Mercurial > projects > ldc
annotate dmd/lexer.c @ 1137:45d73f0a9b43
Automated merge with http://hg.dsource.org/projects/ldc
| author | Christian Kamm <kamm incasoftware de> |
|---|---|
| date | Tue, 24 Mar 2009 14:34:16 +0100 |
| parents | b30fe7e1dbb9 |
| children | 226c07c71967 |
| rev | line source |
|---|---|
| 159 | 1 |
| 2 // Compiler implementation of the D programming language | |
| 3 // Copyright (c) 1999-2008 by Digital Mars | |
| 4 // All Rights Reserved | |
| 5 // written by Walter Bright | |
| 6 // http://www.digitalmars.com | |
| 7 // License for redistribution is by either the Artistic License | |
| 8 // in artistic.txt, or the GNU General Public License in gnu.txt. | |
| 9 // See the included readme.txt for details. | |
| 10 | |
|
872
aa953cc960b6
Apply BlueZeniX's patch for OpenSolaris compatibility. Fixes #158.
Christian Kamm <kamm incasoftware de>
parents:
846
diff
changeset
|
11 #define __C99FEATURES__ 1 |
|
aa953cc960b6
Apply BlueZeniX's patch for OpenSolaris compatibility. Fixes #158.
Christian Kamm <kamm incasoftware de>
parents:
846
diff
changeset
|
12 |
| 159 | 13 /* Lexical Analyzer */ |
| 14 | |
| 15 #include <stdio.h> | |
| 16 #include <string.h> | |
| 17 #include <ctype.h> | |
| 18 #include <stdarg.h> | |
| 19 #include <errno.h> | |
| 20 #include <wchar.h> | |
| 21 #include <stdlib.h> | |
| 22 #include <assert.h> | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
23 #include <math.h> |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
24 |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
25 #if _MSC_VER |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
26 #include <time.h> |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
27 #else |
| 159 | 28 #include <sys/time.h> |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
29 #endif |
| 159 | 30 |
| 31 #ifdef IN_GCC | |
| 32 #include <time.h> | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
33 #elif __GNUC__ |
| 159 | 34 #include <time.h> |
| 35 #endif | |
| 36 | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
37 #include "rmem.h" |
| 159 | 38 |
| 39 #include "stringtable.h" | |
| 40 | |
| 41 #include "lexer.h" | |
| 42 #include "utf.h" | |
| 43 #include "identifier.h" | |
| 44 #include "id.h" | |
| 45 #include "module.h" | |
| 46 | |
| 47 #if _WIN32 && __DMC__ | |
| 48 // from \dm\src\include\setlocal.h | |
| 49 extern "C" char * __cdecl __locale_decpoint; | |
| 50 #endif | |
| 51 | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
52 #if _MSC_VER // workaround VC++ bug, labels and types should be in separate namespaces |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
53 #define Lstring Lstr |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
54 #endif |
|
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
55 |
| 159 | 56 extern int HtmlNamedEntity(unsigned char *p, int length); |
| 57 | |
| 58 #define LS 0x2028 // UTF line separator | |
| 59 #define PS 0x2029 // UTF paragraph separator | |
| 60 | |
| 61 /******************************************** | |
| 62 * Do our own char maps | |
| 63 */ | |
| 64 | |
| 65 static unsigned char cmtable[256]; | |
| 66 | |
| 67 const int CMoctal = 0x1; | |
| 68 const int CMhex = 0x2; | |
| 69 const int CMidchar = 0x4; | |
| 70 | |
| 71 inline unsigned char isoctal (unsigned char c) { return cmtable[c] & CMoctal; } | |
| 72 inline unsigned char ishex (unsigned char c) { return cmtable[c] & CMhex; } | |
| 73 inline unsigned char isidchar(unsigned char c) { return cmtable[c] & CMidchar; } | |
| 74 | |
| 75 static void cmtable_init() | |
| 76 { | |
| 77 for (unsigned c = 0; c < sizeof(cmtable) / sizeof(cmtable[0]); c++) | |
| 78 { | |
| 79 if ('0' <= c && c <= '7') | |
| 80 cmtable[c] |= CMoctal; | |
| 81 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 82 cmtable[c] |= CMhex; | |
| 83 if (isalnum(c) || c == '_') | |
| 84 cmtable[c] |= CMidchar; | |
| 85 } | |
| 86 } | |
| 87 | |
| 88 | |
| 89 /************************* Token **********************************************/ | |
| 90 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
91 const char *Token::tochars[TOKMAX]; |
| 159 | 92 |
| 93 void *Token::operator new(size_t size) | |
| 94 { Token *t; | |
| 95 | |
| 96 if (Lexer::freelist) | |
| 97 { | |
| 98 t = Lexer::freelist; | |
| 99 Lexer::freelist = t->next; | |
| 100 return t; | |
| 101 } | |
| 102 | |
| 103 return ::operator new(size); | |
| 104 } | |
| 105 | |
| 106 #ifdef DEBUG | |
| 107 void Token::print() | |
| 108 { | |
| 109 fprintf(stdmsg, "%s\n", toChars()); | |
| 110 } | |
| 111 #endif | |
| 112 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
113 const char *Token::toChars() |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
114 { const char *p; |
| 159 | 115 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 116 | |
| 117 p = buffer; | |
| 118 switch (value) | |
| 119 { | |
| 120 case TOKint32v: | |
| 121 #if IN_GCC | |
| 122 sprintf(buffer,"%d",(d_int32)int64value); | |
| 123 #else | |
| 124 sprintf(buffer,"%d",int32value); | |
| 125 #endif | |
| 126 break; | |
| 127 | |
| 128 case TOKuns32v: | |
| 129 case TOKcharv: | |
| 130 case TOKwcharv: | |
| 131 case TOKdcharv: | |
| 132 #if IN_GCC | |
| 133 sprintf(buffer,"%uU",(d_uns32)uns64value); | |
| 134 #else | |
| 135 sprintf(buffer,"%uU",uns32value); | |
| 136 #endif | |
| 137 break; | |
| 138 | |
| 139 case TOKint64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
140 sprintf(buffer,"%jdL",int64value); |
| 159 | 141 break; |
| 142 | |
| 143 case TOKuns64v: | |
|
1103
b30fe7e1dbb9
- Updated to DMD frontend 1.041.
Tomas Lindquist Olsen <tomas.l.olsen gmail.com>
parents:
872
diff
changeset
|
144 sprintf(buffer,"%juUL",uns64value); |
| 159 | 145 break; |
| 146 | |
| 147 #if IN_GCC | |
| 148 case TOKfloat32v: | |
| 149 case TOKfloat64v: | |
| 150 case TOKfloat80v: | |
| 151 float80value.format(buffer, sizeof(buffer)); | |
| 152 break; | |
| 153 case TOKimaginary32v: | |
| 154 case TOKimaginary64v: | |
| 155 case TOKimaginary80v: | |
| 156 float80value.format(buffer, sizeof(buffer)); | |
| 157 // %% buffer | |
| 158 strcat(buffer, "i"); | |
| 159 break; | |
| 160 #else | |
| 161 case TOKfloat32v: | |
| 162 sprintf(buffer,"%Lgf", float80value); | |
| 163 break; | |
| 164 | |
| 165 case TOKfloat64v: | |
| 166 sprintf(buffer,"%Lg", float80value); | |
| 167 break; | |
| 168 | |
| 169 case TOKfloat80v: | |
| 170 sprintf(buffer,"%LgL", float80value); | |
| 171 break; | |
| 172 | |
| 173 case TOKimaginary32v: | |
| 174 sprintf(buffer,"%Lgfi", float80value); | |
| 175 break; | |
| 176 | |
| 177 case TOKimaginary64v: | |
| 178 sprintf(buffer,"%Lgi", float80value); | |
| 179 break; | |
| 180 | |
| 181 case TOKimaginary80v: | |
| 182 sprintf(buffer,"%LgLi", float80value); | |
| 183 break; | |
| 184 #endif | |
| 185 | |
| 186 case TOKstring: | |
| 187 #if CSTRINGS | |
| 188 p = string; | |
| 189 #else | |
| 190 { OutBuffer buf; | |
| 191 | |
| 192 buf.writeByte('"'); | |
| 193 for (size_t i = 0; i < len; ) | |
| 194 { unsigned c; | |
| 195 | |
| 196 utf_decodeChar((unsigned char *)ustring, len, &i, &c); | |
| 197 switch (c) | |
| 198 { | |
| 199 case 0: | |
| 200 break; | |
| 201 | |
| 202 case '"': | |
| 203 case '\\': | |
| 204 buf.writeByte('\\'); | |
| 205 default: | |
| 206 if (isprint(c)) | |
| 207 buf.writeByte(c); | |
| 208 else if (c <= 0x7F) | |
| 209 buf.printf("\\x%02x", c); | |
| 210 else if (c <= 0xFFFF) | |
| 211 buf.printf("\\u%04x", c); | |
| 212 else | |
| 213 buf.printf("\\U%08x", c); | |
| 214 continue; | |
| 215 } | |
| 216 break; | |
| 217 } | |
| 218 buf.writeByte('"'); | |
| 219 if (postfix) | |
| 220 buf.writeByte('"'); | |
| 221 buf.writeByte(0); | |
| 222 p = (char *)buf.extractData(); | |
| 223 } | |
| 224 #endif | |
| 225 break; | |
| 226 | |
| 227 case TOKidentifier: | |
| 228 case TOKenum: | |
| 229 case TOKstruct: | |
| 230 case TOKimport: | |
| 231 CASE_BASIC_TYPES: | |
| 232 p = ident->toChars(); | |
| 233 break; | |
| 234 | |
| 235 default: | |
| 236 p = toChars(value); | |
| 237 break; | |
| 238 } | |
| 239 return p; | |
| 240 } | |
| 241 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
242 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
|
243 { const char *p; |
| 159 | 244 static char buffer[3 + 3 * sizeof(value) + 1]; |
| 245 | |
| 246 p = tochars[value]; | |
| 247 if (!p) | |
| 248 { sprintf(buffer,"TOK%d",value); | |
| 249 p = buffer; | |
| 250 } | |
| 251 return p; | |
| 252 } | |
| 253 | |
| 254 /*************************** Lexer ********************************************/ | |
| 255 | |
| 256 Token *Lexer::freelist = NULL; | |
| 257 StringTable Lexer::stringtable; | |
| 258 OutBuffer Lexer::stringbuffer; | |
| 259 | |
| 260 Lexer::Lexer(Module *mod, | |
| 261 unsigned char *base, unsigned begoffset, unsigned endoffset, | |
| 262 int doDocComment, int commentToken) | |
| 263 : loc(mod, 1) | |
| 264 { | |
| 265 //printf("Lexer::Lexer(%p,%d)\n",base,length); | |
| 266 //printf("lexer.mod = %p, %p\n", mod, this->loc.mod); | |
| 267 memset(&token,0,sizeof(token)); | |
| 268 this->base = base; | |
| 269 this->end = base + endoffset; | |
| 270 p = base + begoffset; | |
| 271 this->mod = mod; | |
| 272 this->doDocComment = doDocComment; | |
| 273 this->anyToken = 0; | |
| 274 this->commentToken = commentToken; | |
| 275 //initKeywords(); | |
| 276 | |
| 277 /* If first line starts with '#!', ignore the line | |
| 278 */ | |
| 279 | |
| 280 if (p[0] == '#' && p[1] =='!') | |
| 281 { | |
| 282 p += 2; | |
| 283 while (1) | |
| 284 { unsigned char c = *p; | |
| 285 switch (c) | |
| 286 { | |
| 287 case '\n': | |
| 288 p++; | |
| 289 break; | |
| 290 | |
| 291 case '\r': | |
| 292 p++; | |
| 293 if (*p == '\n') | |
| 294 p++; | |
| 295 break; | |
| 296 | |
| 297 case 0: | |
| 298 case 0x1A: | |
| 299 break; | |
| 300 | |
| 301 default: | |
| 302 if (c & 0x80) | |
| 303 { unsigned u = decodeUTF(); | |
| 304 if (u == PS || u == LS) | |
| 305 break; | |
| 306 } | |
| 307 p++; | |
| 308 continue; | |
| 309 } | |
| 310 break; | |
| 311 } | |
| 312 loc.linnum = 2; | |
| 313 } | |
| 314 } | |
| 315 | |
| 316 | |
| 317 void Lexer::error(const char *format, ...) | |
| 318 { | |
| 319 if (mod && !global.gag) | |
| 320 { | |
| 321 char *p = loc.toChars(); | |
| 322 if (*p) | |
| 323 fprintf(stdmsg, "%s: ", p); | |
| 324 mem.free(p); | |
| 325 | |
| 326 va_list ap; | |
| 327 va_start(ap, format); | |
| 328 vfprintf(stdmsg, format, ap); | |
| 329 va_end(ap); | |
| 330 | |
| 331 fprintf(stdmsg, "\n"); | |
| 332 fflush(stdmsg); | |
| 333 | |
| 334 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 335 fatal(); | |
| 336 } | |
| 337 global.errors++; | |
| 338 } | |
| 339 | |
| 340 void Lexer::error(Loc loc, const char *format, ...) | |
| 341 { | |
| 342 if (mod && !global.gag) | |
| 343 { | |
| 344 char *p = loc.toChars(); | |
| 345 if (*p) | |
| 346 fprintf(stdmsg, "%s: ", p); | |
| 347 mem.free(p); | |
| 348 | |
| 349 va_list ap; | |
| 350 va_start(ap, format); | |
| 351 vfprintf(stdmsg, format, ap); | |
| 352 va_end(ap); | |
| 353 | |
| 354 fprintf(stdmsg, "\n"); | |
| 355 fflush(stdmsg); | |
| 356 | |
| 357 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 358 fatal(); | |
| 359 } | |
| 360 global.errors++; | |
| 361 } | |
| 362 | |
| 363 TOK Lexer::nextToken() | |
| 364 { Token *t; | |
| 365 | |
| 366 if (token.next) | |
| 367 { | |
| 368 t = token.next; | |
| 369 memcpy(&token,t,sizeof(Token)); | |
| 370 t->next = freelist; | |
| 371 freelist = t; | |
| 372 } | |
| 373 else | |
| 374 { | |
| 375 scan(&token); | |
| 376 } | |
| 377 //token.print(); | |
| 378 return token.value; | |
| 379 } | |
| 380 | |
| 381 Token *Lexer::peek(Token *ct) | |
| 382 { Token *t; | |
| 383 | |
| 384 if (ct->next) | |
| 385 t = ct->next; | |
| 386 else | |
| 387 { | |
| 388 t = new Token(); | |
| 389 scan(t); | |
| 390 t->next = NULL; | |
| 391 ct->next = t; | |
| 392 } | |
| 393 return t; | |
| 394 } | |
| 395 | |
|
717
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
396 /*********************** |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
397 * Look ahead at next token's value. |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
398 */ |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
399 |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
400 TOK Lexer::peekNext() |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
401 { |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
402 return peek(&token)->value; |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
403 } |
|
a26b0c5d5942
Merged DMD 1.036.
