Mercurial > projects > ldc
annotate dmd/lexer.c @ 1638:0de4525a9ed6
Apply workaround for #395 by klickverbot.
| author | Christian Kamm <kamm incasoftware de> |
|---|---|
| date | Mon, 08 Mar 2010 20:06:08 +0100 |
| parents | dbf7b54f542f |
| children |
| 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 | |
|
1431
5f6f0929ee4c
Define __C99FEATURES__ in lexer.c for Solaris. Fixes #313.
Christian Kamm <kamm incasoftware de>
parents:
1367
diff
changeset
|
11 #if __sun && __SVR4 |
|
5f6f0929ee4c
Define __C99FEATURES__ in lexer.c for Solaris. Fixes #313.
Christian Kamm <kamm incasoftware de>
parents:
1367
diff
changeset
|
12 #define __C99FEATURES__ 1 // Needed on Solaris for NaN and more, LDC#313 |
|
5f6f0929ee4c
Define __C99FEATURES__ in lexer.c for Solaris. Fixes #313.
Christian Kamm <kamm incasoftware de>
parents:
1367
diff
changeset
|
13 #endif |
|
5f6f0929ee4c
Define __C99FEATURES__ in lexer.c for Solaris. Fixes #313.
Christian Kamm <kamm incasoftware de>
parents:
1367
diff
changeset
|
14 |
|
1228
79758fd2f48a
Added Doxygen file.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1195
diff
changeset
|
15 #if IN_LLVM |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
16 #include <cmath> |
|
1228
79758fd2f48a
Added Doxygen file.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1195
diff
changeset
|
17 #endif |
|
872
aa953cc960b6
Apply BlueZeniX's patch for OpenSolaris compatibility. Fixes #158.
Christian Kamm <kamm incasoftware de>
parents:
846
diff
changeset
|
18 |
| 159 | 19 /* Lexical Analyzer */ |
| 20 | |
| 21 #include <stdio.h> | |
| 22 #include <string.h> | |
| 23 #include <ctype.h> | |
| 24 #include <stdarg.h> | |
| 25 #include <errno.h> | |
| 26 #include <wchar.h> | |
| 27 #include <stdlib.h> | |
| 28 #include <assert.h> | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
29 #include <time.h> // for time() and ctime() |
| 159 | 30 |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
31 #include "rmem.h" |
| 159 | 32 |
| 33 #include "stringtable.h" | |
| 34 | |
| 35 #include "lexer.h" | |
| 36 #include "utf.h" | |
| 37 #include "identifier.h" | |
| 38 #include "id.h" | |
| 39 #include "module.h" | |
| 40 | |
| 41 #if _WIN32 && __DMC__ | |
| 42 // from \dm\src\include\setlocal.h | |
| 43 extern "C" char * __cdecl __locale_decpoint; | |
| 44 #endif | |
| 45 | |
| 46 extern int HtmlNamedEntity(unsigned char *p, int length); | |
| 47 | |
| 48 #define LS 0x2028 // UTF line separator | |
| 49 #define PS 0x2029 // UTF paragraph separator | |
| 50 | |
| 51 /******************************************** | |
| 52 * Do our own char maps | |
| 53 */ | |
| 54 | |
| 55 static unsigned char cmtable[256]; | |
| 56 | |
| 57 const int CMoctal = 0x1; | |
| 58 const int CMhex = 0x2; | |
| 59 const int CMidchar = 0x4; | |
| 60 | |
| 61 inline unsigned char isoctal (unsigned char c) { return cmtable[c] & CMoctal; } | |
| 62 inline unsigned char ishex (unsigned char c) { return cmtable[c] & CMhex; } | |
| 63 inline unsigned char isidchar(unsigned char c) { return cmtable[c] & CMidchar; } | |
| 64 | |
| 65 static void cmtable_init() | |
| 66 { | |
| 67 for (unsigned c = 0; c < sizeof(cmtable) / sizeof(cmtable[0]); c++) | |
| 68 { | |
| 69 if ('0' <= c && c <= '7') | |
| 70 cmtable[c] |= CMoctal; | |
| 71 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 72 cmtable[c] |= CMhex; | |
| 73 if (isalnum(c) || c == '_') | |
| 74 cmtable[c] |= CMidchar; | |
| 75 } | |
| 76 } | |
| 77 | |
| 78 | |
| 79 /************************* Token **********************************************/ | |
| 80 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
81 const char *Token::tochars[TOKMAX]; |
| 159 | 82 |
| 83 void *Token::operator new(size_t size) | |
| 84 { Token *t; | |
| 85 | |
| 86 if (Lexer::freelist) | |
| 87 { | |
| 88 t = Lexer::freelist; | |
| 89 Lexer::freelist = t->next; | |
| 90 return t; | |
| 91 } | |
| 92 | |
| 93 return ::operator new(size); | |
| 94 } | |
| 95 | |
| 96 #ifdef DEBUG | |
| 97 void Token::print() | |
| 98 { | |
| 99 fprintf(stdmsg, "%s\n", toChars()); | |
| 100 } | |
| 101 #endif | |
| 102 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
103 const char *Token::toChars() |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
104 { const char *p; |
| 159 | 105 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 106 | |
| 107 p = buffer; | |
| 108 switch (value) | |
| 109 { | |
| 110 case TOKint32v: | |
| 111 sprintf(buffer,"%d",(d_int32)int64value); | |
| 112 break; | |
| 113 | |
| 114 case TOKuns32v: | |
| 115 case TOKcharv: | |
| 116 case TOKwcharv: | |
| 117 case TOKdcharv: | |
| 118 sprintf(buffer,"%uU",(d_uns32)uns64value); | |
| 119 break; | |
| 120 | |
| 121 case TOKint64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
122 sprintf(buffer,"%jdL",int64value); |
| 159 | 123 break; |
| 124 | |
| 125 case TOKuns64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
126 sprintf(buffer,"%juUL",uns64value); |
| 159 | 127 break; |
| 128 | |
| 129 #if IN_GCC | |
| 130 case TOKfloat32v: | |
| 131 case TOKfloat64v: | |
| 132 case TOKfloat80v: | |
| 133 float80value.format(buffer, sizeof(buffer)); | |
| 134 break; | |
| 135 case TOKimaginary32v: | |
| 136 case TOKimaginary64v: | |
| 137 case TOKimaginary80v: | |
| 138 float80value.format(buffer, sizeof(buffer)); | |
| 139 // %% buffer | |
| 140 strcat(buffer, "i"); | |
| 141 break; | |
| 142 #else | |
| 143 case TOKfloat32v: | |
| 144 sprintf(buffer,"%Lgf", float80value); | |
| 145 break; | |
| 146 | |
| 147 case TOKfloat64v: | |
| 148 sprintf(buffer,"%Lg", float80value); | |
| 149 break; | |
| 150 | |
| 151 case TOKfloat80v: | |
| 152 sprintf(buffer,"%LgL", float80value); | |
| 153 break; | |
| 154 | |
| 155 case TOKimaginary32v: | |
| 156 sprintf(buffer,"%Lgfi", float80value); | |
| 157 break; | |
| 158 | |
| 159 case TOKimaginary64v: | |
| 160 sprintf(buffer,"%Lgi", float80value); | |
| 161 break; | |
| 162 | |
| 163 case TOKimaginary80v: | |
| 164 sprintf(buffer,"%LgLi", float80value); | |
| 165 break; | |
| 166 #endif | |
| 167 | |
| 168 case TOKstring: | |
| 169 #if CSTRINGS | |
| 170 p = string; | |
| 171 #else | |
| 172 { OutBuffer buf; | |
| 173 | |
| 174 buf.writeByte('"'); | |
| 175 for (size_t i = 0; i < len; ) | |
| 176 { unsigned c; | |
| 177 | |
| 178 utf_decodeChar((unsigned char *)ustring, len, &i, &c); | |
| 179 switch (c) | |
| 180 { | |
| 181 case 0: | |
| 182 break; | |
| 183 | |
| 184 case '"': | |
| 185 case '\\': | |
| 186 buf.writeByte('\\'); | |
| 187 default: | |
| 188 if (isprint(c)) | |
| 189 buf.writeByte(c); | |
| 190 else if (c <= 0x7F) | |
| 191 buf.printf("\\x%02x", c); | |
| 192 else if (c <= 0xFFFF) | |
| 193 buf.printf("\\u%04x", c); | |
| 194 else | |
| 195 buf.printf("\\U%08x", c); | |
| 196 continue; | |
| 197 } | |
| 198 break; | |
| 199 } | |
| 200 buf.writeByte('"'); | |
| 201 if (postfix) | |
| 202 buf.writeByte('"'); | |
| 203 buf.writeByte(0); | |
| 204 p = (char *)buf.extractData(); | |
| 205 } | |
| 206 #endif | |
| 207 break; | |
| 208 | |
| 209 case TOKidentifier: | |
| 210 case TOKenum: | |
| 211 case TOKstruct: | |
| 212 case TOKimport: | |
|
1603
eae495e6ae8d
Merge DMD r248: implement Denis Koroskin's macro suggestion
Leandro Lucarella <llucax@gmail.com>
parents:
1587
diff
changeset
|
213 case BASIC_TYPES: |
| 159 | 214 p = ident->toChars(); |
| 215 break; | |
| 216 | |
| 217 default: | |
| 218 p = toChars(value); | |
| 219 break; | |
| 220 } | |
| 221 return p; | |
| 222 } | |
| 223 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
224 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
|
225 { const char *p; |
| 159 | 226 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 227 | |
| 228 p = tochars[value]; | |
| 229 if (!p) | |
| 230 { sprintf(buffer,"TOK%d",value); | |
| 231 p = buffer; | |
| 232 } | |
| 233 return p; | |
| 234 } | |
| 235 | |
| 236 /*************************** Lexer ********************************************/ | |
| 237 | |
| 238 Token *Lexer::freelist = NULL; | |
| 239 StringTable Lexer::stringtable; | |
| 240 OutBuffer Lexer::stringbuffer; | |
| 241 | |
| 242 Lexer::Lexer(Module *mod, | |
| 243 unsigned char *base, unsigned begoffset, unsigned endoffset, | |
| 244 int doDocComment, int commentToken) | |
| 245 : loc(mod, 1) | |
| 246 { | |
| 247 //printf("Lexer::Lexer(%p,%d)\n",base,length); | |
| 248 //printf("lexer.mod = %p, %p\n", mod, this->loc.mod); | |
| 249 memset(&token,0,sizeof(token)); | |
| 250 this->base = base; | |
| 251 this->end = base + endoffset; | |
| 252 p = base + begoffset; | |
| 253 this->mod = mod; | |
| 254 this->doDocComment = doDocComment; | |
| 255 this->anyToken = 0; | |
| 256 this->commentToken = commentToken; | |
| 257 //initKeywords(); | |
| 258 | |
| 259 /* If first line starts with '#!', ignore the line | |
| 260 */ | |
| 261 | |
| 262 if (p[0] == '#' && p[1] =='!') | |
| 263 { | |
| 264 p += 2; | |
| 265 while (1) | |
| 266 { unsigned char c = *p; | |
| 267 switch (c) | |
| 268 { | |
| 269 case '\n': | |
| 270 p++; | |
| 271 break; | |
| 272 | |
| 273 case '\r': | |
| 274 p++; | |
| 275 if (*p == '\n') | |
| 276 p++; | |
| 277 break; | |
| 278 | |
| 279 case 0: | |
| 280 case 0x1A: | |
| 281 break; | |
| 282 | |
| 283 default: | |
| 284 if (c & 0x80) | |
| 285 { unsigned u = decodeUTF(); | |
| 286 if (u == PS || u == LS) | |
| 287 break; | |
| 288 } | |
| 289 p++; | |
| 290 continue; | |
| 291 } | |
| 292 break; | |
| 293 } | |
| 294 loc.linnum = 2; | |
| 295 } | |
| 296 } | |
| 297 | |
| 298 | |
| 299 void Lexer::error(const char *format, ...) | |
| 300 { | |
| 301 if (mod && !global.gag) | |
| 302 { | |
| 303 char *p = loc.toChars(); | |
| 304 if (*p) | |
| 305 fprintf(stdmsg, "%s: ", p); | |
| 306 mem.free(p); | |
| 307 | |
| 308 va_list ap; | |
| 309 va_start(ap, format); | |
| 310 vfprintf(stdmsg, format, ap); | |
| 311 va_end(ap); | |
| 312 | |
| 313 fprintf(stdmsg, "\n"); | |
| 314 fflush(stdmsg); | |
| 315 | |
| 316 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 317 fatal(); | |
| 318 } | |
| 319 global.errors++; | |
| 320 } | |
| 321 | |
| 322 void Lexer::error(Loc loc, const char *format, ...) | |
| 323 { | |
| 324 if (mod && !global.gag) | |
| 325 { | |
| 326 char *p = loc.toChars(); | |
| 327 if (*p) | |
| 328 fprintf(stdmsg, "%s: ", p); | |
| 329 mem.free(p); | |
| 330 | |
| 331 va_list ap; | |
| 332 va_start(ap, format); | |
| 333 vfprintf(stdmsg, format, ap); | |
| 334 va_end(ap); | |
| 335 | |
| 336 fprintf(stdmsg, "\n"); | |
| 337 fflush(stdmsg); | |
| 338 | |
| 339 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 340 fatal(); | |
| 341 } | |
| 342 global.errors++; | |
