality. | |
| 435 * Returns: | |
| 436 * 0 invalid | |
| 437 */ | |
| 438 | |
| 439 int Lexer::isValidIdentifier(char *p) | |
| 440 { | |
| 441 size_t len; | |
| 442 size_t idx; | |
| 443 | |
| 444 if (!p || !*p) | |
| 445 goto Linvalid; | |
| 446 | |
| 447 if (*p >= '0' && *p <= '9') // beware of isdigit() on signed chars | |
| 448 goto Linvalid; | |
| 449 | |
| 450 len = strlen(p); | |
| 451 idx = 0; | |
| 452 while (p[idx]) | |
| 453 { dchar_t dc; | |
| 454 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
455 const char *q = utf_decodeChar((unsigned char *)p, len, &idx, &dc); |
| 159 | 456 if (q) |
| 457 goto Linvalid; | |
| 458 | |
| 459 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) | |
| 460 goto Linvalid; | |
| 461 } | |
| 462 return 1; | |
| 463 | |
| 464 Linvalid: | |
| 465 return 0; | |
| 466 } | |
| 467 | |
| 468 /**************************** | |
| 469 * Turn next token in buffer into a token. | |
| 470 */ | |
| 471 | |
| 472 void Lexer::scan(Token *t) | |
| 473 { | |
| 474 unsigned lastLine = loc.linnum; | |
| 475 unsigned linnum; | |
| 476 | |
| 477 t->blockComment = NULL; | |
| 478 t->lineComment = NULL; | |
| 479 while (1) | |
| 480 { | |
| 481 t->ptr = p; | |
| 482 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 483 switch (*p) | |
| 484 { | |
| 485 case 0: | |
| 486 case 0x1A: | |
| 487 t->value = TOKeof; // end of file | |
| 488 return; | |
| 489 | |
| 490 case ' ': | |
| 491 case '\t': | |
| 492 case '\v': | |
| 493 case '\f': | |
| 494 p++; | |
| 495 continue; // skip white space | |
| 496 | |
| 497 case '\r': | |
| 498 p++; | |
| 499 if (*p != '\n') // if CR stands by itself | |
| 500 loc.linnum++; | |
| 501 continue; // skip white space | |
| 502 | |
| 503 case '\n': | |
| 504 p++; | |
| 505 loc.linnum++; | |
| 506 continue; // skip white space | |
| 507 | |
| 508 case '0': case '1': case '2': case '3': case '4': | |
| 509 case '5': case '6': case '7': case '8': case '9': | |
| 510 t->value = number(t); | |
| 511 return; | |
| 512 | |
| 513 #if CSTRINGS | |
| 514 case '\'': | |
| 515 t->value = charConstant(t, 0); | |
| 516 return; | |
| 517 | |
| 518 case '"': | |
| 519 t->value = stringConstant(t,0); | |
| 520 return; | |
| 521 | |
| 522 case 'l': | |
| 523 case 'L': | |
| 524 if (p[1] == '\'') | |
| 525 { | |
| 526 p++; | |
| 527 t->value = charConstant(t, 1); | |
| 528 return; | |
| 529 } | |
| 530 else if (p[1] == '"') | |
| 531 { | |
| 532 p++; | |
| 533 t->value = stringConstant(t, 1); | |
| 534 return; | |
| 535 } | |
| 536 #else | |
| 537 case '\'': | |
| 538 t->value = charConstant(t,0); | |
| 539 return; | |
| 540 | |
| 541 case 'r': | |
| 542 if (p[1] != '"') | |
| 543 goto case_ident; | |
| 544 p++; | |
| 545 case '`': | |
| 546 t->value = wysiwygStringConstant(t, *p); | |
| 547 return; | |
| 548 | |
| 549 case 'x': | |
| 550 if (p[1] != '"') | |
| 551 goto case_ident; | |
| 552 p++; | |
| 553 t->value = hexStringConstant(t); | |
| 554 return; | |
| 555 | |
| 336 | 556 #if DMDV2 |
| 159 | 557 case 'q': |
| 558 if (p[1] == '"') | |
| 559 { | |
| 560 p++; | |
| 561 t->value = delimitedStringConstant(t); | |
| 562 return; | |
| 563 } | |
| 564 else if (p[1] == '{') | |
| 565 { | |
| 566 p++; | |
| 567 t->value = tokenStringConstant(t); | |
| 568 return; | |
| 569 } | |
| 570 else | |
| 571 goto case_ident; | |
| 572 #endif | |
| 573 | |
| 574 case '"': | |
| 575 t->value = escapeStringConstant(t,0); | |
| 576 return; | |
| 577 | |
| 578 case '\\': // escaped string literal | |
| 579 { unsigned c; | |
| 580 | |
| 581 stringbuffer.reset(); | |
| 582 do | |
| 583 { | |
| 584 p++; | |
| 585 switch (*p) | |
| 586 { | |
| 587 case 'u': | |
| 588 case 'U': | |
| 589 case '&': | |
| 590 c = escapeSequence(); | |
| 591 stringbuffer.writeUTF8(c); | |
| 592 break; | |
| 593 | |
| 594 default: | |
| 595 c = escapeSequence(); | |
| 596 stringbuffer.writeByte(c); | |
| 597 break; | |
| 598 } | |
| 599 } while (*p == '\\'); | |
| 600 t->len = stringbuffer.offset; | |
| 601 stringbuffer.writeByte(0); | |
| 602 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 603 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 604 t->postfix = 0; | |
| 605 t->value = TOKstring; | |
| 606 return; | |
| 607 } | |
| 608 | |
| 609 case 'l': | |
| 610 case 'L': | |
| 611 #endif | |
| 612 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 613 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 614 case 'k': case 'm': case 'n': case 'o': | |
| 336 | 615 #if DMDV2 |
| 159 | 616 case 'p': /*case 'q': case 'r':*/ case 's': case 't': |
| 617 #else | |
| 618 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 619 #endif | |
| 620 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 621 case 'z': | |
| 622 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 623 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 624 case 'K': case 'M': case 'N': case 'O': | |
| 625 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 626 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 627 case 'Z': | |
| 628 case '_': | |
| 629 case_ident: | |
| 630 { unsigned char c; | |
| 631 StringValue *sv; | |
| 632 Identifier *id; | |
| 633 | |
| 634 do | |
| 635 { | |
| 636 c = *++p; | |
| 637 } while (isidchar(c) || (c & 0x80 && isUniAlpha(decodeUTF()))); | |
| 638 sv = stringtable.update((char *)t->ptr, p - t->ptr); | |
| 639 id = (Identifier *) sv->ptrvalue; | |
| 640 if (!id) | |
| 641 { id = new Identifier(sv->lstring.string,TOKidentifier); | |
| 642 sv->ptrvalue = id; | |
| 643 } | |
| 644 t->ident = id; | |
| 645 t->value = (enum TOK) id->value; | |
| 646 anyToken = 1; | |
| 647 if (*t->ptr == '_') // if special identifier token | |
| 648 { | |
| 649 static char date[11+1]; | |
| 650 static char time[8+1]; | |
| 651 static char timestamp[24+1]; | |
| 652 | |
| 653 if (!date[0]) // lazy evaluation | |
| 654 { time_t t; | |
| 655 char *p; | |
| 656 | |
| 657 ::time(&t); | |
| 658 p = ctime(&t); | |
| 659 assert(p); | |
| 660 sprintf(date, "%.6s %.4s", p + 4, p + 20); | |
| 661 sprintf(time, "%.8s", p + 11); | |
| 662 sprintf(timestamp, "%.24s", p); | |
| 663 } | |
| 664 | |
| 336 | 665 #if DMDV1 |
| 159 | 666 if (mod && id == Id::FILE) |
| 667 { | |
| 668 t->ustring = (unsigned char *)(loc.filename ? loc.filename : mod->ident->toChars()); | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
669 goto Lstr; |
| 159 | 670 } |
| 671 else if (mod && id == Id::LINE) | |
| 672 { | |
| 673 t->value = TOKint64v; | |
| 674 t->uns64value = loc.linnum; | |
| 675 } | |
| 336 | 676 else |
| 677 #endif | |
| 678 if (id == Id::DATE) | |
| 159 | 679 { |
| 680 t->ustring = (unsigned char *)date; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
681 goto Lstr; |
| 159 | 682 } |
| 683 else if (id == Id::TIME) | |
| 684 { | |
| 685 t->ustring = (unsigned char *)time; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
686 goto Lstr; |
| 159 | 687 } |
| 688 else if (id == Id::VENDOR) | |
| 689 { | |
|
664
eef8ac26c66c
Some missed LLVMDC -> LDC.
