Mercurial > projects > ldc
annotate dmd/lexer.c @ 1603:eae495e6ae8d
Merge DMD r248: implement Denis Koroskin's macro suggestion
---
dmd/lexer.c | 2 +-
dmd/lexer.h | 8 ++++----
dmd/mars.c | 2 +-
dmd/parse.c | 12 ++++++------
4 files changed, 12 insertions(+), 12 deletions(-)
| author | Leandro Lucarella <llucax@gmail.com> |
|---|---|
| date | Wed, 06 Jan 2010 15:18:19 -0300 |
| parents | def7a1d494fd |
| children | dbf7b54f542f |
| 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; | |
| 637 StringValue *sv; | |
| 638 Identifier *id; | |
| 639 | |
| 640 do | |
| 641 { | |
| 642 c = *++p; | |
| 643 } while (isidchar(c) || (c & 0x80 && isUniAlpha(decodeUTF()))); | |
| 644 sv = stringtable.update((char *)t->ptr, p - t->ptr); | |
| 645 id = (Identifier *) sv->ptrvalue; | |
| 646 if (!id) | |
| 647 { id = new Identifier(sv->lstring.string,TOKidentifier); | |
| 648 sv->ptrvalue = id; | |
| 649 } | |
| 650 t->ident = id; | |
| 651 t->value = (enum TOK) id->value; | |
| 652 anyToken = 1; | |
| 653 if (*t->ptr == '_') // if special identifier token | |
| 654 { | |
| 655 static char date[11+1]; | |
| 656 static char time[8+1]; | |
| 657 static char timestamp[24+1]; | |
| 658 | |
| 659 if (!date[0]) // lazy evaluation | |
| 660 { time_t t; | |
| 661 char *p; | |
| 662 | |
| 663 ::time(&t); | |
| 664 p = ctime(&t); | |
| 665 assert(p); | |
| 666 sprintf(date, "%.6s %.4s", p + 4, p + 20); | |
| 667 sprintf(time, "%.8s", p + 11); | |
| 668 sprintf(timestamp, "%.24s", p); | |
| 669 } | |
| 670 | |
| 336 | 671 #if DMDV1 |
| 159 | 672 if (mod && id == Id::FILE) |
| 673 { | |
| 674 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
|
675 goto Lstr; |
| 159 | 676 } |
| 677 else if (mod && id == Id::LINE) | |
| 678 { | |
| 679 t->value = TOKint64v; | |
| 680 t->uns64value = loc.linnum; | |
| 681 } | |
| 336 | 682 else |
| 683 #endif | |
| 684 if (id == Id::DATE) | |
| 159 | 685 { |
| 686 t->ustring = (unsigned char *)date; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
687 goto Lstr; |
| 159 | 688 } |
| 689 else if (id == Id::TIME) | |
| 690 { | |
| 691 t->ustring = (unsigned char *)time; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
692 goto Lstr; |
| 159 | 693 } |
| 694 else if (id == Id::VENDOR) | |
| 695 { | |
|
664
eef8ac26c66c
Some missed LLVMDC -> LDC.
Christian Kamm <kamm incasoftware de>
parents:
658
diff
changeset
|
696 t->ustring = (unsigned char *)"LDC"; |
|
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::TIMESTAMP) | |
| 700 { | |
| 701 t->ustring = (unsigned char *)timestamp; | |
|
1195
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
702 Lstr: |
| 159 | 703 t->value = TOKstring; |
| 704 Llen: | |
| 705 t->postfix = 0; | |
| 706 t->len = strlen((char *)t->ustring); | |
| 707 } | |
| 708 else if (id == Id::VERSIONX) | |
| 709 { unsigned major = 0; | |
| 710 unsigned minor = 0; | |
| 711 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
712 for (const char *p = global.version + 1; 1; p++) |
| 159 | 713 { |
| 714 char c = *p; | |
| 715 if (isdigit(c)) | |
| 716 minor = minor * 10 + c - '0'; | |
| 717 else if (c == '.') | |
| 718 { major = minor; | |
| 719 minor = 0; | |
| 720 } | |
| 721 else | |
| 722 break; | |
| 723 } | |
| 724 t->value = TOKint64v; | |
| 725 t->uns64value = major * 1000 + minor; | |
| 726 } | |
| 336 | 727 #if DMDV2 |
| 159 | 728 else if (id == Id::EOFX) |
| 729 { | |
| 730 t->value = TOKeof; | |
| 731 // Advance scanner to end of file | |
| 732 while (!(*p == 0 || *p == 0x1A)) | |
| 733 p++; | |
| 734 } | |
| 735 #endif | |
| 736 } | |
| 737 //printf("t->value = %d\n",t->value); | |
| 738 return; | |
| 739 } | |
| 740 | |
| 741 case '/': | |
| 742 p++; | |
| 743 switch (*p) | |
| 744 { | |
| 745 case '=': | |
| 746 p++; | |
| 747 t->value = TOKdivass; | |
| 748 return; | |
| 749 | |
| 750 case '*': | |
| 751 p++; | |
| 752 linnum = loc.linnum; | |
| 753 while (1) | |
| 754 { | |
| 755 while (1) | |
| 756 { unsigned char c = *p; | |
| 757 switch (c) | |
| 758 { | |
| 759 case '/': | |
| 760 break; | |
| 761 | |
| 762 case '\n': | |
| 763 loc.linnum++; | |
| 764 p++; | |
| 765 continue; | |
| 766 | |
| 767 case '\r': | |
| 768 p++; | |
| 769 if (*p != '\n') | |
| 770 loc.linnum++; | |
| 771 continue; | |
| 772 | |
| 773 case 0: | |
| 774 case 0x1A: | |
| 775 error("unterminated /* */ comment"); | |
| 776 p = end; | |
| 777 t->value = TOKeof; | |
| 778 return; | |
| 779 | |
| 780 default: | |
| 781 if (c & 0x80) | |
| 782 { unsigned u = decodeUTF(); | |
| 783 if (u == PS || u == LS) | |
| 784 loc.linnum++; | |
| 785 } | |
| 786 p++; | |
| 787 continue; | |
| 788 } | |
| 789 break; | |
| 790 } | |
| 791 p++; | |
| 792 if (p[-2] == '*' && p - 3 != t->ptr) | |
| 793 break; | |
| 794 } | |
| 795 if (commentToken) | |
| 796 { | |
| 797 t->value = TOKcomment; | |
| 798 return; | |
| 799 } | |
| 800 else if (doDocComment && t->ptr[2] == '*' && p - 4 != t->ptr) | |
| 801 { // if /** but not /**/ | |
| 802 getDocComment(t, lastLine == linnum); | |
| 803 } | |
| 804 continue; | |
| 805 | |
| 806 case '/': // do // style comments | |
| 807 linnum = loc.linnum; | |
| 808 while (1) | |
| 809 { unsigned char c = *++p; | |
| 810 switch (c) | |
| 811 { | |
| 812 case '\n': | |
| 813 break; | |
| 814 | |
| 815 case '\r': | |
| 816 if (p[1] == '\n') | |
| 817 p++; | |
| 818 break; | |
| 819 | |
| 820 case 0: | |
| 821 case 0x1A: | |
| 822 if (commentToken) | |
| 823 { | |
| 824 p = end; | |
| 825 t->value = TOKcomment; | |
| 826 return; | |
| 827 } | |
| 828 if (doDocComment && t->ptr[2] == '/') | |
| 829 getDocComment(t, lastLine == linnum); | |
| 830 p = end; | |
| 831 t->value = TOKeof; | |
| 832 return; | |
| 833 | |
| 834 default: | |
| 835 if (c & 0x80) | |
| 836 { unsigned u = decodeUTF(); | |
| 837 if (u == PS || u == LS) | |
| 838 break; | |
| 839 } | |
| 840 continue; | |
| 841 } | |
| 842 break; | |
| 843 } | |
| 844 | |
| 845 if (commentToken) | |
| 846 { | |
| 847 p++; | |
| 848 loc.linnum++; | |
| 849 t->value = TOKcomment; | |
| 850 return; | |
| 851 } | |
| 852 if (doDocComment && t->ptr[2] == '/') | |
| 853 getDocComment(t, lastLine == linnum); | |
| 854 | |
| 855 p++; | |
| 856 loc.linnum++; | |
| 857 continue; | |
| 858 | |
| 859 case '+': | |
| 860 { int nest; | |
| 861 | |
| 862 linnum = loc.linnum; | |
| 863 p++; | |
| 864 nest = 1; | |
| 865 while (1) | |
| 866 { unsigned char c = *p; | |
| 867 switch (c) | |
| 868 { | |
| 869 case '/': | |
| 870 p++; | |
| 871 if (*p == '+') | |
| 872 { | |
| 873 p++; | |
| 874 nest++; | |
| 875 } | |
| 876 continue; | |
| 877 | |
| 878 case '+': | |
| 879 p++; | |
| 880 if (*p == '/') | |
| 881 { | |
| 882 p++; | |
| 883 if (--nest == 0) | |
| 884 break; | |
| 885 } | |
| 886 continue; | |
| 887 | |
| 888 case '\r': | |
| 889 p++; | |
| 890 if (*p != '\n') | |
| 891 loc.linnum++; | |
| 892 continue; | |
| 893 | |
| 894 case '\n': | |
| 895 loc.linnum++; | |
| 896 p++; | |
| 897 continue; | |
| 898 | |
| 899 case 0: | |
| 900 case 0x1A: | |
| 901 error("unterminated /+ +/ comment"); | |
| 902 p = end; | |
| 903 t->value = TOKeof; | |
| 904 return; | |
| 905 | |
| 906 default: | |
| 907 if (c & 0x80) | |
| 908 { unsigned u = decodeUTF(); | |
| 909 if (u == PS || u == LS) | |
| 910 loc.linnum++; | |
| 911 } | |
| 912 p++; | |
| 913 continue; | |
| 914 } | |
| 915 break; | |
| 916 } | |
| 917 if (commentToken) | |
| 918 { | |
| 919 t->value = TOKcomment; | |
| 920 return; | |
| 921 } | |
| 922 if (doDocComment && t->ptr[2] == '+' && p - 4 != t->ptr) | |
| 923 { // if /++ but not /++/ | |
| 924 getDocComment(t, lastLine == linnum); | |
| 925 } | |
| 926 continue; | |
| 927 } | |
| 928 } | |
| 929 t->value = TOKdiv; | |
| 930 return; | |
| 931 | |
| 932 case '.': | |
| 933 p++; | |
| 934 if (isdigit(*p)) | |
| 935 { /* Note that we don't allow ._1 and ._ as being | |
| 936 * valid floating point numbers. | |
| 937 */ | |
| 938 p--; | |
| 939 t->value = inreal(t); | |
| 940 } | |
| 941 else if (p[0] == '.') | |
| 942 { | |
| 943 if (p[1] == '.') | |
| 944 { p += 2; | |
| 945 t->value = TOKdotdotdot; | |
| 946 } | |
| 947 else | |
| 948 { p++; | |
| 949 t->value = TOKslice; | |
| 950 } | |
| 951 } | |
| 952 else | |
| 953 t->value = TOKdot; | |
| 954 return; | |
| 955 | |
| 956 case '&': | |
| 957 p++; | |
| 958 if (*p == '=') | |
| 959 { p++; | |
| 960 t->value = TOKandass; | |
| 961 } | |
| 962 else if (*p == '&') | |
| 963 { p++; | |
| 964 t->value = TOKandand; | |
| 965 } | |
| 966 else | |
| 967 t->value = TOKand; | |
| 968 return; | |
| 969 | |
| 970 case '|': | |
| 971 p++; | |
| 972 if (*p == '=') | |
| 973 { p++; | |