| 343 } | |
| 344 | |
| 345 TOK Lexer::nextToken() | |
| 346 { Token *t; | |
| 347 | |
| 348 if (token.next) | |
| 349 { | |
| 350 t = token.next; | |
| 351 memcpy(&token,t,sizeof(Token)); | |
| 352 t->next = freelist; | |
| 353 freelist = t; | |
| 354 } | |
| 355 else | |
| 356 { | |
| 357 scan(&token); | |
| 358 } | |
| 359 //token.print(); | |
| 360 return token.value; | |
| 361 } | |
| 362 | |
| 363 Token *Lexer::peek(Token *ct) | |
| 364 { Token *t; | |
| 365 | |
| 366 if (ct->next) | |
| 367 t = ct->next; | |
| 368 else | |
| 369 { | |
| 370 t = new Token(); | |
| 371 scan(t); | |
| 372 t->next = NULL; | |
| 373 ct->next = t; | |
| 374 } | |
| 375 return t; | |
| 376 } | |
| 377 | |
|
717
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
378 /*********************** |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
379 * Look ahead at next token's value. |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
380 */ |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
381 |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
382 TOK Lexer::peekNext() |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
383 { |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
384 return peek(&token)->value; |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
385 } |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
386 |
| 159 | 387 /********************************* |
| 388 * tk is on the opening (. | |
| 389 * Look ahead and return token that is past the closing ). | |
| 390 */ | |
| 391 | |
| 392 Token *Lexer::peekPastParen(Token *tk) | |
| 393 { | |
| 394 //printf("peekPastParen()\n"); | |
| 395 int parens = 1; | |
| 396 int curlynest = 0; | |
| 397 while (1) | |
| 398 { | |
| 399 tk = peek(tk); | |
| 400 //tk->print(); | |
| 401 switch (tk->value) | |
| 402 { | |
| 403 case TOKlparen: | |
| 404 parens++; | |
| 405 continue; | |
| 406 | |
| 407 case TOKrparen: | |
| 408 --parens; | |
| 409 if (parens) | |
| 410 continue; | |
| 411 tk = peek(tk); | |
| 412 break; | |
| 413 | |
| 414 case TOKlcurly: | |
| 415 curlynest++; | |
| 416 continue; | |
| 417 | |
| 418 case TOKrcurly: | |
| 419 if (--curlynest >= 0) | |
| 420 continue; | |
| 421 break; | |
| 422 | |
| 423 case TOKsemicolon: | |
| 424 if (curlynest) | |
| 425 continue; | |
| 426 break; | |
| 427 | |
| 428 case TOKeof: | |
| 429 break; | |
| 430 | |
| 431 default: | |
| 432 continue; | |
| 433 } | |
| 434 return tk; | |
| 435 } | |
| 436 } | |
| 437 | |
| 438 /********************************** | |
| 439 * Determine if string is a valid Identifier. | |
| 440 * Placed here because of commonality with Lexer functionality. | |
| 441 * Returns: | |
| 442 * 0 invalid | |
| 443 */ | |
| 444 | |
| 445 int Lexer::isValidIdentifier(char *p) | |
| 446 { | |
| 447 size_t len; | |
| 448 size_t idx; | |
| 449 | |
| 450 if (!p || !*p) | |
| 451 goto Linvalid; | |
| 452 | |
| 453 if (*p >= '0' && *p <= '9') // beware of isdigit() on signed chars | |
| 454 goto Linvalid; | |
| 455 | |
| 456 len = strlen(p); | |
| 457 idx = 0; | |
| 458 while (p[idx]) | |
| 459 { dchar_t dc; | |
| 460 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
461 const char *q = utf_decodeChar((unsigned char *)p, len, &idx, &dc); |
| 159 | 462 if (q) |
| 463 goto Linvalid; | |
| 464 | |
| 465 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) | |
| 466 goto Linvalid; | |
| 467 } | |
| 468 return 1; | |
| 469 | |
| 470 Linvalid: | |
| 471 return 0; | |
| 472 } | |
| 473 | |
| 474 /**************************** | |
| 475 * Turn next token in buffer into a token. | |
| 476 */ | |
| 477 | |
| 478 void Lexer::scan(Token *t) | |
| 479 { | |
| 480 unsigned lastLine = loc.linnum; | |
| 481 unsigned linnum; | |
| 482 | |
| 483 t->blockComment = NULL; | |
| 484 t->lineComment = NULL; | |
| 485 while (1) | |
| 486 { | |
| 487 t->ptr = p; | |
| 488 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 489 switch (*p) | |
| 490 { | |
| 491 case 0: | |
| 492 case 0x1A: | |
| 493 t->value = TOKeof; // end of file | |
| 494 return; | |
| 495 | |
| 496 case ' ': | |
| 497 case '\t': | |
| 498 case '\v': | |
| 499 case '\f': | |
| 500 p++; | |
| 501 continue; // skip white space | |
| 502 | |
| 503 case '\r': | |
| 504 p++; | |
| 505 if (*p != '\n') // if CR stands by itself | |
| 506 loc.linnum++; | |
| 507 continue; // skip white space | |
| 508 | |
| 509 case '\n': | |
| 510 p++; | |
| 511 loc.linnum++; | |
| 512 continue; // skip white space | |
| 513 | |
| 514 case '0': case '1': case '2': case '3': case '4': | |
| 515 case '5': case '6': case '7': case '8': case '9': | |
| 516 t->value = number(t); | |
| 517 return; | |
| 518 | |
| 519 #if CSTRINGS | |
| 520 case '\'': | |
| 521 t->value = charConstant(t, 0); | |
| 522 return; | |
| 523 | |
| 524 case '"': | |
| 525 t->value = stringConstant(t,0); | |
| 526 return; | |
| 527 | |
| 528 case 'l': | |
| 529 case 'L': | |
| 530 if (p[1] == '\'') | |
| 531 { | |
| 532 p++; | |
| 533 t->value = charConstant(t, 1); | |
| 534 return; | |
| 535 } | |
| 536 else if (p[1] == '"') | |
| 537 { | |
| 538 p++; | |
| 539 t->value = stringConstant(t, 1); | |
| 540 return; | |
| 541 } | |
| 542 #else | |
| 543 case '\'': | |
| 544 t->value = charConstant(t,0); | |
| 545 return; | |
| 546 | |
| 547 case 'r': | |
| 548 if (p[1] != '"') | |
| 549 goto case_ident; | |
| 550 p++; | |
| 551 case '`': | |
| 552 t->value = wysiwygStringConstant(t, *p); | |
| 553 return; | |
| 554 | |
| 555 case 'x': | |
| 556 if (p[1] != '"') | |
| 557 goto case_ident; | |
| 558 p++; | |
| 559 t->value = hexStringConstant(t); | |
| 560 return; | |
| 561 | |
| 336 | 562 #if DMDV2 |
| 159 | 563 case 'q': |
| 564 if (p[1] == '"') | |
| 565 { | |
| 566 p++; | |
| 567 t->value = delimitedStringConstant(t); | |
| 568 return; | |
| 569 } | |
| 570 else if (p[1] == '{') | |
| 571 { | |
| 572 p++; | |
| 573 t->value = tokenStringConstant(t); | |
| 574 return; | |
| 575 } | |
| 576 else | |
| 577 goto case_ident; | |
| 578 #endif | |
| 579 | |
| 580 case '"': | |
| 581 t->value = escapeStringConstant(t,0); | |
| 582 return; | |
| 583 | |
| 584 case '\\': // escaped string literal | |
| 585 { unsigned c; | |
| 586 | |
| 587 stringbuffer.reset(); | |
| 588 do | |
| 589 { | |
| 590 p++; | |
| 591 switch (*p) | |
| 592 { | |
| 593 case 'u': | |
| 594 case 'U': | |
| 595 case '&': | |
| 596 c = escapeSequence(); | |
| 597 stringbuffer.writeUTF8(c); | |
| 598 break; | |
| 599 | |
| 600 default: | |
| 601 c = escapeSequence(); | |
| 602 stringbuffer.writeByte(c); | |
| 603 break; | |
| 604 } | |
| 605 } while (*p == '\\'); | |
| 606 t->len = stringbuffer.offset; | |
| 607 stringbuffer.writeByte(0); | |
| 608 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 609 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 610 t->postfix = 0; | |
| 611 t->value = TOKstring; | |
| 612 return; | |
| 613 } | |
| 614 | |
| 615 case 'l': | |
| 616 case 'L': | |
| 617 #endif | |
| 618 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 619 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 620 case 'k': case 'm': case 'n': case 'o': | |
| 336 | 621 #if DMDV2 |
| 159 | 622 case 'p': /*case 'q': case 'r':*/ case 's': case 't': |
| 623 #else | |
| 624 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 625 #endif | |
| 626 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 627 case 'z': | |
| 628 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 629 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 630 case 'K': case 'M': case 'N': case 'O': | |
| 631 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 632 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 633 case 'Z': | |
| 634 case '_': | |
| 635 case_ident: | |
| 636 { unsigned char c; | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
637 |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
638 while (1) |
| 159 | 639 { |
| 640 c = *++p; | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
641 if (isidchar(c)) |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
642 continue; |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
643 else if (c & 0x80) |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
644 { unsigned char *s = p; |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
645 unsigned u = decodeUTF(); |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
646 if (isUniAlpha(u)) |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
647 continue; |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
648 error("char 0x%04x not allowed in identifier", u); |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
649 p = s; |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
650 } |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
651 break; |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
652 } |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
653 |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
654 StringValue *sv = stringtable.update((char *)t->ptr, p - t->ptr); |
|
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
655 Identifier *id = (Identifier *) sv->ptrvalue; |
| 159 | 656 if (!id) |
| 657 { id = new Identifier(sv->lstring.string,TOKidentifier); | |
| 658 sv->ptrvalue = id; | |
| 659 } | |
| 660 t->ident = id; | |
| 661 t->value = (enum TOK) id->value; | |
| 662 anyToken = 1; | |
| 663 if (*t->ptr == '_') // if special identifier token | |
| 664 { | |
| 665 static char date[11+1]; | |
| 666 static char time[8+1]; | |
| 667 static char timestamp[24+1]; | |
| 668 | |
| 669 if (!date[0]) // lazy evaluation | |
| 670 { time_t t; | |
| 671 char *p; | |
| 672 | |
| 673 ::time(&t); | |
| 674 p = ctime(&t); | |
| 675 assert(p); | |
| 676 sprintf(date, "%.6s %.4s", p + 4, p + 20); | |
| 677 sprintf(time, "%.8s", p + 11); | |
| 678 sprintf(timestamp, "%.24s", p); | |
| 679 } | |
| 680 | |
| 336 | 681 #if DMDV1 |
| 159 | 682 if (mod && id == Id::FILE) |
| 683 { | |
| 684 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
|
685 goto Lstr; |
| 159 | 686 } |
| 687 else if (mod && id == Id::LINE) | |
| 688 { | |
| 689 t->value = TOKint64v; | |
| 690 t->uns64value = loc.linnum; | |
| 691 } | |
| 336 | 692 else |
| 693 #endif | |
| 694 if (id == Id::DATE) | |
| 159 | 695 { |
| 696 t->ustring = (unsigned char *)date; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
697 goto Lstr; |
| 159 | 698 } |
| 699 else if (id == Id::TIME) | |
| 700 { | |
| 701 t->ustring = (unsigned char *)time; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
702 goto Lstr; |
| 159 | 703 } |
| 704 else if (id == Id::VENDOR) | |
| 705 { | |
|
664
eef8ac26c66c
Some missed LLVMDC -> LDC.