Christian Kamm <kamm incasoftware de>
parents:
658
diff
changeset
|
690 t->ustring = (unsigned char *)"LDC"; |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
691 goto Lstr; |
| 159 | 692 } |
| 693 else if (id == Id::TIMESTAMP) | |
| 694 { | |
| 695 t->ustring = (unsigned char *)timestamp; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
696 Lstr: |
| 159 | 697 t->value = TOKstring; |
| 698 Llen: | |
| 699 t->postfix = 0; | |
| 700 t->len = strlen((char *)t->ustring); | |
| 701 } | |
| 702 else if (id == Id::VERSIONX) | |
| 703 { unsigned major = 0; | |
| 704 unsigned minor = 0; | |
| 705 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
706 for (const char *p = global.version + 1; 1; p++) |
| 159 | 707 { |
| 708 char c = *p; | |
| 709 if (isdigit(c)) | |
| 710 minor = minor * 10 + c - '0'; | |
| 711 else if (c == '.') | |
| 712 { major = minor; | |
| 713 minor = 0; | |
| 714 } | |
| 715 else | |
| 716 break; | |
| 717 } | |
| 718 t->value = TOKint64v; | |
| 719 t->uns64value = major * 1000 + minor; | |
| 720 } | |
| 336 | 721 #if DMDV2 |
| 159 | 722 else if (id == Id::EOFX) |
| 723 { | |
| 724 t->value = TOKeof; | |
| 725 // Advance scanner to end of file | |
| 726 while (!(*p == 0 || *p == 0x1A)) | |
| 727 p++; | |
| 728 } | |
| 729 #endif | |
| 730 } | |
| 731 //printf("t->value = %d\n",t->value); | |
| 732 return; | |
| 733 } | |
| 734 | |
| 735 case '/': | |
| 736 p++; | |
| 737 switch (*p) | |
| 738 { | |
| 739 case '=': | |
| 740 p++; | |
| 741 t->value = TOKdivass; | |
| 742 return; | |
| 743 | |
| 744 case '*': | |
| 745 p++; | |
| 746 linnum = loc.linnum; | |
| 747 while (1) | |
| 748 { | |
| 749 while (1) | |
| 750 { unsigned char c = *p; | |
| 751 switch (c) | |
| 752 { | |
| 753 case '/': | |
| 754 break; | |
| 755 | |
| 756 case '\n': | |
| 757 loc.linnum++; | |
| 758 p++; | |
| 759 continue; | |
| 760 | |
| 761 case '\r': | |
| 762 p++; | |
| 763 if (*p != '\n') | |
| 764 loc.linnum++; | |
| 765 continue; | |
| 766 | |
| 767 case 0: | |
| 768 case 0x1A: | |
| 769 error("unterminated /* */ comment"); | |
| 770 p = end; | |
| 771 t->value = TOKeof; | |
| 772 return; | |
| 773 | |
| 774 default: | |
| 775 if (c & 0x80) | |
| 776 { unsigned u = decodeUTF(); | |
| 777 if (u == PS || u == LS) | |
| 778 loc.linnum++; | |
| 779 } | |
| 780 p++; | |
| 781 continue; | |
| 782 } | |
| 783 break; | |
| 784 } | |
| 785 p++; | |
| 786 if (p[-2] == '*' && p - 3 != t->ptr) | |
| 787 break; | |
| 788 } | |
| 789 if (commentToken) | |
| 790 { | |
| 791 t->value = TOKcomment; | |
| 792 return; | |
| 793 } | |
| 794 else if (doDocComment && t->ptr[2] == '*' && p - 4 != t->ptr) | |
| 795 { // if /** but not /**/ | |
| 796 getDocComment(t, lastLine == linnum); | |
| 797 } | |
| 798 continue; | |
| 799 | |
| 800 case '/': // do // style comments | |
| 801 linnum = loc.linnum; | |
| 802 while (1) | |
| 803 { unsigned char c = *++p; | |
| 804 switch (c) | |
| 805 { | |
| 806 case '\n': | |
| 807 break; | |
| 808 | |
| 809 case '\r': | |
| 810 if (p[1] == '\n') | |
| 811 p++; | |
| 812 break; | |
| 813 | |
| 814 case 0: | |
| 815 case 0x1A: | |
| 816 if (commentToken) | |
| 817 { | |
| 818 p = end; | |
| 819 t->value = TOKcomment; | |
| 820 return; | |
| 821 } | |
| 822 if (doDocComment && t->ptr[2] == '/') | |
| 823 getDocComment(t, lastLine == linnum); | |
| 824 p = end; | |
| 825 t->value = TOKeof; | |
| 826 return; | |
| 827 | |
| 828 default: | |
| 829 if (c & 0x80) | |
| 830 { unsigned u = decodeUTF(); | |
| 831 if (u == PS || u == LS) | |
| 832 break; | |
| 833 } | |
| 834 continue; | |
| 835 } | |
| 836 break; | |
| 837 } | |
| 838 | |
| 839 if (commentToken) | |
| 840 { | |
| 841 p++; | |
| 842 loc.linnum++; | |
| 843 t->value = TOKcomment; | |
| 844 return; | |
| 845 } | |
| 846 if (doDocComment && t->ptr[2] == '/') | |
| 847 getDocComment(t, lastLine == linnum); | |
| 848 | |
| 849 p++; | |
| 850 loc.linnum++; | |
| 851 continue; | |
| 852 | |
| 853 case '+': | |
| 854 { int nest; | |
| 855 | |
| 856 linnum = loc.linnum; | |
| 857 p++; | |
| 858 nest = 1; | |
| 859 while (1) | |
| 860 { unsigned char c = *p; | |
| 861 switch (c) | |
| 862 { | |
| 863 case '/': | |
| 864 p++; | |
| 865 if (*p == '+') | |
| 866 { | |
| 867 p++; | |
| 868 nest++; | |
| 869 } | |
| 870 continue; | |
| 871 | |
| 872 case '+': | |
| 873 p++; | |
| 874 if (*p == '/') | |
| 875 { | |
| 876 p++; | |
| 877 if (--nest == 0) | |
| 878 break; | |
| 879 } | |
| 880 continue; | |
| 881 | |
| 882 case '\r': | |
| 883 p++; | |
| 884 if (*p != '\n') | |
| 885 loc.linnum++; | |
| 886 continue; | |
| 887 | |
| 888 case '\n': | |
| 889 loc.linnum++; | |
| 890 p++; | |
| 891 continue; | |
| 892 | |
| 893 case 0: | |
| 894 case 0x1A: | |
| 895 error("unterminated /+ +/ comment"); | |
| 896 p = end; | |
| 897 t->value = TOKeof; | |
| 898 return; | |
| 899 | |
| 900 default: | |
| 901 if (c & 0x80) | |
| 902 { unsigned u = decodeUTF(); | |
| 903 if (u == PS || u == LS) | |
| 904 loc.linnum++; | |
| 905 } | |
| 906 p++; | |
| 907 continue; | |
| 908 } | |
| 909 break; | |
| 910 } | |
| 911 if (commentToken) | |
| 912 { | |
| 913 t->value = TOKcomment; | |
| 914 return; | |
| 915 } | |
| 916 if (doDocComment && t->ptr[2] == '+' && p - 4 != t->ptr) | |
| 917 { // if /++ but not /++/ | |
| 918 getDocComment(t, lastLine == linnum); | |
| 919 } | |
| 920 continue; | |
| 921 } | |
| 922 } | |
| 923 t->value = TOKdiv; | |
| 924 return; | |
| 925 | |
| 926 case '.': | |
| 927 p++; | |
| 928 if (isdigit(*p)) | |
| 929 { /* Note that we don't allow ._1 and ._ as being | |
| 930 * valid floating point numbers. | |
| 931 */ | |
| 932 p--; | |
| 933 t->value = inreal(t); | |
| 934 } | |
| 935 else if (p[0] == '.') | |
| 936 { | |
| 937 if (p[1] == '.') | |
| 938 { p += 2; | |
| 939 t->value = TOKdotdotdot; | |
| 940 } | |
| 941 else | |
| 942 { p++; | |
| 943 t->value = TOKslice; | |
| 944 } | |
| 945 } | |
| 946 else | |
| 947 t->value = TOKdot; | |
| 948 return; | |
| 949 | |
| 950 case '&': | |
| 951 p++; | |
| 952 if (*p == '=') | |
| 953 { p++; | |
| 954 t->value = TOKandass; | |
| 955 } | |
| 956 else if (*p == '&') | |
| 957 { p++; | |
| 958 t->value = TOKandand; | |
| 959 } | |
| 960 else | |
| 961 t->value = TOKand; | |
| 962 return; | |
| 963 | |
| 964 case '|': | |
| 965 p++; | |
| 966 if (*p == '=') | |
| 967 { p++; | |
| 968 t->value = TOKorass; | |
| 969 } | |
| 970 else if (*p == '|') | |
| 971 { p++; | |
| 972 t->value = TOKoror; | |
| 973 } | |
| 974 else | |
| 975 t->value = TOKor; | |
| 976 return; | |
| 977 | |
| 978 case '-': | |
| 979 p++; | |
| 980 if (*p == '=') | |
| 981 { p++; | |
| 982 t->value = TOKminass; | |
| 983 } | |
| 984 #if 0 | |
| 985 else if (*p == '>') | |
| 986 { p++; | |
| 987 t->value = TOKarrow; | |
| 988 } | |
| 989 #endif | |
| 990 else if (*p == '-') | |
| 991 { p++; | |
| 992 t->value = TOKminusminus; | |
| 993 } | |
| 994 else | |
| 995 t->value = TOKmin; | |
| 996 return; | |
| 997 | |
| 998 case '+': | |
| 999 p++; | |
| 1000 if (*p == '=') | |
| 1001 { p++; | |
| 1002 t->value = TOKaddass; | |
| 1003 } | |
| 1004 else if (*p == '+') | |
| 1005 { p++; | |
| 1006 t->value = TOKplusplus; | |
| 1007 } | |
| 1008 else | |
| 1009 t->value = TOKadd; | |
| 1010 return; | |
| 1011 | |
| 1012 case '<': | |
| 1013 p++; | |
| 1014 if (*p == '=') | |
| 1015 { p++; | |
| 1016 t->value = TOKle; // <= | |
| 1017 } | |
| 1018 else if (*p == '<') | |
| 1019 { p++; | |
| 1020 if (*p == '=') | |
| 1021 { p++; | |
| 1022 t->value = TOKshlass; // <<= | |
| 1023 } | |
| 1024 else | |
| 1025 t->value = TOKshl; // << | |
| 1026 } | |
| 1027 else if (*p == '>') | |
| 1028 { p++; | |
| 1029 if (*p == '=') | |
| 1030 { p++; | |
| 1031 t->value = TOKleg; // <>= | |
| 1032 } | |
| 1033 else | |
| 1034 t->value = TOKlg; // <> | |
| 1035 } | |
| 1036 else | |
| 1037 t->value = TOKlt; // < | |
| 1038 return; | |
| 1039 | |
| 1040 case '>': | |
| 1041 p++; | |
| 1042 if (*p == '=') | |
| 1043 { p++; | |
| 1044 t->value = TOKge; // >= | |
| 1045 } | |
| 1046 else if (*p == '>') | |
| 1047 { p++; | |
| 1048 if (*p == '=') | |
| 1049 { p++; | |
| 1050 t->value = TOKshrass; // >>= | |
| 1051 } | |
| 1052 else if (*p == '>') | |
| 1053 { p++; | |
| 1054 if (*p == '=') | |
| 1055 { p++; | |
| 1056 t->value = TOKushrass; // >>>= | |
| 1057 } | |
| 1058 else | |
| 1059 t->value = TOKushr; // >>> | |
| 1060 } | |
| 1061 else | |
| 1062 t->value = TOKshr; // >> | |
| 1063 } | |
| 1064 else | |
| 1065 t->value = TOKgt; // > | |
| 1066 return; | |
| 1067 | |
| 1068 case '!': | |
| 1069 p++; | |
| 1070 if (*p == '=') | |
| 1071 { p++; | |
| 1072 if (*p == '=' && global.params.Dversion == 1) | |
| 1073 { p++; | |
| 1074 t->value = TOKnotidentity; // !== | |
| 1075 } | |
| 1076 else | |
| 1077 t->value = TOKnotequal; // != | |
| 1078 } | |
| 1079 else if (*p == '<') | |
| 1080 { p++; | |
| 1081 if (*p == '>') | |
| 1082 { p++; | |
| 1083 if (*p == '=') | |
| 1084 { p++; | |
| 1085 t->value = TOKunord; // !<>= | |
| 1086 } | |
| 1087 else | |
| 1088 t->value = TOKue; // !<> | |
| 1089 } | |
| 1090 else if (*p == '=') | |
| 1091 { p++; | |
| 1092 t->value = TOKug; // !<= | |
| 1093 } | |
| 1094 else | |
| 1095 t->value = TOKuge; // !< | |
| 1096 } | |
| 1097 else if (*p == '>') | |
| 1098 { p++; | |
| 1099 if (*p == '=') | |
| 1100 { p++; | |
| 1101 t->value = TOKul; // !>= | |
| 1102 } | |
| 1103 else | |
| 1104 t->value = TOKule; // !> | |
| 1105 } | |
| 1106 else | |
| 1107 t->value = TOKnot; // ! | |