| 974 t->value = TOKorass; | |
| 975 } | |
| 976 else if (*p == '|') | |
| 977 { p++; | |
| 978 t->value = TOKoror; | |
| 979 } | |
| 980 else | |
| 981 t->value = TOKor; | |
| 982 return; | |
| 983 | |
| 984 case '-': | |
| 985 p++; | |
| 986 if (*p == '=') | |
| 987 { p++; | |
| 988 t->value = TOKminass; | |
| 989 } | |
| 990 #if 0 | |
| 991 else if (*p == '>') | |
| 992 { p++; | |
| 993 t->value = TOKarrow; | |
| 994 } | |
| 995 #endif | |
| 996 else if (*p == '-') | |
| 997 { p++; | |
| 998 t->value = TOKminusminus; | |
| 999 } | |
| 1000 else | |
| 1001 t->value = TOKmin; | |
| 1002 return; | |
| 1003 | |
| 1004 case '+': | |
| 1005 p++; | |
| 1006 if (*p == '=') | |
| 1007 { p++; | |
| 1008 t->value = TOKaddass; | |
| 1009 } | |
| 1010 else if (*p == '+') | |
| 1011 { p++; | |
| 1012 t->value = TOKplusplus; | |
| 1013 } | |
| 1014 else | |
| 1015 t->value = TOKadd; | |
| 1016 return; | |
| 1017 | |
| 1018 case '<': | |
| 1019 p++; | |
| 1020 if (*p == '=') | |
| 1021 { p++; | |
| 1022 t->value = TOKle; // <= | |
| 1023 } | |
| 1024 else if (*p == '<') | |
| 1025 { p++; | |
| 1026 if (*p == '=') | |
| 1027 { p++; | |
| 1028 t->value = TOKshlass; // <<= | |
| 1029 } | |
| 1030 else | |
| 1031 t->value = TOKshl; // << | |
| 1032 } | |
| 1033 else if (*p == '>') | |
| 1034 { p++; | |
| 1035 if (*p == '=') | |
| 1036 { p++; | |
| 1037 t->value = TOKleg; // <>= | |
| 1038 } | |
| 1039 else | |
| 1040 t->value = TOKlg; // <> | |
| 1041 } | |
| 1042 else | |
| 1043 t->value = TOKlt; // < | |
| 1044 return; | |
| 1045 | |
| 1046 case '>': | |
| 1047 p++; | |
| 1048 if (*p == '=') | |
| 1049 { p++; | |
| 1050 t->value = TOKge; // >= | |
| 1051 } | |
| 1052 else if (*p == '>') | |
| 1053 { p++; | |
| 1054 if (*p == '=') | |
| 1055 { p++; | |
| 1056 t->value = TOKshrass; // >>= | |
| 1057 } | |
| 1058 else if (*p == '>') | |
| 1059 { p++; | |
| 1060 if (*p == '=') | |
| 1061 { p++; | |
| 1062 t->value = TOKushrass; // >>>= | |
| 1063 } | |
| 1064 else | |
| 1065 t->value = TOKushr; // >>> | |
| 1066 } | |
| 1067 else | |
| 1068 t->value = TOKshr; // >> | |
| 1069 } | |
| 1070 else | |
| 1071 t->value = TOKgt; // > | |
| 1072 return; | |
| 1073 | |
| 1074 case '!': | |
| 1075 p++; | |
| 1076 if (*p == '=') | |
| 1077 { p++; | |
| 1078 if (*p == '=' && global.params.Dversion == 1) | |
| 1079 { p++; | |
| 1080 t->value = TOKnotidentity; // !== | |
| 1081 } | |
| 1082 else | |
| 1083 t->value = TOKnotequal; // != | |
| 1084 } | |
| 1085 else if (*p == '<') | |
| 1086 { p++; | |
| 1087 if (*p == '>') | |
| 1088 { p++; | |
| 1089 if (*p == '=') | |
| 1090 { p++; | |
| 1091 t->value = TOKunord; // !<>= | |
| 1092 } | |
| 1093 else | |
| 1094 t->value = TOKue; // !<> | |
| 1095 } | |
| 1096 else if (*p == '=') | |
| 1097 { p++; | |
| 1098 t->value = TOKug; // !<= | |
| 1099 } | |
| 1100 else | |
| 1101 t->value = TOKuge; // !< | |
| 1102 } | |
| 1103 else if (*p == '>') | |
| 1104 { p++; | |
| 1105 if (*p == '=') | |
| 1106 { p++; | |
| 1107 t->value = TOKul; // !>= | |
| 1108 } | |
| 1109 else | |
| 1110 t->value = TOKule; // !> | |
| 1111 } | |
| 1112 else | |
| 1113 t->value = TOKnot; // ! | |
| 1114 return; | |
| 1115 | |
| 1116 case '=': | |
| 1117 p++; | |
| 1118 if (*p == '=') | |
| 1119 { p++; | |
| 1120 if (*p == '=' && global.params.Dversion == 1) | |
| 1121 { p++; | |
| 1122 t->value = TOKidentity; // === | |
| 1123 } | |
| 1124 else | |
| 1125 t->value = TOKequal; // == | |
| 1126 } | |
| 1127 else | |
| 1128 t->value = TOKassign; // = | |
| 1129 return; | |
| 1130 | |
| 1131 case '~': | |
| 1132 p++; | |
| 1133 if (*p == '=') | |
| 1134 { p++; | |
| 1135 t->value = TOKcatass; // ~= | |
| 1136 } | |
| 1137 else | |
| 1138 t->value = TOKtilde; // ~ | |
| 1139 return; | |
| 1140 | |
| 1141 #define SINGLE(c,tok) case c: p++; t->value = tok; return; | |
| 1142 | |
| 1143 SINGLE('(', TOKlparen) | |
| 1144 SINGLE(')', TOKrparen) | |
| 1145 SINGLE('[', TOKlbracket) | |
| 1146 SINGLE(']', TOKrbracket) | |
| 1147 SINGLE('{', TOKlcurly) | |
| 1148 SINGLE('}', TOKrcurly) | |
| 1149 SINGLE('?', TOKquestion) | |
| 1150 SINGLE(',', TOKcomma) | |
| 1151 SINGLE(';', TOKsemicolon) | |
| 1152 SINGLE(':', TOKcolon) | |
| 1153 SINGLE('$', TOKdollar) | |
| 1154 | |
| 1155 #undef SINGLE | |
| 1156 | |
| 1157 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1158 case c1: \ | |
| 1159 p++; \ | |
| 1160 if (*p == c2) \ | |
| 1161 { p++; \ | |
| 1162 t->value = tok2; \ | |
| 1163 } \ | |
| 1164 else \ | |
| 1165 t->value = tok1; \ | |
| 1166 return; | |
| 1167 | |
| 1168 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1169 DOUBLE('%', TOKmod, '=', TOKmodass) | |
| 1170 DOUBLE('^', TOKxor, '=', TOKxorass) | |
| 1171 | |
| 1172 #undef DOUBLE | |
| 1173 | |
| 1174 case '#': | |
| 1175 p++; | |
| 1176 pragma(); | |
| 1177 continue; | |
| 1178 | |
| 1179 default: | |
| 1180 { unsigned char c = *p; | |
| 1181 | |
| 1182 if (c & 0x80) | |
| 1183 { unsigned u = decodeUTF(); | |
| 1184 | |
| 1185 // Check for start of unicode identifier | |
| 1186 if (isUniAlpha(u)) | |
| 1187 goto case_ident; | |
| 1188 | |
| 1189 if (u == PS || u == LS) | |
| 1190 { | |
| 1191 loc.linnum++; | |
| 1192 p++; | |
| 1193 continue; | |
| 1194 } | |
| 1195 } | |
| 1196 if (isprint(c)) | |
| 1197 error("unsupported char '%c'", c); | |
| 1198 else | |
| 1199 error("unsupported char 0x%02x", c); | |
| 1200 p++; | |
| 1201 continue; | |
| 1202 } | |
| 1203 } | |
| 1204 } | |
| 1205 } | |
| 1206 | |
| 1207 /******************************************* | |
| 1208 * Parse escape sequence. | |
| 1209 */ | |
| 1210 | |
| 1211 unsigned Lexer::escapeSequence() | |
| 1212 { unsigned c; | |
| 1213 int n; | |
| 1214 int ndigits; | |
| 1215 | |
| 1216 c = *p; | |
| 1217 switch (c) | |
| 1218 { | |
| 1219 case '\'': | |
| 1220 case '"': | |
| 1221 case '?': | |
| 1222 case '\\': | |
| 1223 Lconsume: | |
| 1224 p++; | |
| 1225 break; | |
| 1226 | |
| 1227 case 'a': c = 7; goto Lconsume; | |
| 1228 case 'b': c = 8; goto Lconsume; | |
| 1229 case 'f': c = 12; goto Lconsume; | |
| 1230 case 'n': c = 10; goto Lconsume; | |
| 1231 case 'r': c = 13; goto Lconsume; | |
| 1232 case 't': c = 9; goto Lconsume; | |
| 1233 case 'v': c = 11; goto Lconsume; | |
| 1234 | |
| 1235 case 'u': | |
| 1236 ndigits = 4; | |
| 1237 goto Lhex; | |
| 1238 case 'U': | |
| 1239 ndigits = 8; | |
| 1240 goto Lhex; | |
| 1241 case 'x': | |
| 1242 ndigits = 2; | |
| 1243 Lhex: | |
| 1244 p++; | |
| 1245 c = *p; | |
| 1246 if (ishex(c)) | |
| 1247 { unsigned v; | |
| 1248 | |
| 1249 n = 0; | |
| 1250 v = 0; | |
| 1251 while (1) | |
| 1252 { | |
| 1253 if (isdigit(c)) | |
| 1254 c -= '0'; | |
| 1255 else if (islower(c)) | |
| 1256 c -= 'a' - 10; | |
| 1257 else | |
| 1258 c -= 'A' - 10; | |
| 1259 v = v * 16 + c; | |
| 1260 c = *++p; | |
| 1261 if (++n == ndigits) | |
| 1262 break; | |
| 1263 if (!ishex(c)) | |
| 1264 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1265 break; | |
| 1266 } | |
| 1267 } | |
| 1268 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1587 | 1269 { error("invalid UTF character \\U%08x", v); |
| 1270 v = '?'; // recover with valid UTF character | |
| 1271 } | |
| 159 | 1272 c = v; |
| 1273 } | |
| 1274 else | |
| 1275 error("undefined escape hex sequence \\%c\n",c); | |
| 1276 break; | |
| 1277 | |
| 1278 case '&': // named character entity | |
| 1279 for (unsigned char *idstart = ++p; 1; p++) | |
| 1280 { | |
| 1281 switch (*p) | |
| 1282 { | |
| 1283 case ';': | |
| 1284 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1285 if (c == ~0) | |
| 1286 { error("unnamed character entity &%.*s;", (int)(p - idstart), idstart); | |
| 1287 c = ' '; | |
| 1288 } | |
| 1289 p++; | |
| 1290 break; | |
| 1291 | |
| 1292 default: | |
| 1293 if (isalpha(*p) || | |
| 1294 (p != idstart + 1 && isdigit(*p))) | |
| 1295 continue; | |
| 1296 error("unterminated named entity"); | |
| 1297 break; | |
| 1298 } | |
| 1299 break; | |
| 1300 } | |
| 1301 break; | |
| 1302 | |
| 1303 case 0: | |
| 1304 case 0x1A: // end of file | |
| 1305 c = '\\'; | |
| 1306 break; | |
| 1307 | |
| 1308 default: | |
| 1309 if (isoctal(c)) | |
| 1310 { unsigned v; | |
| 1311 | |
| 1312 n = 0; | |
| 1313 v = 0; | |
| 1314 do | |
| 1315 { | |
| 1316 v = v * 8 + (c - '0'); | |
| 1317 c = *++p; | |
| 1318 } while (++n < 3 && isoctal(c)); | |
| 1319 c = v; | |
| 1320 if (c > 0xFF) | |
| 1321 error("0%03o is larger than a byte", c); | |
| 1322 } | |
| 1323 else | |
| 1324 error("undefined escape sequence \\%c\n",c); | |
| 1325 break; | |
| 1326 } | |
| 1327 return c; | |
| 1328 } | |
| 1329 | |