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
697
diff
changeset
|
404 |
| 159 | 405 /********************************* |
| 406 * tk is on the opening (. | |
| 407 * Look ahead and return token that is past the closing ). | |
| 408 */ | |
| 409 | |
| 410 Token *Lexer::peekPastParen(Token *tk) | |
| 411 { | |
| 412 //printf("peekPastParen()\n"); | |
| 413 int parens = 1; | |
| 414 int curlynest = 0; | |
| 415 while (1) | |
| 416 { | |
| 417 tk = peek(tk); | |
| 418 //tk->print(); | |
| 419 switch (tk->value) | |
| 420 { | |
| 421 case TOKlparen: | |
| 422 parens++; | |
| 423 continue; | |
| 424 | |
| 425 case TOKrparen: | |
| 426 --parens; | |
| 427 if (parens) | |
| 428 continue; | |
| 429 tk = peek(tk); | |
| 430 break; | |
| 431 | |
| 432 case TOKlcurly: | |
| 433 curlynest++; | |
| 434 continue; | |
| 435 | |
| 436 case TOKrcurly: | |
| 437 if (--curlynest >= 0) | |
| 438 continue; | |
| 439 break; | |
| 440 | |
| 441 case TOKsemicolon: | |
| 442 if (curlynest) | |
| 443 continue; | |
| 444 break; | |
| 445 | |
| 446 case TOKeof: | |
| 447 break; | |
| 448 | |
| 449 default: | |
| 450 continue; | |
| 451 } | |
| 452 return tk; | |
| 453 } | |
| 454 } | |
| 455 | |
| 456 /********************************** | |
| 457 * Determine if string is a valid Identifier. | |
| 458 * Placed here because of commonality with Lexer functionality. | |
| 459 * Returns: | |
| 460 * 0 invalid | |
| 461 */ | |
| 462 | |
| 463 int Lexer::isValidIdentifier(char *p) | |
| 464 { | |
| 465 size_t len; | |
| 466 size_t idx; | |
| 467 | |
| 468 if (!p || !*p) | |
| 469 goto Linvalid; | |
| 470 | |
| 471 if (*p >= '0' && *p <= '9') // beware of isdigit() on signed chars | |
| 472 goto Linvalid; | |
| 473 | |
| 474 len = strlen(p); | |
| 475 idx = 0; | |
| 476 while (p[idx]) | |
| 477 { dchar_t dc; | |
| 478 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
479 const char *q = utf_decodeChar((unsigned char *)p, len, &idx, &dc); |
| 159 | 480 if (q) |
| 481 goto Linvalid; | |
| 482 | |
| 483 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) | |
| 484 goto Linvalid; | |
| 485 } | |
| 486 return 1; | |
| 487 | |
| 488 Linvalid: | |
| 489 return 0; | |
| 490 } | |
| 491 | |
| 492 /**************************** | |
| 493 * Turn next token in buffer into a token. | |
| 494 */ | |
| 495 | |
| 496 void Lexer::scan(Token *t) | |
| 497 { | |
| 498 unsigned lastLine = loc.linnum; | |
| 499 unsigned linnum; | |
| 500 | |
| 501 t->blockComment = NULL; | |
| 502 t->lineComment = NULL; | |
| 503 while (1) | |
| 504 { | |
| 505 t->ptr = p; | |
| 506 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 507 switch (*p) | |
| 508 { | |
| 509 case 0: | |
| 510 case 0x1A: | |
| 511 t->value = TOKeof; // end of file | |
| 512 return; | |
| 513 | |
| 514 case ' ': | |
| 515 case '\t': | |
| 516 case '\v': | |
| 517 case '\f': | |
| 518 p++; | |
| 519 continue; // skip white space | |
| 520 | |
| 521 case '\r': | |
| 522 p++; | |
| 523 if (*p != '\n') // if CR stands by itself | |
| 524 loc.linnum++; | |
| 525 continue; // skip white space | |
| 526 | |
| 527 case '\n': | |
| 528 p++; | |
| 529 loc.linnum++; | |
| 530 continue; // skip white space | |
| 531 | |
| 532 case '0': case '1': case '2': case '3': case '4': | |
| 533 case '5': case '6': case '7': case '8': case '9': | |
| 534 t->value = number(t); | |
| 535 return; | |
| 536 | |
| 537 #if CSTRINGS | |
| 538 case '\'': | |
| 539 t->value = charConstant(t, 0); | |
| 540 return; | |
| 541 | |
| 542 case '"': | |
| 543 t->value = stringConstant(t,0); | |
| 544 return; | |
| 545 | |
| 546 case 'l': | |
| 547 case 'L': | |
| 548 if (p[1] == '\'') | |
| 549 { | |
| 550 p++; | |
| 551 t->value = charConstant(t, 1); | |
| 552 return; | |
| 553 } | |
| 554 else if (p[1] == '"') | |
| 555 { | |
| 556 p++; | |
| 557 t->value = stringConstant(t, 1); | |
| 558 return; | |
| 559 } | |
| 560 #else | |
| 561 case '\'': | |
| 562 t->value = charConstant(t,0); | |
| 563 return; | |
| 564 | |
| 565 case 'r': | |
| 566 if (p[1] != '"') | |
| 567 goto case_ident; | |
| 568 p++; | |
| 569 case '`': | |
| 570 t->value = wysiwygStringConstant(t, *p); | |
| 571 return; | |
| 572 | |
| 573 case 'x': | |
| 574 if (p[1] != '"') | |
| 575 goto case_ident; | |
| 576 p++; | |
| 577 t->value = hexStringConstant(t); | |
| 578 return; | |
| 579 | |
| 336 | 580 #if DMDV2 |
| 159 | 581 case 'q': |
| 582 if (p[1] == '"') | |
| 583 { | |
| 584 p++; | |
| 585 t->value = delimitedStringConstant(t); | |
| 586 return; | |
| 587 } | |
| 588 else if (p[1] == '{') | |
| 589 { | |
| 590 p++; | |
| 591 t->value = tokenStringConstant(t); | |
| 592 return; | |
| 593 } | |
| 594 else | |
| 595 goto case_ident; | |
| 596 #endif | |
| 597 | |
| 598 case '"': | |
| 599 t->value = escapeStringConstant(t,0); | |
| 600 return; | |
| 601 | |
| 602 case '\\': // escaped string literal | |
| 603 { unsigned c; | |
| 604 | |
| 605 stringbuffer.reset(); | |
| 606 do | |
| 607 { | |
| 608 p++; | |
| 609 switch (*p) | |
| 610 { | |
| 611 case 'u': | |
| 612 case 'U': | |
| 613 case '&': | |
| 614 c = escapeSequence(); | |
| 615 stringbuffer.writeUTF8(c); | |
| 616 break; | |
| 617 | |
| 618 default: | |
| 619 c = escapeSequence(); | |
| 620 stringbuffer.writeByte(c); | |
| 621 break; | |
| 622 } | |
| 623 } while (*p == '\\'); | |
| 624 t->len = stringbuffer.offset; | |
| 625 stringbuffer.writeByte(0); | |
| 626 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 627 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 628 t->postfix = 0; | |
| 629 t->value = TOKstring; | |
| 630 return; | |
| 631 } | |
| 632 | |
| 633 case 'l': | |
| 634 case 'L': | |
| 635 #endif | |
| 636 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 637 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 638 case 'k': case 'm': case 'n': case 'o': | |
| 336 | 639 #if DMDV2 |
| 159 | 640 case 'p': /*case 'q': case 'r':*/ case 's': case 't': |
| 641 #else | |
| 642 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 643 #endif | |
| 644 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 645 case 'z': | |
| 646 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 647 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 648 case 'K': case 'M': case 'N': case 'O': | |
| 649 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 650 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 651 case 'Z': | |
| 652 case '_': | |
| 653 case_ident: | |
| 654 { unsigned char c; | |
| 655 StringValue *sv; | |
| 656 Identifier *id; | |
| 657 | |
| 658 do | |
| 659 { | |
| 660 c = *++p; | |
| 661 } while (isidchar(c) || (c & 0x80 && isUniAlpha(decodeUTF()))); | |
| 662 sv = stringtable.update((char *)t->ptr, p - t->ptr); | |
| 663 id = (Identifier *) sv->ptrvalue; | |
| 664 if (!id) | |
| 665 { id = new Identifier(sv->lstring.string,TOKidentifier); | |
| 666 sv->ptrvalue = id; | |
| 667 } | |
| 668 t->ident = id; | |
| 669 t->value = (enum TOK) id->value; | |
| 670 anyToken = 1; | |
| 671 if (*t->ptr == '_') // if special identifier token | |
| 672 { | |
| 673 static char date[11+1]; | |
| 674 static char time[8+1]; | |
| 675 static char timestamp[24+1]; | |
| 676 | |
| 677 if (!date[0]) // lazy evaluation | |
| 678 { time_t t; | |
| 679 char *p; | |
| 680 | |
| 681 ::time(&t); | |
| 682 p = ctime(&t); | |
| 683 assert(p); | |
| 684 sprintf(date, "%.6s %.4s", p + 4, p + 20); | |
| 685 sprintf(time, "%.8s", p + 11); | |
| 686 sprintf(timestamp, "%.24s", p); | |
| 687 } | |
| 688 | |
| 336 | 689 #if DMDV1 |
| 159 | 690 if (mod && id == Id::FILE) |
| 691 { | |
| 692 t->ustring = (unsigned char *)(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 693 goto Lstring; | |
| 694 } | |
| 695 else if (mod && id == Id::LINE) | |
| 696 { | |
| 697 t->value = TOKint64v; | |
| 698 t->uns64value = loc.linnum; | |
| 699 } | |
| 336 | 700 else |
| 701 #endif | |
| 702 if (id == Id::DATE) | |
| 159 | 703 { |
| 704 t->ustring = (unsigned char *)date; | |
| 705 goto Lstring; | |
| 706 } | |
| 707 else if (id == Id::TIME) | |
| 708 { | |
| 709 t->ustring = (unsigned char *)time; | |
| 710 goto Lstring; | |
| 711 } | |
| 712 else if (id == Id::VENDOR) | |
| 713 { | |
|
664
eef8ac26c66c
Some missed LLVMDC -> LDC.