| 1108 return; | |
| 1109 | |
| 1110 case '=': | |
| 1111 p++; | |
| 1112 if (*p == '=') | |
| 1113 { p++; | |
| 1114 if (*p == '=' && global.params.Dversion == 1) | |
| 1115 { p++; | |
| 1116 t->value = TOKidentity; // === | |
| 1117 } | |
| 1118 else | |
| 1119 t->value = TOKequal; // == | |
| 1120 } | |
| 1121 else | |
| 1122 t->value = TOKassign; // = | |
| 1123 return; | |
| 1124 | |
| 1125 case '~': | |
| 1126 p++; | |
| 1127 if (*p == '=') | |
| 1128 { p++; | |
| 1129 t->value = TOKcatass; // ~= | |
| 1130 } | |
| 1131 else | |
| 1132 t->value = TOKtilde; // ~ | |
| 1133 return; | |
| 1134 | |
| 1135 #define SINGLE(c,tok) case c: p++; t->value = tok; return; | |
| 1136 | |
| 1137 SINGLE('(', TOKlparen) | |
| 1138 SINGLE(')', TOKrparen) | |
| 1139 SINGLE('[', TOKlbracket) | |
| 1140 SINGLE(']', TOKrbracket) | |
| 1141 SINGLE('{', TOKlcurly) | |
| 1142 SINGLE('}', TOKrcurly) | |
| 1143 SINGLE('?', TOKquestion) | |
| 1144 SINGLE(',', TOKcomma) | |
| 1145 SINGLE(';', TOKsemicolon) | |
| 1146 SINGLE(':', TOKcolon) | |
| 1147 SINGLE('$', TOKdollar) | |
| 1148 | |
| 1149 #undef SINGLE | |
| 1150 | |
| 1151 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1152 case c1: \ | |
| 1153 p++; \ | |
| 1154 if (*p == c2) \ | |
| 1155 { p++; \ | |
| 1156 t->value = tok2; \ | |
| 1157 } \ | |
| 1158 else \ | |
| 1159 t->value = tok1; \ | |
| 1160 return; | |
| 1161 | |
| 1162 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1163 DOUBLE('%', TOKmod, '=', TOKmodass) | |
| 1164 DOUBLE('^', TOKxor, '=', TOKxorass) | |
| 1165 | |
| 1166 #undef DOUBLE | |
| 1167 | |
| 1168 case '#': | |
| 1169 p++; | |
| 1170 pragma(); | |
| 1171 continue; | |
| 1172 | |
| 1173 default: | |
| 1174 { unsigned char c = *p; | |
| 1175 | |
| 1176 if (c & 0x80) | |
| 1177 { unsigned u = decodeUTF(); | |
| 1178 | |
| 1179 // Check for start of unicode identifier | |
| 1180 if (isUniAlpha(u)) | |
| 1181 goto case_ident; | |
| 1182 | |
| 1183 if (u == PS || u == LS) | |
| 1184 { | |
| 1185 loc.linnum++; | |
| 1186 p++; | |
| 1187 continue; | |
| 1188 } | |
| 1189 } | |
| 1190 if (isprint(c)) | |
| 1191 error("unsupported char '%c'", c); | |
| 1192 else | |
| 1193 error("unsupported char 0x%02x", c); | |
| 1194 p++; | |
| 1195 continue; | |
| 1196 } | |
| 1197 } | |
| 1198 } | |
| 1199 } | |
| 1200 | |
| 1201 /******************************************* | |
| 1202 * Parse escape sequence. | |
| 1203 */ | |
| 1204 | |
| 1205 unsigned Lexer::escapeSequence() | |
| 1206 { unsigned c; | |
| 1207 int n; | |
| 1208 int ndigits; | |
| 1209 | |
| 1210 c = *p; | |
| 1211 switch (c) | |
| 1212 { | |
| 1213 case '\'': | |
| 1214 case '"': | |
| 1215 case '?': | |
| 1216 case '\\': | |
| 1217 Lconsume: | |
| 1218 p++; | |
| 1219 break; | |
| 1220 | |
| 1221 case 'a': c = 7; goto Lconsume; | |
| 1222 case 'b': c = 8; goto Lconsume; | |
| 1223 case 'f': c = 12; goto Lconsume; | |
| 1224 case 'n': c = 10; goto Lconsume; | |
| 1225 case 'r': c = 13; goto Lconsume; | |
| 1226 case 't': c = 9; goto Lconsume; | |
| 1227 case 'v': c = 11; goto Lconsume; | |
| 1228 | |
| 1229 case 'u': | |
| 1230 ndigits = 4; | |
| 1231 goto Lhex; | |
| 1232 case 'U': | |
| 1233 ndigits = 8; | |
| 1234 goto Lhex; | |
| 1235 case 'x': | |
| 1236 ndigits = 2; | |
| 1237 Lhex: | |
| 1238 p++; | |
| 1239 c = *p; | |
| 1240 if (ishex(c)) | |
| 1241 { unsigned v; | |
| 1242 | |
| 1243 n = 0; | |
| 1244 v = 0; | |
| 1245 while (1) | |
| 1246 { | |
| 1247 if (isdigit(c)) | |
| 1248 c -= '0'; | |
| 1249 else if (islower(c)) | |
| 1250 c -= 'a' - 10; | |
| 1251 else | |
| 1252 c -= 'A' - 10; | |
| 1253 v = v * 16 + c; | |
| 1254 c = *++p; | |
| 1255 if (++n == ndigits) | |
| 1256 break; | |
| 1257 if (!ishex(c)) | |
| 1258 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1259 break; | |
| 1260 } | |
| 1261 } | |
| 1262 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1263 error("invalid UTF character \\U%08x", v); | |
| 1264 c = v; | |
| 1265 } | |
| 1266 else | |
| 1267 error("undefined escape hex sequence \\%c\n",c); | |
| 1268 break; | |
| 1269 | |
| 1270 case '&': // named character entity | |
| 1271 for (unsigned char *idstart = ++p; 1; p++) | |
| 1272 { | |
| 1273 switch (*p) | |
| 1274 { | |
| 1275 case ';': | |
| 1276 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1277 if (c == ~0) | |
| 1278 { error("unnamed character entity &%.*s;", (int)(p - idstart), idstart); | |
| 1279 c = ' '; | |
| 1280 } | |
| 1281 p++; | |
| 1282 break; | |
| 1283 | |
| 1284 default: | |
| 1285 if (isalpha(*p) || | |
| 1286 (p != idstart + 1 && isdigit(*p))) | |
| 1287 continue; | |
| 1288 error("unterminated named entity"); | |
| 1289 break; | |
| 1290 } | |
| 1291 break; | |
| 1292 } | |
| 1293 break; | |
| 1294 | |
| 1295 case 0: | |
| 1296 case 0x1A: // end of file | |
| 1297 c = '\\'; | |
| 1298 break; | |
| 1299 | |
| 1300 default: | |
| 1301 if (isoctal(c)) | |
| 1302 { unsigned v; | |
| 1303 | |
| 1304 n = 0; | |
| 1305 v = 0; | |
| 1306 do | |
| 1307 { | |
| 1308 v = v * 8 + (c - '0'); | |
| 1309 c = *++p; | |
| 1310 } while (++n < 3 && isoctal(c)); | |
| 1311 c = v; | |
| 1312 if (c > 0xFF) | |
| 1313 error("0%03o is larger than a byte", c); | |
| 1314 } | |
| 1315 else | |
| 1316 error("undefined escape sequence \\%c\n",c); | |
| 1317 break; | |
| 1318 } | |
| 1319 return c; | |
| 1320 } | |
| 1321 | |
| 1322 /************************************** | |
| 1323 */ | |
| 1324 | |
| 1325 TOK Lexer::wysiwygStringConstant(Token *t, int tc) | |
| 1326 { unsigned c; | |
| 1327 Loc start = loc; | |
| 1328 | |
| 1329 p++; | |
| 1330 stringbuffer.reset(); | |
| 1331 while (1) | |
| 1332 { | |
| 1333 c = *p++; | |
| 1334 switch (c) | |
| 1335 { | |
| 1336 case '\n': | |
| 1337 loc.linnum++; | |
| 1338 break; | |
| 1339 | |
| 1340 case '\r': | |
| 1341 if (*p == '\n') | |
| 1342 continue; // ignore | |
| 1343 c = '\n'; // treat EndOfLine as \n character | |
| 1344 loc.linnum++; | |
| 1345 break; | |
| 1346 | |
| 1347 case 0: | |
| 1348 case 0x1A: | |
| 1349 error("unterminated string constant starting at %s", start.toChars()); | |
| 1350 t->ustring = (unsigned char *)""; | |
| 1351 t->len = 0; | |
| 1352 t->postfix = 0; | |
| 1353 return TOKstring; | |
| 1354 | |
| 1355 case '"': | |
| 1356 case '`': | |
| 1357 if (c == tc) | |
| 1358 { | |
| 1359 t->len = stringbuffer.offset; | |
| 1360 stringbuffer.writeByte(0); | |
| 1361 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1362 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1363 stringPostfix(t); | |
| 1364 return TOKstring; | |
| 1365 } | |
| 1366 break; | |
| 1367 | |
| 1368 default: | |
| 1369 if (c & 0x80) | |
| 1370 { p--; | |
| 1371 unsigned u = decodeUTF(); | |
| 1372 p++; | |
| 1373 if (u == PS || u == LS) | |
| 1374 loc.linnum++; | |
| 1375 stringbuffer.writeUTF8(u); | |
| 1376 continue; | |
| 1377 } | |
| 1378 break; | |
| 1379 } | |
| 1380 stringbuffer.writeByte(c); | |
| 1381 } | |
| 1382 } | |
| 1383 | |
| 1384 /************************************** | |
| 1385 * Lex hex strings: | |
| 1386 * x"0A ae 34FE BD" | |
| 1387 */ | |
| 1388 | |
| 1389 TOK Lexer::hexStringConstant(Token *t) | |
| 1390 { unsigned c; | |
| 1391 Loc start = loc; | |
| 1392 unsigned n = 0; | |
| 1393 unsigned v; | |
| 1394 | |
| 1395 p++; | |
| 1396 stringbuffer.reset(); | |
| 1397 while (1) | |
| 1398 { | |
| 1399 c = *p++; | |
| 1400 switch (c) | |
| 1401 { | |
| 1402 case ' ': | |
| 1403 case '\t': | |
| 1404 case '\v': | |
| 1405 case '\f': | |
| 1406 continue; // skip white space | |
| 1407 | |
| 1408 case '\r': | |
| 1409 if (*p == '\n') | |
| 1410 continue; // ignore | |
| 1411 // Treat isolated '\r' as if it were a '\n' | |
| 1412 case '\n': | |
| 1413 loc.linnum++; | |
| 1414 continue; | |
| 1415 | |
| 1416 case 0: | |
| 1417 case 0x1A: | |
| 1418 error("unterminated string constant starting at %s", start.toChars()); | |
| 1419 t->ustring = (unsigned char *)""; | |
| 1420 t->len = 0; | |
| 1421 t->postfix = 0; | |
| 1422 return TOKstring; | |
| 1423 | |
| 1424 case '"': | |
| 1425 if (n & 1) | |
| 1426 { error("odd number (%d) of hex characters in hex string", n); | |
| 1427 stringbuffer.writeByte(v); | |
| 1428 } | |
| 1429 t->len = stringbuffer.offset; | |
| 1430 stringbuffer.writeByte(0); | |
| 1431 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1432 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1433 stringPostfix(t); | |
| 1434 return TOKstring; | |
| 1435 | |
| 1436 default: | |
| 1437 if (c >= '0' && c <= '9') | |
| 1438 c -= '0'; | |
| 1439 else if (c >= 'a' && c <= 'f') | |
| 1440 c -= 'a' - 10; | |
| 1441 else if (c >= 'A' && c <= 'F') | |
| 1442 c -= 'A' - 10; | |
| 1443 else if (c & 0x80) | |
| 1444 { p--; | |
| 1445 unsigned u = decodeUTF(); | |
| 1446 p++; | |
| 1447 if (u == PS || u == LS) | |
| 1448 loc.linnum++; | |
| 1449 else | |
| 1450 error("non-hex character \\u%x", u); | |
| 1451 } | |
| 1452 else | |
| 1453 error("non-hex character '%c'", c); | |
| 1454 if (n & 1) | |
| 1455 { v = (v << 4) | c; | |
| 1456 stringbuffer.writeByte(v); | |
| 1457 } | |
| 1458 else | |
| 1459 v = c; | |
| 1460 n++; | |
| 1461 break; | |
| 1462 } | |
| 1463 } | |
| 1464 } | |
| 1465 | |
| 1466 | |
| 336 | 1467 #if DMDV2 |
| 159 | 1468 /************************************** |
| 1469 * Lex delimited strings: | |
| 1470 * q"(foo(xxx))" // "foo(xxx)" | |
| 1471 * q"[foo(]" // "foo(" | |
| 1472 * q"/foo]/" // "foo]" | |
| 1473 * q"HERE | |
| 1474 * foo | |