| 1330 /************************************** | |
| 1331 */ | |
| 1332 | |
| 1333 TOK Lexer::wysiwygStringConstant(Token *t, int tc) | |
| 1334 { unsigned c; | |
| 1335 Loc start = loc; | |
| 1336 | |
| 1337 p++; | |
| 1338 stringbuffer.reset(); | |
| 1339 while (1) | |
| 1340 { | |
| 1341 c = *p++; | |
| 1342 switch (c) | |
| 1343 { | |
| 1344 case '\n': | |
| 1345 loc.linnum++; | |
| 1346 break; | |
| 1347 | |
| 1348 case '\r': | |
| 1349 if (*p == '\n') | |
| 1350 continue; // ignore | |
| 1351 c = '\n'; // treat EndOfLine as \n character | |
| 1352 loc.linnum++; | |
| 1353 break; | |
| 1354 | |
| 1355 case 0: | |
| 1356 case 0x1A: | |
| 1357 error("unterminated string constant starting at %s", start.toChars()); | |
| 1358 t->ustring = (unsigned char *)""; | |
| 1359 t->len = 0; | |
| 1360 t->postfix = 0; | |
| 1361 return TOKstring; | |
| 1362 | |
| 1363 case '"': | |
| 1364 case '`': | |
| 1365 if (c == tc) | |
| 1366 { | |
| 1367 t->len = stringbuffer.offset; | |
| 1368 stringbuffer.writeByte(0); | |
| 1369 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1370 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1371 stringPostfix(t); | |
| 1372 return TOKstring; | |
| 1373 } | |
| 1374 break; | |
| 1375 | |
| 1376 default: | |
| 1377 if (c & 0x80) | |
| 1378 { p--; | |
| 1379 unsigned u = decodeUTF(); | |
| 1380 p++; | |
| 1381 if (u == PS || u == LS) | |
| 1382 loc.linnum++; | |
| 1383 stringbuffer.writeUTF8(u); | |
| 1384 continue; | |
| 1385 } | |
| 1386 break; | |
| 1387 } | |
| 1388 stringbuffer.writeByte(c); | |
| 1389 } | |
| 1390 } | |
| 1391 | |
| 1392 /************************************** | |
| 1393 * Lex hex strings: | |
| 1394 * x"0A ae 34FE BD" | |
| 1395 */ | |
| 1396 | |
| 1397 TOK Lexer::hexStringConstant(Token *t) | |
| 1398 { unsigned c; | |
| 1399 Loc start = loc; | |
| 1400 unsigned n = 0; | |
| 1401 unsigned v; | |
| 1402 | |
| 1403 p++; | |
| 1404 stringbuffer.reset(); | |
| 1405 while (1) | |
| 1406 { | |
| 1407 c = *p++; | |
| 1408 switch (c) | |
| 1409 { | |
| 1410 case ' ': | |
| 1411 case '\t': | |
| 1412 case '\v': | |
| 1413 case '\f': | |
| 1414 continue; // skip white space | |
| 1415 | |
| 1416 case '\r': | |
| 1417 if (*p == '\n') | |
| 1418 continue; // ignore | |
| 1419 // Treat isolated '\r' as if it were a '\n' | |
| 1420 case '\n': | |
| 1421 loc.linnum++; | |
| 1422 continue; | |
| 1423 | |
| 1424 case 0: | |
| 1425 case 0x1A: | |
| 1426 error("unterminated string constant starting at %s", start.toChars()); | |
| 1427 t->ustring = (unsigned char *)""; | |
| 1428 t->len = 0; | |
| 1429 t->postfix = 0; | |
| 1430 return TOKstring; | |
| 1431 | |
| 1432 case '"': | |
| 1433 if (n & 1) | |
| 1434 { error("odd number (%d) of hex characters in hex string", n); | |
| 1435 stringbuffer.writeByte(v); | |
| 1436 } | |
| 1437 t->len = stringbuffer.offset; | |
| 1438 stringbuffer.writeByte(0); | |
| 1439 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1440 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1441 stringPostfix(t); | |
| 1442 return TOKstring; | |
| 1443 | |
| 1444 default: | |
| 1445 if (c >= '0' && c <= '9') | |
| 1446 c -= '0'; | |
| 1447 else if (c >= 'a' && c <= 'f') | |
| 1448 c -= 'a' - 10; | |
| 1449 else if (c >= 'A' && c <= 'F') | |
| 1450 c -= 'A' - 10; | |
| 1451 else if (c & 0x80) | |
| 1452 { p--; | |
| 1453 unsigned u = decodeUTF(); | |
| 1454 p++; | |
| 1455 if (u == PS || u == LS) | |
| 1456 loc.linnum++; | |
| 1457 else | |
| 1458 error("non-hex character \\u%x", u); | |
| 1459 } | |
| 1460 else | |
| 1461 error("non-hex character '%c'", c); | |
| 1462 if (n & 1) | |
| 1463 { v = (v << 4) | c; | |
| 1464 stringbuffer.writeByte(v); | |
| 1465 } | |
| 1466 else | |
| 1467 v = c; | |
| 1468 n++; | |
| 1469 break; | |
| 1470 } | |
| 1471 } | |
| 1472 } | |
| 1473 | |
| 1474 | |
| 336 | 1475 #if DMDV2 |
| 159 | 1476 /************************************** |
| 1477 * Lex delimited strings: | |
| 1478 * q"(foo(xxx))" // "foo(xxx)" | |
| 1479 * q"[foo(]" // "foo(" | |
| 1480 * q"/foo]/" // "foo]" | |
| 1481 * q"HERE | |
| 1482 * foo | |
| 1483 * HERE" // "foo\n" | |
| 1484 * Input: | |
| 1485 * p is on the " | |
| 1486 */ | |
| 1487 | |
| 1488 TOK Lexer::delimitedStringConstant(Token *t) | |
| 1489 { unsigned c; | |
| 1490 Loc start = loc; | |
| 1491 unsigned delimleft = 0; | |
| 1492 unsigned delimright = 0; | |
| 1493 unsigned nest = 1; | |
| 1494 unsigned nestcount; | |
| 1495 Identifier *hereid = NULL; | |
| 1496 unsigned blankrol = 0; | |
| 1497 unsigned startline = 0; | |
| 1498 | |
| 1499 p++; | |
| 1500 stringbuffer.reset(); | |
| 1501 while (1) | |
| 1502 { | |
| 1503 c = *p++; | |
| 1504 //printf("c = '%c'\n", c); | |
| 1505 switch (c) | |
| 1506 { | |
| 1507 case '\n': | |
| 1508 Lnextline: | |
| 1509 loc.linnum++; | |
| 1510 startline = 1; | |
| 1511 if (blankrol) | |
| 1512 { blankrol = 0; | |
| 1513 continue; | |
| 1514 } | |
| 1515 if (hereid) | |
| 1516 { | |
| 1517 stringbuffer.writeUTF8(c); | |
| 1518 continue; | |
| 1519 } | |
| 1520 break; | |
| 1521 | |
| 1522 case '\r': | |
| 1523 if (*p == '\n') | |
| 1524 continue; // ignore | |
| 1525 c = '\n'; // treat EndOfLine as \n character | |
| 1526 goto Lnextline; | |
| 1527 | |
| 1528 case 0: | |
| 1529 case 0x1A: | |
| 1530 goto Lerror; | |
| 1531 | |
| 1532 default: | |
| 1533 if (c & 0x80) | |
| 1534 { p--; | |
| 1535 c = decodeUTF(); | |
| 1536 p++; | |
| 1537 if (c == PS || c == LS) | |
| 1538 goto Lnextline; | |
| 1539 } | |
| 1540 break; | |
| 1541 } | |
| 1542 if (delimleft == 0) | |
| 1543 { delimleft = c; | |
| 1544 nest = 1; | |
| 1545 nestcount = 1; | |
| 1546 if (c == '(') | |
| 1547 delimright = ')'; | |
| 1548 else if (c == '{') | |
| 1549 delimright = '}'; | |
| 1550 else if (c == '[') | |
| 1551 delimright = ']'; | |
| 1552 else if (c == '<') | |
| 1553 delimright = '>'; | |
| 1554 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1555 { // Start of identifier; must be a heredoc | |
| 1556 Token t; | |
| 1557 p--; | |
| 1558 scan(&t); // read in heredoc identifier | |
| 1559 if (t.value != TOKidentifier) | |
| 1560 { error("identifier expected for heredoc, not %s", t.toChars()); | |
| 1561 delimright = c; | |
| 1562 } | |
| 1563 else | |
| 1564 { hereid = t.ident; | |
| 1565 //printf("hereid = '%s'\n", hereid->toChars()); | |
| 1566 blankrol = 1; | |
| 1567 } | |
| 1568 nest = 0; | |
| 1569 } | |
| 1570 else | |
| 1571 { delimright = c; | |
| 1572 nest = 0; | |
| 1573 } | |
| 1574 } | |
| 1575 else | |
| 1576 { | |
| 1577 if (blankrol) | |
| 1578 { error("heredoc rest of line should be blank"); | |
| 1579 blankrol = 0; | |
| 1580 continue; | |
| 1581 } | |
| 1582 if (nest == 1) | |
| 1583 { | |
| 1584 if (c == delimleft) | |
| 1585 nestcount++; | |
| 1586 else if (c == delimright) | |
| 1587 { nestcount--; | |
| 1588 if (nestcount == 0) | |
| 1589 goto Ldone; | |
| 1590 } | |
| 1591 } | |
| 1592 else if (c == delimright) | |
| 1593 goto Ldone; | |
| 1594 if (startline && isalpha(c)) | |
| 1595 { Token t; | |
| 1596 unsigned char *psave = p; | |
| 1597 p--; | |
| 1598 scan(&t); // read in possible heredoc identifier | |
| 1599 //printf("endid = '%s'\n", t.ident->toChars()); | |
| 1600 if (t.value == TOKidentifier && t.ident->equals(hereid)) | |
| 1601 { /* should check that rest of line is blank | |
| 1602 */ | |
| 1603 goto Ldone; | |
| 1604 } | |
| 1605 p = psave; | |
| 1606 } | |
| 1607 stringbuffer.writeUTF8(c); | |
| 1608 startline = 0; | |
| 1609 } | |
| 1610 } | |
| 1611 | |
| 1612 Ldone: | |
| 1613 if (*p == '"') | |
| 1614 p++; | |
| 1615 else | |
| 1616 error("delimited string must end in %c\"", delimright); | |
| 1617 t->len = stringbuffer.offset; | |
| 1618 stringbuffer.writeByte(0); | |
| 1619 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1620 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1621 stringPostfix(t); | |
| 1622 return TOKstring; | |
| 1623 | |
| 1624 Lerror: | |
| 1625 error("unterminated string constant starting at %s", start.toChars()); | |
| 1626 t->ustring = (unsigned char *)""; | |
| 1627 t->len = 0; | |
| 1628 t->postfix = 0; | |
| 1629 return TOKstring; | |
| 1630 } | |
| 1631 | |
| 1632 /************************************** | |
| 1633 * Lex delimited strings: | |
| 1634 * q{ foo(xxx) } // " foo(xxx) " | |
| 1635 * q{foo(} // "foo(" | |
| 1636 * q{{foo}"}"} // "{foo}"}"" | |
| 1637 * Input: | |
| 1638 * p is on the q | |
| 1639 */ | |
| 1640 | |
| 1641 TOK Lexer::tokenStringConstant(Token *t) | |
| 1642 { | |
| 1643 unsigned nest = 1; | |
| 1644 Loc start = loc; | |
| 1645 unsigned char *pstart = ++p; | |
| 1646 | |
| 1647 while (1) | |