Christian Kamm <kamm incasoftware de>
parents:
658
diff
changeset
|
706 t->ustring = (unsigned char *)"LDC"; |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
707 goto Lstr; |
| 159 | 708 } |
| 709 else if (id == Id::TIMESTAMP) | |
| 710 { | |
| 711 t->ustring = (unsigned char *)timestamp; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
712 Lstr: |
| 159 | 713 t->value = TOKstring; |
| 714 Llen: | |
| 715 t->postfix = 0; | |
| 716 t->len = strlen((char *)t->ustring); | |
| 717 } | |
| 718 else if (id == Id::VERSIONX) | |
| 719 { unsigned major = 0; | |
| 720 unsigned minor = 0; | |
| 721 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
722 for (const char *p = global.version + 1; 1; p++) |
| 159 | 723 { |
| 724 char c = *p; | |
| 725 if (isdigit(c)) | |
| 726 minor = minor * 10 + c - '0'; | |
| 727 else if (c == '.') | |
| 728 { major = minor; | |
| 729 minor = 0; | |
| 730 } | |
| 731 else | |
| 732 break; | |
| 733 } | |
| 734 t->value = TOKint64v; | |
| 735 t->uns64value = major * 1000 + minor; | |
| 736 } | |
| 336 | 737 #if DMDV2 |
| 159 | 738 else if (id == Id::EOFX) |
| 739 { | |
| 740 t->value = TOKeof; | |
| 741 // Advance scanner to end of file | |
| 742 while (!(*p == 0 || *p == 0x1A)) | |
| 743 p++; | |
| 744 } | |
| 745 #endif | |
| 746 } | |
| 747 //printf("t->value = %d\n",t->value); | |
| 748 return; | |
| 749 } | |
| 750 | |
| 751 case '/': | |
| 752 p++; | |
| 753 switch (*p) | |
| 754 { | |
| 755 case '=': | |
| 756 p++; | |
| 757 t->value = TOKdivass; | |
| 758 return; | |
| 759 | |
| 760 case '*': | |
| 761 p++; | |
| 762 linnum = loc.linnum; | |
| 763 while (1) | |
| 764 { | |
| 765 while (1) | |
| 766 { unsigned char c = *p; | |
| 767 switch (c) | |
| 768 { | |
| 769 case '/': | |
| 770 break; | |
| 771 | |
| 772 case '\n': | |
| 773 loc.linnum++; | |
| 774 p++; | |
| 775 continue; | |
| 776 | |
| 777 case '\r': | |
| 778 p++; | |
| 779 if (*p != '\n') | |
| 780 loc.linnum++; | |
| 781 continue; | |
| 782 | |
| 783 case 0: | |
| 784 case 0x1A: | |
| 785 error("unterminated /* */ comment"); | |
| 786 p = end; | |
| 787 t->value = TOKeof; | |
| 788 return; | |
| 789 | |
| 790 default: | |
| 791 if (c & 0x80) | |
| 792 { unsigned u = decodeUTF(); | |
| 793 if (u == PS || u == LS) | |
| 794 loc.linnum++; | |
| 795 } | |
| 796 p++; | |
| 797 continue; | |
| 798 } | |
| 799 break; | |
| 800 } | |
| 801 p++; | |
| 802 if (p[-2] == '*' && p - 3 != t->ptr) | |
| 803 break; | |
| 804 } | |
| 805 if (commentToken) | |
| 806 { | |
| 807 t->value = TOKcomment; | |
| 808 return; | |
| 809 } | |
| 810 else if (doDocComment && t->ptr[2] == '*' && p - 4 != t->ptr) | |
| 811 { // if /** but not /**/ | |
| 812 getDocComment(t, lastLine == linnum); | |
| 813 } | |
| 814 continue; | |
| 815 | |
| 816 case '/': // do // style comments | |
| 817 linnum = loc.linnum; | |
| 818 while (1) | |
| 819 { unsigned char c = *++p; | |
| 820 switch (c) | |
| 821 { | |
| 822 case '\n': | |
| 823 break; | |
| 824 | |
| 825 case '\r': | |
| 826 if (p[1] == '\n') | |
| 827 p++; | |
| 828 break; | |
| 829 | |
| 830 case 0: | |
| 831 case 0x1A: | |
| 832 if (commentToken) | |
| 833 { | |
| 834 p = end; | |
| 835 t->value = TOKcomment; | |
| 836 return; | |
| 837 } | |
| 838 if (doDocComment && t->ptr[2] == '/') | |
| 839 getDocComment(t, lastLine == linnum); | |
| 840 p = end; | |
| 841 t->value = TOKeof; | |
| 842 return; | |
| 843 | |
| 844 default: | |
| 845 if (c & 0x80) | |
| 846 { unsigned u = decodeUTF(); | |
| 847 if (u == PS || u == LS) | |
| 848 break; | |
| 849 } | |
| 850 continue; | |
| 851 } | |
| 852 break; | |
| 853 } | |
| 854 | |
| 855 if (commentToken) | |
| 856 { | |
| 857 p++; | |
| 858 loc.linnum++; | |
| 859 t->value = TOKcomment; | |
| 860 return; | |
| 861 } | |
| 862 if (doDocComment && t->ptr[2] == '/') | |
| 863 getDocComment(t, lastLine == linnum); | |
| 864 | |
| 865 p++; | |
| 866 loc.linnum++; | |
| 867 continue; | |
| 868 | |
| 869 case '+': | |
| 870 { int nest; | |
| 871 | |
| 872 linnum = loc.linnum; | |
| 873 p++; | |
| 874 nest = 1; | |
| 875 while (1) | |
| 876 { unsigned char c = *p; | |
| 877 switch (c) | |
| 878 { | |
| 879 case '/': | |
| 880 p++; | |
| 881 if (*p == '+') | |
| 882 { | |
| 883 p++; | |
| 884 nest++; | |
| 885 } | |
| 886 continue; | |
| 887 | |
| 888 case '+': | |
| 889 p++; | |
| 890 if (*p == '/') | |
| 891 { | |
| 892 p++; | |
| 893 if (--nest == 0) | |
| 894 break; | |
| 895 } | |
| 896 continue; | |
| 897 | |
| 898 case '\r': | |
| 899 p++; | |
| 900 if (*p != '\n') | |
| 901 loc.linnum++; | |
| 902 continue; | |
| 903 | |
| 904 case '\n': | |
| 905 loc.linnum++; | |
| 906 p++; | |
| 907 continue; | |
| 908 | |
| 909 case 0: | |
| 910 case 0x1A: | |
| 911 error("unterminated /+ +/ comment"); | |
| 912 p = end; | |
| 913 t->value = TOKeof; | |
| 914 return; | |
| 915 | |
| 916 default: | |
| 917 if (c & 0x80) | |
| 918 { unsigned u = decodeUTF(); | |
| 919 if (u == PS || u == LS) | |
| 920 loc.linnum++; | |
| 921 } | |
| 922 p++; | |
| 923 continue; | |
| 924 } | |
| 925 break; | |
| 926 } | |
| 927 if (commentToken) | |
| 928 { | |
| 929 t->value = TOKcomment; | |
| 930 return; | |
| 931 } | |
| 932 if (doDocComment && t->ptr[2] == '+' && p - 4 != t->ptr) | |
| 933 { // if /++ but not /++/ | |
| 934 getDocComment(t, lastLine == linnum); | |
| 935 } | |
| 936 continue; | |
| 937 } | |
| 938 } | |
| 939 t->value = TOKdiv; | |
| 940 return; | |
| 941 | |
| 942 case '.': | |
| 943 p++; | |
| 944 if (isdigit(*p)) | |
| 945 { /* Note that we don't allow ._1 and ._ as being | |
| 946 * valid floating point numbers. | |
| 947 */ | |
| 948 p--; | |
| 949 t->value = inreal(t); | |
| 950 } | |
| 951 else if (p[0] == '.') | |
| 952 { | |
| 953 if (p[1] == '.') | |
| 954 { p += 2; | |
| 955 t->value = TOKdotdotdot; | |
| 956 } | |
| 957 else | |
| 958 { p++; | |
| 959 t->value = TOKslice; | |
| 960 } | |
| 961 } | |
| 962 else | |
| 963 t->value = TOKdot; | |
| 964 return; | |
| 965 | |
| 966 case '&': | |
| 967 p++; | |
| 968 if (*p == '=') | |
| 969 { p++; | |
| 970 t->value = TOKandass; | |
| 971 } | |
| 972 else if (*p == '&') | |
| 973 { p++; | |
| 974 t->value = TOKandand; | |
| 975 } | |
| 976 else | |
| 977 t->value = TOKand; | |
| 978 return; | |
| 979 | |
| 980 case '|': | |
| 981 p++; | |
| 982 if (*p == '=') | |
| 983 { p++; | |
| 984 t->value = TOKorass; | |
| 985 } | |
| 986 else if (*p == '|') | |
| 987 { p++; | |
| 988 t->value = TOKoror; | |
| 989 } | |
| 990 else | |
| 991 t->value = TOKor; | |
| 992 return; | |
| 993 | |
| 994 case '-': | |
| 995 p++; | |
| 996 if (*p == '=') | |
| 997 { p++; | |
| 998 t->value = TOKminass; | |
| 999 } | |
| 1000 #if 0 | |
| 1001 else if (*p == '>') | |
| 1002 { p++; | |
| 1003 t->value = TOKarrow; | |
| 1004 } | |
| 1005 #endif | |
| 1006 else if (*p == '-') | |
| 1007 { p++; | |
| 1008 t->value = TOKminusminus; | |
| 1009 } | |
| 1010 else | |
| 1011 t->value = TOKmin; | |
| 1012 return; | |
| 1013 | |
| 1014 case '+': | |
| 1015 p++; | |
| 1016 if (*p == '=') | |
| 1017 { p++; | |
| 1018 t->value = TOKaddass; | |
| 1019 } | |
| 1020 else if (*p == '+') | |
| 1021 { p++; | |
| 1022 t->value = TOKplusplus; | |
| 1023 } | |
| 1024 else | |
| 1025 t->value = TOKadd; | |
| 1026 return; | |
| 1027 | |
| 1028 case '<': | |
| 1029 p++; | |
| 1030 if (*p == '=') | |
| 1031 { p++; | |
| 1032 t->value = TOKle; // <= | |
| 1033 } | |
| 1034 else if (*p == '<') | |
| 1035 { p++; | |
| 1036 if (*p == '=') | |
| 1037 { p++; | |
| 1038 t->value = TOKshlass; // <<= | |
| 1039 } | |
| 1040 else | |
| 1041 t->value = TOKshl; // << | |
| 1042 } | |
| 1043 else if (*p == '>') | |
| 1044 { p++; | |
| 1045 if (*p == '=') | |
| 1046 { p++; | |
| 1047 t->value = TOKleg; // <>= | |
| 1048 } | |
| 1049 else | |
| 1050 t->value = TOKlg; // <> | |
| 1051 } | |
| 1052 else | |
| 1053 t->value = TOKlt; // < | |
| 1054 return; | |
| 1055 | |
| 1056 case '>': | |
| 1057 p++; | |
| 1058 if (*p == '=') | |
| 1059 { p++; | |
| 1060 t->value = TOKge; // >= | |
| 1061 } | |
| 1062 else if (*p == '>') | |
| 1063 { p++; | |
| 1064 if (*p == '=') | |
| 1065 { p++; | |
| 1066 t->value = TOKshrass; // >>= | |
| 1067 } | |
| 1068 else if (*p == '>') | |
| 1069 { p++; | |
| 1070 if (*p == '=') | |
| 1071 { p++; | |
| 1072 t->value = TOKushrass; // >>>= | |
| 1073 } | |
| 1074 else | |
| 1075 t->value = TOKushr; // >>> | |
| 1076 } | |
| 1077 else | |
| 1078 t->value = TOKshr; // >> | |
| 1079 } | |
| 1080 else | |
| 1081 t->value = TOKgt; // > | |
| 1082 return; | |
| 1083 | |
| 1084 case '!': | |
| 1085 p++; | |
| 1086 if (*p == '=') | |
| 1087 { p++; | |
| 1088 if (*p == '=' && global.params.Dversion == 1) | |
| 1089 { p++; | |
| 1090 t->value = TOKnotidentity; // !== | |
| 1091 } | |
| 1092 else | |
| 1093 t->value = TOKnotequal; // != | |
| 1094 } | |
| 1095 else if (*p == '<') | |
| 1096 { p++; | |
| 1097 if (*p == '>') | |
| 1098 { p++; | |
| 1099 if (*p == '=') | |
| 1100 { p++; | |
| 1101 t->value = TOKunord; // !<>= | |
| 1102 } | |
| 1103 else | |
| 1104 t->value = TOKue; // !<> | |
| 1105 } | |
| 1106 else if (*p == '=') | |
| 1107 { p++; | |
| 1108 t->value = TOKug; // !<= | |
| 1109 } | |
| 1110 else | |
| 1111 t->value = TOKuge; // !< | |
| 1112 } | |
| 1113 else if (*p == '>') | |
| 1114 { p++; | |
| 1115 if (*p == '=') | |
| 1116 { p++; | |
| 1117 t->value = TOKul; // !>= | |
| 1118 } | |
| 1119 else | |
| 1120 t->value = TOKule; // !> | |
| 1121 } | |
| 1122 else | |
| 1123 t->value = TOKnot; // ! | |
| 1124 return; | |
| 1125 | |
| 1126 case '=': | |
| 1127 p++; | |
| 1128 if (*p == '=') | |
| 1129 { p++; | |
| 1130 if (*p == '=' && global.params.Dversion == 1) | |
| 1131 { p++; | |
| 1132 t->value = TOKidentity; // === | |
| 1133 } | |
| 1134 else | |
| 1135 t->value = TOKequal; // == | |
| 1136 } | |
| 1137 else | |
| 1138 t->value = TOKassign; // = | |
| 1139 return; | |
| 1140 | |
| 1141 case '~': | |
| 1142 p++; | |
| 1143 if (*p == '=') | |
| 1144 { p++; | |
| 1145 t->value = TOKcatass; // ~= | |
| 1146 } | |
| 1147 else | |
| 1148 t->value = TOKtilde; // ~ | |
| 1149 return; | |
| 1150 | |
| 1151 #define SINGLE(c,tok) case c: p++; t->value = tok; return; | |
| 1152 | |
| 1153 SINGLE('(', TOKlparen) | |
| 1154 SINGLE(')', TOKrparen) | |
| 1155 SINGLE('[', TOKlbracket) | |
| 1156 SINGLE(']', TOKrbracket) | |
| 1157 SINGLE('{', TOKlcurly) | |
| 1158 SINGLE('}', TOKrcurly) | |
| 1159 SINGLE('?', TOKquestion) | |
| 1160 SINGLE(',', TOKcomma) | |
| 1161 SINGLE(';', TOKsemicolon) | |
| 1162 SINGLE(':', TOKcolon) | |
| 1163 SINGLE('$', TOKdollar) | |
| 1164 | |
| 1165 #undef SINGLE | |
| 1166 | |
| 1167 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1168 case c1: \ | |
| 1169 p++; \ | |
| 1170 if (*p == c2) \ | |
| 1171 { p++; \ | |
| 1172 t->value = tok2; \ | |
| 1173 } \ | |
| 1174 else \ | |
| 1175 t->value = tok1; \ | |
| 1176 return; | |
| 1177 | |
| 1178 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1179 DOUBLE('%', TOKmod, '=', TOKmodass) | |
| 1180 DOUBLE('^', TOKxor, '=', TOKxorass) | |
| 1181 | |
| 1182 #undef DOUBLE | |
| 1183 | |
| 1184 case '#': | |
| 1185 p++; | |
| 1186 pragma(); | |
| 1187 continue; | |