| 1475 * HERE" // "foo\n" | |
| 1476 * Input: | |
| 1477 * p is on the " | |
| 1478 */ | |
| 1479 | |
| 1480 TOK Lexer::delimitedStringConstant(Token *t) | |
| 1481 { unsigned c; | |
| 1482 Loc start = loc; | |
| 1483 unsigned delimleft = 0; | |
| 1484 unsigned delimright = 0; | |
| 1485 unsigned nest = 1; | |
| 1486 unsigned nestcount; | |
| 1487 Identifier *hereid = NULL; | |
| 1488 unsigned blankrol = 0; | |
| 1489 unsigned startline = 0; | |
| 1490 | |
| 1491 p++; | |
| 1492 stringbuffer.reset(); | |
| 1493 while (1) | |
| 1494 { | |
| 1495 c = *p++; | |
| 1496 //printf("c = '%c'\n", c); | |
| 1497 switch (c) | |
| 1498 { | |
| 1499 case '\n': | |
| 1500 Lnextline: | |
| 1501 loc.linnum++; | |
| 1502 startline = 1; | |
| 1503 if (blankrol) | |
| 1504 { blankrol = 0; | |
| 1505 continue; | |
| 1506 } | |
| 1507 if (hereid) | |
| 1508 { | |
| 1509 stringbuffer.writeUTF8(c); | |
| 1510 continue; | |
| 1511 } | |
| 1512 break; | |
| 1513 | |
| 1514 case '\r': | |
| 1515 if (*p == '\n') | |
| 1516 continue; // ignore | |
| 1517 c = '\n'; // treat EndOfLine as \n character | |
| 1518 goto Lnextline; | |
| 1519 | |
| 1520 case 0: | |
| 1521 case 0x1A: | |
| 1522 goto Lerror; | |
| 1523 | |
| 1524 default: | |
| 1525 if (c & 0x80) | |
| 1526 { p--; | |
| 1527 c = decodeUTF(); | |
| 1528 p++; | |
| 1529 if (c == PS || c == LS) | |
| 1530 goto Lnextline; | |
| 1531 } | |
| 1532 break; | |
| 1533 } | |
| 1534 if (delimleft == 0) | |
| 1535 { delimleft = c; | |
| 1536 nest = 1; | |
| 1537 nestcount = 1; | |
| 1538 if (c == '(') | |
| 1539 delimright = ')'; | |
| 1540 else if (c == '{') | |
| 1541 delimright = '}'; | |
| 1542 else if (c == '[') | |
| 1543 delimright = ']'; | |
| 1544 else if (c == '<') | |
| 1545 delimright = '>'; | |
| 1546 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1547 { // Start of identifier; must be a heredoc | |
| 1548 Token t; | |
| 1549 p--; | |
| 1550 scan(&t); // read in heredoc identifier | |
| 1551 if (t.value != TOKidentifier) | |
| 1552 { error("identifier expected for heredoc, not %s", t.toChars()); | |
| 1553 delimright = c; | |
| 1554 } | |
| 1555 else | |
| 1556 { hereid = t.ident; | |
| 1557 //printf("hereid = '%s'\n", hereid->toChars()); | |
| 1558 blankrol = 1; | |
| 1559 } | |
| 1560 nest = 0; | |
| 1561 } | |
| 1562 else | |
| 1563 { delimright = c; | |
| 1564 nest = 0; | |
| 1565 } | |
| 1566 } | |
| 1567 else | |
| 1568 { | |
| 1569 if (blankrol) | |
| 1570 { error("heredoc rest of line should be blank"); | |
| 1571 blankrol = 0; | |
| 1572 continue; | |
| 1573 } | |
| 1574 if (nest == 1) | |
| 1575 { | |
| 1576 if (c == delimleft) | |
| 1577 nestcount++; | |
| 1578 else if (c == delimright) | |
| 1579 { nestcount--; | |
| 1580 if (nestcount == 0) | |
| 1581 goto Ldone; | |
| 1582 } | |
| 1583 } | |
| 1584 else if (c == delimright) | |
| 1585 goto Ldone; | |
| 1586 if (startline && isalpha(c)) | |
| 1587 { Token t; | |
| 1588 unsigned char *psave = p; | |
| 1589 p--; | |
| 1590 scan(&t); // read in possible heredoc identifier | |
| 1591 //printf("endid = '%s'\n", t.ident->toChars()); | |
| 1592 if (t.value == TOKidentifier && t.ident->equals(hereid)) | |
| 1593 { /* should check that rest of line is blank | |
| 1594 */ | |
| 1595 goto Ldone; | |
| 1596 } | |
| 1597 p = psave; | |
| 1598 } | |
| 1599 stringbuffer.writeUTF8(c); | |
| 1600 startline = 0; | |
| 1601 } | |
| 1602 } | |
| 1603 | |
| 1604 Ldone: | |
| 1605 if (*p == '"') | |
| 1606 p++; | |
| 1607 else | |
| 1608 error("delimited string must end in %c\"", delimright); | |
| 1609 t->len = stringbuffer.offset; | |
| 1610 stringbuffer.writeByte(0); | |
| 1611 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1612 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1613 stringPostfix(t); | |
| 1614 return TOKstring; | |
| 1615 | |
| 1616 Lerror: | |
| 1617 error("unterminated string constant starting at %s", start.toChars()); | |
| 1618 t->ustring = (unsigned char *)""; | |
| 1619 t->len = 0; | |
| 1620 t->postfix = 0; | |
| 1621 return TOKstring; | |
| 1622 } | |
| 1623 | |
| 1624 /************************************** | |
| 1625 * Lex delimited strings: | |
| 1626 * q{ foo(xxx) } // " foo(xxx) " | |
| 1627 * q{foo(} // "foo(" | |
| 1628 * q{{foo}"}"} // "{foo}"}"" | |
| 1629 * Input: | |
| 1630 * p is on the q | |
| 1631 */ | |
| 1632 | |
| 1633 TOK Lexer::tokenStringConstant(Token *t) | |
| 1634 { | |
| 1635 unsigned nest = 1; | |
| 1636 Loc start = loc; | |
| 1637 unsigned char *pstart = ++p; | |
| 1638 | |
| 1639 while (1) | |
| 1640 { Token tok; | |
| 1641 | |
| 1642 scan(&tok); | |
| 1643 switch (tok.value) | |
| 1644 { | |
| 1645 case TOKlcurly: | |
| 1646 nest++; | |
| 1647 continue; | |
| 1648 | |
| 1649 case TOKrcurly: | |
| 1650 if (--nest == 0) | |
| 1651 goto Ldone; | |
| 1652 continue; | |
| 1653 | |
| 1654 case TOKeof: | |
| 1655 goto Lerror; | |
| 1656 | |
| 1657 default: | |
| 1658 continue; | |
| 1659 } | |
| 1660 } | |
| 1661 | |
| 1662 Ldone: | |
| 1663 t->len = p - 1 - pstart; | |
| 1664 t->ustring = (unsigned char *)mem.malloc(t->len + 1); | |
| 1665 memcpy(t->ustring, pstart, t->len); | |
| 1666 t->ustring[t->len] = 0; | |
| 1667 stringPostfix(t); | |
| 1668 return TOKstring; | |
| 1669 | |
| 1670 Lerror: | |
| 1671 error("unterminated token string constant starting at %s", start.toChars()); | |
| 1672 t->ustring = (unsigned char *)""; | |
| 1673 t->len = 0; | |
| 1674 t->postfix = 0; | |
| 1675 return TOKstring; | |
| 1676 } | |
| 1677 | |
| 1678 #endif | |
| 1679 | |
| 1680 | |
| 1681 /************************************** | |
| 1682 */ | |
| 1683 | |
| 1684 TOK Lexer::escapeStringConstant(Token *t, int wide) | |
| 1685 { unsigned c; | |
| 1686 Loc start = loc; | |
| 1687 | |
| 1688 p++; | |
| 1689 stringbuffer.reset(); | |
| 1690 while (1) | |
| 1691 { | |
| 1692 c = *p++; | |
| 1693 switch (c) | |
| 1694 { | |
| 1695 case '\\': | |
| 1696 switch (*p) | |
| 1697 { | |
| 1698 case 'u': | |
| 1699 case 'U': | |
| 1700 case '&': | |
| 1701 c = escapeSequence(); | |
| 1702 stringbuffer.writeUTF8(c); | |
| 1703 continue; | |
| 1704 | |
| 1705 default: | |
| 1706 c = escapeSequence(); | |
| 1707 break; | |
| 1708 } | |
| 1709 break; | |
| 1710 | |
| 1711 case '\n': | |
| 1712 loc.linnum++; | |
| 1713 break; | |
| 1714 | |
| 1715 case '\r': | |
| 1716 if (*p == '\n') | |
| 1717 continue; // ignore | |
| 1718 c = '\n'; // treat EndOfLine as \n character | |
| 1719 loc.linnum++; | |
| 1720 break; | |
| 1721 | |
| 1722 case '"': | |
| 1723 t->len = stringbuffer.offset; | |
| 1724 stringbuffer.writeByte(0); | |
| 1725 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1726 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1727 stringPostfix(t); | |
| 1728 return TOKstring; | |
| 1729 | |
| 1730 case 0: | |
| 1731 case 0x1A: | |
| 1732 p--; | |
| 1733 error("unterminated string constant starting at %s", start.toChars()); | |
| 1734 t->ustring = (unsigned char *)""; | |
| 1735 t->len = 0; | |
| 1736 t->postfix = 0; | |
| 1737 return TOKstring; | |
| 1738 | |
| 1739 default: | |
| 1740 if (c & 0x80) | |
| 1741 { | |
| 1742 p--; | |
| 1743 c = decodeUTF(); | |
| 1744 if (c == LS || c == PS) | |
| 1745 { c = '\n'; | |
| 1746 loc.linnum++; | |
| 1747 } | |
| 1748 p++; | |
| 1749 stringbuffer.writeUTF8(c); | |
| 1750 continue; | |
| 1751 } | |
| 1752 break; | |
| 1753 } | |
| 1754 stringbuffer.writeByte(c); | |
| 1755 } | |
| 1756 } | |
| 1757 | |
| 1758 /************************************** | |
| 1759 */ | |
| 1760 | |
| 1761 TOK Lexer::charConstant(Token *t, int wide) | |
| 1762 { | |
| 1763 unsigned c; | |
| 1764 TOK tk = TOKcharv; | |
| 1765 | |
| 1766 //printf("Lexer::charConstant\n"); | |
| 1767 p++; | |
| 1768 c = *p++; | |
| 1769 switch (c) | |
| 1770 { | |
| 1771 case '\\': | |
| 1772 switch (*p) | |
| 1773 { | |
| 1774 case 'u': | |
| 1775 t->uns64value = escapeSequence(); | |
| 1776 tk = TOKwcharv; | |
| 1777 break; | |
| 1778 | |
| 1779 case 'U': | |
| 1780 case '&': | |
| 1781 t->uns64value = escapeSequence(); | |
| 1782 tk = TOKdcharv; | |
| 1783 break; | |
| 1784 | |
| 1785 default: | |
| 1786 t->uns64value = escapeSequence(); | |
| 1787 break; | |
| 1788 } | |
| 1789 break; | |
| 1790 | |
| 1791 case '\n': | |
| 1792 L1: | |
| 1793 loc.linnum++; | |
| 1794 case '\r': | |
| 1795 case 0: | |
| 1796 case 0x1A: | |
| 1797 case '\'': | |
| 1798 error("unterminated character constant"); | |
| 1799 return tk; | |
| 1800 | |
| 1801 default: | |
| 1802 if (c & 0x80) | |
| 1803 { | |
| 1804 p--; | |
| 1805 c = decodeUTF(); | |
| 1806 p++; | |
| 1807 if (c == LS || c == PS) | |
| 1808 goto L1; | |
| 1809 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 1810 tk = TOKwcharv; | |
| 1811 else | |
| 1812 tk = TOKdcharv; | |
| 1813 } | |
| 1814 t->uns64value = c; | |
| 1815 break; | |
| 1816 } | |
| 1817 | |
| 1818 if (*p != '\'') | |
| 1819 { error("unterminated character constant"); | |
| 1820 return tk; | |
| 1821 } | |
| 1822 p++; | |
| 1823 return tk; | |
| 1824 } | |
| 1825 | |
| 1826 /*************************************** | |
| 1827 * Get postfix of string literal. | |
| 1828 */ | |
| 1829 | |
| 1830 void Lexer::stringPostfix(Token *t) | |
| 1831 { | |
| 1832 switch (*p) | |
| 1833 { | |
| 1834 case 'c': | |
| 1835 case 'w': | |
| 1836 case 'd': | |
| 1837 t->postfix = *p; | |
| 1838 p++; | |
| 1839 break; | |
| 1840 | |
| 1841 default: | |
| 1842 t->postfix = 0; | |
| 1843 break; | |
| 1844 } | |
| 1845 } | |
| 1846 | |