| 1648 { Token tok; | |
| 1649 | |
| 1650 scan(&tok); | |
| 1651 switch (tok.value) | |
| 1652 { | |
| 1653 case TOKlcurly: | |
| 1654 nest++; | |
| 1655 continue; | |
| 1656 | |
| 1657 case TOKrcurly: | |
| 1658 if (--nest == 0) | |
| 1659 goto Ldone; | |
| 1660 continue; | |
| 1661 | |
| 1662 case TOKeof: | |
| 1663 goto Lerror; | |
| 1664 | |
| 1665 default: | |
| 1666 continue; | |
| 1667 } | |
| 1668 } | |
| 1669 | |
| 1670 Ldone: | |
| 1671 t->len = p - 1 - pstart; | |
| 1672 t->ustring = (unsigned char *)mem.malloc(t->len + 1); | |
| 1673 memcpy(t->ustring, pstart, t->len); | |
| 1674 t->ustring[t->len] = 0; | |
| 1675 stringPostfix(t); | |
| 1676 return TOKstring; | |
| 1677 | |
| 1678 Lerror: | |
| 1679 error("unterminated token string constant starting at %s", start.toChars()); | |
| 1680 t->ustring = (unsigned char *)""; | |
| 1681 t->len = 0; | |
| 1682 t->postfix = 0; | |
| 1683 return TOKstring; | |
| 1684 } | |
| 1685 | |
| 1686 #endif | |
| 1687 | |
| 1688 | |
| 1689 /************************************** | |
| 1690 */ | |
| 1691 | |
| 1692 TOK Lexer::escapeStringConstant(Token *t, int wide) | |
| 1693 { unsigned c; | |
| 1694 Loc start = loc; | |
| 1695 | |
| 1696 p++; | |
| 1697 stringbuffer.reset(); | |
| 1698 while (1) | |
| 1699 { | |
| 1700 c = *p++; | |
| 1701 switch (c) | |
| 1702 { | |
| 1703 case '\\': | |
| 1704 switch (*p) | |
| 1705 { | |
| 1706 case 'u': | |
| 1707 case 'U': | |
| 1708 case '&': | |
| 1709 c = escapeSequence(); | |
| 1710 stringbuffer.writeUTF8(c); | |
| 1711 continue; | |
| 1712 | |
| 1713 default: | |
| 1714 c = escapeSequence(); | |
| 1715 break; | |
| 1716 } | |
| 1717 break; | |
| 1718 | |
| 1719 case '\n': | |
| 1720 loc.linnum++; | |
| 1721 break; | |
| 1722 | |
| 1723 case '\r': | |
| 1724 if (*p == '\n') | |
| 1725 continue; // ignore | |
| 1726 c = '\n'; // treat EndOfLine as \n character | |
| 1727 loc.linnum++; | |
| 1728 break; | |
| 1729 | |
| 1730 case '"': | |
| 1731 t->len = stringbuffer.offset; | |
| 1732 stringbuffer.writeByte(0); | |
| 1733 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1734 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1735 stringPostfix(t); | |
| 1736 return TOKstring; | |
| 1737 | |
| 1738 case 0: | |
| 1739 case 0x1A: | |
| 1740 p--; | |
| 1741 error("unterminated string constant starting at %s", start.toChars()); | |
| 1742 t->ustring = (unsigned char *)""; | |
| 1743 t->len = 0; | |
| 1744 t->postfix = 0; | |
| 1745 return TOKstring; | |
| 1746 | |
| 1747 default: | |
| 1748 if (c & 0x80) | |
| 1749 { | |
| 1750 p--; | |
| 1751 c = decodeUTF(); | |
| 1752 if (c == LS || c == PS) | |
| 1753 { c = '\n'; | |
| 1754 loc.linnum++; | |
| 1755 } | |
| 1756 p++; | |
| 1757 stringbuffer.writeUTF8(c); | |
| 1758 continue; | |
| 1759 } | |
| 1760 break; | |
| 1761 } | |
| 1762 stringbuffer.writeByte(c); | |
| 1763 } | |
| 1764 } | |
| 1765 | |
| 1766 /************************************** | |
| 1767 */ | |
| 1768 | |
| 1769 TOK Lexer::charConstant(Token *t, int wide) | |
| 1770 { | |
| 1771 unsigned c; | |
| 1772 TOK tk = TOKcharv; | |
| 1773 | |
| 1774 //printf("Lexer::charConstant\n"); | |
| 1775 p++; | |
| 1776 c = *p++; | |
| 1777 switch (c) | |
| 1778 { | |
| 1779 case '\\': | |
| 1780 switch (*p) | |
| 1781 { | |
| 1782 case 'u': | |
| 1783 t->uns64value = escapeSequence(); | |
| 1784 tk = TOKwcharv; | |
| 1785 break; | |
| 1786 | |
| 1787 case 'U': | |
| 1788 case '&': | |
| 1789 t->uns64value = escapeSequence(); | |
| 1790 tk = TOKdcharv; | |
| 1791 break; | |
| 1792 | |
| 1793 default: | |
| 1794 t->uns64value = escapeSequence(); | |
| 1795 break; | |
| 1796 } | |
| 1797 break; | |
| 1798 | |
| 1799 case '\n': | |
| 1800 L1: | |
| 1801 loc.linnum++; | |
| 1802 case '\r': | |
| 1803 case 0: | |
| 1804 case 0x1A: | |
| 1805 case '\'': | |
| 1806 error("unterminated character constant"); | |
| 1807 return tk; | |
| 1808 | |
| 1809 default: | |
| 1810 if (c & 0x80) | |
| 1811 { | |
| 1812 p--; | |
| 1813 c = decodeUTF(); | |
| 1814 p++; | |
| 1815 if (c == LS || c == PS) | |
| 1816 goto L1; | |
| 1817 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 1818 tk = TOKwcharv; | |
| 1819 else | |
| 1820 tk = TOKdcharv; | |
| 1821 } | |
| 1822 t->uns64value = c; | |
| 1823 break; | |
| 1824 } | |
| 1825 | |
| 1826 if (*p != '\'') | |
| 1827 { error("unterminated character constant"); | |
| 1828 return tk; | |
| 1829 } | |
| 1830 p++; | |
| 1831 return tk; | |
| 1832 } | |
| 1833 | |
| 1834 /*************************************** | |
| 1835 * Get postfix of string literal. | |
| 1836 */ | |
| 1837 | |
| 1838 void Lexer::stringPostfix(Token *t) | |
| 1839 { | |
| 1840 switch (*p) | |
| 1841 { | |
| 1842 case 'c': | |
| 1843 case 'w': | |
| 1844 case 'd': | |
| 1845 t->postfix = *p; | |
| 1846 p++; | |
| 1847 break; | |
| 1848 | |
| 1849 default: | |
| 1850 t->postfix = 0; | |
| 1851 break; | |
| 1852 } | |
| 1853 } | |
| 1854 | |
| 1855 /*************************************** | |
| 1856 * Read \u or \U unicode sequence | |
| 1857 * Input: | |
| 1858 * u 'u' or 'U' | |
| 1859 */ | |
| 1860 | |
| 1861 #if 0 | |
| 1862 unsigned Lexer::wchar(unsigned u) | |
| 1863 { | |
| 1864 unsigned value; | |
| 1865 unsigned n; | |
| 1866 unsigned char c; | |
| 1867 unsigned nchars; | |
| 1868 | |
| 1869 nchars = (u == 'U') ? 8 : 4; | |
| 1870 value = 0; | |
| 1871 for (n = 0; 1; n++) | |
| 1872 { | |
| 1873 ++p; | |
| 1874 if (n == nchars) | |
| 1875 break; | |
| 1876 c = *p; | |
| 1877 if (!ishex(c)) | |
| 1878 { error("\\%c sequence must be followed by %d hex characters", u, nchars); | |
| 1879 break; | |
| 1880 } | |
| 1881 if (isdigit(c)) | |
| 1882 c -= '0'; | |
| 1883 else if (islower(c)) | |
| 1884 c -= 'a' - 10; | |
| 1885 else | |
| 1886 c -= 'A' - 10; | |
| 1887 value <<= 4; | |
| 1888 value |= c; | |
| 1889 } | |
| 1890 return value; | |
| 1891 } | |
| 1892 #endif | |
| 1893 | |
| 1894 /************************************** | |
| 1895 * Read in a number. | |
| 1896 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 1897 * integers can be decimal, octal or hex | |
| 1898 * Handle the suffixes U, UL, LU, L, etc. | |
| 1899 * If it's double, store it in tok.TKutok.Vdouble. | |
| 1900 * Returns: | |
| 1901 * TKnum | |
| 1902 * TKdouble,... | |
| 1903 */ | |
| 1904 | |
| 1905 TOK Lexer::number(Token *t) | |
| 1906 { | |
| 1907 // We use a state machine to collect numbers | |
| 1908 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 1909 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 1910 STATE_hexh, STATE_error }; | |
| 1911 enum STATE state; | |
| 1912 | |
| 1913 enum FLAGS | |
| 1914 { FLAGS_decimal = 1, // decimal | |
| 1915 FLAGS_unsigned = 2, // u or U suffix | |
| 1916 FLAGS_long = 4, // l or L suffix | |
| 1917 }; | |
| 1918 enum FLAGS flags = FLAGS_decimal; | |
| 1919 | |
| 1920 int i; | |
| 1921 int base; | |
| 1922 unsigned c; | |
| 1923 unsigned char *start; | |
| 1924 TOK result; | |
| 1925 | |
| 1926 //printf("Lexer::number()\n"); | |
| 1927 state = STATE_initial; | |
| 1928 base = 0; | |
| 1929 stringbuffer.reset(); | |
| 1930 start = p; | |
| 1931 while (1) | |
| 1932 { | |
| 1933 c = *p; | |
| 1934 switch (state) | |
| 1935 { | |
| 1936 case STATE_initial: // opening state | |
| 1937 if (c == '0') | |
| 1938 state = STATE_0; | |
| 1939 else | |
| 1940 state = STATE_decimal; | |
| 1941 break; | |
| 1942 | |
| 1943 case STATE_0: | |
| 1944 flags = (FLAGS) (flags & ~FLAGS_decimal); | |
| 1945 switch (c) | |
| 1946 { | |
| 1947 #if ZEROH | |
| 1948 case 'H': // 0h | |
| 1949 case 'h': | |
| 1950 goto hexh; | |
| 1951 #endif | |
| 1952 case 'X': | |
| 1953 case 'x': | |
| 1954 state = STATE_hex0; | |
| 1955 break; | |
| 1956 | |
| 1957 case '.': | |
| 1958 if (p[1] == '.') // .. is a separate token | |
| 1959 goto done; | |
| 1960 case 'i': | |
| 1961 case 'f': | |
| 1962 case 'F': | |
| 1963 goto real; | |
| 1964 #if ZEROH | |
| 1965 case 'E': | |
| 1966 case 'e': | |
| 1967 goto case_hex; | |
| 1968 #endif | |
| 1969 case 'B': | |
| 1970 case 'b': | |
| 1971 state = STATE_binary0; | |
| 1972 break; | |
| 1973 | |
| 1974 case '0': case '1': case '2': case '3': | |
| 1975 case '4': case '5': case '6': case '7': | |
| 1976 state = STATE_octal; | |
| 1977 break; | |
| 1978 | |
| 1979 #if ZEROH | |
| 1980 case '8': case '9': case 'A': | |
| 1981 case 'C': case 'D': case 'F': | |
| 1982 case 'a': case 'c': case 'd': case 'f': | |
| 1983 case_hex: | |
| 1984 state = STATE_hexh; | |
| 1985 break; | |
| 1986 #endif | |
| 1987 case '_': | |
| 1988 state = STATE_octal; | |
| 1989 p++; | |
| 1990 continue; | |