| 1188 | |
| 1189 default: | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1190 { unsigned c = *p; |
| 159 | 1191 |
| 1192 if (c & 0x80) | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1193 { c = decodeUTF(); |
| 159 | 1194 |
| 1195 // Check for start of unicode identifier | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1196 if (isUniAlpha(c)) |
| 159 | 1197 goto case_ident; |
| 1198 | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1199 if (c == PS || c == LS) |
| 159 | 1200 { |
| 1201 loc.linnum++; | |
| 1202 p++; | |
| 1203 continue; | |
| 1204 } | |
| 1205 } | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1206 if (c < 0x80 && isprint(c)) |
| 159 | 1207 error("unsupported char '%c'", c); |
| 1208 else | |
| 1209 error("unsupported char 0x%02x", c); | |
| 1210 p++; | |
| 1211 continue; | |
| 1212 } | |
| 1213 } | |
| 1214 } | |
| 1215 } | |
| 1216 | |
| 1217 /******************************************* | |
| 1218 * Parse escape sequence. | |
| 1219 */ | |
| 1220 | |
| 1221 unsigned Lexer::escapeSequence() | |
| 1222 { unsigned c; | |
| 1223 int n; | |
| 1224 int ndigits; | |
| 1225 | |
| 1226 c = *p; | |
| 1227 switch (c) | |
| 1228 { | |
| 1229 case '\'': | |
| 1230 case '"': | |
| 1231 case '?': | |
| 1232 case '\\': | |
| 1233 Lconsume: | |
| 1234 p++; | |
| 1235 break; | |
| 1236 | |
| 1237 case 'a': c = 7; goto Lconsume; | |
| 1238 case 'b': c = 8; goto Lconsume; | |
| 1239 case 'f': c = 12; goto Lconsume; | |
| 1240 case 'n': c = 10; goto Lconsume; | |
| 1241 case 'r': c = 13; goto Lconsume; | |
| 1242 case 't': c = 9; goto Lconsume; | |
| 1243 case 'v': c = 11; goto Lconsume; | |
| 1244 | |
| 1245 case 'u': | |
| 1246 ndigits = 4; | |
| 1247 goto Lhex; | |
| 1248 case 'U': | |
| 1249 ndigits = 8; | |
| 1250 goto Lhex; | |
| 1251 case 'x': | |
| 1252 ndigits = 2; | |
| 1253 Lhex: | |
| 1254 p++; | |
| 1255 c = *p; | |
| 1256 if (ishex(c)) | |
| 1257 { unsigned v; | |
| 1258 | |
| 1259 n = 0; | |
| 1260 v = 0; | |
| 1261 while (1) | |
| 1262 { | |
| 1263 if (isdigit(c)) | |
| 1264 c -= '0'; | |
| 1265 else if (islower(c)) | |
| 1266 c -= 'a' - 10; | |
| 1267 else | |
| 1268 c -= 'A' - 10; | |
| 1269 v = v * 16 + c; | |
| 1270 c = *++p; | |
| 1271 if (++n == ndigits) | |
| 1272 break; | |
| 1273 if (!ishex(c)) | |
| 1274 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1275 break; | |
| 1276 } | |
| 1277 } | |
| 1278 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1587 | 1279 { error("invalid UTF character \\U%08x", v); |
| 1280 v = '?'; // recover with valid UTF character | |
| 1281 } | |
| 159 | 1282 c = v; |
| 1283 } | |
| 1284 else | |
| 1285 error("undefined escape hex sequence \\%c\n",c); | |
| 1286 break; | |
| 1287 | |
| 1288 case '&': // named character entity | |
| 1289 for (unsigned char *idstart = ++p; 1; p++) | |
| 1290 { | |
| 1291 switch (*p) | |
| 1292 { | |
| 1293 case ';': | |
| 1294 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1295 if (c == ~0) | |
| 1296 { error("unnamed character entity &%.*s;", (int)(p - idstart), idstart); | |
| 1297 c = ' '; | |
| 1298 } | |
| 1299 p++; | |
| 1300 break; | |
| 1301 | |
| 1302 default: | |
| 1303 if (isalpha(*p) || | |
| 1304 (p != idstart + 1 && isdigit(*p))) | |
| 1305 continue; | |
| 1306 error("unterminated named entity"); | |
| 1307 break; | |
| 1308 } | |
| 1309 break; | |
| 1310 } | |
| 1311 break; | |
| 1312 | |
| 1313 case 0: | |
| 1314 case 0x1A: // end of file | |
| 1315 c = '\\'; | |
| 1316 break; | |
| 1317 | |
| 1318 default: | |
| 1319 if (isoctal(c)) | |
| 1320 { unsigned v; | |
| 1321 | |
| 1322 n = 0; | |
| 1323 v = 0; | |
| 1324 do | |
| 1325 { | |
| 1326 v = v * 8 + (c - '0'); | |
| 1327 c = *++p; | |
| 1328 } while (++n < 3 && isoctal(c)); | |
| 1329 c = v; | |
| 1330 if (c > 0xFF) | |
| 1331 error("0%03o is larger than a byte", c); | |
| 1332 } | |
| 1333 else | |
| 1334 error("undefined escape sequence \\%c\n",c); | |
| 1335 break; | |
| 1336 } | |
| 1337 return c; | |
| 1338 } | |
| 1339 | |
| 1340 /************************************** | |
| 1341 */ | |
| 1342 | |
| 1343 TOK Lexer::wysiwygStringConstant(Token *t, int tc) | |
| 1344 { unsigned c; | |
| 1345 Loc start = loc; | |
| 1346 | |
| 1347 p++; | |
| 1348 stringbuffer.reset(); | |
| 1349 while (1) | |
| 1350 { | |
| 1351 c = *p++; | |
| 1352 switch (c) | |
| 1353 { | |
| 1354 case '\n': | |
| 1355 loc.linnum++; | |
| 1356 break; | |
| 1357 | |
| 1358 case '\r': | |
| 1359 if (*p == '\n') | |
| 1360 continue; // ignore | |
| 1361 c = '\n'; // treat EndOfLine as \n character | |
| 1362 loc.linnum++; | |
| 1363 break; | |
| 1364 | |
| 1365 case 0: | |
| 1366 case 0x1A: | |
| 1367 error("unterminated string constant starting at %s", start.toChars()); | |
| 1368 t->ustring = (unsigned char *)""; | |
| 1369 t->len = 0; | |
| 1370 t->postfix = 0; | |
| 1371 return TOKstring; | |
| 1372 | |
| 1373 case '"': | |
| 1374 case '`': | |
| 1375 if (c == tc) | |
| 1376 { | |
| 1377 t->len = stringbuffer.offset; | |
| 1378 stringbuffer.writeByte(0); | |
| 1379 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1380 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1381 stringPostfix(t); | |
| 1382 return TOKstring; | |
| 1383 } | |
| 1384 break; | |
| 1385 | |
| 1386 default: | |
| 1387 if (c & 0x80) | |
| 1388 { p--; | |
| 1389 unsigned u = decodeUTF(); | |
| 1390 p++; | |
| 1391 if (u == PS || u == LS) | |
| 1392 loc.linnum++; | |
| 1393 stringbuffer.writeUTF8(u); | |
| 1394 continue; | |
| 1395 } | |
| 1396 break; | |
| 1397 } | |
| 1398 stringbuffer.writeByte(c); | |
| 1399 } | |
| 1400 } | |
| 1401 | |
| 1402 /************************************** | |
| 1403 * Lex hex strings: | |
| 1404 * x"0A ae 34FE BD" | |
| 1405 */ | |
| 1406 | |
| 1407 TOK Lexer::hexStringConstant(Token *t) | |
| 1408 { unsigned c; | |
| 1409 Loc start = loc; | |
| 1410 unsigned n = 0; | |
| 1411 unsigned v; | |
| 1412 | |
| 1413 p++; | |
| 1414 stringbuffer.reset(); | |
| 1415 while (1) | |
| 1416 { | |
| 1417 c = *p++; | |
| 1418 switch (c) | |
| 1419 { | |
| 1420 case ' ': | |
| 1421 case '\t': | |
| 1422 case '\v': | |
| 1423 case '\f': | |
| 1424 continue; // skip white space | |
| 1425 | |
| 1426 case '\r': | |
| 1427 if (*p == '\n') | |
| 1428 continue; // ignore | |
| 1429 // Treat isolated '\r' as if it were a '\n' | |
| 1430 case '\n': | |
| 1431 loc.linnum++; | |
| 1432 continue; | |
| 1433 | |
| 1434 case 0: | |
| 1435 case 0x1A: | |
| 1436 error("unterminated string constant starting at %s", start.toChars()); | |
| 1437 t->ustring = (unsigned char *)""; | |
| 1438 t->len = 0; | |
| 1439 t->postfix = 0; | |
| 1440 return TOKstring; | |
| 1441 | |
| 1442 case '"': | |
| 1443 if (n & 1) | |
| 1444 { error("odd number (%d) of hex characters in hex string", n); | |
| 1445 stringbuffer.writeByte(v); | |
| 1446 } | |
| 1447 t->len = stringbuffer.offset; | |
| 1448 stringbuffer.writeByte(0); | |
| 1449 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1450 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1451 stringPostfix(t); | |
| 1452 return TOKstring; | |
| 1453 | |
| 1454 default: | |
| 1455 if (c >= '0' && c <= '9') | |
| 1456 c -= '0'; | |
| 1457 else if (c >= 'a' && c <= 'f') | |
| 1458 c -= 'a' - 10; | |
| 1459 else if (c >= 'A' && c <= 'F') | |
| 1460 c -= 'A' - 10; | |
| 1461 else if (c & 0x80) | |
| 1462 { p--; | |
| 1463 unsigned u = decodeUTF(); | |
| 1464 p++; | |
| 1465 if (u == PS || u == LS) | |
| 1466 loc.linnum++; | |
| 1467 else | |
|
1614
dbf7b54f542f
Merge DMD r292: bugzilla 3455 Some Unicode characters not allowed...
Leandro Lucarella <llucax@gmail.com>
parents:
1603
diff
changeset
|
1468 error("non-hex character \\u%04x", u); |
| 159 | 1469 } |
| 1470 else | |
| 1471 error("non-hex character '%c'", c); | |
| 1472 if (n & 1) | |
| 1473 { v = (v << 4) | c; | |
| 1474 stringbuffer.writeByte(v); | |
| 1475 } | |
| 1476 else | |
| 1477 v = c; | |
| 1478 n++; | |
| 1479 break; | |
| 1480 } | |
| 1481 } | |
| 1482 } | |
| 1483 | |
| 1484 | |
| 336 | 1485 #if DMDV2 |
| 159 | 1486 /************************************** |
| 1487 * Lex delimited strings: | |
| 1488 * q"(foo(xxx))" // "foo(xxx)" | |
| 1489 * q"[foo(]" // "foo(" | |
| 1490 * q"/foo]/" // "foo]" | |
| 1491 * q"HERE | |
| 1492 * foo | |
| 1493 * HERE" // "foo\n" | |
| 1494 * Input: | |
| 1495 * p is on the " | |
| 1496 */ | |
| 1497 | |
| 1498 TOK Lexer::delimitedStringConstant(Token *t) | |
| 1499 { unsigned c; | |
| 1500 Loc start = loc; | |
| 1501 unsigned delimleft = 0; | |
| 1502 unsigned delimright = 0; | |
| 1503 unsigned nest = 1; | |
| 1504 unsigned nestcount; | |
| 1505 Identifier *hereid = NULL; | |
| 1506 unsigned blankrol = 0; | |
| 1507 unsigned startline = 0; | |
| 1508 | |
| 1509 p++; | |
| 1510 stringbuffer.reset(); | |
| 1511 while (1) | |
| 1512 { | |
| 1513 c = *p++; | |
| 1514 //printf("c = '%c'\n", c); | |
| 1515 switch (c) | |
| 1516 { | |
| 1517 case '\n': | |
| 1518 Lnextline: | |
| 1519 loc.linnum++; | |
| 1520 startline = 1; | |
| 1521 if (blankrol) | |
| 1522 { blankrol = 0; | |
| 1523 continue; | |
| 1524 } | |
| 1525 if (hereid) | |
| 1526 { | |
| 1527 stringbuffer.writeUTF8(c); | |
| 1528 continue; | |
| 1529 } | |
| 1530 break; | |
| 1531 | |
| 1532 case '\r': | |
| 1533 if (*p == '\n') | |
| 1534 continue; // ignore | |
| 1535 c = '\n'; // treat EndOfLine as \n character | |
| 1536 goto Lnextline; | |
| 1537 | |
| 1538 case 0: | |
| 1539 case 0x1A: | |
| 1540 goto Lerror; | |
| 1541 | |
| 1542 default: | |
| 1543 if (c & 0x80) | |
| 1544 { p--; | |
| 1545 c = decodeUTF(); | |
| 1546 p++; | |
| 1547 if (c == PS || c == LS) | |
| 1548 goto Lnextline; | |
| 1549 } | |
| 1550 break; | |
| 1551 } | |
| 1552 if (delimleft == 0) | |
| 1553 { delimleft = c; | |
| 1554 nest = 1; | |
| 1555 nestcount = 1; | |
| 1556 if (c == '(') | |
| 1557 delimright = ')'; | |
| 1558 else if (c == '{') | |
| 1559 delimright = '}'; | |
| 1560 else if (c == '[') | |
| 1561 delimright = ']'; | |
| 1562 else if (c == '<') | |
| 1563 delimright = '>'; | |
| 1564 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1565 { // Start of identifier; must be a heredoc | |
| 1566 Token t; | |
| 1567 p--; | |
| 1568 scan(&t); // read in heredoc identifier | |
| 1569 if (t.value != TOKidentifier) | |
| 1570 { error("identifier expected for heredoc, not %s", t.toChars()); | |
| 1571 delimright = c; | |
| 1572 } | |
| 1573 else | |
| 1574 { hereid = t.ident; | |
| 1575 //printf("hereid = '%s'\n", hereid->toChars()); | |
| 1576 blankrol = 1; | |
| 1577 } | |
| 1578 nest = 0; | |
| 1579 } | |
| 1580 else | |
| 1581 { delimright = c; | |
| 1582 nest = 0; | |
| 1583 } | |
| 1584 } | |
| 1585 else | |
| 1586 { | |
| 1587 if (blankrol) | |
| 1588 { error("heredoc rest of line should be blank"); | |
| 1589 blankrol = 0; | |
| 1590 continue; | |
| 1591 } | |
| 1592 if (nest == 1) | |
| 1593 { | |
| 1594 if (c == delimleft) | |
| 1595 nestcount++; | |
| 1596 else if (c == delimright) | |
| 1597 { nestcount--; | |
| 1598 if (nestcount == 0) | |
| 1599 goto Ldone; | |
| 1600 } | |
| 1601 } | |
| 1602 else if (c == delimright) | |
| 1603 goto Ldone; | |