| 1847 /*************************************** | |
| 1848 * Read \u or \U unicode sequence | |
| 1849 * Input: | |
| 1850 * u 'u' or 'U' | |
| 1851 */ | |
| 1852 | |
| 1853 #if 0 | |
| 1854 unsigned Lexer::wchar(unsigned u) | |
| 1855 { | |
| 1856 unsigned value; | |
| 1857 unsigned n; | |
| 1858 unsigned char c; | |
| 1859 unsigned nchars; | |
| 1860 | |
| 1861 nchars = (u == 'U') ? 8 : 4; | |
| 1862 value = 0; | |
| 1863 for (n = 0; 1; n++) | |
| 1864 { | |
| 1865 ++p; | |
| 1866 if (n == nchars) | |
| 1867 break; | |
| 1868 c = *p; | |
| 1869 if (!ishex(c)) | |
| 1870 { error("\\%c sequence must be followed by %d hex characters", u, nchars); | |
| 1871 break; | |
| 1872 } | |
| 1873 if (isdigit(c)) | |
| 1874 c -= '0'; | |
| 1875 else if (islower(c)) | |
| 1876 c -= 'a' - 10; | |
| 1877 else | |
| 1878 c -= 'A' - 10; | |
| 1879 value <<= 4; | |
| 1880 value |= c; | |
| 1881 } | |
| 1882 return value; | |
| 1883 } | |
| 1884 #endif | |
| 1885 | |
| 1886 /************************************** | |
| 1887 * Read in a number. | |
| 1888 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 1889 * integers can be decimal, octal or hex | |
| 1890 * Handle the suffixes U, UL, LU, L, etc. | |
| 1891 * If it's double, store it in tok.TKutok.Vdouble. | |
| 1892 * Returns: | |
| 1893 * TKnum | |
| 1894 * TKdouble,... | |
| 1895 */ | |
| 1896 | |
| 1897 TOK Lexer::number(Token *t) | |
| 1898 { | |
| 1899 // We use a state machine to collect numbers | |
| 1900 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 1901 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 1902 STATE_hexh, STATE_error }; | |
| 1903 enum STATE state; | |
| 1904 | |
| 1905 enum FLAGS | |
| 1906 { FLAGS_decimal = 1, // decimal | |
| 1907 FLAGS_unsigned = 2, // u or U suffix | |
| 1908 FLAGS_long = 4, // l or L suffix | |
| 1909 }; | |
| 1910 enum FLAGS flags = FLAGS_decimal; | |
| 1911 | |
| 1912 int i; | |
| 1913 int base; | |
| 1914 unsigned c; | |
| 1915 unsigned char *start; | |
| 1916 TOK result; | |
| 1917 | |
| 1918 //printf("Lexer::number()\n"); | |
| 1919 state = STATE_initial; | |
| 1920 base = 0; | |
| 1921 stringbuffer.reset(); | |
| 1922 start = p; | |
| 1923 while (1) | |
| 1924 { | |
| 1925 c = *p; | |
| 1926 switch (state) | |
| 1927 { | |
| 1928 case STATE_initial: // opening state | |
| 1929 if (c == '0') | |
| 1930 state = STATE_0; | |
| 1931 else | |
| 1932 state = STATE_decimal; | |
| 1933 break; | |
| 1934 | |
| 1935 case STATE_0: | |
| 1936 flags = (FLAGS) (flags & ~FLAGS_decimal); | |
| 1937 switch (c) | |
| 1938 { | |
| 1939 #if ZEROH | |
| 1940 case 'H': // 0h | |
| 1941 case 'h': | |
| 1942 goto hexh; | |
| 1943 #endif | |
| 1944 case 'X': | |
| 1945 case 'x': | |
| 1946 state = STATE_hex0; | |
| 1947 break; | |
| 1948 | |
| 1949 case '.': | |
| 1950 if (p[1] == '.') // .. is a separate token | |
| 1951 goto done; | |
| 1952 case 'i': | |
| 1953 case 'f': | |
| 1954 case 'F': | |
| 1955 goto real; | |
| 1956 #if ZEROH | |
| 1957 case 'E': | |
| 1958 case 'e': | |
| 1959 goto case_hex; | |
| 1960 #endif | |
| 1961 case 'B': | |
| 1962 case 'b': | |
| 1963 state = STATE_binary0; | |
| 1964 break; | |
| 1965 | |
| 1966 case '0': case '1': case '2': case '3': | |
| 1967 case '4': case '5': case '6': case '7': | |
| 1968 state = STATE_octal; | |
| 1969 break; | |
| 1970 | |
| 1971 #if ZEROH | |
| 1972 case '8': case '9': case 'A': | |
| 1973 case 'C': case 'D': case 'F': | |
| 1974 case 'a': case 'c': case 'd': case 'f': | |
| 1975 case_hex: | |
| 1976 state = STATE_hexh; | |
| 1977 break; | |
| 1978 #endif | |
| 1979 case '_': | |
| 1980 state = STATE_octal; | |
| 1981 p++; | |
| 1982 continue; | |
| 1983 | |
| 1984 case 'L': | |
| 1985 if (p[1] == 'i') | |
| 1986 goto real; | |
| 1987 goto done; | |
| 1988 | |
| 1989 default: | |
| 1990 goto done; | |
| 1991 } | |
| 1992 break; | |
| 1993 | |
| 1994 case STATE_decimal: // reading decimal number | |
| 1995 if (!isdigit(c)) | |
| 1996 { | |
| 1997 #if ZEROH | |
| 1998 if (ishex(c) | |
| 1999 || c == 'H' || c == 'h' | |
| 2000 ) | |
| 2001 goto hexh; | |
| 2002 #endif | |
| 2003 if (c == '_') // ignore embedded _ | |
| 2004 { p++; | |
| 2005 continue; | |
| 2006 } | |
| 2007 if (c == '.' && p[1] != '.') | |
| 2008 goto real; | |
| 2009 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2010 c == 'e' || c == 'E') | |
| 2011 { | |
| 2012 real: // It's a real number. Back up and rescan as a real | |
| 2013 p = start; | |
| 2014 return inreal(t); | |
| 2015 } | |
| 2016 else if (c == 'L' && p[1] == 'i') | |
| 2017 goto real; | |
| 2018 goto done; | |
| 2019 } | |
| 2020 break; | |
| 2021 | |
| 2022 case STATE_hex0: // reading hex number | |
| 2023 case STATE_hex: | |
| 2024 if (!ishex(c)) | |
| 2025 { | |
| 2026 if (c == '_') // ignore embedded _ | |
| 2027 { p++; | |
| 2028 continue; | |
| 2029 } | |
| 2030 if (c == '.' && p[1] != '.') | |
| 2031 goto real; | |
| 2032 if (c == 'P' || c == 'p' || c == 'i') | |
| 2033 goto real; | |
| 2034 if (state == STATE_hex0) | |
| 2035 error("Hex digit expected, not '%c'", c); | |
| 2036 goto done; | |
| 2037 } | |
| 2038 state = STATE_hex; | |
| 2039 break; | |
| 2040 | |
| 2041 #if ZEROH | |
| 2042 hexh: | |
| 2043 state = STATE_hexh; | |
| 2044 case STATE_hexh: // parse numbers like 0FFh | |
| 2045 if (!ishex(c)) | |
| 2046 { | |
| 2047 if (c == 'H' || c == 'h') | |
| 2048 { | |
| 2049 p++; | |
| 2050 base = 16; | |
| 2051 goto done; | |
| 2052 } | |
| 2053 else | |
| 2054 { | |
| 2055 // Check for something like 1E3 or 0E24 | |
| 2056 if (memchr((char *)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2057 memchr((char *)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2058 goto real; | |
| 2059 error("Hex digit expected, not '%c'", c); | |
| 2060 goto done; | |
| 2061 } | |
| 2062 } | |
| 2063 break; | |
| 2064 #endif | |
| 2065 | |
| 2066 case STATE_octal: // reading octal number | |
| 2067 case STATE_octale: // reading octal number with non-octal digits | |
| 2068 if (!isoctal(c)) | |
| 2069 { | |
| 2070 #if ZEROH | |
| 2071 if (ishex(c) | |
| 2072 || c == 'H' || c == 'h' | |
| 2073 ) | |
| 2074 goto hexh; | |
| 2075 #endif | |
| 2076 if (c == '_') // ignore embedded _ | |
| 2077 { p++; | |
| 2078 continue; | |
| 2079 } | |
| 2080 if (c == '.' && p[1] != '.') | |
| 2081 goto real; | |
| 2082 if (c == 'i') | |
| 2083 goto real; | |
| 2084 if (isdigit(c)) | |
| 2085 { | |
| 2086 state = STATE_octale; | |
| 2087 } | |
| 2088 else | |
| 2089 goto done; | |
| 2090 } | |
| 2091 break; | |
| 2092 | |
| 2093 case STATE_binary0: // starting binary number | |
| 2094 case STATE_binary: // reading binary number | |
| 2095 if (c != '0' && c != '1') | |
| 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 (state == STATE_binary0) | |
| 2108 { error("binary digit expected"); | |
| 2109 state = STATE_error; | |
| 2110 break; | |
| 2111 } | |
| 2112 else | |
| 2113 goto done; | |
| 2114 } | |
| 2115 state = STATE_binary; | |
| 2116 break; | |
| 2117 | |
| 2118 case STATE_error: // for error recovery | |
| 2119 if (!isdigit(c)) // scan until non-digit | |
| 2120 goto done; | |
| 2121 break; | |
| 2122 | |
| 2123 default: | |
| 2124 assert(0); | |
| 2125 } | |
| 2126 stringbuffer.writeByte(c); | |
| 2127 p++; | |
| 2128 } | |
| 2129 done: | |
| 2130 stringbuffer.writeByte(0); // terminate string | |
| 2131 if (state == STATE_octale) | |
| 2132 error("Octal digit expected"); | |
| 2133 | |
| 2134 uinteger_t n; // unsigned >=64 bit integer type | |
| 2135 | |
| 2136 if (stringbuffer.offset == 2 && (state == STATE_decimal || state == STATE_0)) | |
| 2137 n = stringbuffer.data[0] - '0'; | |
| 2138 else | |
| 2139 { | |
| 2140 // Convert string to integer | |
| 2141 #if __DMC__ | |
| 2142 errno = 0; | |
| 2143 n = strtoull((char *)stringbuffer.data,NULL,base); | |
| 2144 if (errno == ERANGE) | |
| 2145 error("integer overflow"); | |
| 2146 #else | |
| 2147 // Not everybody implements strtoull() | |
| 2148 char *p = (char *)stringbuffer.data; | |
| 2149 int r = 10, d; | |
| 2150 | |
| 2151 if (*p == '0') | |
| 2152 { | |
| 2153 if (p[1] == 'x' || p[1] == 'X') | |
| 2154 p += 2, r = 16; | |
| 2155 else if (p[1] == 'b' || p[1] == 'B') | |
| 2156 p += 2, r = 2; | |
| 2157 else if (isdigit(p[1])) | |
| 2158 p += 1, r = 8; | |
| 2159 } | |
| 2160 | |
| 2161 n = 0; | |
| 2162 while (1) | |
| 2163 { | |
| 2164 if (*p >= '0' && *p <= '9') | |
| 2165 d = *p - '0'; | |
| 2166 else if (*p >= 'a' && *p <= 'z') | |
| 2167 d = *p - 'a' + 10; | |
| 2168 else if (*p >= 'A' && *p <= 'Z') | |
| 2169 d = *p - 'A' + 10; | |
| 2170 else | |
| 2171 break; | |
| 2172 if (d >= r) | |
| 2173 break; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2174 uinteger_t n2 = n * r; |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2175 //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
|
2176 if (n2 / r != n || n2 + d < n) |
| 159 | 2177 { |
| 2178 error ("integer overflow"); | |
| 2179 break; | |
| 2180 } | |
| 2181 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2182 n = n2 + d; |
| 159 | 2183 p++; |
| 2184 } | |
| 2185 #endif | |
| 2186 if (sizeof(n) > 8 && | |
| 2187 n > 0xFFFFFFFFFFFFFFFFULL) // if n needs more than 64 bits | |
| 2188 error("integer overflow"); | |
| 2189 } | |
| 2190 | |
| 2191 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2192 while (1) | |
| 2193 { unsigned char f; | |
| 2194 | |
| 2195 switch (*p) | |
| 2196 { case 'U': | |
| 2197 case 'u': | |
| 2198 f = FLAGS_unsigned; | |