Christian Kamm <kamm incasoftware de>
parents:
658
diff
changeset
|
714 t->ustring = (unsigned char *)"LDC"; |
| 159 | 715 goto Lstring; |
| 716 } | |
| 717 else if (id == Id::TIMESTAMP) | |
| 718 { | |
| 719 t->ustring = (unsigned char *)timestamp; | |
| 720 Lstring: | |
| 721 t->value = TOKstring; | |
| 722 Llen: | |
| 723 t->postfix = 0; | |
| 724 t->len = strlen((char *)t->ustring); | |
| 725 } | |
| 726 else if (id == Id::VERSIONX) | |
| 727 { unsigned major = 0; | |
| 728 unsigned minor = 0; | |
| 729 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
730 for (const char *p = global.version + 1; 1; p++) |
| 159 | 731 { |
| 732 char c = *p; | |
| 733 if (isdigit(c)) | |
| 734 minor = minor * 10 + c - '0'; | |
| 735 else if (c == '.') | |
| 736 { major = minor; | |
| 737 minor = 0; | |
| 738 } | |
| 739 else | |
| 740 break; | |
| 741 } | |
| 742 t->value = TOKint64v; | |
| 743 t->uns64value = major * 1000 + minor; | |
| 744 } | |
| 336 | 745 #if DMDV2 |
| 159 | 746 else if (id == Id::EOFX) |
| 747 { | |
| 748 t->value = TOKeof; | |
| 749 // Advance scanner to end of file | |
| 750 while (!(*p == 0 || *p == 0x1A)) | |
| 751 p++; | |
| 752 } | |
| 753 #endif | |
| 754 } | |
| 755 //printf("t->value = %d\n",t->value); | |
| 756 return; | |
| 757 } | |
| 758 | |
| 759 case '/': | |
| 760 p++; | |
| 761 switch (*p) | |
| 762 { | |
| 763 case '=': | |
| 764 p++; | |
| 765 t->value = TOKdivass; | |
| 766 return; | |
| 767 | |
| 768 case '*': | |
| 769 p++; | |
| 770 linnum = loc.linnum; | |
| 771 while (1) | |
| 772 { | |
| 773 while (1) | |
| 774 { unsigned char c = *p; | |
| 775 switch (c) | |
| 776 { | |
| 777 case '/': | |
| 778 break; | |
| 779 | |
| 780 case '\n': | |
| 781 loc.linnum++; | |
| 782 p++; | |
| 783 continue; | |
| 784 | |
| 785 case '\r': | |
| 786 p++; | |
| 787 if (*p != '\n') | |
| 788 loc.linnum++; | |
| 789 continue; | |
| 790 | |
| 791 case 0: | |
| 792 case 0x1A: | |
| 793 error("unterminated /* */ comment"); | |
| 794 p = end; | |
| 795 t->value = TOKeof; | |
| 796 return; | |
| 797 | |
| 798 default: | |
| 799 if (c & 0x80) | |
| 800 { unsigned u = decodeUTF(); | |
| 801 if (u == PS || u == LS) | |
| 802 loc.linnum++; | |
| 803 } | |
| 804 p++; | |
| 805 continue; | |
| 806 } | |
| 807 break; | |
| 808 } | |
| 809 p++; | |
| 810 if (p[-2] == '*' && p - 3 != t->ptr) | |
| 811 break; | |
| 812 } | |
| 813 if (commentToken) | |
| 814 { | |
| 815 t->value = TOKcomment; | |
| 816 return; | |
| 817 } | |
| 818 else if (doDocComment && t->ptr[2] == '*' && p - 4 != t->ptr) | |
| 819 { // if /** but not /**/ | |
| 820 getDocComment(t, lastLine == linnum); | |
| 821 } | |
| 822 continue; | |
| 823 | |
| 824 case '/': // do // style comments | |
| 825 linnum = loc.linnum; | |
| 826 while (1) | |
| 827 { unsigned char c = *++p; | |
| 828 switch (c) | |
| 829 { | |
| 830 case '\n': | |
| 831 break; | |
| 832 | |
| 833 case '\r': | |
| 834 if (p[1] == '\n') | |
| 835 p++; | |
| 836 break; | |
| 837 | |
| 838 case 0: | |
| 839 case 0x1A: | |
| 840 if (commentToken) | |
| 841 { | |
| 842 p = end; | |
| 843 t->value = TOKcomment; | |
| 844 return; | |
| 845 } | |
| 846 if (doDocComment && t->ptr[2] == '/') | |
| 847 getDocComment(t, lastLine == linnum); | |
| 848 p = end; | |
| 849 t->value = TOKeof; | |
| 850 return; | |
| 851 | |
| 852 default: | |
| 853 if (c & 0x80) | |
| 854 { unsigned u = decodeUTF(); | |
| 855 if (u == PS || u == LS) | |
| 856 break; | |
| 857 } | |
| 858 continue; | |
| 859 } | |
| 860 break; | |
| 861 } | |
| 862 | |
| 863 if (commentToken) | |
| 864 { | |
| 865 p++; | |
| 866 loc.linnum++; | |
| 867 t->value = TOKcomment; | |
| 868 return; | |
| 869 } | |
| 870 if (doDocComment && t->ptr[2] == '/') | |
| 871 getDocComment(t, lastLine == linnum); | |
| 872 | |
| 873 p++; | |
| 874 loc.linnum++; | |
| 875 continue; | |
| 876 | |
| 877 case '+': | |
| 878 { int nest; | |
| 879 | |
| 880 linnum = loc.linnum; | |
| 881 p++; | |
| 882 nest = 1; | |
| 883 while (1) | |
| 884 { unsigned char c = *p; | |
| 885 switch (c) | |
| 886 { | |
| 887 case '/': | |
| 888 p++; | |
| 889 if (*p == '+') | |
| 890 { | |
| 891 p++; | |
| 892 nest++; | |
| 893 } | |
| 894 continue; | |
| 895 | |
| 896 case '+': | |
| 897 p++; | |
| 898 if (*p == '/') | |
| 899 { | |
| 900 p++; | |
| 901 if (--nest == 0) | |
| 902 break; | |
| 903 } | |
| 904 continue; | |
| 905 | |
| 906 case '\r': | |
| 907 p++; | |
| 908 if (*p != '\n') | |
| 909 loc.linnum++; | |
| 910 continue; | |
| 911 | |
| 912 case '\n': | |
| 913 loc.linnum++; | |
| 914 p++; | |
| 915 continue; | |
| 916 | |
| 917 case 0: | |
| 918 case 0x1A: | |
| 919 error("unterminated /+ +/ comment"); | |
| 920 p = end; | |
| 921 t->value = TOKeof; | |
| 922 return; | |
| 923 | |
| 924 default: | |
| 925 if (c & 0x80) | |
| 926 { unsigned u = decodeUTF(); | |
| 927 if (u == PS || u == LS) | |
| 928 loc.linnum++; | |
| 929 } | |
| 930 p++; | |
| 931 continue; | |
| 932 } | |
| 933 break; | |
| 934 } | |
| 935 if (commentToken) | |
| 936 { | |
| 937 t->value = TOKcomment; | |
| 938 return; | |
| 939 } | |
| 940 if (doDocComment && t->ptr[2] == '+' && p - 4 != t->ptr) | |
| 941 { // if /++ but not /++/ | |
| 942 getDocComment(t, lastLine == linnum); | |
| 943 } | |
| 944 continue; | |
| 945 } | |
| 946 } | |
| 947 t->value = TOKdiv; | |
| 948 return; | |
| 949 | |
| 950 case '.': | |
| 951 p++; | |
| 952 if (isdigit(*p)) | |
| 953 { /* Note that we don't allow ._1 and ._ as being | |
| 954 * valid floating point numbers. | |
| 955 */ | |
| 956 p--; | |
| 957 t->value = inreal(t); | |
| 958 } | |
| 959 else if (p[0] == '.') | |
| 960 { | |
| 961 if (p[1] == '.') | |
| 962 { p += 2; | |
| 963 t->value = TOKdotdotdot; | |
| 964 } | |
| 965 else | |
| 966 { p++; | |
| 967 t->value = TOKslice; | |
| 968 } | |
| 969 } | |
| 970 else | |
| 971 t->value = TOKdot; | |
| 972 return; | |
| 973 | |
| 974 case '&': | |
| 975 p++; | |
| 976 if (*p == '=') | |
| 977 { p++; | |
| 978 t->value = TOKandass; | |
| 979 } | |
| 980 else if (*p == '&') | |
| 981 { p++; | |
| 982 t->value = TOKandand; | |
| 983 } | |
| 984 else | |
| 985 t->value = TOKand; | |
| 986 return; | |
| 987 | |
| 988 case '|': | |
| 989 p++; | |
| 990 if (*p == '=') | |
| 991 { p++; | |
| 992 t->value = TOKorass; | |
| 993 } | |
| 994 else if (*p == '|') | |
| 995 { p++; | |
| 996 t->value = TOKoror; | |
| 997 } | |
| 998 else | |
| 999 t->value = TOKor; | |
| 1000 return; | |
| 1001 | |
| 1002 case '-': | |
| 1003 p++; | |
| 1004 if (*p == '=') | |
| 1005 { p++; | |
| 1006 t->value = TOKminass; | |
| 1007 } | |
| 1008 #if 0 | |
| 1009 else if (*p == '>') | |
| 1010 { p++; | |
| 1011 t->value = TOKarrow; | |
| 1012 } | |
| 1013 #endif | |
| 1014 else if (*p == '-') | |
| 1015 { p++; | |
| 1016 t->value = TOKminusminus; | |
| 1017 } | |
| 1018 else | |
| 1019 t->value = TOKmin; | |
| 1020 return; | |
| 1021 | |
| 1022 case '+': | |
| 1023 p++; | |
| 1024 if (*p == '=') | |
| 1025 { p++; | |
| 1026 t->value = TOKaddass; | |
| 1027 } | |
| 1028 else if (*p == '+') | |
| 1029 { p++; | |
| 1030 t->value = TOKplusplus; | |
| 1031 } | |
| 1032 else | |
| 1033 t->value = TOKadd; | |
| 1034 return; | |
| 1035 | |
| 1036 case '<': | |
| 1037 p++; | |
| 1038 if (*p == '=') | |
| 1039 { p++; | |
| 1040 t->value = TOKle; // <= | |
| 1041 } | |
| 1042 else if (*p == '<') | |
| 1043 { p++; | |
| 1044 if (*p == '=') | |
| 1045 { p++; | |
| 1046 t->value = TOKshlass; // <<= | |
| 1047 } | |
| 1048 else | |
| 1049 t->value = TOKshl; // << | |
| 1050 } | |
| 1051 else if (*p == '>') | |
| 1052 { p++; | |
| 1053 if (*p == '=') | |
| 1054 { p++; | |
| 1055 t->value = TOKleg; // <>= | |
| 1056 } | |
| 1057 else | |
| 1058 t->value = TOKlg; // <> | |
| 1059 } | |
| 1060 else | |
| 1061 t->value = TOKlt; // < | |
| 1062 return; | |
| 1063 | |
| 1064 case '>': | |
| 1065 p++; | |
| 1066 if (*p == '=') | |
| 1067 { p++; | |
| 1068 t->value = TOKge; // >= | |
| 1069 } | |
| 1070 else if (*p == '>') | |
| 1071 { p++; | |
| 1072 if (*p == '=') | |
| 1073 { p++; | |
| 1074 t->value = TOKshrass; // >>= | |
| 1075 } | |
| 1076 else if (*p == '>') | |
| 1077 { p++; | |
| 1078 if (*p == '=') | |
| 1079 { p++; | |
| 1080 t->value = TOKushrass; // >>>= | |
| 1081 } | |
| 1082 else | |
| 1083 t->value = TOKushr; // >>> | |
| 1084 } | |
| 1085 else | |
| 1086 t->value = TOKshr; // >> | |
| 1087 } | |
| 1088 else | |
| 1089 t->value = TOKgt; // > | |
| 1090 return; | |
| 1091 | |
| 1092 case '!': | |
| 1093 p++; | |
| 1094 if (*p == '=') | |
| 1095 { p++; | |
| 1096 if (*p == '=' && global.params.Dversion == 1) | |
| 1097 { p++; | |
| 1098 t->value = TOKnotidentity; // !== | |