| 1604 if (startline && isalpha(c)) | |
| 1605 { Token t; | |
| 1606 unsigned char *psave = p; | |
| 1607 p--; | |
| 1608 scan(&t); // read in possible heredoc identifier | |
| 1609 //printf("endid = '%s'\n", t.ident->toChars()); | |
| 1610 if (t.value == TOKidentifier && t.ident->equals(hereid)) | |
| 1611 { /* should check that rest of line is blank | |
| 1612 */ | |
| 1613 goto Ldone; | |
| 1614 } | |
| 1615 p = psave; | |
| 1616 } | |
| 1617 stringbuffer.writeUTF8(c); | |
| 1618 startline = 0; | |
| 1619 } | |
| 1620 } | |
| 1621 | |
| 1622 Ldone: | |
| 1623 if (*p == '"') | |
| 1624 p++; | |
| 1625 else | |
| 1626 error("delimited string must end in %c\"", delimright); | |
| 1627 t->len = stringbuffer.offset; | |
| 1628 stringbuffer.writeByte(0); | |
| 1629 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1630 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1631 stringPostfix(t); | |
| 1632 return TOKstring; | |
| 1633 | |
| 1634 Lerror: | |
| 1635 error("unterminated string constant starting at %s", start.toChars()); | |
| 1636 t->ustring = (unsigned char *)""; | |
| 1637 t->len = 0; | |
| 1638 t->postfix = 0; | |
| 1639 return TOKstring; | |
| 1640 } | |
| 1641 | |
| 1642 /************************************** | |
| 1643 * Lex delimited strings: | |
| 1644 * q{ foo(xxx) } // " foo(xxx) " | |
| 1645 * q{foo(} // "foo(" | |
| 1646 * q{{foo}"}"} // "{foo}"}"" | |
| 1647 * Input: | |
| 1648 * p is on the q | |
| 1649 */ | |
| 1650 | |
| 1651 TOK Lexer::tokenStringConstant(Token *t) | |
| 1652 { | |
| 1653 unsigned nest = 1; | |
| 1654 Loc start = loc; | |
| 1655 unsigned char *pstart = ++p; | |
| 1656 | |
| 1657 while (1) | |
| 1658 { Token tok; | |
| 1659 | |
| 1660 scan(&tok); | |
| 1661 switch (tok.value) | |
| 1662 { | |
| 1663 case TOKlcurly: | |
| 1664 nest++; | |
| 1665 continue; | |
| 1666 | |
| 1667 case TOKrcurly: | |
| 1668 if (--nest == 0) | |
| 1669 goto Ldone; | |
| 1670 continue; | |
| 1671 | |
| 1672 case TOKeof: | |
| 1673 goto Lerror; | |
| 1674 | |
| 1675 default: | |
| 1676 continue; | |
| 1677 } | |
| 1678 } | |
| 1679 | |
| 1680 Ldone: | |
| 1681 t->len = p - 1 - pstart; | |
| 1682 t->ustring = (unsigned char *)mem.malloc(t->len + 1); | |
| 1683 memcpy(t->ustring, pstart, t->len); | |
| 1684 t->ustring[t->len] = 0; | |
| 1685 stringPostfix(t); | |
| 1686 return TOKstring; | |
| 1687 | |
| 1688 Lerror: | |
| 1689 error("unterminated token string constant starting at %s", start.toChars()); | |
| 1690 t->ustring = (unsigned char *)""; | |
| 1691 t->len = 0; | |
| 1692 t->postfix = 0; | |
| 1693 return TOKstring; | |
| 1694 } | |
| 1695 | |
| 1696 #endif | |
| 1697 | |
| 1698 | |
| 1699 /************************************** | |
| 1700 */ | |
| 1701 | |
| 1702 TOK Lexer::escapeStringConstant(Token *t, int wide) | |
| 1703 { unsigned c; | |
| 1704 Loc start = loc; | |
| 1705 | |
| 1706 p++; | |
| 1707 stringbuffer.reset(); | |
| 1708 while (1) | |
| 1709 { | |
| 1710 c = *p++; | |
| 1711 switch (c) | |
| 1712 { | |
| 1713 case '\\': | |
| 1714 switch (*p) | |
| 1715 { | |
| 1716 case 'u': | |
| 1717 case 'U': | |
| 1718 case '&': | |
| 1719 c = escapeSequence(); | |
| 1720 stringbuffer.writeUTF8(c); | |
| 1721 continue; | |
| 1722 | |
| 1723 default: | |
| 1724 c = escapeSequence(); | |
| 1725 break; | |
| 1726 } | |
| 1727 break; | |
| 1728 | |
| 1729 case '\n': | |
| 1730 loc.linnum++; | |
| 1731 break; | |
| 1732 | |
| 1733 case '\r': | |
| 1734 if (*p == '\n') | |
| 1735 continue; // ignore | |
| 1736 c = '\n'; // treat EndOfLine as \n character | |
| 1737 loc.linnum++; | |
| 1738 break; | |
| 1739 | |
| 1740 case '"': | |
| 1741 t->len = stringbuffer.offset; | |
| 1742 stringbuffer.writeByte(0); | |
| 1743 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1744 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1745 stringPostfix(t); | |
| 1746 return TOKstring; | |
| 1747 | |
| 1748 case 0: | |
| 1749 case 0x1A: | |
| 1750 p--; | |
| 1751 error("unterminated string constant starting at %s", start.toChars()); | |
| 1752 t->ustring = (unsigned char *)""; | |
| 1753 t->len = 0; | |
| 1754 t->postfix = 0; | |
| 1755 return TOKstring; | |
| 1756 | |
| 1757 default: | |
| 1758 if (c & 0x80) | |
| 1759 { | |
| 1760 p--; | |
| 1761 c = decodeUTF(); | |
| 1762 if (c == LS || c == PS) | |
| 1763 { c = '\n'; | |
| 1764 loc.linnum++; | |
| 1765 } | |
| 1766 p++; | |
| 1767 stringbuffer.writeUTF8(c); | |
| 1768 continue; | |
| 1769 } | |
| 1770 break; | |
| 1771 } | |
| 1772 stringbuffer.writeByte(c); | |
| 1773 } | |
| 1774 } | |
| 1775 | |
| 1776 /************************************** | |
| 1777 */ | |
| 1778 | |
| 1779 TOK Lexer::charConstant(Token *t, int wide) | |
| 1780 { | |
| 1781 unsigned c; | |
| 1782 TOK tk = TOKcharv; | |
| 1783 | |
| 1784 //printf("Lexer::charConstant\n"); | |
| 1785 p++; | |
| 1786 c = *p++; | |
| 1787 switch (c) | |
| 1788 { | |
| 1789 case '\\': | |
| 1790 switch (*p) | |
| 1791 { | |
| 1792 case 'u': | |
| 1793 t->uns64value = escapeSequence(); | |
| 1794 tk = TOKwcharv; | |
| 1795 break; | |
| 1796 | |
| 1797 case 'U': | |
| 1798 case '&': | |
| 1799 t->uns64value = escapeSequence(); | |
| 1800 tk = TOKdcharv; | |
| 1801 break; | |
| 1802 | |
| 1803 default: | |
| 1804 t->uns64value = escapeSequence(); | |
| 1805 break; | |
| 1806 } | |
| 1807 break; | |
| 1808 | |
| 1809 case '\n': | |
| 1810 L1: | |
| 1811 loc.linnum++; | |
| 1812 case '\r': | |
| 1813 case 0: | |
| 1814 case 0x1A: | |
| 1815 case '\'': | |
| 1816 error("unterminated character constant"); | |
| 1817 return tk; | |
| 1818 | |
| 1819 default: | |
| 1820 if (c & 0x80) | |
| 1821 { | |
| 1822 p--; | |
| 1823 c = decodeUTF(); | |
| 1824 p++; | |
| 1825 if (c == LS || c == PS) | |
| 1826 goto L1; | |
| 1827 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 1828 tk = TOKwcharv; | |
| 1829 else | |
| 1830 tk = TOKdcharv; | |
| 1831 } | |
| 1832 t->uns64value = c; | |
| 1833 break; | |
| 1834 } | |
| 1835 | |
| 1836 if (*p != '\'') | |
| 1837 { error("unterminated character constant"); | |
| 1838 return tk; | |
| 1839 } | |
| 1840 p++; | |
| 1841 return tk; | |
| 1842 } | |
| 1843 | |
| 1844 /*************************************** | |
| 1845 * Get postfix of string literal. | |
| 1846 */ | |
| 1847 | |
| 1848 void Lexer::stringPostfix(Token *t) | |
| 1849 { | |
| 1850 switch (*p) | |
| 1851 { | |
| 1852 case 'c': | |
| 1853 case 'w': | |
| 1854 case 'd': | |
| 1855 t->postfix = *p; | |
| 1856 p++; | |
| 1857 break; | |
| 1858 | |
| 1859 default: | |
| 1860 t->postfix = 0; | |
| 1861 break; | |
| 1862 } | |
| 1863 } | |
| 1864 | |
| 1865 /*************************************** | |
| 1866 * Read \u or \U unicode sequence | |
| 1867 * Input: | |
| 1868 * u 'u' or 'U' | |
| 1869 */ | |
| 1870 | |
| 1871 #if 0 | |
| 1872 unsigned Lexer::wchar(unsigned u) | |
| 1873 { | |
| 1874 unsigned value; | |
| 1875 unsigned n; | |
| 1876 unsigned char c; | |
| 1877 unsigned nchars; | |
| 1878 | |
| 1879 nchars = (u == 'U') ? 8 : 4; | |
| 1880 value = 0; | |
| 1881 for (n = 0; 1; n++) | |
| 1882 { | |
| 1883 ++p; | |
| 1884 if (n == nchars) | |
| 1885 break; | |
| 1886 c = *p; | |
| 1887 if (!ishex(c)) | |
| 1888 { error("\\%c sequence must be followed by %d hex characters", u, nchars); | |
| 1889 break; | |
| 1890 } | |
| 1891 if (isdigit(c)) | |
| 1892 c -= '0'; | |
| 1893 else if (islower(c)) | |
| 1894 c -= 'a' - 10; | |
| 1895 else | |
| 1896 c -= 'A' - 10; | |
| 1897 value <<= 4; | |
| 1898 value |= c; | |
| 1899 } | |
| 1900 return value; | |
| 1901 } | |
| 1902 #endif | |
| 1903 | |
| 1904 /************************************** | |
| 1905 * Read in a number. | |
| 1906 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 1907 * integers can be decimal, octal or hex | |
| 1908 * Handle the suffixes U, UL, LU, L, etc. | |
| 1909 * If it's double, store it in tok.TKutok.Vdouble. | |
| 1910 * Returns: | |
| 1911 * TKnum | |
| 1912 * TKdouble,... | |
| 1913 */ | |
| 1914 | |
| 1915 TOK Lexer::number(Token *t) | |
| 1916 { | |
| 1917 // We use a state machine to collect numbers | |
| 1918 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 1919 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 1920 STATE_hexh, STATE_error }; | |
| 1921 enum STATE state; | |
| 1922 | |
| 1923 enum FLAGS | |
| 1924 { FLAGS_decimal = 1, // decimal | |
| 1925 FLAGS_unsigned = 2, // u or U suffix | |
| 1926 FLAGS_long = 4, // l or L suffix | |
| 1927 }; | |
| 1928 enum FLAGS flags = FLAGS_decimal; | |
| 1929 | |
| 1930 int i; | |
| 1931 int base; | |
| 1932 unsigned c; | |
| 1933 unsigned char *start; | |
| 1934 TOK result; | |
| 1935 | |
| 1936 //printf("Lexer::number()\n"); | |
| 1937 state = STATE_initial; | |
| 1938 base = 0; | |
| 1939 stringbuffer.reset(); | |
| 1940 start = p; | |
| 1941 while (1) | |
| 1942 { | |
| 1943 c = *p; | |
| 1944 switch (state) | |
| 1945 { | |
| 1946 case STATE_initial: // opening state | |
| 1947 if (c == '0') | |
| 1948 state = STATE_0; | |
| 1949 else | |
| 1950 state = STATE_decimal; | |
| 1951 break; | |
| 1952 | |
| 1953 case STATE_0: | |
| 1954 flags = (FLAGS) (flags & ~FLAGS_decimal); | |
| 1955 switch (c) | |
| 1956 { | |
| 1957 #if ZEROH | |
| 1958 case 'H': // 0h | |
| 1959 case 'h': | |
| 1960 goto hexh; | |
| 1961 #endif | |
| 1962 case 'X': | |
| 1963 case 'x': | |
| 1964 state = STATE_hex0; | |
| 1965 break; | |
| 1966 | |
| 1967 case '.': | |
| 1968 if (p[1] == '.') // .. is a separate token | |
| 1969 goto done; | |
| 1970 case 'i': | |
| 1971 case 'f': | |
| 1972 case 'F': | |
| 1973 goto real; | |
| 1974 #if ZEROH | |
| 1975 case 'E': | |
| 1976 case 'e': | |
| 1977 goto case_hex; | |
| 1978 #endif | |
| 1979 case 'B': | |
| 1980 case 'b': | |
| 1981 state = STATE_binary0; | |
| 1982 break; | |
| 1983 | |
| 1984 case '0': case '1': case '2': case '3': | |
| 1985 case '4': case '5': case '6': case '7': | |
| 1986 state = STATE_octal; | |
| 1987 break; | |
| 1988 | |
| 1989 #if ZEROH | |
| 1990 case '8': case '9': case 'A': | |
| 1991 case 'C': case 'D': case 'F': | |
| 1992 case 'a': case 'c': case 'd': case 'f': | |
| 1993 case_hex: | |
| 1994 state = STATE_hexh; | |
| 1995 break; | |
| 1996 #endif | |
| 1997 case '_': | |
| 1998 state = STATE_octal; | |
| 1999 p++; | |
| 2000 continue; | |
| 2001 | |
| 2002 case 'L': | |
| 2003 if (p[1] == 'i') | |
| 2004 goto real; | |
| 2005 goto done; | |
| 2006 | |
| 2007 default: | |
| 2008 goto done; | |
| 2009 } | |
| 2010 break; | |
| 2011 | |
| 2012 case STATE_decimal: // reading decimal number | |
| 2013 if (!isdigit(c)) | |
| 2014 { | |
| 2015 #if ZEROH | |
| 2016 if (ishex(c) | |
| 2017 || c == 'H' || c == 'h' | |
| 2018 ) | |
| 2019 goto hexh; | |
| 2020 #endif | |
| 2021 if (c == '_') // ignore embedded _ | |
| 2022 { p++; | |
| 2023 continue; | |
| 2024 } | |
| 2025 if (c == '.' && p[1] != '.') | |
| 2026 goto real; | |
| 2027 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2028 c == 'e' || c == 'E') | |
| 2029 { | |
| 2030 real: // It's a real number. Back up and rescan as a real | |
| 2031 p = start; | |
| 2032 return inreal(t); | |
| 2033 } | |
| 2034 else if (c == 'L' && p[1] == 'i') | |
| 2035 goto real; | |