| 1991 | |
| 1992 case 'L': | |
| 1993 if (p[1] == 'i') | |
| 1994 goto real; | |
| 1995 goto done; | |
| 1996 | |
| 1997 default: | |
| 1998 goto done; | |
| 1999 } | |
| 2000 break; | |
| 2001 | |
| 2002 case STATE_decimal: // reading decimal number | |
| 2003 if (!isdigit(c)) | |
| 2004 { | |
| 2005 #if ZEROH | |
| 2006 if (ishex(c) | |
| 2007 || c == 'H' || c == 'h' | |
| 2008 ) | |
| 2009 goto hexh; | |
| 2010 #endif | |
| 2011 if (c == '_') // ignore embedded _ | |
| 2012 { p++; | |
| 2013 continue; | |
| 2014 } | |
| 2015 if (c == '.' && p[1] != '.') | |
| 2016 goto real; | |
| 2017 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2018 c == 'e' || c == 'E') | |
| 2019 { | |
| 2020 real: // It's a real number. Back up and rescan as a real | |
| 2021 p = start; | |
| 2022 return inreal(t); | |
| 2023 } | |
| 2024 else if (c == 'L' && p[1] == 'i') | |
| 2025 goto real; | |
| 2026 goto done; | |
| 2027 } | |
| 2028 break; | |
| 2029 | |
| 2030 case STATE_hex0: // reading hex number | |
| 2031 case STATE_hex: | |
| 2032 if (!ishex(c)) | |
| 2033 { | |
| 2034 if (c == '_') // ignore embedded _ | |
| 2035 { p++; | |
| 2036 continue; | |
| 2037 } | |
| 2038 if (c == '.' && p[1] != '.') | |
| 2039 goto real; | |
| 2040 if (c == 'P' || c == 'p' || c == 'i') | |
| 2041 goto real; | |
| 2042 if (state == STATE_hex0) | |
| 2043 error("Hex digit expected, not '%c'", c); | |
| 2044 goto done; | |
| 2045 } | |
| 2046 state = STATE_hex; | |
| 2047 break; | |
| 2048 | |
| 2049 #if ZEROH | |
| 2050 hexh: | |
| 2051 state = STATE_hexh; | |
| 2052 case STATE_hexh: // parse numbers like 0FFh | |
| 2053 if (!ishex(c)) | |
| 2054 { | |
| 2055 if (c == 'H' || c == 'h') | |
| 2056 { | |
| 2057 p++; | |
| 2058 base = 16; | |
| 2059 goto done; | |
| 2060 } | |
| 2061 else | |
| 2062 { | |
| 2063 // Check for something like 1E3 or 0E24 | |
| 2064 if (memchr((char *)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2065 memchr((char *)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2066 goto real; | |
| 2067 error("Hex digit expected, not '%c'", c); | |
| 2068 goto done; | |
| 2069 } | |
| 2070 } | |
| 2071 break; | |
| 2072 #endif | |
| 2073 | |
| 2074 case STATE_octal: // reading octal number | |
| 2075 case STATE_octale: // reading octal number with non-octal digits | |
| 2076 if (!isoctal(c)) | |
| 2077 { | |
| 2078 #if ZEROH | |
| 2079 if (ishex(c) | |
| 2080 || c == 'H' || c == 'h' | |
| 2081 ) | |
| 2082 goto hexh; | |
| 2083 #endif | |
| 2084 if (c == '_') // ignore embedded _ | |
| 2085 { p++; | |
| 2086 continue; | |
| 2087 } | |
| 2088 if (c == '.' && p[1] != '.') | |
| 2089 goto real; | |
| 2090 if (c == 'i') | |
| 2091 goto real; | |
| 2092 if (isdigit(c)) | |
| 2093 { | |
| 2094 state = STATE_octale; | |
| 2095 } | |
| 2096 else | |
| 2097 goto done; | |
| 2098 } | |
| 2099 break; | |
| 2100 | |
| 2101 case STATE_binary0: // starting binary number | |
| 2102 case STATE_binary: // reading binary number | |
| 2103 if (c != '0' && c != '1') | |
| 2104 { | |
| 2105 #if ZEROH | |
| 2106 if (ishex(c) | |
| 2107 || c == 'H' || c == 'h' | |
| 2108 ) | |
| 2109 goto hexh; | |
| 2110 #endif | |
| 2111 if (c == '_') // ignore embedded _ | |
| 2112 { p++; | |
| 2113 continue; | |
| 2114 } | |
| 2115 if (state == STATE_binary0) | |
| 2116 { error("binary digit expected"); | |
| 2117 state = STATE_error; | |
| 2118 break; | |
| 2119 } | |
| 2120 else | |
| 2121 goto done; | |
| 2122 } | |
| 2123 state = STATE_binary; | |
| 2124 break; | |
| 2125 | |
| 2126 case STATE_error: // for error recovery | |
| 2127 if (!isdigit(c)) // scan until non-digit | |
| 2128 goto done; | |
| 2129 break; | |
| 2130 | |
| 2131 default: | |
| 2132 assert(0); | |
| 2133 } | |
| 2134 stringbuffer.writeByte(c); | |
| 2135 p++; | |
| 2136 } | |
| 2137 done: | |
| 2138 stringbuffer.writeByte(0); // terminate string | |
| 2139 if (state == STATE_octale) | |
| 2140 error("Octal digit expected"); | |
| 2141 | |
| 2142 uinteger_t n; // unsigned >=64 bit integer type | |
| 2143 | |
| 2144 if (stringbuffer.offset == 2 && (state == STATE_decimal || state == STATE_0)) | |
| 2145 n = stringbuffer.data[0] - '0'; | |
| 2146 else | |
| 2147 { | |
| 2148 // Convert string to integer | |
| 2149 #if __DMC__ | |
| 2150 errno = 0; | |
| 2151 n = strtoull((char *)stringbuffer.data,NULL,base); | |
| 2152 if (errno == ERANGE) | |
| 2153 error("integer overflow"); | |
| 2154 #else | |
| 2155 // Not everybody implements strtoull() | |
| 2156 char *p = (char *)stringbuffer.data; | |
| 2157 int r = 10, d; | |
| 2158 | |
| 2159 if (*p == '0') | |
| 2160 { | |
| 2161 if (p[1] == 'x' || p[1] == 'X') | |
| 2162 p += 2, r = 16; | |
| 2163 else if (p[1] == 'b' || p[1] == 'B') | |
| 2164 p += 2, r = 2; | |
| 2165 else if (isdigit(p[1])) | |
| 2166 p += 1, r = 8; | |
| 2167 } | |
| 2168 | |
| 2169 n = 0; | |
| 2170 while (1) | |
| 2171 { | |
| 2172 if (*p >= '0' && *p <= '9') | |
| 2173 d = *p - '0'; | |
| 2174 else if (*p >= 'a' && *p <= 'z') | |
| 2175 d = *p - 'a' + 10; | |
| 2176 else if (*p >= 'A' && *p <= 'Z') | |
| 2177 d = *p - 'A' + 10; | |
| 2178 else | |
| 2179 break; | |
| 2180 if (d >= r) | |
| 2181 break; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2182 uinteger_t n2 = n * r; |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2183 //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
|
2184 if (n2 / r != n || n2 + d < n) |
| 159 | 2185 { |
| 2186 error ("integer overflow"); | |
| 2187 break; | |
| 2188 } | |
| 2189 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2190 n = n2 + d; |
| 159 | 2191 p++; |
| 2192 } | |
| 2193 #endif | |
| 2194 if (sizeof(n) > 8 && | |
| 2195 n > 0xFFFFFFFFFFFFFFFFULL) // if n needs more than 64 bits | |
| 2196 error("integer overflow"); | |
| 2197 } | |
| 2198 | |
| 2199 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2200 while (1) | |
| 2201 { unsigned char f; | |
| 2202 | |
| 2203 switch (*p) | |
| 2204 { case 'U': | |
| 2205 case 'u': | |
| 2206 f = FLAGS_unsigned; | |
| 2207 goto L1; | |
| 2208 | |
| 2209 case 'l': | |
| 2210 if (1 || !global.params.useDeprecated) | |
| 2211 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2212 case 'L': | |
| 2213 f = FLAGS_long; | |
| 2214 L1: | |
| 2215 p++; | |
| 2216 if (flags & f) | |
| 2217 error("unrecognized token"); | |
| 2218 flags = (FLAGS) (flags | f); | |
| 2219 continue; | |
| 2220 default: | |
| 2221 break; | |
| 2222 } | |
| 2223 break; | |
| 2224 } | |
| 2225 | |
| 2226 switch (flags) | |
| 2227 { | |
| 2228 case 0: | |
| 2229 /* Octal or Hexadecimal constant. | |
| 2230 * First that fits: int, uint, long, ulong | |
| 2231 */ | |
| 2232 if (n & 0x8000000000000000LL) | |
| 2233 result = TOKuns64v; | |
| 2234 else if (n & 0xFFFFFFFF00000000LL) | |
| 2235 result = TOKint64v; | |
| 2236 else if (n & 0x80000000) | |
| 2237 result = TOKuns32v; | |
| 2238 else | |
| 2239 result = TOKint32v; | |
| 2240 break; | |
| 2241 | |
| 2242 case FLAGS_decimal: | |
| 2243 /* First that fits: int, long, long long | |
| 2244 */ | |
| 2245 if (n & 0x8000000000000000LL) | |
| 2246 { error("signed integer overflow"); | |
| 2247 result = TOKuns64v; | |
| 2248 } | |
| 2249 else if (n & 0xFFFFFFFF80000000LL) | |
| 2250 result = TOKint64v; | |
| 2251 else | |
| 2252 result = TOKint32v; | |
| 2253 break; | |
| 2254 | |
| 2255 case FLAGS_unsigned: | |
| 2256 case FLAGS_decimal | FLAGS_unsigned: | |
| 2257 /* First that fits: uint, ulong | |
| 2258 */ | |
| 2259 if (n & 0xFFFFFFFF00000000LL) | |
| 2260 result = TOKuns64v; | |
| 2261 else | |
| 2262 result = TOKuns32v; | |
| 2263 break; | |
| 2264 | |
| 2265 case FLAGS_decimal | FLAGS_long: | |
| 2266 if (n & 0x8000000000000000LL) | |
| 2267 { error("signed integer overflow"); | |
| 2268 result = TOKuns64v; | |
| 2269 } | |
| 2270 else | |
| 2271 result = TOKint64v; | |
| 2272 break; | |
| 2273 | |
| 2274 case FLAGS_long: | |
| 2275 if (n & 0x8000000000000000LL) | |
| 2276 result = TOKuns64v; | |
| 2277 else | |
| 2278 result = TOKint64v; | |
| 2279 break; | |
| 2280 | |
| 2281 case FLAGS_unsigned | FLAGS_long: | |
| 2282 case FLAGS_decimal | FLAGS_unsigned | FLAGS_long: | |
| 2283 result = TOKuns64v; | |
| 2284 break; | |
| 2285 | |
| 2286 default: | |
| 2287 #ifdef DEBUG | |
| 2288 printf("%x\n",flags); | |
| 2289 #endif | |
| 2290 assert(0); | |
| 2291 } | |
| 2292 t->uns64value = n; | |
| 2293 return result; | |
| 2294 } | |
| 2295 | |
| 2296 /************************************** | |
| 2297 * Read in characters, converting them to real. | |
| 2298 * Bugs: | |
| 2299 * Exponent overflow not detected. | |