| 1099 } | |
| 1100 else | |
| 1101 t->value = TOKnotequal; // != | |
| 1102 } | |
| 1103 else if (*p == '<') | |
| 1104 { p++; | |
| 1105 if (*p == '>') | |
| 1106 { p++; | |
| 1107 if (*p == '=') | |
| 1108 { p++; | |
| 1109 t->value = TOKunord; // !<>= | |
| 1110 } | |
| 1111 else | |
| 1112 t->value = TOKue; // !<> | |
| 1113 } | |
| 1114 else if (*p == '=') | |
| 1115 { p++; | |
| 1116 t->value = TOKug; // !<= | |
| 1117 } | |
| 1118 else | |
| 1119 t->value = TOKuge; // !< | |
| 1120 } | |
| 1121 else if (*p == '>') | |
| 1122 { p++; | |
| 1123 if (*p == '=') | |
| 1124 { p++; | |
| 1125 t->value = TOKul; // !>= | |
| 1126 } | |
| 1127 else | |
| 1128 t->value = TOKule; // !> | |
| 1129 } | |
| 1130 else | |
| 1131 t->value = TOKnot; // ! | |
| 1132 return; | |
| 1133 | |
| 1134 case '=': | |
| 1135 p++; | |
| 1136 if (*p == '=') | |
| 1137 { p++; | |
| 1138 if (*p == '=' && global.params.Dversion == 1) | |
| 1139 { p++; | |
| 1140 t->value = TOKidentity; // === | |
| 1141 } | |
| 1142 else | |
| 1143 t->value = TOKequal; // == | |
| 1144 } | |
| 1145 else | |
| 1146 t->value = TOKassign; // = | |
| 1147 return; | |
| 1148 | |
| 1149 case '~': | |
| 1150 p++; | |
| 1151 if (*p == '=') | |
| 1152 { p++; | |
| 1153 t->value = TOKcatass; // ~= | |
| 1154 } | |
| 1155 else | |
| 1156 t->value = TOKtilde; // ~ | |
| 1157 return; | |
| 1158 | |
| 1159 #define SINGLE(c,tok) case c: p++; t->value = tok; return; | |
| 1160 | |
| 1161 SINGLE('(', TOKlparen) | |
| 1162 SINGLE(')', TOKrparen) | |
| 1163 SINGLE('[', TOKlbracket) | |
| 1164 SINGLE(']', TOKrbracket) | |
| 1165 SINGLE('{', TOKlcurly) | |
| 1166 SINGLE('}', TOKrcurly) | |
| 1167 SINGLE('?', TOKquestion) | |
| 1168 SINGLE(',', TOKcomma) | |
| 1169 SINGLE(';', TOKsemicolon) | |
| 1170 SINGLE(':', TOKcolon) | |
| 1171 SINGLE('$', TOKdollar) | |
| 1172 | |
| 1173 #undef SINGLE | |
| 1174 | |
| 1175 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1176 case c1: \ | |
| 1177 p++; \ | |
| 1178 if (*p == c2) \ | |
| 1179 { p++; \ | |
| 1180 t->value = tok2; \ | |
| 1181 } \ | |
| 1182 else \ | |
| 1183 t->value = tok1; \ | |
| 1184 return; | |
| 1185 | |
| 1186 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1187 DOUBLE('%', TOKmod, '=', TOKmodass) | |
| 1188 DOUBLE('^', TOKxor, '=', TOKxorass) | |
| 1189 | |
| 1190 #undef DOUBLE | |
| 1191 | |
| 1192 case '#': | |
| 1193 p++; | |
| 1194 pragma(); | |
| 1195 continue; | |
| 1196 | |
| 1197 default: | |
| 1198 { unsigned char c = *p; | |
| 1199 | |
| 1200 if (c & 0x80) | |
| 1201 { unsigned u = decodeUTF(); | |
| 1202 | |
| 1203 // Check for start of unicode identifier | |
| 1204 if (isUniAlpha(u)) | |
| 1205 goto case_ident; | |
| 1206 | |
| 1207 if (u == PS || u == LS) | |
| 1208 { | |
| 1209 loc.linnum++; | |
| 1210 p++; | |
| 1211 continue; | |
| 1212 } | |
| 1213 } | |
| 1214 if (isprint(c)) | |
| 1215 error("unsupported char '%c'", c); | |
| 1216 else | |
| 1217 error("unsupported char 0x%02x", c); | |
| 1218 p++; | |
| 1219 continue; | |
| 1220 } | |
| 1221 } | |
| 1222 } | |
| 1223 } | |
| 1224 | |
| 1225 /******************************************* | |
| 1226 * Parse escape sequence. | |
| 1227 */ | |
| 1228 | |
| 1229 unsigned Lexer::escapeSequence() | |
| 1230 { unsigned c; | |
| 1231 int n; | |
| 1232 int ndigits; | |
| 1233 | |
| 1234 c = *p; | |
| 1235 switch (c) | |
| 1236 { | |
| 1237 case '\'': | |
| 1238 case '"': | |
| 1239 case '?': | |
| 1240 case '\\': | |
| 1241 Lconsume: | |
| 1242 p++; | |
| 1243 break; | |
| 1244 | |
| 1245 case 'a': c = 7; goto Lconsume; | |
| 1246 case 'b': c = 8; goto Lconsume; | |
| 1247 case 'f': c = 12; goto Lconsume; | |
| 1248 case 'n': c = 10; goto Lconsume; | |
| 1249 case 'r': c = 13; goto Lconsume; | |
| 1250 case 't': c = 9; goto Lconsume; | |
| 1251 case 'v': c = 11; goto Lconsume; | |
| 1252 | |
| 1253 case 'u': | |
| 1254 ndigits = 4; | |
| 1255 goto Lhex; | |
| 1256 case 'U': | |
| 1257 ndigits = 8; | |
| 1258 goto Lhex; | |
| 1259 case 'x': | |
| 1260 ndigits = 2; | |
| 1261 Lhex: | |
| 1262 p++; | |
| 1263 c = *p; | |
| 1264 if (ishex(c)) | |
| 1265 { unsigned v; | |
| 1266 | |
| 1267 n = 0; | |
| 1268 v = 0; | |
| 1269 while (1) | |
| 1270 { | |
| 1271 if (isdigit(c)) | |
| 1272 c -= '0'; | |
| 1273 else if (islower(c)) | |
| 1274 c -= 'a' - 10; | |
| 1275 else | |
| 1276 c -= 'A' - 10; | |
| 1277 v = v * 16 + c; | |
| 1278 c = *++p; | |
| 1279 if (++n == ndigits) | |
| 1280 break; | |
| 1281 if (!ishex(c)) | |
| 1282 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1283 break; | |
| 1284 } | |
| 1285 } | |
| 1286 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1287 error("invalid UTF character \\U%08x", v); | |
| 1288 c = v; | |
| 1289 } | |
| 1290 else | |
| 1291 error("undefined escape hex sequence \\%c\n",c); | |
| 1292 break; | |
| 1293 | |
| 1294 case '&': // named character entity | |
| 1295 for (unsigned char *idstart = ++p; 1; p++) | |
| 1296 { | |
| 1297 switch (*p) | |
| 1298 { | |
| 1299 case ';': | |
| 1300 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1301 if (c == ~0) | |
| 1302 { error("unnamed character entity &%.*s;", (int)(p - idstart), idstart); | |
| 1303 c = ' '; | |
| 1304 } | |
| 1305 p++; | |
| 1306 break; | |
| 1307 | |
| 1308 default: | |
| 1309 if (isalpha(*p) || | |
| 1310 (p != idstart + 1 && isdigit(*p))) | |
| 1311 continue; | |
| 1312 error("unterminated named entity"); | |
| 1313 break; | |
| 1314 } | |
| 1315 break; | |
| 1316 } | |
| 1317 break; | |
| 1318 | |
| 1319 case 0: | |
| 1320 case 0x1A: // end of file | |
| 1321 c = '\\'; | |
| 1322 break; | |
| 1323 | |
| 1324 default: | |
| 1325 if (isoctal(c)) | |
| 1326 { unsigned v; | |
| 1327 | |
| 1328 n = 0; | |
| 1329 v = 0; | |
| 1330 do | |
| 1331 { | |
| 1332 v = v * 8 + (c - '0'); | |
| 1333 c = *++p; | |
| 1334 } while (++n < 3 && isoctal(c)); | |
| 1335 c = v; | |
| 1336 if (c > 0xFF) | |
| 1337 error("0%03o is larger than a byte", c); | |
| 1338 } | |
| 1339 else | |
| 1340 error("undefined escape sequence \\%c\n",c); | |
| 1341 break; | |
| 1342 } | |
| 1343 return c; | |
| 1344 } | |
| 1345 | |
| 1346 /************************************** | |
| 1347 */ | |
| 1348 | |
| 1349 TOK Lexer::wysiwygStringConstant(Token *t, int tc) | |
| 1350 { unsigned c; | |
| 1351 Loc start = loc; | |
| 1352 | |
| 1353 p++; | |
| 1354 stringbuffer.reset(); | |
| 1355 while (1) | |
| 1356 { | |
| 1357 c = *p++; | |
| 1358 switch (c) | |
| 1359 { | |
| 1360 case '\n': | |
| 1361 loc.linnum++; | |
| 1362 break; | |
| 1363 | |
| 1364 case '\r': | |
| 1365 if (*p == '\n') | |
| 1366 continue; // ignore | |
| 1367 c = '\n'; // treat EndOfLine as \n character | |
| 1368 loc.linnum++; | |
| 1369 break; | |
| 1370 | |
| 1371 case 0: | |
| 1372 case 0x1A: | |
| 1373 error("unterminated string constant starting at %s", start.toChars()); | |
| 1374 t->ustring = (unsigned char *)""; | |
| 1375 t->len = 0; | |
| 1376 t->postfix = 0; | |
| 1377 return TOKstring; | |
| 1378 | |
| 1379 case '"': | |
| 1380 case '`': | |
| 1381 if (c == tc) | |
| 1382 { | |
| 1383 t->len = stringbuffer.offset; | |
| 1384 stringbuffer.writeByte(0); | |
| 1385 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1386 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1387 stringPostfix(t); | |
| 1388 return TOKstring; | |
| 1389 } | |
| 1390 break; | |
| 1391 | |
| 1392 default: | |
| 1393 if (c & 0x80) | |
| 1394 { p--; | |
| 1395 unsigned u = decodeUTF(); | |
| 1396 p++; | |
| 1397 if (u == PS || u == LS) | |
| 1398 loc.linnum++; | |
| 1399 stringbuffer.writeUTF8(u); | |
| 1400 continue; | |
| 1401 } | |
| 1402 break; | |
| 1403 } | |
| 1404 stringbuffer.writeByte(c); | |
| 1405 } | |
| 1406 } | |
| 1407 | |
| 1408 /************************************** | |
| 1409 * Lex hex strings: | |
| 1410 * x"0A ae 34FE BD" | |
| 1411 */ | |
| 1412 | |
| 1413 TOK Lexer::hexStringConstant(Token *t) | |
| 1414 { unsigned c; | |
| 1415 Loc start = loc; | |
| 1416 unsigned n = 0; | |
| 1417 unsigned v; | |
| 1418 | |
| 1419 p++; | |
| 1420 stringbuffer.reset(); | |
| 1421 while (1) | |
| 1422 { | |
| 1423 c = *p++; | |
| 1424 switch (c) | |
| 1425 { | |
| 1426 case ' ': | |
| 1427 case '\t': | |
| 1428 case '\v': | |
| 1429 case '\f': | |
| 1430 continue; // skip white space | |
| 1431 | |
| 1432 case '\r': | |
| 1433 if (*p == '\n') | |
| 1434 continue; // ignore | |
| 1435 // Treat isolated '\r' as if it were a '\n' | |
| 1436 case '\n': | |
| 1437 loc.linnum++; | |
| 1438 continue; | |
| 1439 | |
| 1440 case 0: | |
| 1441 case 0x1A: | |
| 1442 error("unterminated string constant starting at %s", start.toChars()); | |