| 2036 goto done; | |
| 2037 } | |
| 2038 break; | |
| 2039 | |
| 2040 case STATE_hex0: // reading hex number | |
| 2041 case STATE_hex: | |
| 2042 if (!ishex(c)) | |
| 2043 { | |
| 2044 if (c == '_') // ignore embedded _ | |
| 2045 { p++; | |
| 2046 continue; | |
| 2047 } | |
| 2048 if (c == '.' && p[1] != '.') | |
| 2049 goto real; | |
| 2050 if (c == 'P' || c == 'p' || c == 'i') | |
| 2051 goto real; | |
| 2052 if (state == STATE_hex0) | |
| 2053 error("Hex digit expected, not '%c'", c); | |
| 2054 goto done; | |
| 2055 } | |
| 2056 state = STATE_hex; | |
| 2057 break; | |
| 2058 | |
| 2059 #if ZEROH | |
| 2060 hexh: | |
| 2061 state = STATE_hexh; | |
| 2062 case STATE_hexh: // parse numbers like 0FFh | |
| 2063 if (!ishex(c)) | |
| 2064 { | |
| 2065 if (c == 'H' || c == 'h') | |
| 2066 { | |
| 2067 p++; | |
| 2068 base = 16; | |
| 2069 goto done; | |
| 2070 } | |
| 2071 else | |
| 2072 { | |
| 2073 // Check for something like 1E3 or 0E24 | |
| 2074 if (memchr((char *)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2075 memchr((char *)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2076 goto real; | |
| 2077 error("Hex digit expected, not '%c'", c); | |
| 2078 goto done; | |
| 2079 } | |
| 2080 } | |
| 2081 break; | |
| 2082 #endif | |
| 2083 | |
| 2084 case STATE_octal: // reading octal number | |
| 2085 case STATE_octale: // reading octal number with non-octal digits | |
| 2086 if (!isoctal(c)) | |
| 2087 { | |
| 2088 #if ZEROH | |
| 2089 if (ishex(c) | |
| 2090 || c == 'H' || c == 'h' | |
| 2091 ) | |
| 2092 goto hexh; | |
| 2093 #endif | |
| 2094 if (c == '_') // ignore embedded _ | |
| 2095 { p++; | |
| 2096 continue; | |
| 2097 } | |
| 2098 if (c == '.' && p[1] != '.') | |
| 2099 goto real; | |
| 2100 if (c == 'i') | |
| 2101 goto real; | |
| 2102 if (isdigit(c)) | |
| 2103 { | |
| 2104 state = STATE_octale; | |
| 2105 } | |
| 2106 else | |
| 2107 goto done; | |
| 2108 } | |
| 2109 break; | |
| 2110 | |
| 2111 case STATE_binary0: // starting binary number | |
| 2112 case STATE_binary: // reading binary number | |
| 2113 if (c != '0' && c != '1') | |
| 2114 { | |
| 2115 #if ZEROH | |
| 2116 if (ishex(c) | |
| 2117 || c == 'H' || c == 'h' | |
| 2118 ) | |
| 2119 goto hexh; | |
| 2120 #endif | |
| 2121 if (c == '_') // ignore embedded _ | |
| 2122 { p++; | |
| 2123 continue; | |
| 2124 } | |
| 2125 if (state == STATE_binary0) | |
| 2126 { error("binary digit expected"); | |
| 2127 state = STATE_error; | |
| 2128 break; | |
| 2129 } | |
| 2130 else | |
| 2131 goto done; | |
| 2132 } | |
| 2133 state = STATE_binary; | |
| 2134 break; | |
| 2135 | |
| 2136 case STATE_error: // for error recovery | |
| 2137 if (!isdigit(c)) // scan until non-digit | |
| 2138 goto done; | |
| 2139 break; | |
| 2140 | |
| 2141 default: | |
| 2142 assert(0); | |
| 2143 } | |
| 2144 stringbuffer.writeByte(c); | |
| 2145 p++; | |
| 2146 } | |
| 2147 done: | |
| 2148 stringbuffer.writeByte(0); // terminate string | |
| 2149 if (state == STATE_octale) | |
| 2150 error("Octal digit expected"); | |
| 2151 | |
| 2152 uinteger_t n; // unsigned >=64 bit integer type | |
| 2153 | |
| 2154 if (stringbuffer.offset == 2 && (state == STATE_decimal || state == STATE_0)) | |
| 2155 n = stringbuffer.data[0] - '0'; | |
| 2156 else | |
| 2157 { | |
| 2158 // Convert string to integer | |
| 2159 #if __DMC__ | |
| 2160 errno = 0; | |
| 2161 n = strtoull((char *)stringbuffer.data,NULL,base); | |
| 2162 if (errno == ERANGE) | |
| 2163 error("integer overflow"); | |
| 2164 #else | |
| 2165 // Not everybody implements strtoull() | |
| 2166 char *p = (char *)stringbuffer.data; | |
| 2167 int r = 10, d; | |
| 2168 | |
| 2169 if (*p == '0') | |
| 2170 { | |
| 2171 if (p[1] == 'x' || p[1] == 'X') | |
| 2172 p += 2, r = 16; | |
| 2173 else if (p[1] == 'b' || p[1] == 'B') | |
| 2174 p += 2, r = 2; | |
| 2175 else if (isdigit(p[1])) | |
| 2176 p += 1, r = 8; | |
| 2177 } | |
| 2178 | |
| 2179 n = 0; | |
| 2180 while (1) | |
| 2181 { | |
| 2182 if (*p >= '0' && *p <= '9') | |
| 2183 d = *p - '0'; | |
| 2184 else if (*p >= 'a' && *p <= 'z') | |
| 2185 d = *p - 'a' + 10; | |
| 2186 else if (*p >= 'A' && *p <= 'Z') | |
| 2187 d = *p - 'A' + 10; | |
| 2188 else | |
| 2189 break; | |
| 2190 if (d >= r) | |
| 2191 break; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2192 uinteger_t n2 = n * r; |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2193 //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
|
2194 if (n2 / r != n || n2 + d < n) |
| 159 | 2195 { |
| 2196 error ("integer overflow"); | |
| 2197 break; | |
| 2198 } | |
| 2199 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2200 n = n2 + d; |
| 159 | 2201 p++; |
| 2202 } | |
| 2203 #endif | |
| 2204 if (sizeof(n) > 8 && | |
| 2205 n > 0xFFFFFFFFFFFFFFFFULL) // if n needs more than 64 bits | |
| 2206 error("integer overflow"); | |
| 2207 } | |
| 2208 | |
| 2209 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2210 while (1) | |
| 2211 { unsigned char f; | |
| 2212 | |
| 2213 switch (*p) | |
| 2214 { case 'U': | |
| 2215 case 'u': | |
| 2216 f = FLAGS_unsigned; | |
| 2217 goto L1; | |
| 2218 | |
| 2219 case 'l': | |
| 2220 if (1 || !global.params.useDeprecated) | |
| 2221 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2222 case 'L': | |
| 2223 f = FLAGS_long; | |
| 2224 L1: | |
| 2225 p++; | |
| 2226 if (flags & f) | |
| 2227 error("unrecognized token"); | |
| 2228 flags = (FLAGS) (flags | f); | |
| 2229 continue; | |
| 2230 default: | |
| 2231 break; | |
| 2232 } | |
| 2233 break; | |
| 2234 } | |
| 2235 | |
| 2236 switch (flags) | |
| 2237 { | |
| 2238 case 0: | |
| 2239 /* Octal or Hexadecimal constant. | |
| 2240 * First that fits: int, uint, long, ulong | |
| 2241 */ | |
| 2242 if (n & 0x8000000000000000LL) | |
| 2243 result = TOKuns64v; | |
| 2244 else if (n & 0xFFFFFFFF00000000LL) | |
| 2245 result = TOKint64v; | |
| 2246 else if (n & 0x80000000) | |
| 2247 result = TOKuns32v; | |
| 2248 else | |
| 2249 result = TOKint32v; | |
| 2250 break; | |
| 2251 | |
| 2252 case FLAGS_decimal: | |
| 2253 /* First that fits: int, long, long long | |
| 2254 */ | |
| 2255 if (n & 0x8000000000000000LL) | |
| 2256 { error("signed integer overflow"); | |
| 2257 result = TOKuns64v; | |
| 2258 } | |
| 2259 else if (n & 0xFFFFFFFF80000000LL) | |
| 2260 result = TOKint64v; | |
| 2261 else | |
| 2262 result = TOKint32v; | |
| 2263 break; | |
| 2264 | |
| 2265 case FLAGS_unsigned: | |
| 2266 case FLAGS_decimal | FLAGS_unsigned: | |
| 2267 /* First that fits: uint, ulong | |
| 2268 */ | |
| 2269 if (n & 0xFFFFFFFF00000000LL) | |
| 2270 result = TOKuns64v; | |
| 2271 else | |
| 2272 result = TOKuns32v; | |
| 2273 break; | |
| 2274 | |
| 2275 case FLAGS_decimal | FLAGS_long: | |
| 2276 if (n & 0x8000000000000000LL) | |
| 2277 { error("signed integer overflow"); | |
| 2278 result = TOKuns64v; | |
| 2279 } | |
| 2280 else | |
| 2281 result = TOKint64v; | |
| 2282 break; | |
| 2283 | |
| 2284 case FLAGS_long: | |
| 2285 if (n & 0x8000000000000000LL) | |
| 2286 result = TOKuns64v; | |
| 2287 else | |
| 2288 result = TOKint64v; | |
| 2289 break; | |
| 2290 | |
| 2291 case FLAGS_unsigned | FLAGS_long: | |
| 2292 case FLAGS_decimal | FLAGS_unsigned | FLAGS_long: | |
| 2293 result = TOKuns64v; | |
| 2294 break; | |
| 2295 | |
| 2296 default: | |
| 2297 #ifdef DEBUG | |
| 2298 printf("%x\n",flags); | |
| 2299 #endif | |
| 2300 assert(0); | |
| 2301 } | |
| 2302 t->uns64value = n; | |
| 2303 return result; | |
| 2304 } | |
| 2305 | |
| 2306 /************************************** | |
| 2307 * Read in characters, converting them to real. | |
| 2308 * Bugs: | |
| 2309 * Exponent overflow not detected. | |
| 2310 * Too much requested precision is not detected. | |
| 2311 */ | |
| 2312 | |
| 2313 TOK Lexer::inreal(Token *t) | |
| 2314 #ifdef __DMC__ | |
| 2315 __in | |
| 2316 { | |
| 2317 assert(*p == '.' || isdigit(*p)); | |
| 2318 } | |
| 2319 __out (result) | |
| 2320 { | |
| 2321 switch (result) | |
| 2322 { | |
| 2323 case TOKfloat32v: | |
| 2324 case TOKfloat64v: | |
| 2325 case TOKfloat80v: | |
| 2326 case TOKimaginary32v: | |
| 2327 case TOKimaginary64v: | |
| 2328 case TOKimaginary80v: | |
| 2329 break; | |
| 2330 | |
| 2331 default: | |
| 2332 assert(0); | |
| 2333 } | |
| 2334 } | |
| 2335 __body | |
| 2336 #endif /* __DMC__ */ | |
| 2337 { int dblstate; | |
| 2338 unsigned c; | |
| 2339 char hex; // is this a hexadecimal-floating-constant? | |
| 2340 TOK result; | |
| 2341 | |
| 2342 //printf("Lexer::inreal()\n"); | |
| 2343 stringbuffer.reset(); | |
| 2344 dblstate = 0; | |
| 2345 hex = 0; | |
| 2346 Lnext: | |
| 2347 while (1) | |
| 2348 { | |
| 2349 // Get next char from input | |
| 2350 c = *p++; | |
| 2351 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2352 while (1) | |
| 2353 { | |
| 2354 switch (dblstate) | |
| 2355 { | |
| 2356 case 0: // opening state | |
| 2357 if (c == '0') | |
| 2358 dblstate = 9; | |
| 2359 else if (c == '.') | |
| 2360 dblstate = 3; | |
| 2361 else | |
| 2362 dblstate = 1; | |
| 2363 break; | |
| 2364 | |
| 2365 case 9: | |
| 2366 dblstate = 1; | |
| 2367 if (c == 'X' || c == 'x') | |
| 2368 { hex++; | |
| 2369 break; | |
| 2370 } | |
| 2371 case 1: // digits to left of . | |
| 2372 case 3: // digits to right of . | |
| 2373 case 7: // continuing exponent digits | |
| 2374 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2375 { | |
| 2376 if (c == '_') | |
| 2377 goto Lnext; // ignore embedded '_' | |
| 2378 dblstate++; | |
| 2379 continue; | |
| 2380 } | |
| 2381 break; | |
| 2382 | |
| 2383 case 2: // no more digits to left of . | |
| 2384 if (c == '.') | |
| 2385 { dblstate++; | |
| 2386 break; | |
| 2387 } | |
| 2388 case 4: // no more digits to right of . | |
| 2389 if ((c == 'E' || c == 'e') || | |
| 2390 hex && (c == 'P' || c == 'p')) | |
| 2391 { dblstate = 5; | |
| 2392 hex = 0; // exponent is always decimal | |
| 2393 break; | |
| 2394 } | |
| 2395 if (hex) | |
| 2396 error("binary-exponent-part required"); | |
| 2397 goto done; | |
| 2398 | |
| 2399 case 5: // looking immediately to right of E | |
| 2400 dblstate++; | |
| 2401 if (c == '-' || c == '+') | |
| 2402 break; | |
| 2403 case 6: // 1st exponent digit expected | |
| 2404 if (!isdigit(c)) | |
| 2405 error("exponent expected"); | |
| 2406 dblstate++; | |
| 2407 break; | |
| 2408 | |
| 2409 case 8: // past end of exponent digits | |
| 2410 goto done; | |
| 2411 } | |
| 2412 break; | |
| 2413 } | |
| 2414 stringbuffer.writeByte(c); | |
| 2415 } | |
| 2416 done: | |
| 2417 p--; | |
| 2418 | |
| 2419 stringbuffer.writeByte(0); | |
| 2420 | |
| 2421 #if _WIN32 && __DMC__ | |
| 2422 char *save = __locale_decpoint; | |
| 2423 __locale_decpoint = "."; | |
| 2424 #endif | |
| 2425 #ifdef IN_GCC | |
| 2426 t->float80value = real_t::parse((char *)stringbuffer.data, real_t::LongDouble); | |
| 2427 #else | |
| 2428 t->float80value = strtold((char *)stringbuffer.data, NULL); | |
| 2429 #endif | |
| 2430 errno = 0; | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2431 float strtofres; |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2432 double strtodres; |
| 159 | 2433 switch (*p) |