| 2199 goto L1; | |
| 2200 | |
| 2201 case 'l': | |
| 2202 if (1 || !global.params.useDeprecated) | |
| 2203 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2204 case 'L': | |
| 2205 f = FLAGS_long; | |
| 2206 L1: | |
| 2207 p++; | |
| 2208 if (flags & f) | |
| 2209 error("unrecognized token"); | |
| 2210 flags = (FLAGS) (flags | f); | |
| 2211 continue; | |
| 2212 default: | |
| 2213 break; | |
| 2214 } | |
| 2215 break; | |
| 2216 } | |
| 2217 | |
| 2218 switch (flags) | |
| 2219 { | |
| 2220 case 0: | |
| 2221 /* Octal or Hexadecimal constant. | |
| 2222 * First that fits: int, uint, long, ulong | |
| 2223 */ | |
| 2224 if (n & 0x8000000000000000LL) | |
| 2225 result = TOKuns64v; | |
| 2226 else if (n & 0xFFFFFFFF00000000LL) | |
| 2227 result = TOKint64v; | |
| 2228 else if (n & 0x80000000) | |
| 2229 result = TOKuns32v; | |
| 2230 else | |
| 2231 result = TOKint32v; | |
| 2232 break; | |
| 2233 | |
| 2234 case FLAGS_decimal: | |
| 2235 /* First that fits: int, long, long long | |
| 2236 */ | |
| 2237 if (n & 0x8000000000000000LL) | |
| 2238 { error("signed integer overflow"); | |
| 2239 result = TOKuns64v; | |
| 2240 } | |
| 2241 else if (n & 0xFFFFFFFF80000000LL) | |
| 2242 result = TOKint64v; | |
| 2243 else | |
| 2244 result = TOKint32v; | |
| 2245 break; | |
| 2246 | |
| 2247 case FLAGS_unsigned: | |
| 2248 case FLAGS_decimal | FLAGS_unsigned: | |
| 2249 /* First that fits: uint, ulong | |
| 2250 */ | |
| 2251 if (n & 0xFFFFFFFF00000000LL) | |
| 2252 result = TOKuns64v; | |
| 2253 else | |
| 2254 result = TOKuns32v; | |
| 2255 break; | |
| 2256 | |
| 2257 case FLAGS_decimal | FLAGS_long: | |
| 2258 if (n & 0x8000000000000000LL) | |
| 2259 { error("signed integer overflow"); | |
| 2260 result = TOKuns64v; | |
| 2261 } | |
| 2262 else | |
| 2263 result = TOKint64v; | |
| 2264 break; | |
| 2265 | |
| 2266 case FLAGS_long: | |
| 2267 if (n & 0x8000000000000000LL) | |
| 2268 result = TOKuns64v; | |
| 2269 else | |
| 2270 result = TOKint64v; | |
| 2271 break; | |
| 2272 | |
| 2273 case FLAGS_unsigned | FLAGS_long: | |
| 2274 case FLAGS_decimal | FLAGS_unsigned | FLAGS_long: | |
| 2275 result = TOKuns64v; | |
| 2276 break; | |
| 2277 | |
| 2278 default: | |
| 2279 #ifdef DEBUG | |
| 2280 printf("%x\n",flags); | |
| 2281 #endif | |
| 2282 assert(0); | |
| 2283 } | |
| 2284 t->uns64value = n; | |
| 2285 return result; | |
| 2286 } | |
| 2287 | |
| 2288 /************************************** | |
| 2289 * Read in characters, converting them to real. | |
| 2290 * Bugs: | |
| 2291 * Exponent overflow not detected. | |
| 2292 * Too much requested precision is not detected. | |
| 2293 */ | |
| 2294 | |
| 2295 TOK Lexer::inreal(Token *t) | |
| 2296 #ifdef __DMC__ | |
| 2297 __in | |
| 2298 { | |
| 2299 assert(*p == '.' || isdigit(*p)); | |
| 2300 } | |
| 2301 __out (result) | |
| 2302 { | |
| 2303 switch (result) | |
| 2304 { | |
| 2305 case TOKfloat32v: | |
| 2306 case TOKfloat64v: | |
| 2307 case TOKfloat80v: | |
| 2308 case TOKimaginary32v: | |
| 2309 case TOKimaginary64v: | |
| 2310 case TOKimaginary80v: | |
| 2311 break; | |
| 2312 | |
| 2313 default: | |
| 2314 assert(0); | |
| 2315 } | |
| 2316 } | |
| 2317 __body | |
| 2318 #endif /* __DMC__ */ | |
| 2319 { int dblstate; | |
| 2320 unsigned c; | |
| 2321 char hex; // is this a hexadecimal-floating-constant? | |
| 2322 TOK result; | |
| 2323 | |
| 2324 //printf("Lexer::inreal()\n"); | |
| 2325 stringbuffer.reset(); | |
| 2326 dblstate = 0; | |
| 2327 hex = 0; | |
| 2328 Lnext: | |
| 2329 while (1) | |
| 2330 { | |
| 2331 // Get next char from input | |
| 2332 c = *p++; | |
| 2333 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2334 while (1) | |
| 2335 { | |
| 2336 switch (dblstate) | |
| 2337 { | |
| 2338 case 0: // opening state | |
| 2339 if (c == '0') | |
| 2340 dblstate = 9; | |
| 2341 else if (c == '.') | |
| 2342 dblstate = 3; | |
| 2343 else | |
| 2344 dblstate = 1; | |
| 2345 break; | |
| 2346 | |
| 2347 case 9: | |
| 2348 dblstate = 1; | |
| 2349 if (c == 'X' || c == 'x') | |
| 2350 { hex++; | |
| 2351 break; | |
| 2352 } | |
| 2353 case 1: // digits to left of . | |
| 2354 case 3: // digits to right of . | |
| 2355 case 7: // continuing exponent digits | |
| 2356 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2357 { | |
| 2358 if (c == '_') | |
| 2359 goto Lnext; // ignore embedded '_' | |
| 2360 dblstate++; | |
| 2361 continue; | |
| 2362 } | |
| 2363 break; | |
| 2364 | |
| 2365 case 2: // no more digits to left of . | |
| 2366 if (c == '.') | |
| 2367 { dblstate++; | |
| 2368 break; | |
| 2369 } | |
| 2370 case 4: // no more digits to right of . | |
| 2371 if ((c == 'E' || c == 'e') || | |
| 2372 hex && (c == 'P' || c == 'p')) | |
| 2373 { dblstate = 5; | |
| 2374 hex = 0; // exponent is always decimal | |
| 2375 break; | |
| 2376 } | |
| 2377 if (hex) | |
| 2378 error("binary-exponent-part required"); | |
| 2379 goto done; | |
| 2380 | |
| 2381 case 5: // looking immediately to right of E | |
| 2382 dblstate++; | |
| 2383 if (c == '-' || c == '+') | |
| 2384 break; | |
| 2385 case 6: // 1st exponent digit expected | |
| 2386 if (!isdigit(c)) | |
| 2387 error("exponent expected"); | |
| 2388 dblstate++; | |
| 2389 break; | |
| 2390 | |
| 2391 case 8: // past end of exponent digits | |
| 2392 goto done; | |
| 2393 } | |
| 2394 break; | |
| 2395 } | |
| 2396 stringbuffer.writeByte(c); | |
| 2397 } | |
| 2398 done: | |
| 2399 p--; | |
| 2400 | |
| 2401 stringbuffer.writeByte(0); | |
| 2402 | |
| 2403 #if _WIN32 && __DMC__ | |
| 2404 char *save = __locale_decpoint; | |
| 2405 __locale_decpoint = "."; | |
| 2406 #endif | |
| 2407 #ifdef IN_GCC | |
| 2408 t->float80value = real_t::parse((char *)stringbuffer.data, real_t::LongDouble); | |
| 2409 #else | |
| 2410 t->float80value = strtold((char *)stringbuffer.data, NULL); | |
| 2411 #endif | |
| 2412 errno = 0; | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2413 float strtofres; |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2414 double strtodres; |
| 159 | 2415 switch (*p) |
| 2416 { | |
| 2417 case 'F': | |
| 2418 case 'f': | |
| 2419 #ifdef IN_GCC | |
| 2420 real_t::parse((char *)stringbuffer.data, real_t::Float); | |
| 2421 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2422 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
|
2423 // 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
|
2424 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2425 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
|
2426 errno = 0; |
| 159 | 2427 #endif |
| 2428 result = TOKfloat32v; | |
| 2429 p++; | |
| 2430 break; | |
| 2431 | |
| 2432 default: | |
| 2433 #ifdef IN_GCC | |
| 2434 real_t::parse((char *)stringbuffer.data, real_t::Double); | |
| 2435 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2436 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
|
2437 // 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
|
2438 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2439 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
|
2440 errno = 0; |
| 159 | 2441 #endif |
| 2442 result = TOKfloat64v; | |
| 2443 break; | |
| 2444 | |
| 2445 case 'l': | |
| 2446 if (!global.params.useDeprecated) | |
| 2447 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2448 case 'L': | |
| 2449 result = TOKfloat80v; | |
| 2450 p++; | |
| 2451 break; | |
| 2452 } | |
| 2453 if (*p == 'i' || *p == 'I') | |
| 2454 { | |
| 2455 if (!global.params.useDeprecated && *p == 'I') | |
| 2456 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2457 p++; | |
| 2458 switch (result) | |
| 2459 { | |
| 2460 case TOKfloat32v: | |
| 2461 result = TOKimaginary32v; | |
| 2462 break; | |
| 2463 case TOKfloat64v: | |
| 2464 result = TOKimaginary64v; | |
| 2465 break; | |
| 2466 case TOKfloat80v: | |
| 2467 result = TOKimaginary80v; | |
| 2468 break; | |
| 2469 } | |
| 2470 } | |
| 2471 #if _WIN32 && __DMC__ | |
| 2472 __locale_decpoint = save; | |
| 2473 #endif | |
| 2474 if (errno == ERANGE) | |
| 2475 error("number is not representable"); | |
| 2476 return result; | |
| 2477 } | |
| 2478 | |
| 2479 /********************************************* | |
| 2480 * Do pragma. | |
| 2481 * Currently, the only pragma supported is: | |
| 2482 * #line linnum [filespec] | |
| 2483 */ | |
| 2484 | |
| 2485 void Lexer::pragma() | |
| 2486 { | |
| 2487 Token tok; | |
| 2488 int linnum; | |
| 2489 char *filespec = NULL; | |
| 2490 Loc loc = this->loc; | |
| 2491 | |
| 2492 scan(&tok); | |
| 2493 if (tok.value != TOKidentifier || tok.ident != Id::line) | |
| 2494 goto Lerr; | |
| 2495 | |
| 2496 scan(&tok); | |
| 2497 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 2498 linnum = tok.uns64value - 1; | |
| 2499 else | |
| 2500 goto Lerr; | |
| 2501 | |
| 2502 while (1) | |
| 2503 { | |
| 2504 switch (*p) | |
| 2505 { | |
| 2506 case 0: | |
| 2507 case 0x1A: | |
| 2508 case '\n': | |
| 2509 Lnewline: | |
| 2510 this->loc.linnum = linnum; | |
| 2511 if (filespec) | |
| 2512 this->loc.filename = filespec; | |
| 2513 return; | |
| 2514 | |
| 2515 case '\r': | |
| 2516 p++; | |
| 2517 if (*p != '\n') | |
| 2518 { p--; | |
| 2519 goto Lnewline; | |
| 2520 } | |
| 2521 continue; | |
| 2522 | |
| 2523 case ' ': | |
| 2524 case '\t': | |
| 2525 case '\v': | |
| 2526 case '\f': | |
| 2527 p++; | |
| 2528 continue; // skip white space | |
| 2529 | |
| 2530 case '_': | |
| 2531 if (mod && memcmp(p, "__FILE__", 8) == 0) | |
| 2532 { | |
| 2533 p += 8; | |