| 1443 t->ustring = (unsigned char *)""; | |
| 1444 t->len = 0; | |
| 1445 t->postfix = 0; | |
| 1446 return TOKstring; | |
| 1447 | |
| 1448 case '"': | |
| 1449 if (n & 1) | |
| 1450 { error("odd number (%d) of hex characters in hex string", n); | |
| 1451 stringbuffer.writeByte(v); | |
| 1452 } | |
| 1453 t->len = stringbuffer.offset; | |
| 1454 stringbuffer.writeByte(0); | |
| 1455 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1456 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1457 stringPostfix(t); | |
| 1458 return TOKstring; | |
| 1459 | |
| 1460 default: | |
| 1461 if (c >= '0' && c <= '9') | |
| 1462 c -= '0'; | |
| 1463 else if (c >= 'a' && c <= 'f') | |
| 1464 c -= 'a' - 10; | |
| 1465 else if (c >= 'A' && c <= 'F') | |
| 1466 c -= 'A' - 10; | |
| 1467 else if (c & 0x80) | |
| 1468 { p--; | |
| 1469 unsigned u = decodeUTF(); | |
| 1470 p++; | |
| 1471 if (u == PS || u == LS) | |
| 1472 loc.linnum++; | |
| 1473 else | |
| 1474 error("non-hex character \\u%x", u); | |
| 1475 } | |
| 1476 else | |
| 1477 error("non-hex character '%c'", c); | |
| 1478 if (n & 1) | |
| 1479 { v = (v << 4) | c; | |
| 1480 stringbuffer.writeByte(v); | |
| 1481 } | |
| 1482 else | |
| 1483 v = c; | |
| 1484 n++; | |
| 1485 break; | |
| 1486 } | |
| 1487 } | |
| 1488 } | |
| 1489 | |
| 1490 | |
| 336 | 1491 #if DMDV2 |
| 159 | 1492 /************************************** |
| 1493 * Lex delimited strings: | |
| 1494 * q"(foo(xxx))" // "foo(xxx)" | |
| 1495 * q"[foo(]" // "foo(" | |
| 1496 * q"/foo]/" // "foo]" | |
| 1497 * q"HERE | |
| 1498 * foo | |
| 1499 * HERE" // "foo\n" | |
| 1500 * Input: | |
| 1501 * p is on the " | |
| 1502 */ | |
| 1503 | |
| 1504 TOK Lexer::delimitedStringConstant(Token *t) | |
| 1505 { unsigned c; | |
| 1506 Loc start = loc; | |
| 1507 unsigned delimleft = 0; | |
| 1508 unsigned delimright = 0; | |
| 1509 unsigned nest = 1; | |
| 1510 unsigned nestcount; | |
| 1511 Identifier *hereid = NULL; | |
| 1512 unsigned blankrol = 0; | |
| 1513 unsigned startline = 0; | |
| 1514 | |
| 1515 p++; | |
| 1516 stringbuffer.reset(); | |
| 1517 while (1) | |
| 1518 { | |
| 1519 c = *p++; | |
| 1520 //printf("c = '%c'\n", c); | |
| 1521 switch (c) | |
| 1522 { | |
| 1523 case '\n': | |
| 1524 Lnextline: | |
| 1525 loc.linnum++; | |
| 1526 startline = 1; | |
| 1527 if (blankrol) | |
| 1528 { blankrol = 0; | |
| 1529 continue; | |
| 1530 } | |
| 1531 if (hereid) | |
| 1532 { | |
| 1533 stringbuffer.writeUTF8(c); | |
| 1534 continue; | |
| 1535 } | |
| 1536 break; | |
| 1537 | |
| 1538 case '\r': | |
| 1539 if (*p == '\n') | |
| 1540 continue; // ignore | |
| 1541 c = '\n'; // treat EndOfLine as \n character | |
| 1542 goto Lnextline; | |
| 1543 | |
| 1544 case 0: | |
| 1545 case 0x1A: | |
| 1546 goto Lerror; | |
| 1547 | |
| 1548 default: | |
| 1549 if (c & 0x80) | |
| 1550 { p--; | |
| 1551 c = decodeUTF(); | |
| 1552 p++; | |
| 1553 if (c == PS || c == LS) | |
| 1554 goto Lnextline; | |
| 1555 } | |
| 1556 break; | |
| 1557 } | |
| 1558 if (delimleft == 0) | |
| 1559 { delimleft = c; | |
| 1560 nest = 1; | |
| 1561 nestcount = 1; | |
| 1562 if (c == '(') | |
| 1563 delimright = ')'; | |
| 1564 else if (c == '{') | |
| 1565 delimright = '}'; | |
| 1566 else if (c == '[') | |
| 1567 delimright = ']'; | |
| 1568 else if (c == '<') | |
| 1569 delimright = '>'; | |
| 1570 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1571 { // Start of identifier; must be a heredoc | |
| 1572 Token t; | |
| 1573 p--; | |
| 1574 scan(&t); // read in heredoc identifier | |
| 1575 if (t.value != TOKidentifier) | |
| 1576 { error("identifier expected for heredoc, not %s", t.toChars()); | |
| 1577 delimright = c; | |
| 1578 } | |
| 1579 else | |
| 1580 { hereid = t.ident; | |
| 1581 //printf("hereid = '%s'\n", hereid->toChars()); | |
| 1582 blankrol = 1; | |
| 1583 } | |
| 1584 nest = 0; | |
| 1585 } | |
| 1586 else | |
| 1587 { delimright = c; | |
| 1588 nest = 0; | |
| 1589 } | |
| 1590 } | |
| 1591 else | |
| 1592 { | |
| 1593 if (blankrol) | |
| 1594 { error("heredoc rest of line should be blank"); | |
| 1595 blankrol = 0; | |
| 1596 continue; | |
| 1597 } | |
| 1598 if (nest == 1) | |
| 1599 { | |
| 1600 if (c == delimleft) | |
| 1601 nestcount++; | |
| 1602 else if (c == delimright) | |
| 1603 { nestcount--; | |
| 1604 if (nestcount == 0) | |
| 1605 goto Ldone; | |
| 1606 } | |
| 1607 } | |
| 1608 else if (c == delimright) | |
| 1609 goto Ldone; | |
| 1610 if (startline && isalpha(c)) | |
| 1611 { Token t; | |
| 1612 unsigned char *psave = p; | |
| 1613 p--; | |
| 1614 scan(&t); // read in possible heredoc identifier | |
| 1615 //printf("endid = '%s'\n", t.ident->toChars()); | |
| 1616 if (t.value == TOKidentifier && t.ident->equals(hereid)) | |
| 1617 { /* should check that rest of line is blank | |
| 1618 */ | |
| 1619 goto Ldone; | |
| 1620 } | |
| 1621 p = psave; | |
| 1622 } | |
| 1623 stringbuffer.writeUTF8(c); | |
| 1624 startline = 0; | |
| 1625 } | |
| 1626 } | |
| 1627 | |
| 1628 Ldone: | |
| 1629 if (*p == '"') | |
| 1630 p++; | |
| 1631 else | |
| 1632 error("delimited string must end in %c\"", delimright); | |
| 1633 t->len = stringbuffer.offset; | |
| 1634 stringbuffer.writeByte(0); | |
| 1635 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1636 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1637 stringPostfix(t); | |
| 1638 return TOKstring; | |
| 1639 | |
| 1640 Lerror: | |
| 1641 error("unterminated string constant starting at %s", start.toChars()); | |
| 1642 t->ustring = (unsigned char *)""; | |
| 1643 t->len = 0; | |
| 1644 t->postfix = 0; | |
| 1645 return TOKstring; | |
| 1646 } | |
| 1647 | |
| 1648 /************************************** | |
| 1649 * Lex delimited strings: | |
| 1650 * q{ foo(xxx) } // " foo(xxx) " | |
| 1651 * q{foo(} // "foo(" | |
| 1652 * q{{foo}"}"} // "{foo}"}"" | |
| 1653 * Input: | |
| 1654 * p is on the q | |
| 1655 */ | |
| 1656 | |
| 1657 TOK Lexer::tokenStringConstant(Token *t) | |
| 1658 { | |
| 1659 unsigned nest = 1; | |
| 1660 Loc start = loc; | |
| 1661 unsigned char *pstart = ++p; | |
| 1662 | |
| 1663 while (1) | |
| 1664 { Token tok; | |
| 1665 | |
| 1666 scan(&tok); | |
| 1667 switch (tok.value) | |
| 1668 { | |
| 1669 case TOKlcurly: | |
| 1670 nest++; | |
| 1671 continue; | |
| 1672 | |
| 1673 case TOKrcurly: | |
| 1674 if (--nest == 0) | |
| 1675 goto Ldone; | |
| 1676 continue; | |
| 1677 | |
| 1678 case TOKeof: | |
| 1679 goto Lerror; | |
| 1680 | |
| 1681 default: | |
| 1682 continue; | |
| 1683 } | |
| 1684 } | |
| 1685 | |
| 1686 Ldone: | |
| 1687 t->len = p - 1 - pstart; | |
| 1688 t->ustring = (unsigned char *)mem.malloc(t->len + 1); | |
| 1689 memcpy(t->ustring, pstart, t->len); | |
| 1690 t->ustring[t->len] = 0; | |
| 1691 stringPostfix(t); | |
| 1692 return TOKstring; | |
| 1693 | |
| 1694 Lerror: | |
| 1695 error("unterminated token string constant starting at %s", start.toChars()); | |
| 1696 t->ustring = (unsigned char *)""; | |
| 1697 t->len = 0; | |
| 1698 t->postfix = 0; | |
| 1699 return TOKstring; | |
| 1700 } | |
| 1701 | |
| 1702 #endif | |
| 1703 | |
| 1704 | |
| 1705 /************************************** | |
| 1706 */ | |
| 1707 | |
| 1708 TOK Lexer::escapeStringConstant(Token *t, int wide) | |
| 1709 { unsigned c; | |
| 1710 Loc start = loc; | |
| 1711 | |
| 1712 p++; | |
| 1713 stringbuffer.reset(); | |
| 1714 while (1) | |
| 1715 { | |
| 1716 c = *p++; | |
| 1717 switch (c) | |
| 1718 { | |
| 1719 case '\\': | |
| 1720 switch (*p) | |
| 1721 { | |
| 1722 case 'u': | |
| 1723 case 'U': | |
| 1724 case '&': | |
| 1725 c = escapeSequence(); | |
| 1726 stringbuffer.writeUTF8(c); | |
| 1727 continue; | |
| 1728 | |
| 1729 default: | |
| 1730 c = escapeSequence(); | |
| 1731 break; | |
| 1732 } | |
| 1733 break; | |
| 1734 | |
| 1735 case '\n': | |
| 1736 loc.linnum++; | |
| 1737 break; | |
| 1738 | |
| 1739 case '\r': | |
| 1740 if (*p == '\n') | |
| 1741 continue; // ignore | |
| 1742 c = '\n'; // treat EndOfLine as \n character | |
| 1743 loc.linnum++; | |
| 1744 break; | |
| 1745 | |
| 1746 case '"': | |
| 1747 t->len = stringbuffer.offset; | |
| 1748 stringbuffer.writeByte(0); | |
| 1749 t->ustring = (unsigned char *)mem.malloc(stringbuffer.offset); | |
| 1750 memcpy(t->ustring, stringbuffer.data, stringbuffer.offset); | |
| 1751 stringPostfix(t); | |
| 1752 return TOKstring; | |
| 1753 | |
| 1754 case 0: | |
| 1755 case 0x1A: | |
| 1756 p--; | |
| 1757 error("unterminated string constant starting at %s", start.toChars()); | |
| 1758 t->ustring = (unsigned char *)""; | |
| 1759 t->len = 0; | |
| 1760 t->postfix = 0; | |