| 2434 { | |
| 2435 case 'F': | |
| 2436 case 'f': | |
| 2437 #ifdef IN_GCC | |
| 2438 real_t::parse((char *)stringbuffer.data, real_t::Float); | |
| 2439 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2440 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
|
2441 // 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
|
2442 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2443 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
|
2444 errno = 0; |
| 159 | 2445 #endif |
| 2446 result = TOKfloat32v; | |
| 2447 p++; | |
| 2448 break; | |
| 2449 | |
| 2450 default: | |
| 2451 #ifdef IN_GCC | |
| 2452 real_t::parse((char *)stringbuffer.data, real_t::Double); | |
| 2453 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2454 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
|
2455 // 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
|
2456 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2457 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
|
2458 errno = 0; |
| 159 | 2459 #endif |
| 2460 result = TOKfloat64v; | |
| 2461 break; | |
| 2462 | |
| 2463 case 'l': | |
| 2464 if (!global.params.useDeprecated) | |
| 2465 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2466 case 'L': | |
| 2467 result = TOKfloat80v; | |
| 2468 p++; | |
| 2469 break; | |
| 2470 } | |
| 2471 if (*p == 'i' || *p == 'I') | |
| 2472 { | |
| 2473 if (!global.params.useDeprecated && *p == 'I') | |
| 2474 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2475 p++; | |
| 2476 switch (result) | |
| 2477 { | |
| 2478 case TOKfloat32v: | |
| 2479 result = TOKimaginary32v; | |
| 2480 break; | |
| 2481 case TOKfloat64v: | |
| 2482 result = TOKimaginary64v; | |
| 2483 break; | |
| 2484 case TOKfloat80v: | |
| 2485 result = TOKimaginary80v; | |
| 2486 break; | |
| 2487 } | |
| 2488 } | |
| 2489 #if _WIN32 && __DMC__ | |
| 2490 __locale_decpoint = save; | |
| 2491 #endif | |
| 2492 if (errno == ERANGE) | |
| 2493 error("number is not representable"); | |
| 2494 return result; | |
| 2495 } | |
| 2496 | |
| 2497 /********************************************* | |
| 2498 * Do pragma. | |
| 2499 * Currently, the only pragma supported is: | |
| 2500 * #line linnum [filespec] | |
| 2501 */ | |
| 2502 | |
| 2503 void Lexer::pragma() | |
| 2504 { | |
| 2505 Token tok; | |
| 2506 int linnum; | |
| 2507 char *filespec = NULL; | |
| 2508 Loc loc = this->loc; | |
| 2509 | |
| 2510 scan(&tok); | |
| 2511 if (tok.value != TOKidentifier || tok.ident != Id::line) | |
| 2512 goto Lerr; | |
| 2513 | |
| 2514 scan(&tok); | |
| 2515 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 2516 linnum = tok.uns64value - 1; | |
| 2517 else | |
| 2518 goto Lerr; | |
| 2519 | |
| 2520 while (1) | |
| 2521 { | |
| 2522 switch (*p) | |
| 2523 { | |
| 2524 case 0: | |
| 2525 case 0x1A: | |
| 2526 case '\n': | |
| 2527 Lnewline: | |
| 2528 this->loc.linnum = linnum; | |
| 2529 if (filespec) | |
| 2530 this->loc.filename = filespec; | |
| 2531 return; | |
| 2532 | |
| 2533 case '\r': | |
| 2534 p++; | |
| 2535 if (*p != '\n') | |
| 2536 { p--; | |
| 2537 goto Lnewline; | |
| 2538 } | |
| 2539 continue; | |
| 2540 | |
| 2541 case ' ': | |
| 2542 case '\t': | |
| 2543 case '\v': | |
| 2544 case '\f': | |
| 2545 p++; | |
| 2546 continue; // skip white space | |
| 2547 | |
| 2548 case '_': | |
| 2549 if (mod && memcmp(p, "__FILE__", 8) == 0) | |
| 2550 { | |
| 2551 p += 8; | |
| 2552 filespec = mem.strdup(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 2553 } | |
| 2554 continue; | |
| 2555 | |
| 2556 case '"': | |
| 2557 if (filespec) | |
| 2558 goto Lerr; | |
| 2559 stringbuffer.reset(); | |
| 2560 p++; | |
| 2561 while (1) | |
| 2562 { unsigned c; | |
| 2563 | |
| 2564 c = *p; | |
| 2565 switch (c) | |
| 2566 { | |
| 2567 case '\n': | |
| 2568 case '\r': | |
| 2569 case 0: | |
| 2570 case 0x1A: | |
| 2571 goto Lerr; | |
| 2572 | |
| 2573 case '"': | |
| 2574 stringbuffer.writeByte(0); | |
| 2575 filespec = mem.strdup((char *)stringbuffer.data); | |
| 2576 p++; | |
| 2577 break; | |
| 2578 | |
| 2579 default: | |
| 2580 if (c & 0x80) | |
| 2581 { unsigned u = decodeUTF(); | |
| 2582 if (u == PS || u == LS) | |
| 2583 goto Lerr; | |
| 2584 } | |
| 2585 stringbuffer.writeByte(c); | |
| 2586 p++; | |
| 2587 continue; | |
| 2588 } | |
| 2589 break; | |
| 2590 } | |
| 2591 continue; | |
| 2592 | |
| 2593 default: | |
| 2594 if (*p & 0x80) | |
| 2595 { unsigned u = decodeUTF(); | |
| 2596 if (u == PS || u == LS) | |
| 2597 goto Lnewline; | |
| 2598 } | |
| 2599 goto Lerr; | |
| 2600 } | |
| 2601 } | |
| 2602 | |
| 2603 Lerr: | |
| 2604 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 2605 } | |
| 2606 | |
| 2607 | |
| 2608 /******************************************** | |
| 2609 * Decode UTF character. | |
| 2610 * Issue error messages for invalid sequences. | |
| 2611 * Return decoded character, advance p to last character in UTF sequence. | |
| 2612 */ | |
| 2613 | |
| 2614 unsigned Lexer::decodeUTF() | |
| 2615 { | |
| 2616 dchar_t u; | |
| 2617 unsigned char c; | |
| 2618 unsigned char *s = p; | |
| 2619 size_t len; | |
| 2620 size_t idx; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2621 const char *msg; |
| 159 | 2622 |
| 2623 c = *s; | |
| 2624 assert(c & 0x80); | |
| 2625 | |
| 2626 // Check length of remaining string up to 6 UTF-8 characters | |
| 2627 for (len = 1; len < 6 && s[len]; len++) | |
| 2628 ; | |
| 2629 | |
| 2630 idx = 0; | |
| 2631 msg = utf_decodeChar(s, len, &idx, &u); | |
| 2632 p += idx - 1; | |
| 2633 if (msg) | |
| 2634 { | |
| 2635 error("%s", msg); | |
| 2636 } | |
| 2637 return u; | |
| 2638 } | |
| 2639 | |
| 2640 | |
| 2641 /*************************************************** | |
| 2642 * Parse doc comment embedded between t->ptr and p. | |
| 2643 * Remove trailing blanks and tabs from lines. | |
| 2644 * Replace all newlines with \n. | |
| 2645 * Remove leading comment character from each line. | |
| 2646 * Decide if it's a lineComment or a blockComment. | |
| 2647 * Append to previous one for this token. | |
| 2648 */ | |
| 2649 | |
| 2650 void Lexer::getDocComment(Token *t, unsigned lineComment) | |
| 2651 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2652 /* 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
|
2653 */ |
| 159 | 2654 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
|
2655 |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2656 /* 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
|
2657 */ |
| 159 | 2658 unsigned char *q = t->ptr + 3; // start of comment text |
| 2659 | |
| 2660 unsigned char *qend = p; | |
| 2661 if (ct == '*' || ct == '+') | |
| 2662 qend -= 2; | |
| 2663 | |
| 2664 /* Scan over initial row of ****'s or ++++'s or ////'s | |
| 2665 */ | |
| 2666 for (; q < qend; q++) | |
| 2667 { | |
| 2668 if (*q != ct) | |
| 2669 break; | |
| 2670 } | |
| 2671 | |
| 2672 /* Remove trailing row of ****'s or ++++'s | |
| 2673 */ | |
| 2674 if (ct != '/') | |
| 2675 { | |
| 2676 for (; q < qend; qend--) | |
| 2677 { | |
| 2678 if (qend[-1] != ct) | |
| 2679 break; | |
| 2680 } | |
| 2681 } | |
| 2682 | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2683 /* Comment is now [q .. qend]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2684 * Canonicalize it into buf[]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2685 */ |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2686 OutBuffer buf; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2687 int linestart = 0; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2688 |
| 159 | 2689 for (; q < qend; q++) |
| 2690 { | |
| 2691 unsigned char c = *q; | |
| 2692 | |
| 2693 switch (c) | |
| 2694 { | |
| 2695 case '*': | |
| 2696 case '+': | |
| 2697 if (linestart && c == ct) | |
| 2698 { linestart = 0; | |
| 2699 /* Trim preceding whitespace up to preceding \n | |
| 2700 */ | |
| 2701 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2702 buf.offset--; | |
| 2703 continue; | |
| 2704 } | |
| 2705 break; | |
| 2706 | |
| 2707 case ' ': | |
| 2708 case '\t': | |
| 2709 break; | |
| 2710 | |
| 2711 case '\r': | |
| 2712 if (q[1] == '\n') | |
| 2713 continue; // skip the \r | |
| 2714 goto Lnewline; | |
| 2715 | |
| 2716 default: | |
| 2717 if (c == 226) | |
| 2718 { | |
| 2719 // If LS or PS | |
| 2720 if (q[1] == 128 && | |
| 2721 (q[2] == 168 || q[2] == 169)) | |
| 2722 { | |
| 2723 q += 2; | |
| 2724 goto Lnewline; | |
| 2725 } | |
| 2726 } | |
| 2727 linestart = 0; | |
| 2728 break; | |
| 2729 | |
| 2730 Lnewline: | |
| 2731 c = '\n'; // replace all newlines with \n | |
| 2732 case '\n': | |
| 2733 linestart = 1; | |
| 2734 | |
| 2735 /* Trim trailing whitespace | |
| 2736 */ | |
| 2737 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2738 buf.offset--; | |
| 2739 | |
| 2740 break; | |
| 2741 } | |
| 2742 buf.writeByte(c); | |
| 2743 } | |
| 2744 | |
| 2745 // Always end with a newline | |
| 2746 if (!buf.offset || buf.data[buf.offset - 1] != '\n') | |
| 2747 buf.writeByte('\n'); | |
| 2748 | |
| 2749 buf.writeByte(0); | |
| 2750 | |
| 2751 // It's a line comment if the start of the doc comment comes | |
| 2752 // after other non-whitespace on the same line. | |
| 2753 unsigned char** dc = (lineComment && anyToken) | |
| 2754 ? &t->lineComment | |
| 2755 : &t->blockComment; | |
| 2756 | |
| 2757 // Combine with previous doc comment, if any | |
| 2758 if (*dc) | |
| 2759 *dc = combineComments(*dc, (unsigned char *)buf.data); | |
| 2760 else | |
| 2761 *dc = (unsigned char *)buf.extractData(); | |
| 2762 } | |
| 2763 | |
| 2764 /******************************************** | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2765 * Combine two document comments into one, |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2766 * separated by a newline. |
| 159 | 2767 */ |
| 2768 | |
| 2769 unsigned char *Lexer::combineComments(unsigned char *c1, unsigned char *c2) | |
| 2770 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2771 //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
|
2772 |
| 159 | 2773 unsigned char *c = c2; |
| 2774 | |
| 2775 if (c1) | |
| 2776 { c = c1; | |
| 2777 if (c2) | |
| 2778 { size_t len1 = strlen((char *)c1); | |
| 2779 size_t len2 = strlen((char *)c2); | |
| 2780 | |
| 2781 c = (unsigned char *)mem.malloc(len1 + 1 + len2 + 1); | |
| 2782 memcpy(c, c1, len1); | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2783 if (len1 && c1[len1 - 1] != '\n') |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2784 { c[len1] = '\n'; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2785 len1++; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2786 } |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2787 memcpy(c + len1, c2, len2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2788 c[len1 + len2] = 0; |
| 159 | 2789 } |
| 2790 } | |
| 2791 return c; | |
| 2792 } | |
| 2793 | |
| 2794 /******************************************** | |
| 2795 * Create an identifier in the string table. | |
| 2796 */ | |
| 2797 | |
| 2798 Identifier *Lexer::idPool(const char *s) | |
| 2799 { | |
| 2800 size_t len = strlen(s); | |