| 2534 filespec = mem.strdup(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 2535 } | |
| 2536 continue; | |
| 2537 | |
| 2538 case '"': | |
| 2539 if (filespec) | |
| 2540 goto Lerr; | |
| 2541 stringbuffer.reset(); | |
| 2542 p++; | |
| 2543 while (1) | |
| 2544 { unsigned c; | |
| 2545 | |
| 2546 c = *p; | |
| 2547 switch (c) | |
| 2548 { | |
| 2549 case '\n': | |
| 2550 case '\r': | |
| 2551 case 0: | |
| 2552 case 0x1A: | |
| 2553 goto Lerr; | |
| 2554 | |
| 2555 case '"': | |
| 2556 stringbuffer.writeByte(0); | |
| 2557 filespec = mem.strdup((char *)stringbuffer.data); | |
| 2558 p++; | |
| 2559 break; | |
| 2560 | |
| 2561 default: | |
| 2562 if (c & 0x80) | |
| 2563 { unsigned u = decodeUTF(); | |
| 2564 if (u == PS || u == LS) | |
| 2565 goto Lerr; | |
| 2566 } | |
| 2567 stringbuffer.writeByte(c); | |
| 2568 p++; | |
| 2569 continue; | |
| 2570 } | |
| 2571 break; | |
| 2572 } | |
| 2573 continue; | |
| 2574 | |
| 2575 default: | |
| 2576 if (*p & 0x80) | |
| 2577 { unsigned u = decodeUTF(); | |
| 2578 if (u == PS || u == LS) | |
| 2579 goto Lnewline; | |
| 2580 } | |
| 2581 goto Lerr; | |
| 2582 } | |
| 2583 } | |
| 2584 | |
| 2585 Lerr: | |
| 2586 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 2587 } | |
| 2588 | |
| 2589 | |
| 2590 /******************************************** | |
| 2591 * Decode UTF character. | |
| 2592 * Issue error messages for invalid sequences. | |
| 2593 * Return decoded character, advance p to last character in UTF sequence. | |
| 2594 */ | |
| 2595 | |
| 2596 unsigned Lexer::decodeUTF() | |
| 2597 { | |
| 2598 dchar_t u; | |
| 2599 unsigned char c; | |
| 2600 unsigned char *s = p; | |
| 2601 size_t len; | |
| 2602 size_t idx; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2603 const char *msg; |
| 159 | 2604 |
| 2605 c = *s; | |
| 2606 assert(c & 0x80); | |
| 2607 | |
| 2608 // Check length of remaining string up to 6 UTF-8 characters | |
| 2609 for (len = 1; len < 6 && s[len]; len++) | |
| 2610 ; | |
| 2611 | |
| 2612 idx = 0; | |
| 2613 msg = utf_decodeChar(s, len, &idx, &u); | |
| 2614 p += idx - 1; | |
| 2615 if (msg) | |
| 2616 { | |
| 2617 error("%s", msg); | |
| 2618 } | |
| 2619 return u; | |
| 2620 } | |
| 2621 | |
| 2622 | |
| 2623 /*************************************************** | |
| 2624 * Parse doc comment embedded between t->ptr and p. | |
| 2625 * Remove trailing blanks and tabs from lines. | |
| 2626 * Replace all newlines with \n. | |
| 2627 * Remove leading comment character from each line. | |
| 2628 * Decide if it's a lineComment or a blockComment. | |
| 2629 * Append to previous one for this token. | |
| 2630 */ | |
| 2631 | |
| 2632 void Lexer::getDocComment(Token *t, unsigned lineComment) | |
| 2633 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2634 /* 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
|
2635 */ |
| 159 | 2636 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
|
2637 |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2638 /* 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
|
2639 */ |
| 159 | 2640 unsigned char *q = t->ptr + 3; // start of comment text |
| 2641 | |
| 2642 unsigned char *qend = p; | |
| 2643 if (ct == '*' || ct == '+') | |
| 2644 qend -= 2; | |
| 2645 | |
| 2646 /* Scan over initial row of ****'s or ++++'s or ////'s | |
| 2647 */ | |
| 2648 for (; q < qend; q++) | |
| 2649 { | |
| 2650 if (*q != ct) | |
| 2651 break; | |
| 2652 } | |
| 2653 | |
| 2654 /* Remove trailing row of ****'s or ++++'s | |
| 2655 */ | |
| 2656 if (ct != '/') | |
| 2657 { | |
| 2658 for (; q < qend; qend--) | |
| 2659 { | |
| 2660 if (qend[-1] != ct) | |
| 2661 break; | |
| 2662 } | |
| 2663 } | |
| 2664 | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2665 /* Comment is now [q .. qend]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2666 * Canonicalize it into buf[]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2667 */ |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2668 OutBuffer buf; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2669 int linestart = 0; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2670 |
| 159 | 2671 for (; q < qend; q++) |
| 2672 { | |
| 2673 unsigned char c = *q; | |
| 2674 | |
| 2675 switch (c) | |
| 2676 { | |
| 2677 case '*': | |
| 2678 case '+': | |
| 2679 if (linestart && c == ct) | |
| 2680 { linestart = 0; | |
| 2681 /* Trim preceding whitespace up to preceding \n | |
| 2682 */ | |
| 2683 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2684 buf.offset--; | |
| 2685 continue; | |
| 2686 } | |
| 2687 break; | |
| 2688 | |
| 2689 case ' ': | |
| 2690 case '\t': | |
| 2691 break; | |
| 2692 | |
| 2693 case '\r': | |
| 2694 if (q[1] == '\n') | |
| 2695 continue; // skip the \r | |
| 2696 goto Lnewline; | |
| 2697 | |
| 2698 default: | |
| 2699 if (c == 226) | |
| 2700 { | |
| 2701 // If LS or PS | |
| 2702 if (q[1] == 128 && | |
| 2703 (q[2] == 168 || q[2] == 169)) | |
| 2704 { | |
| 2705 q += 2; | |
| 2706 goto Lnewline; | |
| 2707 } | |
| 2708 } | |
| 2709 linestart = 0; | |
| 2710 break; | |
| 2711 | |
| 2712 Lnewline: | |
| 2713 c = '\n'; // replace all newlines with \n | |
| 2714 case '\n': | |
| 2715 linestart = 1; | |
| 2716 | |
| 2717 /* Trim trailing whitespace | |
| 2718 */ | |
| 2719 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2720 buf.offset--; | |
| 2721 | |
| 2722 break; | |
| 2723 } | |
| 2724 buf.writeByte(c); | |
| 2725 } | |
| 2726 | |
| 2727 // Always end with a newline | |
| 2728 if (!buf.offset || buf.data[buf.offset - 1] != '\n') | |
| 2729 buf.writeByte('\n'); | |
| 2730 | |
| 2731 buf.writeByte(0); | |
| 2732 | |
| 2733 // It's a line comment if the start of the doc comment comes | |
| 2734 // after other non-whitespace on the same line. | |
| 2735 unsigned char** dc = (lineComment && anyToken) | |
| 2736 ? &t->lineComment | |
| 2737 : &t->blockComment; | |
| 2738 | |
| 2739 // Combine with previous doc comment, if any | |
| 2740 if (*dc) | |
| 2741 *dc = combineComments(*dc, (unsigned char *)buf.data); | |
| 2742 else | |
| 2743 *dc = (unsigned char *)buf.extractData(); | |
| 2744 } | |
| 2745 | |
| 2746 /******************************************** | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2747 * Combine two document comments into one, |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2748 * separated by a newline. |
| 159 | 2749 */ |
| 2750 | |
| 2751 unsigned char *Lexer::combineComments(unsigned char *c1, unsigned char *c2) | |
| 2752 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2753 //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
|
2754 |
| 159 | 2755 unsigned char *c = c2; |
| 2756 | |
| 2757 if (c1) | |
| 2758 { c = c1; | |
| 2759 if (c2) | |
| 2760 { size_t len1 = strlen((char *)c1); | |
| 2761 size_t len2 = strlen((char *)c2); | |
| 2762 | |
| 2763 c = (unsigned char *)mem.malloc(len1 + 1 + len2 + 1); | |
| 2764 memcpy(c, c1, len1); | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2765 if (len1 && c1[len1 - 1] != '\n') |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2766 { c[len1] = '\n'; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2767 len1++; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2768 } |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2769 memcpy(c + len1, c2, len2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2770 c[len1 + len2] = 0; |
| 159 | 2771 } |
| 2772 } | |
| 2773 return c; | |
| 2774 } | |
| 2775 | |
| 2776 /******************************************** | |
| 2777 * Create an identifier in the string table. | |
| 2778 */ | |
| 2779 | |
| 2780 Identifier *Lexer::idPool(const char *s) | |
| 2781 { | |
| 2782 size_t len = strlen(s); | |
| 2783 StringValue *sv = stringtable.update(s, len); | |
| 2784 Identifier *id = (Identifier *) sv->ptrvalue; | |
| 2785 if (!id) | |
| 2786 { | |
| 2787 id = new Identifier(sv->lstring.string, TOKidentifier); | |
| 2788 sv->ptrvalue = id; | |
| 2789 } | |
| 2790 return id; | |
| 2791 } | |
| 2792 | |
| 2793 /********************************************* | |
| 2794 * Create a unique identifier using the prefix s. | |
| 2795 */ | |
| 2796 | |
| 2797 Identifier *Lexer::uniqueId(const char *s, int num) | |
| 2798 { char buffer[32]; | |
| 2799 size_t slen = strlen(s); | |
| 2800 | |
| 2801 assert(slen + sizeof(num) * 3 + 1 <= sizeof(buffer)); | |
| 2802 sprintf(buffer, "%s%d", s, num); | |
| 2803 return idPool(buffer); | |
| 2804 } | |
| 2805 | |
| 2806 Identifier *Lexer::uniqueId(const char *s) | |
| 2807 { | |
| 2808 static int num; | |
| 2809 return uniqueId(s, ++num); | |
| 2810 } | |
| 2811 | |
| 2812 /**************************************** | |
| 2813 */ | |
| 2814 | |
| 2815 struct Keyword | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2816 { const char *name; |
| 159 | 2817 enum TOK value; |
| 2818 }; | |
| 2819 | |
| 2820 static Keyword keywords[] = | |
| 2821 { | |
| 2822 // { "", TOK }, | |
| 2823 | |
| 2824 { "this", TOKthis }, | |
| 2825 { "super", TOKsuper }, | |
| 2826 { "assert", TOKassert }, | |
| 2827 { "null", TOKnull }, | |
| 2828 { "true", TOKtrue }, | |
| 2829 { "false", TOKfalse }, | |
| 2830 { "cast", TOKcast }, | |
| 2831 { "new", TOKnew }, | |
| 2832 { "delete", TOKdelete }, | |
| 2833 { "throw", TOKthrow }, | |
| 2834 { "module", TOKmodule }, | |
| 2835 { "pragma", TOKpragma }, | |
| 2836 { "typeof", TOKtypeof }, | |
| 2837 { "typeid", TOKtypeid }, | |
| 2838 | |
| 2839 { "template", TOKtemplate }, | |