| 2300 * Too much requested precision is not detected. | |
| 2301 */ | |
| 2302 | |
| 2303 TOK Lexer::inreal(Token *t) | |
| 2304 #ifdef __DMC__ | |
| 2305 __in | |
| 2306 { | |
| 2307 assert(*p == '.' || isdigit(*p)); | |
| 2308 } | |
| 2309 __out (result) | |
| 2310 { | |
| 2311 switch (result) | |
| 2312 { | |
| 2313 case TOKfloat32v: | |
| 2314 case TOKfloat64v: | |
| 2315 case TOKfloat80v: | |
| 2316 case TOKimaginary32v: | |
| 2317 case TOKimaginary64v: | |
| 2318 case TOKimaginary80v: | |
| 2319 break; | |
| 2320 | |
| 2321 default: | |
| 2322 assert(0); | |
| 2323 } | |
| 2324 } | |
| 2325 __body | |
| 2326 #endif /* __DMC__ */ | |
| 2327 { int dblstate; | |
| 2328 unsigned c; | |
| 2329 char hex; // is this a hexadecimal-floating-constant? | |
| 2330 TOK result; | |
| 2331 | |
| 2332 //printf("Lexer::inreal()\n"); | |
| 2333 stringbuffer.reset(); | |
| 2334 dblstate = 0; | |
| 2335 hex = 0; | |
| 2336 Lnext: | |
| 2337 while (1) | |
| 2338 { | |
| 2339 // Get next char from input | |
| 2340 c = *p++; | |
| 2341 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2342 while (1) | |
| 2343 { | |
| 2344 switch (dblstate) | |
| 2345 { | |
| 2346 case 0: // opening state | |
| 2347 if (c == '0') | |
| 2348 dblstate = 9; | |
| 2349 else if (c == '.') | |
| 2350 dblstate = 3; | |
| 2351 else | |
| 2352 dblstate = 1; | |
| 2353 break; | |
| 2354 | |
| 2355 case 9: | |
| 2356 dblstate = 1; | |
| 2357 if (c == 'X' || c == 'x') | |
| 2358 { hex++; | |
| 2359 break; | |
| 2360 } | |
| 2361 case 1: // digits to left of . | |
| 2362 case 3: // digits to right of . | |
| 2363 case 7: // continuing exponent digits | |
| 2364 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2365 { | |
| 2366 if (c == '_') | |
| 2367 goto Lnext; // ignore embedded '_' | |
| 2368 dblstate++; | |
| 2369 continue; | |
| 2370 } | |
| 2371 break; | |
| 2372 | |
| 2373 case 2: // no more digits to left of . | |
| 2374 if (c == '.') | |
| 2375 { dblstate++; | |
| 2376 break; | |
| 2377 } | |
| 2378 case 4: // no more digits to right of . | |
| 2379 if ((c == 'E' || c == 'e') || | |
| 2380 hex && (c == 'P' || c == 'p')) | |
| 2381 { dblstate = 5; | |
| 2382 hex = 0; // exponent is always decimal | |
| 2383 break; | |
| 2384 } | |
| 2385 if (hex) | |
| 2386 error("binary-exponent-part required"); | |
| 2387 goto done; | |
| 2388 | |
| 2389 case 5: // looking immediately to right of E | |
| 2390 dblstate++; | |
| 2391 if (c == '-' || c == '+') | |
| 2392 break; | |
| 2393 case 6: // 1st exponent digit expected | |
| 2394 if (!isdigit(c)) | |
| 2395 error("exponent expected"); | |
| 2396 dblstate++; | |
| 2397 break; | |
| 2398 | |
| 2399 case 8: // past end of exponent digits | |
| 2400 goto done; | |
| 2401 } | |
| 2402 break; | |
| 2403 } | |
| 2404 stringbuffer.writeByte(c); | |
| 2405 } | |
| 2406 done: | |
| 2407 p--; | |
| 2408 | |
| 2409 stringbuffer.writeByte(0); | |
| 2410 | |
| 2411 #if _WIN32 && __DMC__ | |
| 2412 char *save = __locale_decpoint; | |
| 2413 __locale_decpoint = "."; | |
| 2414 #endif | |
| 2415 #ifdef IN_GCC | |
| 2416 t->float80value = real_t::parse((char *)stringbuffer.data, real_t::LongDouble); | |
| 2417 #else | |
| 2418 t->float80value = strtold((char *)stringbuffer.data, NULL); | |
| 2419 #endif | |
| 2420 errno = 0; | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2421 float strtofres; |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2422 double strtodres; |
| 159 | 2423 switch (*p) |
| 2424 { | |
| 2425 case 'F': | |
| 2426 case 'f': | |
| 2427 #ifdef IN_GCC | |
| 2428 real_t::parse((char *)stringbuffer.data, real_t::Float); | |
| 2429 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2430 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
|
2431 // 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
|
2432 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2433 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
|
2434 errno = 0; |
| 159 | 2435 #endif |
| 2436 result = TOKfloat32v; | |
| 2437 p++; | |
| 2438 break; | |
| 2439 | |
| 2440 default: | |
| 2441 #ifdef IN_GCC | |
| 2442 real_t::parse((char *)stringbuffer.data, real_t::Double); | |
| 2443 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2444 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
|
2445 // 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
|
2446 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2447 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
|
2448 errno = 0; |
| 159 | 2449 #endif |
| 2450 result = TOKfloat64v; | |
| 2451 break; | |
| 2452 | |
| 2453 case 'l': | |
| 2454 if (!global.params.useDeprecated) | |
| 2455 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2456 case 'L': | |
| 2457 result = TOKfloat80v; | |
| 2458 p++; | |
| 2459 break; | |
| 2460 } | |
| 2461 if (*p == 'i' || *p == 'I') | |
| 2462 { | |
| 2463 if (!global.params.useDeprecated && *p == 'I') | |
| 2464 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2465 p++; | |
| 2466 switch (result) | |
| 2467 { | |
| 2468 case TOKfloat32v: | |
| 2469 result = TOKimaginary32v; | |
| 2470 break; | |
| 2471 case TOKfloat64v: | |
| 2472 result = TOKimaginary64v; | |
| 2473 break; | |
| 2474 case TOKfloat80v: | |
| 2475 result = TOKimaginary80v; | |
| 2476 break; | |
| 2477 } | |
| 2478 } | |
| 2479 #if _WIN32 && __DMC__ | |
| 2480 __locale_decpoint = save; | |
| 2481 #endif | |
| 2482 if (errno == ERANGE) | |
| 2483 error("number is not representable"); | |
| 2484 return result; | |
| 2485 } | |
| 2486 | |
| 2487 /********************************************* | |
| 2488 * Do pragma. | |
| 2489 * Currently, the only pragma supported is: | |
| 2490 * #line linnum [filespec] | |
| 2491 */ | |
| 2492 | |
| 2493 void Lexer::pragma() | |
| 2494 { | |
| 2495 Token tok; | |
| 2496 int linnum; | |
| 2497 char *filespec = NULL; | |
| 2498 Loc loc = this->loc; | |
| 2499 | |
| 2500 scan(&tok); | |
| 2501 if (tok.value != TOKidentifier || tok.ident != Id::line) | |
| 2502 goto Lerr; | |
| 2503 | |
| 2504 scan(&tok); | |
| 2505 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 2506 linnum = tok.uns64value - 1; | |
| 2507 else | |
| 2508 goto Lerr; | |
| 2509 | |
| 2510 while (1) | |
| 2511 { | |
| 2512 switch (*p) | |
| 2513 { | |
| 2514 case 0: | |
| 2515 case 0x1A: | |
| 2516 case '\n': | |
| 2517 Lnewline: | |
| 2518 this->loc.linnum = linnum; | |
| 2519 if (filespec) | |
| 2520 this->loc.filename = filespec; | |
| 2521 return; | |
| 2522 | |
| 2523 case '\r': | |
| 2524 p++; | |
| 2525 if (*p != '\n') | |
| 2526 { p--; | |
| 2527 goto Lnewline; | |
| 2528 } | |
| 2529 continue; | |
| 2530 | |
| 2531 case ' ': | |
| 2532 case '\t': | |
| 2533 case '\v': | |
| 2534 case '\f': | |
| 2535 p++; | |
| 2536 continue; // skip white space | |
| 2537 | |
| 2538 case '_': | |
| 2539 if (mod && memcmp(p, "__FILE__", 8) == 0) | |
| 2540 { | |
| 2541 p += 8; | |
| 2542 filespec = mem.strdup(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 2543 } | |
| 2544 continue; | |
| 2545 | |
| 2546 case '"': | |
| 2547 if (filespec) | |
| 2548 goto Lerr; | |
| 2549 stringbuffer.reset(); | |
| 2550 p++; | |
| 2551 while (1) | |
| 2552 { unsigned c; | |
| 2553 | |
| 2554 c = *p; | |
| 2555 switch (c) | |
| 2556 { | |
| 2557 case '\n': | |
| 2558 case '\r': | |
| 2559 case 0: | |
| 2560 case 0x1A: | |
| 2561 goto Lerr; | |
| 2562 | |
| 2563 case '"': | |
| 2564 stringbuffer.writeByte(0); | |
| 2565 filespec = mem.strdup((char *)stringbuffer.data); | |
| 2566 p++; | |
| 2567 break; | |
| 2568 | |
| 2569 default: | |
| 2570 if (c & 0x80) | |
| 2571 { unsigned u = decodeUTF(); | |
| 2572 if (u == PS || u == LS) | |
| 2573 goto Lerr; | |
| 2574 } | |
| 2575 stringbuffer.writeByte(c); | |
| 2576 p++; | |
| 2577 continue; | |
| 2578 } | |
| 2579 break; | |
| 2580 } | |
| 2581 continue; | |
| 2582 | |
| 2583 default: | |
| 2584 if (*p & 0x80) | |
| 2585 { unsigned u = decodeUTF(); | |
| 2586 if (u == PS || u == LS) | |
| 2587 goto Lnewline; | |
| 2588 } | |
| 2589 goto Lerr; | |
| 2590 } | |
| 2591 } | |
| 2592 | |
| 2593 Lerr: | |
| 2594 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 2595 } | |
| 2596 | |
| 2597 | |
| 2598 /******************************************** | |
| 2599 * Decode UTF character. | |
| 2600 * Issue error messages for invalid sequences. | |
| 2601 * Return decoded character, advance p to last character in UTF sequence. | |
| 2602 */ | |
| 2603 | |
| 2604 unsigned Lexer::decodeUTF() | |
| 2605 { | |
| 2606 dchar_t u; | |
| 2607 unsigned char c; | |
| 2608 unsigned char *s = p; | |
| 2609 size_t len; | |