| 1761 return TOKstring; | |
| 1762 | |
| 1763 default: | |
| 1764 if (c & 0x80) | |
| 1765 { | |
| 1766 p--; | |
| 1767 c = decodeUTF(); | |
| 1768 if (c == LS || c == PS) | |
| 1769 { c = '\n'; | |
| 1770 loc.linnum++; | |
| 1771 } | |
| 1772 p++; | |
| 1773 stringbuffer.writeUTF8(c); | |
| 1774 continue; | |
| 1775 } | |
| 1776 break; | |
| 1777 } | |
| 1778 stringbuffer.writeByte(c); | |
| 1779 } | |
| 1780 } | |
| 1781 | |
| 1782 /************************************** | |
| 1783 */ | |
| 1784 | |
| 1785 TOK Lexer::charConstant(Token *t, int wide) | |
| 1786 { | |
| 1787 unsigned c; | |
| 1788 TOK tk = TOKcharv; | |
| 1789 | |
| 1790 //printf("Lexer::charConstant\n"); | |
| 1791 p++; | |
| 1792 c = *p++; | |
| 1793 switch (c) | |
| 1794 { | |
| 1795 case '\\': | |
| 1796 switch (*p) | |
| 1797 { | |
| 1798 case 'u': | |
| 1799 t->uns64value = escapeSequence(); | |
| 1800 tk = TOKwcharv; | |
| 1801 break; | |
| 1802 | |
| 1803 case 'U': | |
| 1804 case '&': | |
| 1805 t->uns64value = escapeSequence(); | |
| 1806 tk = TOKdcharv; | |
| 1807 break; | |
| 1808 | |
| 1809 default: | |
| 1810 t->uns64value = escapeSequence(); | |
| 1811 break; | |
| 1812 } | |
| 1813 break; | |
| 1814 | |
| 1815 case '\n': | |
| 1816 L1: | |
| 1817 loc.linnum++; | |
| 1818 case '\r': | |
| 1819 case 0: | |
| 1820 case 0x1A: | |
| 1821 case '\'': | |
| 1822 error("unterminated character constant"); | |
| 1823 return tk; | |
| 1824 | |
| 1825 default: | |
| 1826 if (c & 0x80) | |
| 1827 { | |
| 1828 p--; | |
| 1829 c = decodeUTF(); | |
| 1830 p++; | |
| 1831 if (c == LS || c == PS) | |
| 1832 goto L1; | |
| 1833 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 1834 tk = TOKwcharv; | |
| 1835 else | |
| 1836 tk = TOKdcharv; | |
| 1837 } | |
| 1838 t->uns64value = c; | |
| 1839 break; | |
| 1840 } | |
| 1841 | |
| 1842 if (*p != '\'') | |
| 1843 { error("unterminated character constant"); | |
| 1844 return tk; | |
| 1845 } | |
| 1846 p++; | |
| 1847 return tk; | |
| 1848 } | |
| 1849 | |
| 1850 /*************************************** | |
| 1851 * Get postfix of string literal. | |
| 1852 */ | |
| 1853 | |
| 1854 void Lexer::stringPostfix(Token *t) | |
| 1855 { | |
| 1856 switch (*p) | |
| 1857 { | |
| 1858 case 'c': | |
| 1859 case 'w': | |
| 1860 case 'd': | |
| 1861 t->postfix = *p; | |
| 1862 p++; | |
| 1863 break; | |
| 1864 | |
| 1865 default: | |
| 1866 t->postfix = 0; | |
| 1867 break; | |
| 1868 } | |
| 1869 } | |
| 1870 | |
| 1871 /*************************************** | |
| 1872 * Read \u or \U unicode sequence | |
| 1873 * Input: | |
| 1874 * u 'u' or 'U' | |
| 1875 */ | |
| 1876 | |
| 1877 #if 0 | |
| 1878 unsigned Lexer::wchar(unsigned u) | |
| 1879 { | |
| 1880 unsigned value; | |
| 1881 unsigned n; | |
| 1882 unsigned char c; | |
| 1883 unsigned nchars; | |
| 1884 | |
| 1885 nchars = (u == 'U') ? 8 : 4; | |
| 1886 value = 0; | |
| 1887 for (n = 0; 1; n++) | |
| 1888 { | |
| 1889 ++p; | |
| 1890 if (n == nchars) | |
| 1891 break; | |
| 1892 c = *p; | |
| 1893 if (!ishex(c)) | |
| 1894 { error("\\%c sequence must be followed by %d hex characters", u, nchars); | |
| 1895 break; | |
| 1896 } | |
| 1897 if (isdigit(c)) | |
| 1898 c -= '0'; | |
| 1899 else if (islower(c)) | |
| 1900 c -= 'a' - 10; | |
| 1901 else | |
| 1902 c -= 'A' - 10; | |
| 1903 value <<= 4; | |
| 1904 value |= c; | |
| 1905 } | |
| 1906 return value; | |
| 1907 } | |
| 1908 #endif | |
| 1909 | |
| 1910 /************************************** | |
| 1911 * Read in a number. | |
| 1912 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 1913 * integers can be decimal, octal or hex | |
| 1914 * Handle the suffixes U, UL, LU, L, etc. | |
| 1915 * If it's double, store it in tok.TKutok.Vdouble. | |
| 1916 * Returns: | |
| 1917 * TKnum | |
| 1918 * TKdouble,... | |
| 1919 */ | |
| 1920 | |
| 1921 TOK Lexer::number(Token *t) | |
| 1922 { | |
| 1923 // We use a state machine to collect numbers | |
| 1924 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 1925 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 1926 STATE_hexh, STATE_error }; | |
| 1927 enum STATE state; | |
| 1928 | |
| 1929 enum FLAGS | |
| 1930 { FLAGS_decimal = 1, // decimal | |
| 1931 FLAGS_unsigned = 2, // u or U suffix | |
| 1932 FLAGS_long = 4, // l or L suffix | |
| 1933 }; | |
| 1934 enum FLAGS flags = FLAGS_decimal; | |
| 1935 | |
| 1936 int i; | |
| 1937 int base; | |
| 1938 unsigned c; | |
| 1939 unsigned char *start; | |
| 1940 TOK result; | |
| 1941 | |
| 1942 //printf("Lexer::number()\n"); | |
| 1943 state = STATE_initial; | |
| 1944 base = 0; | |
| 1945 stringbuffer.reset(); | |
| 1946 start = p; | |
| 1947 while (1) | |
| 1948 { | |
| 1949 c = *p; | |
| 1950 switch (state) | |
| 1951 { | |
| 1952 case STATE_initial: // opening state | |
| 1953 if (c == '0') | |
| 1954 state = STATE_0; | |
| 1955 else | |
| 1956 state = STATE_decimal; | |
| 1957 break; | |
| 1958 | |
| 1959 case STATE_0: | |
| 1960 flags = (FLAGS) (flags & ~FLAGS_decimal); | |
| 1961 switch (c) | |
| 1962 { | |
| 1963 #if ZEROH | |
| 1964 case 'H': // 0h | |
| 1965 case 'h': | |
| 1966 goto hexh; | |
| 1967 #endif | |
| 1968 case 'X': | |
| 1969 case 'x': | |
| 1970 state = STATE_hex0; | |
| 1971 break; | |
| 1972 | |
| 1973 case '.': | |
| 1974 if (p[1] == '.') // .. is a separate token | |
| 1975 goto done; | |
| 1976 case 'i': | |
| 1977 case 'f': | |
| 1978 case 'F': | |
| 1979 goto real; | |
| 1980 #if ZEROH | |
| 1981 case 'E': | |
| 1982 case 'e': | |
| 1983 goto case_hex; | |
| 1984 #endif | |
| 1985 case 'B': | |
| 1986 case 'b': | |
| 1987 state = STATE_binary0; | |
| 1988 break; | |
| 1989 | |
| 1990 case '0': case '1': case '2': case '3': | |
| 1991 case '4': case '5': case '6': case '7': | |
| 1992 state = STATE_octal; | |
| 1993 break; | |
| 1994 | |
| 1995 #if ZEROH | |
| 1996 case '8': case '9': case 'A': | |
| 1997 case 'C': case 'D': case 'F': | |
| 1998 case 'a': case 'c': case 'd': case 'f': | |
| 1999 case_hex: | |
| 2000 state = STATE_hexh; | |
| 2001 break; | |
| 2002 #endif | |
| 2003 case '_': | |
| 2004 state = STATE_octal; | |
| 2005 p++; | |
| 2006 continue; | |
| 2007 | |
| 2008 case 'L': | |
| 2009 if (p[1] == 'i') | |
| 2010 goto real; | |
| 2011 goto done; | |
| 2012 | |
| 2013 default: | |
| 2014 goto done; | |
| 2015 } | |
| 2016 break; | |
| 2017 | |
| 2018 case STATE_decimal: // reading decimal number | |
| 2019 if (!isdigit(c)) | |
| 2020 { | |
| 2021 #if ZEROH | |
| 2022 if (ishex(c) | |
| 2023 || c == 'H' || c == 'h' | |
| 2024 ) | |
| 2025 goto hexh; | |
| 2026 #endif | |
| 2027 if (c == '_') // ignore embedded _ | |
| 2028 { p++; | |
| 2029 continue; | |
| 2030 } | |
| 2031 if (c == '.' && p[1] != '.') | |
| 2032 goto real; | |
| 2033 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2034 c == 'e' || c == 'E') | |
| 2035 { | |
| 2036 real: // It's a real number. Back up and rescan as a real | |
| 2037 p = start; | |
| 2038 return inreal(t); | |
| 2039 } | |
| 2040 else if (c == 'L' && p[1] == 'i') | |
| 2041 goto real; | |
| 2042 goto done; | |
| 2043 } | |
| 2044 break; | |
| 2045 | |
| 2046 case STATE_hex0: // reading hex number | |
| 2047 case STATE_hex: | |
| 2048 if (!ishex(c)) | |
| 2049 { | |
| 2050 if (c == '_') // ignore embedded _ | |
| 2051 { p++; | |
| 2052 continue; | |
| 2053 } | |
| 2054 if (c == '.' && p[1] != '.') | |
| 2055 goto real; | |
| 2056 if (c == 'P' || c == 'p' || c == 'i') | |
| 2057 goto real; | |
| 2058 if (state == STATE_hex0) | |
| 2059 error("Hex digit expected, not '%c'", c); | |
| 2060 goto done; | |
| 2061 } | |
| 2062 state = STATE_hex; | |
| 2063 break; | |
| 2064 | |
| 2065 #if ZEROH | |
| 2066 hexh: | |
| 2067 state = STATE_hexh; | |
| 2068 case STATE_hexh: // parse numbers like 0FFh | |
| 2069 if (!ishex(c)) | |
| 2070 { | |
| 2071 if (c == 'H' || c == 'h') | |
| 2072 { | |
| 2073 p++; | |
| 2074 base = 16; | |
| 2075 goto done; | |
| 2076 } | |
| 2077 else | |
| 2078 { | |
| 2079 // Check for something like 1E3 or 0E24 | |
| 2080 if (memchr((char *)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2081 memchr((char *)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2082 goto real; | |
| 2083 error("Hex digit expected, not '%c'", c); | |
| 2084 goto done; | |
| 2085 } | |
| 2086 } | |
| 2087 break; | |
| 2088 #endif | |
| 2089 | |
| 2090 case STATE_octal: // reading octal number | |
| 2091 case STATE_octale: // reading octal number with non-octal digits | |
| 2092 if (!isoctal(c)) | |
| 2093 { | |
| 2094 #if ZEROH | |
| 2095 if (ishex(c) | |
| 2096 || c == 'H' || c == 'h' | |
| 2097 ) | |
| 2098 goto hexh; | |
| 2099 #endif | |