| 2840 | |
| 2841 { "void", TOKvoid }, | |
| 2842 { "byte", TOKint8 }, | |
| 2843 { "ubyte", TOKuns8 }, | |
| 2844 { "short", TOKint16 }, | |
| 2845 { "ushort", TOKuns16 }, | |
| 2846 { "int", TOKint32 }, | |
| 2847 { "uint", TOKuns32 }, | |
| 2848 { "long", TOKint64 }, | |
| 2849 { "ulong", TOKuns64 }, | |
| 2850 { "cent", TOKcent, }, | |
| 2851 { "ucent", TOKucent, }, | |
| 2852 { "float", TOKfloat32 }, | |
| 2853 { "double", TOKfloat64 }, | |
| 2854 { "real", TOKfloat80 }, | |
| 2855 | |
| 2856 { "bool", TOKbool }, | |
| 2857 { "char", TOKchar }, | |
| 2858 { "wchar", TOKwchar }, | |
| 2859 { "dchar", TOKdchar }, | |
| 2860 | |
| 2861 { "ifloat", TOKimaginary32 }, | |
| 2862 { "idouble", TOKimaginary64 }, | |
| 2863 { "ireal", TOKimaginary80 }, | |
| 2864 | |
| 2865 { "cfloat", TOKcomplex32 }, | |
| 2866 { "cdouble", TOKcomplex64 }, | |
| 2867 { "creal", TOKcomplex80 }, | |
| 2868 | |
| 2869 { "delegate", TOKdelegate }, | |
| 2870 { "function", TOKfunction }, | |
| 2871 | |
| 2872 { "is", TOKis }, | |
| 2873 { "if", TOKif }, | |
| 2874 { "else", TOKelse }, | |
| 2875 { "while", TOKwhile }, | |
| 2876 { "for", TOKfor }, | |
| 2877 { "do", TOKdo }, | |
| 2878 { "switch", TOKswitch }, | |
| 2879 { "case", TOKcase }, | |
| 2880 { "default", TOKdefault }, | |
| 2881 { "break", TOKbreak }, | |
| 2882 { "continue", TOKcontinue }, | |
| 2883 { "synchronized", TOKsynchronized }, | |
| 2884 { "return", TOKreturn }, | |
| 2885 { "goto", TOKgoto }, | |
| 2886 { "try", TOKtry }, | |
| 2887 { "catch", TOKcatch }, | |
| 2888 { "finally", TOKfinally }, | |
| 2889 { "with", TOKwith }, | |
| 2890 { "asm", TOKasm }, | |
| 2891 { "foreach", TOKforeach }, | |
| 2892 { "foreach_reverse", TOKforeach_reverse }, | |
| 2893 { "scope", TOKscope }, | |
| 2894 | |
| 2895 { "struct", TOKstruct }, | |
| 2896 { "class", TOKclass }, | |
| 2897 { "interface", TOKinterface }, | |
| 2898 { "union", TOKunion }, | |
| 2899 { "enum", TOKenum }, | |
| 2900 { "import", TOKimport }, | |
| 2901 { "mixin", TOKmixin }, | |
| 2902 { "static", TOKstatic }, | |
| 2903 { "final", TOKfinal }, | |
| 2904 { "const", TOKconst }, | |
| 2905 { "typedef", TOKtypedef }, | |
| 2906 { "alias", TOKalias }, | |
| 2907 { "override", TOKoverride }, | |
| 2908 { "abstract", TOKabstract }, | |
| 2909 { "volatile", TOKvolatile }, | |
| 2910 { "debug", TOKdebug }, | |
| 2911 { "deprecated", TOKdeprecated }, | |
| 2912 { "in", TOKin }, | |
| 2913 { "out", TOKout }, | |
| 2914 { "inout", TOKinout }, | |
| 2915 { "lazy", TOKlazy }, | |
| 2916 { "auto", TOKauto }, | |
| 2917 | |
| 2918 { "align", TOKalign }, | |
| 2919 { "extern", TOKextern }, | |
| 2920 { "private", TOKprivate }, | |
| 2921 { "package", TOKpackage }, | |
| 2922 { "protected", TOKprotected }, | |
| 2923 { "public", TOKpublic }, | |
| 2924 { "export", TOKexport }, | |
| 2925 | |
| 2926 { "body", TOKbody }, | |
| 2927 { "invariant", TOKinvariant }, | |
| 2928 { "unittest", TOKunittest }, | |
| 2929 { "version", TOKversion }, | |
| 2930 //{ "manifest", TOKmanifest }, | |
| 2931 | |
| 2932 // Added after 1.0 | |
| 2933 { "ref", TOKref }, | |
| 2934 { "macro", TOKmacro }, | |
| 336 | 2935 #if DMDV2 |
| 159 | 2936 { "pure", TOKpure }, |
| 2937 { "nothrow", TOKnothrow }, | |
| 336 | 2938 { "__thread", TOKtls }, |
| 159 | 2939 { "__traits", TOKtraits }, |
| 2940 { "__overloadset", TOKoverloadset }, | |
| 336 | 2941 { "__FILE__", TOKfile }, |
| 2942 { "__LINE__", TOKline }, | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2943 { "shared", TOKshared }, |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2944 { "immutable", TOKimmutable }, |
| 159 | 2945 #endif |
| 2946 }; | |
| 2947 | |
| 2948 int Token::isKeyword() | |
| 2949 { | |
| 2950 for (unsigned u = 0; u < sizeof(keywords) / sizeof(keywords[0]); u++) | |
| 2951 { | |
| 2952 if (keywords[u].value == value) | |
| 2953 return 1; | |
| 2954 } | |
| 2955 return 0; | |
| 2956 } | |
| 2957 | |
| 2958 void Lexer::initKeywords() | |
| 2959 { StringValue *sv; | |
| 2960 unsigned u; | |
| 2961 enum TOK v; | |
| 2962 unsigned nkeywords = sizeof(keywords) / sizeof(keywords[0]); | |
| 2963 | |
| 2964 if (global.params.Dversion == 1) | |
| 2965 nkeywords -= 2; | |
| 2966 | |
| 2967 cmtable_init(); | |
| 2968 | |
| 2969 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
|
2970 { const char *s; |
| 159 | 2971 |
| 2972 //printf("keyword[%d] = '%s'\n",u, keywords[u].name); | |
| 2973 s = keywords[u].name; | |
| 2974 v = keywords[u].value; | |
| 2975 sv = stringtable.insert(s, strlen(s)); | |
| 2976 sv->ptrvalue = (void *) new Identifier(sv->lstring.string,v); | |
| 2977 | |
| 2978 //printf("tochars[%d] = '%s'\n",v, s); | |
| 2979 Token::tochars[v] = s; | |
| 2980 } | |
| 2981 | |
| 2982 Token::tochars[TOKeof] = "EOF"; | |
| 2983 Token::tochars[TOKlcurly] = "{"; | |
| 2984 Token::tochars[TOKrcurly] = "}"; | |
| 2985 Token::tochars[TOKlparen] = "("; | |
| 2986 Token::tochars[TOKrparen] = ")"; | |
| 2987 Token::tochars[TOKlbracket] = "["; | |
| 2988 Token::tochars[TOKrbracket] = "]"; | |
| 2989 Token::tochars[TOKsemicolon] = ";"; | |
| 2990 Token::tochars[TOKcolon] = ":"; | |
| 2991 Token::tochars[TOKcomma] = ","; | |
| 2992 Token::tochars[TOKdot] = "."; | |
| 2993 Token::tochars[TOKxor] = "^"; | |
| 2994 Token::tochars[TOKxorass] = "^="; | |
| 2995 Token::tochars[TOKassign] = "="; | |
| 2996 Token::tochars[TOKconstruct] = "="; | |
| 336 | 2997 #if DMDV2 |
| 159 | 2998 Token::tochars[TOKblit] = "="; |
| 2999 #endif | |
| 3000 Token::tochars[TOKlt] = "<"; | |
| 3001 Token::tochars[TOKgt] = ">"; | |
| 3002 Token::tochars[TOKle] = "<="; | |
| 3003 Token::tochars[TOKge] = ">="; | |
| 3004 Token::tochars[TOKequal] = "=="; | |
| 3005 Token::tochars[TOKnotequal] = "!="; | |
| 3006 Token::tochars[TOKnotidentity] = "!is"; | |
| 3007 Token::tochars[TOKtobool] = "!!"; | |
| 3008 | |
| 3009 Token::tochars[TOKunord] = "!<>="; | |
| 3010 Token::tochars[TOKue] = "!<>"; | |
| 3011 Token::tochars[TOKlg] = "<>"; | |
| 3012 Token::tochars[TOKleg] = "<>="; | |
| 3013 Token::tochars[TOKule] = "!>"; | |
| 3014 Token::tochars[TOKul] = "!>="; | |
| 3015 Token::tochars[TOKuge] = "!<"; | |
| 3016 Token::tochars[TOKug] = "!<="; | |
| 3017 | |
| 3018 Token::tochars[TOKnot] = "!"; | |
| 3019 Token::tochars[TOKtobool] = "!!"; | |
| 3020 Token::tochars[TOKshl] = "<<"; | |
| 3021 Token::tochars[TOKshr] = ">>"; | |
| 3022 Token::tochars[TOKushr] = ">>>"; | |
| 3023 Token::tochars[TOKadd] = "+"; | |
| 3024 Token::tochars[TOKmin] = "-"; | |
| 3025 Token::tochars[TOKmul] = "*"; | |
| 3026 Token::tochars[TOKdiv] = "/"; | |
| 3027 Token::tochars[TOKmod] = "%"; | |
| 3028 Token::tochars[TOKslice] = ".."; | |
| 3029 Token::tochars[TOKdotdotdot] = "..."; | |
| 3030 Token::tochars[TOKand] = "&"; | |
| 3031 Token::tochars[TOKandand] = "&&"; | |
| 3032 Token::tochars[TOKor] = "|"; | |
| 3033 Token::tochars[TOKoror] = "||"; | |
| 3034 Token::tochars[TOKarray] = "[]"; | |
| 3035 Token::tochars[TOKindex] = "[i]"; | |
| 3036 Token::tochars[TOKaddress] = "&"; | |
| 3037 Token::tochars[TOKstar] = "*"; | |
| 3038 Token::tochars[TOKtilde] = "~"; | |
| 3039 Token::tochars[TOKdollar] = "$"; | |
| 3040 Token::tochars[TOKcast] = "cast"; | |
| 3041 Token::tochars[TOKplusplus] = "++"; | |
| 3042 Token::tochars[TOKminusminus] = "--"; | |
| 3043 Token::tochars[TOKtype] = "type"; | |
| 3044 Token::tochars[TOKquestion] = "?"; | |
| 3045 Token::tochars[TOKneg] = "-"; | |
| 3046 Token::tochars[TOKuadd] = "+"; | |
| 3047 Token::tochars[TOKvar] = "var"; | |
| 3048 Token::tochars[TOKaddass] = "+="; | |
| 3049 Token::tochars[TOKminass] = "-="; | |
| 3050 Token::tochars[TOKmulass] = "*="; | |
| 3051 Token::tochars[TOKdivass] = "/="; | |
| 3052 Token::tochars[TOKmodass] = "%="; | |
| 3053 Token::tochars[TOKshlass] = "<<="; | |
| 3054 Token::tochars[TOKshrass] = ">>="; | |
| 3055 Token::tochars[TOKushrass] = ">>>="; | |
| 3056 Token::tochars[TOKandass] = "&="; | |
| 3057 Token::tochars[TOKorass] = "|="; | |
| 3058 Token::tochars[TOKcatass] = "~="; | |
| 3059 Token::tochars[TOKcat] = "~"; | |
| 3060 Token::tochars[TOKcall] = "call"; | |
| 3061 Token::tochars[TOKidentity] = "is"; | |
| 3062 Token::tochars[TOKnotidentity] = "!is"; | |
| 3063 | |
| 3064 Token::tochars[TOKorass] = "|="; | |
| 3065 Token::tochars[TOKidentifier] = "identifier"; | |
| 3066 | |
| 3067 // For debugging | |
| 3068 Token::tochars[TOKdotexp] = "dotexp"; | |
| 3069 Token::tochars[TOKdotti] = "dotti"; | |
| 3070 Token::tochars[TOKdotvar] = "dotvar"; | |
| 3071 Token::tochars[TOKdottype] = "dottype"; | |
| 3072 Token::tochars[TOKsymoff] = "symoff"; | |
| 3073 Token::tochars[TOKtypedot] = "typedot"; | |
| 3074 Token::tochars[TOKarraylength] = "arraylength"; | |
| 3075 Token::tochars[TOKarrayliteral] = "arrayliteral"; | |
| 3076 Token::tochars[TOKassocarrayliteral] = "assocarrayliteral"; | |
| 3077 Token::tochars[TOKstructliteral] = "structliteral"; | |
| 3078 Token::tochars[TOKstring] = "string"; | |
| 3079 Token::tochars[TOKdsymbol] = "symbol"; | |
| 3080 Token::tochars[TOKtuple] = "tuple"; | |
| 3081 Token::tochars[TOKdeclaration] = "declaration"; | |
| 3082 Token::tochars[TOKdottd] = "dottd"; | |
| 3083 Token::tochars[TOKon_scope_exit] = "scope(exit)"; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
3084 Token::tochars[TOKon_scope_success] = "scope(success)"; |
|
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
3085 Token::tochars[TOKon_scope_failure] = "scope(failure)"; |
| 159 | 3086 } |