| 2610 size_t idx; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2611 const char *msg; |
| 159 | 2612 |
| 2613 c = *s; | |
| 2614 assert(c & 0x80); | |
| 2615 | |
| 2616 // Check length of remaining string up to 6 UTF-8 characters | |
| 2617 for (len = 1; len < 6 && s[len]; len++) | |
| 2618 ; | |
| 2619 | |
| 2620 idx = 0; | |
| 2621 msg = utf_decodeChar(s, len, &idx, &u); | |
| 2622 p += idx - 1; | |
| 2623 if (msg) | |
| 2624 { | |
| 2625 error("%s", msg); | |
| 2626 } | |
| 2627 return u; | |
| 2628 } | |
| 2629 | |
| 2630 | |
| 2631 /*************************************************** | |
| 2632 * Parse doc comment embedded between t->ptr and p. | |
| 2633 * Remove trailing blanks and tabs from lines. | |
| 2634 * Replace all newlines with \n. | |
| 2635 * Remove leading comment character from each line. | |
| 2636 * Decide if it's a lineComment or a blockComment. | |
| 2637 * Append to previous one for this token. | |
| 2638 */ | |
| 2639 | |
| 2640 void Lexer::getDocComment(Token *t, unsigned lineComment) | |
| 2641 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2642 /* 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
|
2643 */ |
| 159 | 2644 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
|
2645 |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2646 /* 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
|
2647 */ |
| 159 | 2648 unsigned char *q = t->ptr + 3; // start of comment text |
| 2649 | |
| 2650 unsigned char *qend = p; | |
| 2651 if (ct == '*' || ct == '+') | |
| 2652 qend -= 2; | |
| 2653 | |
| 2654 /* Scan over initial row of ****'s or ++++'s or ////'s | |
| 2655 */ | |
| 2656 for (; q < qend; q++) | |
| 2657 { | |
| 2658 if (*q != ct) | |
| 2659 break; | |
| 2660 } | |
| 2661 | |
| 2662 /* Remove trailing row of ****'s or ++++'s | |
| 2663 */ | |
| 2664 if (ct != '/') | |
| 2665 { | |
| 2666 for (; q < qend; qend--) | |
| 2667 { | |
| 2668 if (qend[-1] != ct) | |
| 2669 break; | |
| 2670 } | |
| 2671 } | |
| 2672 | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2673 /* Comment is now [q .. qend]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2674 * Canonicalize it into buf[]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2675 */ |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2676 OutBuffer buf; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2677 int linestart = 0; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2678 |
| 159 | 2679 for (; q < qend; q++) |
| 2680 { | |
| 2681 unsigned char c = *q; | |
| 2682 | |
| 2683 switch (c) | |
| 2684 { | |
| 2685 case '*': | |
| 2686 case '+': | |
| 2687 if (linestart && c == ct) | |
| 2688 { linestart = 0; | |
| 2689 /* Trim preceding whitespace up to preceding \n | |
| 2690 */ | |
| 2691 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2692 buf.offset--; | |
| 2693 continue; | |
| 2694 } | |
| 2695 break; | |
| 2696 | |
| 2697 case ' ': | |
| 2698 case '\t': | |
| 2699 break; | |
| 2700 | |
| 2701 case '\r': | |
| 2702 if (q[1] == '\n') | |
| 2703 continue; // skip the \r | |
| 2704 goto Lnewline; | |
| 2705 | |
| 2706 default: | |
| 2707 if (c == 226) | |
| 2708 { | |
| 2709 // If LS or PS | |
| 2710 if (q[1] == 128 && | |
| 2711 (q[2] == 168 || q[2] == 169)) | |
| 2712 { | |
| 2713 q += 2; | |
| 2714 goto Lnewline; | |
| 2715 } | |
| 2716 } | |
| 2717 linestart = 0; | |
| 2718 break; | |
| 2719 | |
| 2720 Lnewline: | |
| 2721 c = '\n'; // replace all newlines with \n | |
| 2722 case '\n': | |
| 2723 linestart = 1; | |
| 2724 | |
| 2725 /* Trim trailing whitespace | |
| 2726 */ | |
| 2727 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2728 buf.offset--; | |
| 2729 | |
| 2730 break; | |
| 2731 } | |
| 2732 buf.writeByte(c); | |
| 2733 } | |
| 2734 | |
| 2735 // Always end with a newline | |
| 2736 if (!buf.offset || buf.data[buf.offset - 1] != '\n') | |
| 2737 buf.writeByte('\n'); | |
| 2738 | |
| 2739 buf.writeByte(0); | |
| 2740 | |
| 2741 // It's a line comment if the start of the doc comment comes | |
| 2742 // after other non-whitespace on the same line. | |
| 2743 unsigned char** dc = (lineComment && anyToken) | |
| 2744 ? &t->lineComment | |
| 2745 : &t->blockComment; | |
| 2746 | |
| 2747 // Combine with previous doc comment, if any | |
| 2748 if (*dc) | |
| 2749 *dc = combineComments(*dc, (unsigned char *)buf.data); | |
| 2750 else | |
| 2751 *dc = (unsigned char *)buf.extractData(); | |
| 2752 } | |
| 2753 | |
| 2754 /******************************************** | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2755 * Combine two document comments into one, |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2756 * separated by a newline. |
| 159 | 2757 */ |
| 2758 | |
| 2759 unsigned char *Lexer::combineComments(unsigned char *c1, unsigned char *c2) | |
| 2760 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2761 //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
|
2762 |
| 159 | 2763 unsigned char *c = c2; |
| 2764 | |
| 2765 if (c1) | |
| 2766 { c = c1; | |
| 2767 if (c2) | |
| 2768 { size_t len1 = strlen((char *)c1); | |
| 2769 size_t len2 = strlen((char *)c2); | |
| 2770 | |
| 2771 c = (unsigned char *)mem.malloc(len1 + 1 + len2 + 1); | |
| 2772 memcpy(c, c1, len1); | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2773 if (len1 && c1[len1 - 1] != '\n') |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2774 { c[len1] = '\n'; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2775 len1++; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2776 } |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2777 memcpy(c + len1, c2, len2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2778 c[len1 + len2] = 0; |
| 159 | 2779 } |
| 2780 } | |
| 2781 return c; | |
| 2782 } | |
| 2783 | |
| 2784 /******************************************** | |
| 2785 * Create an identifier in the string table. | |
| 2786 */ | |
| 2787 | |
| 2788 Identifier *Lexer::idPool(const char *s) | |
| 2789 { | |
| 2790 size_t len = strlen(s); | |
| 2791 StringValue *sv = stringtable.update(s, len); | |
| 2792 Identifier *id = (Identifier *) sv->ptrvalue; | |
| 2793 if (!id) | |
| 2794 { | |
| 2795 id = new Identifier(sv->lstring.string, TOKidentifier); | |
| 2796 sv->ptrvalue = id; | |
| 2797 } | |
| 2798 return id; | |
| 2799 } | |
| 2800 | |
| 2801 /********************************************* | |
| 2802 * Create a unique identifier using the prefix s. | |
| 2803 */ | |
| 2804 | |
| 2805 Identifier *Lexer::uniqueId(const char *s, int num) | |
| 2806 { char buffer[32]; | |
| 2807 size_t slen = strlen(s); | |
| 2808 | |
| 2809 assert(slen + sizeof(num) * 3 + 1 <= sizeof(buffer)); | |
| 2810 sprintf(buffer, "%s%d", s, num); | |
| 2811 return idPool(buffer); | |
| 2812 } | |
| 2813 | |
| 2814 Identifier *Lexer::uniqueId(const char *s) | |
| 2815 { | |
| 2816 static int num; | |
| 2817 return uniqueId(s, ++num); | |
| 2818 } | |
| 2819 | |
| 2820 /**************************************** | |
| 2821 */ | |
| 2822 | |
| 2823 struct Keyword | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2824 { const char *name; |
| 159 | 2825 enum TOK value; |
| 2826 }; | |
| 2827 | |
| 2828 static Keyword keywords[] = | |
| 2829 { | |
| 2830 // { "", TOK }, | |
| 2831 | |
| 2832 { "this", TOKthis }, | |
| 2833 { "super", TOKsuper }, | |
| 2834 { "assert", TOKassert }, | |
| 2835 { "null", TOKnull }, | |
| 2836 { "true", TOKtrue }, | |
| 2837 { "false", TOKfalse }, | |
| 2838 { "cast", TOKcast }, | |
| 2839 { "new", TOKnew }, | |
| 2840 { "delete", TOKdelete }, | |
| 2841 { "throw", TOKthrow }, | |
| 2842 { "module", TOKmodule }, | |
| 2843 { "pragma", TOKpragma }, | |
| 2844 { "typeof", TOKtypeof }, | |
| 2845 { "typeid", TOKtypeid }, | |
| 2846 | |
| 2847 { "template", TOKtemplate }, | |
| 2848 | |
| 2849 { "void", TOKvoid }, | |
| 2850 { "byte", TOKint8 }, | |
| 2851 { "ubyte", TOKuns8 }, | |
| 2852 { "short", TOKint16 }, | |
| 2853 { "ushort", TOKuns16 }, | |
| 2854 { "int", TOKint32 }, | |
| 2855 { "uint", TOKuns32 }, | |
| 2856 { "long", TOKint64 }, | |
| 2857 { "ulong", TOKuns64 }, | |
| 2858 { "cent", TOKcent, }, | |
| 2859 { "ucent", TOKucent, }, | |
| 2860 { "float", TOKfloat32 }, | |
| 2861 { "double", TOKfloat64 }, | |
| 2862 { "real", TOKfloat80 }, | |
| 2863 | |
| 2864 { "bool", TOKbool }, | |
| 2865 { "char", TOKchar }, | |
| 2866 { "wchar", TOKwchar }, | |
| 2867 { "dchar", TOKdchar }, | |
| 2868 | |
| 2869 { "ifloat", TOKimaginary32 }, | |
| 2870 { "idouble", TOKimaginary64 }, | |
| 2871 { "ireal", TOKimaginary80 }, | |
| 2872 | |
| 2873 { "cfloat", TOKcomplex32 }, | |
| 2874 { "cdouble", TOKcomplex64 }, | |