| 2100 if (c == '_') // ignore embedded _ | |
| 2101 { p++; | |
| 2102 continue; | |
| 2103 } | |
| 2104 if (c == '.' && p[1] != '.') | |
| 2105 goto real; | |
| 2106 if (c == 'i') | |
| 2107 goto real; | |
| 2108 if (isdigit(c)) | |
| 2109 { | |
| 2110 state = STATE_octale; | |
| 2111 } | |
| 2112 else | |
| 2113 goto done; | |
| 2114 } | |
| 2115 break; | |
| 2116 | |
| 2117 case STATE_binary0: // starting binary number | |
| 2118 case STATE_binary: // reading binary number | |
| 2119 if (c != '0' && c != '1') | |
| 2120 { | |
| 2121 #if ZEROH | |
| 2122 if (ishex(c) | |
| 2123 || c == 'H' || c == 'h' | |
| 2124 ) | |
| 2125 goto hexh; | |
| 2126 #endif | |
| 2127 if (c == '_') // ignore embedded _ | |
| 2128 { p++; | |
| 2129 continue; | |
| 2130 } | |
| 2131 if (state == STATE_binary0) | |
| 2132 { error("binary digit expected"); | |
| 2133 state = STATE_error; | |
| 2134 break; | |
| 2135 } | |
| 2136 else | |
| 2137 goto done; | |
| 2138 } | |
| 2139 state = STATE_binary; | |
| 2140 break; | |
| 2141 | |
| 2142 case STATE_error: // for error recovery | |
| 2143 if (!isdigit(c)) // scan until non-digit | |
| 2144 goto done; | |
| 2145 break; | |
| 2146 | |
| 2147 default: | |
| 2148 assert(0); | |
| 2149 } | |
| 2150 stringbuffer.writeByte(c); | |
| 2151 p++; | |
| 2152 } | |
| 2153 done: | |
| 2154 stringbuffer.writeByte(0); // terminate string | |
| 2155 if (state == STATE_octale) | |
| 2156 error("Octal digit expected"); | |
| 2157 | |
| 2158 uinteger_t n; // unsigned >=64 bit integer type | |
| 2159 | |
| 2160 if (stringbuffer.offset == 2 && (state == STATE_decimal || state == STATE_0)) | |
| 2161 n = stringbuffer.data[0] - '0'; | |
| 2162 else | |
| 2163 { | |
| 2164 // Convert string to integer | |
| 2165 #if __DMC__ | |
| 2166 errno = 0; | |
| 2167 n = strtoull((char *)stringbuffer.data,NULL,base); | |
| 2168 if (errno == ERANGE) | |
| 2169 error("integer overflow"); | |
| 2170 #else | |
| 2171 // Not everybody implements strtoull() | |
| 2172 char *p = (char *)stringbuffer.data; | |
| 2173 int r = 10, d; | |
| 2174 | |
| 2175 if (*p == '0') | |
| 2176 { | |
| 2177 if (p[1] == 'x' || p[1] == 'X') | |
| 2178 p += 2, r = 16; | |
| 2179 else if (p[1] == 'b' || p[1] == 'B') | |
| 2180 p += 2, r = 2; | |
| 2181 else if (isdigit(p[1])) | |
| 2182 p += 1, r = 8; | |
| 2183 } | |
| 2184 | |
| 2185 n = 0; | |
| 2186 while (1) | |
| 2187 { | |
| 2188 if (*p >= '0' && *p <= '9') | |
| 2189 d = *p - '0'; | |
| 2190 else if (*p >= 'a' && *p <= 'z') | |
| 2191 d = *p - 'a' + 10; | |
| 2192 else if (*p >= 'A' && *p <= 'Z') | |
| 2193 d = *p - 'A' + 10; | |
| 2194 else | |
| 2195 break; | |
| 2196 if (d >= r) | |
| 2197 break; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2198 uinteger_t n2 = n * r; |
|
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2199 //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
|
2200 if (n2 / r != n || n2 + d < n) |
| 159 | 2201 { |
| 2202 error ("integer overflow"); | |
| 2203 break; | |
| 2204 } | |
| 2205 | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2206 n = n2 + d; |
| 159 | 2207 p++; |
| 2208 } | |
| 2209 #endif | |
| 2210 if (sizeof(n) > 8 && | |
| 2211 n > 0xFFFFFFFFFFFFFFFFULL) // if n needs more than 64 bits | |
| 2212 error("integer overflow"); | |
| 2213 } | |
| 2214 | |
| 2215 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2216 while (1) | |
| 2217 { unsigned char f; | |
| 2218 | |
| 2219 switch (*p) | |
| 2220 { case 'U': | |
| 2221 case 'u': | |
| 2222 f = FLAGS_unsigned; | |
| 2223 goto L1; | |
| 2224 | |
| 2225 case 'l': | |
| 2226 if (1 || !global.params.useDeprecated) | |
| 2227 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2228 case 'L': | |
| 2229 f = FLAGS_long; | |
| 2230 L1: | |
| 2231 p++; | |
| 2232 if (flags & f) | |
| 2233 error("unrecognized token"); | |
| 2234 flags = (FLAGS) (flags | f); | |
| 2235 continue; | |
| 2236 default: | |
| 2237 break; | |
| 2238 } | |
| 2239 break; | |
| 2240 } | |
| 2241 | |
| 2242 switch (flags) | |
| 2243 { | |
| 2244 case 0: | |
| 2245 /* Octal or Hexadecimal constant. | |
| 2246 * First that fits: int, uint, long, ulong | |
| 2247 */ | |
| 2248 if (n & 0x8000000000000000LL) | |
| 2249 result = TOKuns64v; | |
| 2250 else if (n & 0xFFFFFFFF00000000LL) | |
| 2251 result = TOKint64v; | |
| 2252 else if (n & 0x80000000) | |
| 2253 result = TOKuns32v; | |
| 2254 else | |
| 2255 result = TOKint32v; | |
| 2256 break; | |
| 2257 | |
| 2258 case FLAGS_decimal: | |
| 2259 /* First that fits: int, long, long long | |
| 2260 */ | |
| 2261 if (n & 0x8000000000000000LL) | |
| 2262 { error("signed integer overflow"); | |
| 2263 result = TOKuns64v; | |
| 2264 } | |
| 2265 else if (n & 0xFFFFFFFF80000000LL) | |
| 2266 result = TOKint64v; | |
| 2267 else | |
| 2268 result = TOKint32v; | |
| 2269 break; | |
| 2270 | |
| 2271 case FLAGS_unsigned: | |
| 2272 case FLAGS_decimal | FLAGS_unsigned: | |
| 2273 /* First that fits: uint, ulong | |
| 2274 */ | |
| 2275 if (n & 0xFFFFFFFF00000000LL) | |
| 2276 result = TOKuns64v; | |
| 2277 else | |
| 2278 result = TOKuns32v; | |
| 2279 break; | |
| 2280 | |
| 2281 case FLAGS_decimal | FLAGS_long: | |
| 2282 if (n & 0x8000000000000000LL) | |
| 2283 { error("signed integer overflow"); | |
| 2284 result = TOKuns64v; | |
| 2285 } | |
| 2286 else | |
| 2287 result = TOKint64v; | |
| 2288 break; | |
| 2289 | |
| 2290 case FLAGS_long: | |
| 2291 if (n & 0x8000000000000000LL) | |
| 2292 result = TOKuns64v; | |
| 2293 else | |
| 2294 result = TOKint64v; | |
| 2295 break; | |
| 2296 | |
| 2297 case FLAGS_unsigned | FLAGS_long: | |
| 2298 case FLAGS_decimal | FLAGS_unsigned | FLAGS_long: | |
| 2299 result = TOKuns64v; | |
| 2300 break; | |
| 2301 | |
| 2302 default: | |
| 2303 #ifdef DEBUG | |
| 2304 printf("%x\n",flags); | |
| 2305 #endif | |
| 2306 assert(0); | |
| 2307 } | |
| 2308 t->uns64value = n; | |
| 2309 return result; | |
| 2310 } | |
| 2311 | |
| 2312 /************************************** | |
| 2313 * Read in characters, converting them to real. | |
| 2314 * Bugs: | |
| 2315 * Exponent overflow not detected. | |
| 2316 * Too much requested precision is not detected. | |
| 2317 */ | |
| 2318 | |
| 2319 TOK Lexer::inreal(Token *t) | |
| 2320 #ifdef __DMC__ | |
| 2321 __in | |
| 2322 { | |
| 2323 assert(*p == '.' || isdigit(*p)); | |
| 2324 } | |
| 2325 __out (result) | |
| 2326 { | |
| 2327 switch (result) | |
| 2328 { | |
| 2329 case TOKfloat32v: | |
| 2330 case TOKfloat64v: | |
| 2331 case TOKfloat80v: | |
| 2332 case TOKimaginary32v: | |
| 2333 case TOKimaginary64v: | |
| 2334 case TOKimaginary80v: | |
| 2335 break; | |
| 2336 | |
| 2337 default: | |
| 2338 assert(0); | |
| 2339 } | |
| 2340 } | |
| 2341 __body | |
| 2342 #endif /* __DMC__ */ | |
| 2343 { int dblstate; | |
| 2344 unsigned c; | |
| 2345 char hex; // is this a hexadecimal-floating-constant? | |
| 2346 TOK result; | |
| 2347 | |
| 2348 //printf("Lexer::inreal()\n"); | |
| 2349 stringbuffer.reset(); | |
| 2350 dblstate = 0; | |
| 2351 hex = 0; | |
| 2352 Lnext: | |
| 2353 while (1) | |
| 2354 { | |
| 2355 // Get next char from input | |
| 2356 c = *p++; | |
| 2357 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2358 while (1) | |
| 2359 { | |
| 2360 switch (dblstate) | |
| 2361 { | |
| 2362 case 0: // opening state | |
| 2363 if (c == '0') | |
| 2364 dblstate = 9; | |
| 2365 else if (c == '.') | |
| 2366 dblstate = 3; | |
| 2367 else | |
| 2368 dblstate = 1; | |
| 2369 break; | |
| 2370 | |
| 2371 case 9: | |
| 2372 dblstate = 1; | |
| 2373 if (c == 'X' || c == 'x') | |
| 2374 { hex++; | |
| 2375 break; | |
| 2376 } | |
| 2377 case 1: // digits to left of . | |
| 2378 case 3: // digits to right of . | |
| 2379 case 7: // continuing exponent digits | |
| 2380 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2381 { | |
| 2382 if (c == '_') | |
| 2383 goto Lnext; // ignore embedded '_' | |
| 2384 dblstate++; | |
| 2385 continue; | |
| 2386 } | |
| 2387 break; | |
| 2388 | |
| 2389 case 2: // no more digits to left of . | |
| 2390 if (c == '.') | |
| 2391 { dblstate++; | |
| 2392 break; | |
| 2393 } | |
| 2394 case 4: // no more digits to right of . | |
| 2395 if ((c == 'E' || c == 'e') || | |
| 2396 hex && (c == 'P' || c == 'p')) | |
| 2397 { dblstate = 5; | |
| 2398 hex = 0; // exponent is always decimal | |
| 2399 break; | |
| 2400 } | |
| 2401 if (hex) | |
| 2402 error("binary-exponent-part required"); | |
| 2403 goto done; | |
| 2404 | |
| 2405 case 5: // looking immediately to right of E | |
| 2406 dblstate++; | |
| 2407 if (c == '-' || c == '+') | |
| 2408 break; | |
| 2409 case 6: // 1st exponent digit expected | |