| 2801 StringValue *sv = stringtable.update(s, len); | |
| 2802 Identifier *id = (Identifier *) sv->ptrvalue; | |
| 2803 if (!id) | |
| 2804 { | |
| 2805 id = new Identifier(sv->lstring.string, TOKidentifier); | |
| 2806 sv->ptrvalue = id; | |
| 2807 } | |
| 2808 return id; | |
| 2809 } | |
| 2810 | |
| 2811 /********************************************* | |
| 2812 * Create a unique identifier using the prefix s. | |
| 2813 */ | |
| 2814 | |
| 2815 Identifier *Lexer::uniqueId(const char *s, int num) | |
| 2816 { char buffer[32]; | |
| 2817 size_t slen = strlen(s); | |
| 2818 | |
| 2819 assert(slen + sizeof(num) * 3 + 1 <= sizeof(buffer)); | |
| 2820 sprintf(buffer, "%s%d", s, num); | |
| 2821 return idPool(buffer); | |
| 2822 } | |
| 2823 | |
| 2824 Identifier *Lexer::uniqueId(const char *s) | |
| 2825 { | |
| 2826 static int num; | |
| 2827 return uniqueId(s, ++num); | |
| 2828 } | |
| 2829 | |
| 2830 /**************************************** | |
| 2831 */ | |
| 2832 | |
| 2833 struct Keyword | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2834 { const char *name; |
| 159 | 2835 enum TOK value; |
| 2836 }; | |
| 2837 | |
| 2838 static Keyword keywords[] = | |
| 2839 { | |
| 2840 // { "", TOK }, | |
| 2841 | |
| 2842 { "this", TOKthis }, | |
| 2843 { "super", TOKsuper }, | |
| 2844 { "assert", TOKassert }, | |
| 2845 { "null", TOKnull }, | |
| 2846 { "true", TOKtrue }, | |
| 2847 { "false", TOKfalse }, | |
| 2848 { "cast", TOKcast }, | |
| 2849 { "new", TOKnew }, | |
| 2850 { "delete", TOKdelete }, | |
| 2851 { "throw", TOKthrow }, | |
| 2852 { "module", TOKmodule }, | |
| 2853 { "pragma", TOKpragma }, | |
| 2854 { "typeof", TOKtypeof }, | |
| 2855 { "typeid", TOKtypeid }, | |
| 2856 | |
| 2857 { "template", TOKtemplate }, | |
| 2858 | |
| 2859 { "void", TOKvoid }, | |
| 2860 { "byte", TOKint8 }, | |
| 2861 { "ubyte", TOKuns8 }, | |
| 2862 { "short", TOKint16 }, | |
| 2863 { "ushort", TOKuns16 }, | |
| 2864 { "int", TOKint32 }, | |
| 2865 { "uint", TOKuns32 }, | |
| 2866 { "long", TOKint64 }, | |
| 2867 { "ulong", TOKuns64 }, | |
| 2868 { "cent", TOKcent, }, | |
| 2869 { "ucent", TOKucent, }, | |
| 2870 { "float", TOKfloat32 }, | |
| 2871 { "double", TOKfloat64 }, | |
| 2872 { "real", TOKfloat80 }, | |
| 2873 | |
| 2874 { "bool", TOKbool }, | |
| 2875 { "char", TOKchar }, | |
| 2876 { "wchar", TOKwchar }, | |
| 2877 { "dchar", TOKdchar }, | |
| 2878 | |
| 2879 { "ifloat", TOKimaginary32 }, | |
| 2880 { "idouble", TOKimaginary64 }, | |
| 2881 { "ireal", TOKimaginary80 }, | |
| 2882 | |
| 2883 { "cfloat", TOKcomplex32 }, | |
| 2884 { "cdouble", TOKcomplex64 }, | |
| 2885 { "creal", TOKcomplex80 }, | |
| 2886 | |
| 2887 { "delegate", TOKdelegate }, | |
| 2888 { "function", TOKfunction }, | |
| 2889 | |
| 2890 { "is", TOKis }, | |
| 2891 { "if", TOKif }, | |
| 2892 { "else", TOKelse }, | |
| 2893 { "while", TOKwhile }, | |
| 2894 { "for", TOKfor }, | |
| 2895 { "do", TOKdo }, | |
| 2896 { "switch", TOKswitch }, | |
| 2897 { "case", TOKcase }, | |
| 2898 { "default", TOKdefault }, | |
| 2899 { "break", TOKbreak }, | |
| 2900 { "continue", TOKcontinue }, | |
| 2901 { "synchronized", TOKsynchronized }, | |
| 2902 { "return", TOKreturn }, | |
| 2903 { "goto", TOKgoto }, | |
| 2904 { "try", TOKtry }, | |
| 2905 { "catch", TOKcatch }, | |
| 2906 { "finally", TOKfinally }, | |
| 2907 { "with", TOKwith }, | |
| 2908 { "asm", TOKasm }, | |
| 2909 { "foreach", TOKforeach }, | |
| 2910 { "foreach_reverse", TOKforeach_reverse }, | |
| 2911 { "scope", TOKscope }, | |
| 2912 | |
| 2913 { "struct", TOKstruct }, | |
| 2914 { "class", TOKclass }, | |
| 2915 { "interface", TOKinterface }, | |
| 2916 { "union", TOKunion }, | |
| 2917 { "enum", TOKenum }, | |
| 2918 { "import", TOKimport }, | |
| 2919 { "mixin", TOKmixin }, | |
| 2920 { "static", TOKstatic }, | |
| 2921 { "final", TOKfinal }, | |
| 2922 { "const", TOKconst }, | |
| 2923 { "typedef", TOKtypedef }, | |
| 2924 { "alias", TOKalias }, | |
| 2925 { "override", TOKoverride }, | |
| 2926 { "abstract", TOKabstract }, | |
| 2927 { "volatile", TOKvolatile }, | |
| 2928 { "debug", TOKdebug }, | |
| 2929 { "deprecated", TOKdeprecated }, | |
| 2930 { "in", TOKin }, | |
| 2931 { "out", TOKout }, | |
| 2932 { "inout", TOKinout }, | |
| 2933 { "lazy", TOKlazy }, | |
| 2934 { "auto", TOKauto }, | |
| 2935 | |
| 2936 { "align", TOKalign }, | |
| 2937 { "extern", TOKextern }, | |
| 2938 { "private", TOKprivate }, | |
| 2939 { "package", TOKpackage }, | |
| 2940 { "protected", TOKprotected }, | |
| 2941 { "public", TOKpublic }, | |
| 2942 { "export", TOKexport }, | |
| 2943 | |
| 2944 { "body", TOKbody }, | |
| 2945 { "invariant", TOKinvariant }, | |
| 2946 { "unittest", TOKunittest }, | |
| 2947 { "version", TOKversion }, | |
| 2948 //{ "manifest", TOKmanifest }, | |
| 2949 | |
| 2950 // Added after 1.0 | |
| 2951 { "ref", TOKref }, | |
| 2952 { "macro", TOKmacro }, | |
| 336 | 2953 #if DMDV2 |
| 159 | 2954 { "pure", TOKpure }, |
| 2955 { "nothrow", TOKnothrow }, | |
| 336 | 2956 { "__thread", TOKtls }, |
| 159 | 2957 { "__traits", TOKtraits }, |
| 2958 { "__overloadset", TOKoverloadset }, | |
| 336 | 2959 { "__FILE__", TOKfile }, |
| 2960 { "__LINE__", TOKline }, | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2961 { "shared", TOKshared }, |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2962 { "immutable", TOKimmutable }, |
| 159 | 2963 #endif |
| 2964 }; | |
| 2965 | |
| 2966 int Token::isKeyword() | |
| 2967 { | |
| 2968 for (unsigned u = 0; u < sizeof(keywords) / sizeof(keywords[0]); u++) | |
| 2969 { | |
| 2970 if (keywords[u].value == value) | |
| 2971 return 1; | |
| 2972 } | |
| 2973 return 0; | |
| 2974 } | |
| 2975 | |
| 2976 void Lexer::initKeywords() | |
| 2977 { StringValue *sv; | |
| 2978 unsigned u; | |
| 2979 enum TOK v; | |
| 2980 unsigned nkeywords = sizeof(keywords) / sizeof(keywords[0]); | |
| 2981 | |
| 2982 if (global.params.Dversion == 1) | |
| 2983 nkeywords -= 2; | |
| 2984 | |
| 2985 cmtable_init(); | |
| 2986 | |
| 2987 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
|
2988 { const char *s; |
| 159 | 2989 |
| 2990 //printf("keyword[%d] = '%s'\n",u, keywords[u].name); | |
| 2991 s = keywords[u].name; | |
| 2992 v = keywords[u].value; | |
| 2993 sv = stringtable.insert(s, strlen(s)); | |
| 2994 sv->ptrvalue = (void *) new Identifier(sv->lstring.string,v); | |
| 2995 | |
| 2996 //printf("tochars[%d] = '%s'\n",v, s); | |
| 2997 Token::tochars[v] = s; | |
| 2998 } | |
| 2999 | |
| 3000 Token::tochars[TOKeof] = "EOF"; | |
| 3001 Token::tochars[TOKlcurly] = "{"; | |
| 3002 Token::tochars[TOKrcurly] = "}"; | |
| 3003 Token::tochars[TOKlparen] = "("; | |
| 3004 Token::tochars[TOKrparen] = ")"; | |
| 3005 Token::tochars[TOKlbracket] = "["; | |
| 3006 Token::tochars[TOKrbracket] = "]"; | |
| 3007 Token::tochars[TOKsemicolon] = ";"; | |
| 3008 Token::tochars[TOKcolon] = ":"; | |
| 3009 Token::tochars[TOKcomma] = ","; | |
| 3010 Token::tochars[TOKdot] = "."; | |
| 3011 Token::tochars[TOKxor] = "^"; | |
| 3012 Token::tochars[TOKxorass] = "^="; | |
| 3013 Token::tochars[TOKassign] = "="; | |
| 3014 Token::tochars[TOKconstruct] = "="; | |
| 336 | 3015 #if DMDV2 |
| 159 | 3016 Token::tochars[TOKblit] = "="; |
| 3017 #endif | |
| 3018 Token::tochars[TOKlt] = "<"; | |
| 3019 Token::tochars[TOKgt] = ">"; | |
| 3020 Token::tochars[TOKle] = "<="; | |
| 3021 Token::tochars[TOKge] = ">="; | |
| 3022 Token::tochars[TOKequal] = "=="; | |
| 3023 Token::tochars[TOKnotequal] = "!="; | |
| 3024 Token::tochars[TOKnotidentity] = "!is"; | |
| 3025 Token::tochars[TOKtobool] = "!!"; | |
| 3026 | |
| 3027 Token::tochars[TOKunord] = "!<>="; | |
| 3028 Token::tochars[TOKue] = "!<>"; | |
| 3029 Token::tochars[TOKlg] = "<>"; | |
| 3030 Token::tochars[TOKleg] = "<>="; | |
| 3031 Token::tochars[TOKule] = "!>"; | |
| 3032 Token::tochars[TOKul] = "!>="; | |
| 3033 Token::tochars[TOKuge] = "!<"; | |
| 3034 Token::tochars[TOKug] = "!<="; | |
| 3035 | |
| 3036 Token::tochars[TOKnot] = "!"; | |
| 3037 Token::tochars[TOKtobool] = "!!"; | |
| 3038 Token::tochars[TOKshl] = "<<"; | |
| 3039 Token::tochars[TOKshr] = ">>"; | |
| 3040 Token::tochars[TOKushr] = ">>>"; | |
| 3041 Token::tochars[TOKadd] = "+"; | |
| 3042 Token::tochars[TOKmin] = "-"; | |
| 3043 Token::tochars[TOKmul] = "*"; | |
| 3044 Token::tochars[TOKdiv] = "/"; | |
| 3045 Token::tochars[TOKmod] = "%"; | |
| 3046 Token::tochars[TOKslice] = ".."; | |
| 3047 Token::tochars[TOKdotdotdot] = "..."; | |
| 3048 Token::tochars[TOKand] = "&"; | |
| 3049 Token::tochars[TOKandand] = "&&"; | |
| 3050 Token::tochars[TOKor] = "|"; | |
| 3051 Token::tochars[TOKoror] = "||"; | |
| 3052 Token::tochars[TOKarray] = "[]"; | |
| 3053 Token::tochars[TOKindex] = "[i]"; | |
| 3054 Token::tochars[TOKaddress] = "&"; | |
| 3055 Token::tochars[TOKstar] = "*"; | |
| 3056 Token::tochars[TOKtilde] = "~"; | |
| 3057 Token::tochars[TOKdollar] = "$"; | |
| 3058 Token::tochars[TOKcast] = "cast"; | |
| 3059 Token::tochars[TOKplusplus] = "++"; | |
| 3060 Token::tochars[TOKminusminus] = "--"; | |
| 3061 Token::tochars[TOKtype] = "type"; | |
| 3062 Token::tochars[TOKquestion] = "?"; | |
| 3063 Token::tochars[TOKneg] = "-"; | |
| 3064 Token::tochars[TOKuadd] = "+"; | |
| 3065 Token::tochars[TOKvar] = "var"; | |
| 3066 Token::tochars[TOKaddass] = "+="; | |
| 3067 Token::tochars[TOKminass] = "-="; | |
| 3068 Token::tochars[TOKmulass] = "*="; | |
| 3069 Token::tochars[TOKdivass] = "/="; | |
| 3070 Token::tochars[TOKmodass] = "%="; | |
| 3071 Token::tochars[TOKshlass] = "<<="; | |
| 3072 Token::tochars[TOKshrass] = ">>="; | |
| 3073 Token::tochars[TOKushrass] = ">>>="; | |
| 3074 Token::tochars[TOKandass] = "&="; | |
| 3075 Token::tochars[TOKorass] = "|="; | |
| 3076 Token::tochars[TOKcatass] = "~="; | |
| 3077 Token::tochars[TOKcat] = "~"; | |
| 3078 Token::tochars[TOKcall] = "call"; | |
| 3079 Token::tochars[TOKidentity] = "is"; | |
| 3080 Token::tochars[TOKnotidentity] = "!is"; | |
| 3081 | |
| 3082 Token::tochars[TOKorass] = "|="; | |
| 3083 Token::tochars[TOKidentifier] = "identifier"; | |
| 3084 | |
| 3085 // For debugging | |
| 1587 | 3086 Token::tochars[TOKerror] = "error"; |
| 159 | 3087 Token::tochars[TOKdotexp] = "dotexp"; |
| 3088 Token::tochars[TOKdotti] = "dotti"; | |
| 3089 Token::tochars[TOKdotvar] = "dotvar"; | |
| 3090 Token::tochars[TOKdottype] = "dottype"; | |
| 3091 Token::tochars[TOKsymoff] = "symoff"; | |
| 3092 Token::tochars[TOKarraylength] = "arraylength"; | |
| 3093 Token::tochars[TOKarrayliteral] = "arrayliteral"; | |
| 3094 Token::tochars[TOKassocarrayliteral] = "assocarrayliteral"; | |
| 3095 Token::tochars[TOKstructliteral] = "structliteral"; | |
| 3096 Token::tochars[TOKstring] = "string"; | |
| 3097 Token::tochars[TOKdsymbol] = "symbol"; | |
| 3098 Token::tochars[TOKtuple] = "tuple"; | |
| 3099 Token::tochars[TOKdeclaration] = "declaration"; | |
| 3100 Token::tochars[TOKdottd] = "dottd"; | |
| 3101 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
|
3102 Token::tochars[TOKon_scope_success] = "scope(success)"; |
|
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
3103 Token::tochars[TOKon_scope_failure] = "scope(failure)"; |
| 159 | 3104 } |