| 2875 { "creal", TOKcomplex80 }, | |
| 2876 | |
| 2877 { "delegate", TOKdelegate }, | |
| 2878 { "function", TOKfunction }, | |
| 2879 | |
| 2880 { "is", TOKis }, | |
| 2881 { "if", TOKif }, | |
| 2882 { "else", TOKelse }, | |
| 2883 { "while", TOKwhile }, | |
| 2884 { "for", TOKfor }, | |
| 2885 { "do", TOKdo }, | |
| 2886 { "switch", TOKswitch }, | |
| 2887 { "case", TOKcase }, | |
| 2888 { "default", TOKdefault }, | |
| 2889 { "break", TOKbreak }, | |
| 2890 { "continue", TOKcontinue }, | |
| 2891 { "synchronized", TOKsynchronized }, | |
| 2892 { "return", TOKreturn }, | |
| 2893 { "goto", TOKgoto }, | |
| 2894 { "try", TOKtry }, | |
| 2895 { "catch", TOKcatch }, | |
| 2896 { "finally", TOKfinally }, | |
| 2897 { "with", TOKwith }, | |
| 2898 { "asm", TOKasm }, | |
| 2899 { "foreach", TOKforeach }, | |
| 2900 { "foreach_reverse", TOKforeach_reverse }, | |
| 2901 { "scope", TOKscope }, | |
| 2902 | |
| 2903 { "struct", TOKstruct }, | |
| 2904 { "class", TOKclass }, | |
| 2905 { "interface", TOKinterface }, | |
| 2906 { "union", TOKunion }, | |
| 2907 { "enum", TOKenum }, | |
| 2908 { "import", TOKimport }, | |
| 2909 { "mixin", TOKmixin }, | |
| 2910 { "static", TOKstatic }, | |
| 2911 { "final", TOKfinal }, | |
| 2912 { "const", TOKconst }, | |
| 2913 { "typedef", TOKtypedef }, | |
| 2914 { "alias", TOKalias }, | |
| 2915 { "override", TOKoverride }, | |
| 2916 { "abstract", TOKabstract }, | |
| 2917 { "volatile", TOKvolatile }, | |
| 2918 { "debug", TOKdebug }, | |
| 2919 { "deprecated", TOKdeprecated }, | |
| 2920 { "in", TOKin }, | |
| 2921 { "out", TOKout }, | |
| 2922 { "inout", TOKinout }, | |
| 2923 { "lazy", TOKlazy }, | |
| 2924 { "auto", TOKauto }, | |
| 2925 | |
| 2926 { "align", TOKalign }, | |
| 2927 { "extern", TOKextern }, | |
| 2928 { "private", TOKprivate }, | |
| 2929 { "package", TOKpackage }, | |
| 2930 { "protected", TOKprotected }, | |
| 2931 { "public", TOKpublic }, | |
| 2932 { "export", TOKexport }, | |
| 2933 | |
| 2934 { "body", TOKbody }, | |
| 2935 { "invariant", TOKinvariant }, | |
| 2936 { "unittest", TOKunittest }, | |
| 2937 { "version", TOKversion }, | |
| 2938 //{ "manifest", TOKmanifest }, | |
| 2939 | |
| 2940 // Added after 1.0 | |
| 2941 { "ref", TOKref }, | |
| 2942 { "macro", TOKmacro }, | |
| 336 | 2943 #if DMDV2 |
| 159 | 2944 { "pure", TOKpure }, |
| 2945 { "nothrow", TOKnothrow }, | |
| 336 | 2946 { "__thread", TOKtls }, |
| 159 | 2947 { "__traits", TOKtraits }, |
| 2948 { "__overloadset", TOKoverloadset }, | |
| 336 | 2949 { "__FILE__", TOKfile }, |
| 2950 { "__LINE__", TOKline }, | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2951 { "shared", TOKshared }, |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2952 { "immutable", TOKimmutable }, |
| 159 | 2953 #endif |
| 2954 }; | |
| 2955 | |
| 2956 int Token::isKeyword() | |
| 2957 { | |
| 2958 for (unsigned u = 0; u < sizeof(keywords) / sizeof(keywords[0]); u++) | |
| 2959 { | |
| 2960 if (keywords[u].value == value) | |
| 2961 return 1; | |
| 2962 } | |
| 2963 return 0; | |
| 2964 } | |
| 2965 | |
| 2966 void Lexer::initKeywords() | |
| 2967 { StringValue *sv; | |
| 2968 unsigned u; | |
| 2969 enum TOK v; | |
| 2970 unsigned nkeywords = sizeof(keywords) / sizeof(keywords[0]); | |
| 2971 | |
| 2972 if (global.params.Dversion == 1) | |
| 2973 nkeywords -= 2; | |
| 2974 | |
| 2975 cmtable_init(); | |
| 2976 | |
| 2977 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
|
2978 { const char *s; |
| 159 | 2979 |
| 2980 //printf("keyword[%d] = '%s'\n",u, keywords[u].name); | |
| 2981 s = keywords[u].name; | |
| 2982 v = keywords[u].value; | |
| 2983 sv = stringtable.insert(s, strlen(s)); | |
| 2984 sv->ptrvalue = (void *) new Identifier(sv->lstring.string,v); | |
| 2985 | |
| 2986 //printf("tochars[%d] = '%s'\n",v, s); | |
| 2987 Token::tochars[v] = s; | |
| 2988 } | |
| 2989 | |
| 2990 Token::tochars[TOKeof] = "EOF"; | |
| 2991 Token::tochars[TOKlcurly] = "{"; | |
| 2992 Token::tochars[TOKrcurly] = "}"; | |
| 2993 Token::tochars[TOKlparen] = "("; | |
| 2994 Token::tochars[TOKrparen] = ")"; | |
| 2995 Token::tochars[TOKlbracket] = "["; | |
| 2996 Token::tochars[TOKrbracket] = "]"; | |
| 2997 Token::tochars[TOKsemicolon] = ";"; | |
| 2998 Token::tochars[TOKcolon] = ":"; | |
| 2999 Token::tochars[TOKcomma] = ","; | |
| 3000 Token::tochars[TOKdot] = "."; | |
| 3001 Token::tochars[TOKxor] = "^"; | |
| 3002 Token::tochars[TOKxorass] = "^="; | |
| 3003 Token::tochars[TOKassign] = "="; | |
| 3004 Token::tochars[TOKconstruct] = "="; | |
| 336 | 3005 #if DMDV2 |
| 159 | 3006 Token::tochars[TOKblit] = "="; |
| 3007 #endif | |
| 3008 Token::tochars[TOKlt] = "<"; | |
| 3009 Token::tochars[TOKgt] = ">"; | |
| 3010 Token::tochars[TOKle] = "<="; | |
| 3011 Token::tochars[TOKge] = ">="; | |
| 3012 Token::tochars[TOKequal] = "=="; | |
| 3013 Token::tochars[TOKnotequal] = "!="; | |
| 3014 Token::tochars[TOKnotidentity] = "!is"; | |
| 3015 Token::tochars[TOKtobool] = "!!"; | |
| 3016 | |
| 3017 Token::tochars[TOKunord] = "!<>="; | |
| 3018 Token::tochars[TOKue] = "!<>"; | |
| 3019 Token::tochars[TOKlg] = "<>"; | |
| 3020 Token::tochars[TOKleg] = "<>="; | |
| 3021 Token::tochars[TOKule] = "!>"; | |
| 3022 Token::tochars[TOKul] = "!>="; | |
| 3023 Token::tochars[TOKuge] = "!<"; | |
| 3024 Token::tochars[TOKug] = "!<="; | |
| 3025 | |
| 3026 Token::tochars[TOKnot] = "!"; | |
| 3027 Token::tochars[TOKtobool] = "!!"; | |
| 3028 Token::tochars[TOKshl] = "<<"; | |
| 3029 Token::tochars[TOKshr] = ">>"; | |
| 3030 Token::tochars[TOKushr] = ">>>"; | |
| 3031 Token::tochars[TOKadd] = "+"; | |
| 3032 Token::tochars[TOKmin] = "-"; | |
| 3033 Token::tochars[TOKmul] = "*"; | |
| 3034 Token::tochars[TOKdiv] = "/"; | |
| 3035 Token::tochars[TOKmod] = "%"; | |
| 3036 Token::tochars[TOKslice] = ".."; | |
| 3037 Token::tochars[TOKdotdotdot] = "..."; | |
| 3038 Token::tochars[TOKand] = "&"; | |
| 3039 Token::tochars[TOKandand] = "&&"; | |
| 3040 Token::tochars[TOKor] = "|"; | |
| 3041 Token::tochars[TOKoror] = "||"; | |
| 3042 Token::tochars[TOKarray] = "[]"; | |
| 3043 Token::tochars[TOKindex] = "[i]"; | |
| 3044 Token::tochars[TOKaddress] = "&"; | |
| 3045 Token::tochars[TOKstar] = "*"; | |
| 3046 Token::tochars[TOKtilde] = "~"; | |
| 3047 Token::tochars[TOKdollar] = "$"; | |
| 3048 Token::tochars[TOKcast] = "cast"; | |
| 3049 Token::tochars[TOKplusplus] = "++"; | |
| 3050 Token::tochars[TOKminusminus] = "--"; | |
| 3051 Token::tochars[TOKtype] = "type"; | |
| 3052 Token::tochars[TOKquestion] = "?"; | |
| 3053 Token::tochars[TOKneg] = "-"; | |
| 3054 Token::tochars[TOKuadd] = "+"; | |
| 3055 Token::tochars[TOKvar] = "var"; | |
| 3056 Token::tochars[TOKaddass] = "+="; | |
| 3057 Token::tochars[TOKminass] = "-="; | |
| 3058 Token::tochars[TOKmulass] = "*="; | |
| 3059 Token::tochars[TOKdivass] = "/="; | |
| 3060 Token::tochars[TOKmodass] = "%="; | |
| 3061 Token::tochars[TOKshlass] = "<<="; | |
| 3062 Token::tochars[TOKshrass] = ">>="; | |
| 3063 Token::tochars[TOKushrass] = ">>>="; | |
| 3064 Token::tochars[TOKandass] = "&="; | |
| 3065 Token::tochars[TOKorass] = "|="; | |
| 3066 Token::tochars[TOKcatass] = "~="; | |
| 3067 Token::tochars[TOKcat] = "~"; | |
| 3068 Token::tochars[TOKcall] = "call"; | |
| 3069 Token::tochars[TOKidentity] = "is"; | |
| 3070 Token::tochars[TOKnotidentity] = "!is"; | |
| 3071 | |
| 3072 Token::tochars[TOKorass] = "|="; | |
| 3073 Token::tochars[TOKidentifier] = "identifier"; | |
| 3074 | |
| 3075 // For debugging | |
| 1587 | 3076 Token::tochars[TOKerror] = "error"; |
| 159 | 3077 Token::tochars[TOKdotexp] = "dotexp"; |
| 3078 Token::tochars[TOKdotti] = "dotti"; | |
| 3079 Token::tochars[TOKdotvar] = "dotvar"; | |
| 3080 Token::tochars[TOKdottype] = "dottype"; | |
| 3081 Token::tochars[TOKsymoff] = "symoff"; | |
| 3082 Token::tochars[TOKarraylength] = "arraylength"; | |
| 3083 Token::tochars[TOKarrayliteral] = "arrayliteral"; | |
| 3084 Token::tochars[TOKassocarrayliteral] = "assocarrayliteral"; | |
| 3085 Token::tochars[TOKstructliteral] = "structliteral"; | |
| 3086 Token::tochars[TOKstring] = "string"; | |
| 3087 Token::tochars[TOKdsymbol] = "symbol"; | |
| 3088 Token::tochars[TOKtuple] = "tuple"; | |
| 3089 Token::tochars[TOKdeclaration] = "declaration"; | |
| 3090 Token::tochars[TOKdottd] = "dottd"; | |
| 3091 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
|
3092 Token::tochars[TOKon_scope_success] = "scope(success)"; |
|
e961851fb8be
Merged DMD 1.042.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
1165
diff
changeset
|
3093 Token::tochars[TOKon_scope_failure] = "scope(failure)"; |
| 159 | 3094 } |