| 2410 if (!isdigit(c)) | |
| 2411 error("exponent expected"); | |
| 2412 dblstate++; | |
| 2413 break; | |
| 2414 | |
| 2415 case 8: // past end of exponent digits | |
| 2416 goto done; | |
| 2417 } | |
| 2418 break; | |
| 2419 } | |
| 2420 stringbuffer.writeByte(c); | |
| 2421 } | |
| 2422 done: | |
| 2423 p--; | |
| 2424 | |
| 2425 stringbuffer.writeByte(0); | |
| 2426 | |
| 2427 #if _WIN32 && __DMC__ | |
| 2428 char *save = __locale_decpoint; | |
| 2429 __locale_decpoint = "."; | |
| 2430 #endif | |
| 2431 #ifdef IN_GCC | |
| 2432 t->float80value = real_t::parse((char *)stringbuffer.data, real_t::LongDouble); | |
| 2433 #else | |
| 2434 t->float80value = strtold((char *)stringbuffer.data, NULL); | |
| 2435 #endif | |
| 2436 errno = 0; | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2437 float strtofres; |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2438 double strtodres; |
| 159 | 2439 switch (*p) |
| 2440 { | |
| 2441 case 'F': | |
| 2442 case 'f': | |
| 2443 #ifdef IN_GCC | |
| 2444 real_t::parse((char *)stringbuffer.data, real_t::Float); | |
| 2445 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2446 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
|
2447 // 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
|
2448 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2449 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
|
2450 errno = 0; |
| 159 | 2451 #endif |
| 2452 result = TOKfloat32v; | |
| 2453 p++; | |
| 2454 break; | |
| 2455 | |
| 2456 default: | |
| 2457 #ifdef IN_GCC | |
| 2458 real_t::parse((char *)stringbuffer.data, real_t::Double); | |
| 2459 #else | |
|
696
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2460 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
|
2461 // 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
|
2462 if (errno == ERANGE && |
|
48f462341528
Fix issues with gradual underflow and strtof, strtod on Mac.
Christian Kamm <kamm incasoftware de>
parents:
664
diff
changeset
|
2463 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
|
2464 errno = 0; |
| 159 | 2465 #endif |
| 2466 result = TOKfloat64v; | |
| 2467 break; | |
| 2468 | |
| 2469 case 'l': | |
| 2470 if (!global.params.useDeprecated) | |
| 2471 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2472 case 'L': | |
| 2473 result = TOKfloat80v; | |
| 2474 p++; | |
| 2475 break; | |
| 2476 } | |
| 2477 if (*p == 'i' || *p == 'I') | |
| 2478 { | |
| 2479 if (!global.params.useDeprecated && *p == 'I') | |
| 2480 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2481 p++; | |
| 2482 switch (result) | |
| 2483 { | |
| 2484 case TOKfloat32v: | |
| 2485 result = TOKimaginary32v; | |
| 2486 break; | |
| 2487 case TOKfloat64v: | |
| 2488 result = TOKimaginary64v; | |
| 2489 break; | |
| 2490 case TOKfloat80v: | |
| 2491 result = TOKimaginary80v; | |
| 2492 break; | |
| 2493 } | |
| 2494 } | |
| 2495 #if _WIN32 && __DMC__ | |
| 2496 __locale_decpoint = save; | |
| 2497 #endif | |
| 2498 if (errno == ERANGE) | |
| 2499 error("number is not representable"); | |
| 2500 return result; | |
| 2501 } | |
| 2502 | |
| 2503 /********************************************* | |
| 2504 * Do pragma. | |
| 2505 * Currently, the only pragma supported is: | |
| 2506 * #line linnum [filespec] | |
| 2507 */ | |
| 2508 | |
| 2509 void Lexer::pragma() | |
| 2510 { | |
| 2511 Token tok; | |
| 2512 int linnum; | |
| 2513 char *filespec = NULL; | |
| 2514 Loc loc = this->loc; | |
| 2515 | |
| 2516 scan(&tok); | |
| 2517 if (tok.value != TOKidentifier || tok.ident != Id::line) | |
| 2518 goto Lerr; | |
| 2519 | |
| 2520 scan(&tok); | |
| 2521 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 2522 linnum = tok.uns64value - 1; | |
| 2523 else | |
| 2524 goto Lerr; | |
| 2525 | |
| 2526 while (1) | |
| 2527 { | |
| 2528 switch (*p) | |
| 2529 { | |
| 2530 case 0: | |
| 2531 case 0x1A: | |
| 2532 case '\n': | |
| 2533 Lnewline: | |
| 2534 this->loc.linnum = linnum; | |
| 2535 if (filespec) | |
| 2536 this->loc.filename = filespec; | |
| 2537 return; | |
| 2538 | |
| 2539 case '\r': | |
| 2540 p++; | |
| 2541 if (*p != '\n') | |
| 2542 { p--; | |
| 2543 goto Lnewline; | |
| 2544 } | |
| 2545 continue; | |
| 2546 | |
| 2547 case ' ': | |
| 2548 case '\t': | |
| 2549 case '\v': | |
| 2550 case '\f': | |
| 2551 p++; | |
| 2552 continue; // skip white space | |
| 2553 | |
| 2554 case '_': | |
| 2555 if (mod && memcmp(p, "__FILE__", 8) == 0) | |
| 2556 { | |
| 2557 p += 8; | |
| 2558 filespec = mem.strdup(loc.filename ? loc.filename : mod->ident->toChars()); | |
| 2559 } | |
| 2560 continue; | |
| 2561 | |
| 2562 case '"': | |
| 2563 if (filespec) | |
| 2564 goto Lerr; | |
| 2565 stringbuffer.reset(); | |
| 2566 p++; | |
| 2567 while (1) | |
| 2568 { unsigned c; | |
| 2569 | |
| 2570 c = *p; | |
| 2571 switch (c) | |
| 2572 { | |
| 2573 case '\n': | |
| 2574 case '\r': | |
| 2575 case 0: | |
| 2576 case 0x1A: | |
| 2577 goto Lerr; | |
| 2578 | |
| 2579 case '"': | |
| 2580 stringbuffer.writeByte(0); | |
| 2581 filespec = mem.strdup((char *)stringbuffer.data); | |
| 2582 p++; | |
| 2583 break; | |
| 2584 | |
| 2585 default: | |
| 2586 if (c & 0x80) | |
| 2587 { unsigned u = decodeUTF(); | |
| 2588 if (u == PS || u == LS) | |
| 2589 goto Lerr; | |
| 2590 } | |
| 2591 stringbuffer.writeByte(c); | |
| 2592 p++; | |
| 2593 continue; | |
| 2594 } | |
| 2595 break; | |
| 2596 } | |
| 2597 continue; | |
| 2598 | |
| 2599 default: | |
| 2600 if (*p & 0x80) | |
| 2601 { unsigned u = decodeUTF(); | |
| 2602 if (u == PS || u == LS) | |
| 2603 goto Lnewline; | |
| 2604 } | |
| 2605 goto Lerr; | |
| 2606 } | |
| 2607 } | |
| 2608 | |
| 2609 Lerr: | |
| 2610 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 2611 } | |
| 2612 | |
| 2613 | |
| 2614 /******************************************** | |
| 2615 * Decode UTF character. | |
| 2616 * Issue error messages for invalid sequences. | |
| 2617 * Return decoded character, advance p to last character in UTF sequence. | |
| 2618 */ | |
| 2619 | |
| 2620 unsigned Lexer::decodeUTF() | |
| 2621 { | |
| 2622 dchar_t u; | |
| 2623 unsigned char c; | |
| 2624 unsigned char *s = p; | |
| 2625 size_t len; | |
| 2626 size_t idx; | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2627 const char *msg; |
| 159 | 2628 |
| 2629 c = *s; | |
| 2630 assert(c & 0x80); | |
| 2631 | |
| 2632 // Check length of remaining string up to 6 UTF-8 characters | |
| 2633 for (len = 1; len < 6 && s[len]; len++) | |
| 2634 ; | |
| 2635 | |
| 2636 idx = 0; | |
| 2637 msg = utf_decodeChar(s, len, &idx, &u); | |
| 2638 p += idx - 1; | |
| 2639 if (msg) | |
| 2640 { | |
| 2641 error("%s", msg); | |
| 2642 } | |
| 2643 return u; | |
| 2644 } | |
| 2645 | |
| 2646 | |
| 2647 /*************************************************** | |
| 2648 * Parse doc comment embedded between t->ptr and p. | |
| 2649 * Remove trailing blanks and tabs from lines. | |
| 2650 * Replace all newlines with \n. | |
| 2651 * Remove leading comment character from each line. | |
| 2652 * Decide if it's a lineComment or a blockComment. | |
| 2653 * Append to previous one for this token. | |
| 2654 */ | |
| 2655 | |
| 2656 void Lexer::getDocComment(Token *t, unsigned lineComment) | |
| 2657 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2658 /* 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
|
2659 */ |
| 159 | 2660 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
|
2661 |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2662 /* 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
|
2663 */ |
| 159 | 2664 unsigned char *q = t->ptr + 3; // start of comment text |
| 2665 | |
| 2666 unsigned char *qend = p; | |
| 2667 if (ct == '*' || ct == '+') | |
| 2668 qend -= 2; | |
| 2669 | |
| 2670 /* Scan over initial row of ****'s or ++++'s or ////'s | |
| 2671 */ | |
| 2672 for (; q < qend; q++) | |
| 2673 { | |
| 2674 if (*q != ct) | |
| 2675 break; | |
| 2676 } | |
| 2677 | |
| 2678 /* Remove trailing row of ****'s or ++++'s | |
| 2679 */ | |
| 2680 if (ct != '/') | |
| 2681 { | |
| 2682 for (; q < qend; qend--) | |
| 2683 { | |
| 2684 if (qend[-1] != ct) | |
| 2685 break; | |
| 2686 } | |
| 2687 } | |
| 2688 | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2689 /* Comment is now [q .. qend]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2690 * Canonicalize it into buf[]. |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2691 */ |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2692 OutBuffer buf; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2693 int linestart = 0; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2694 |
| 159 | 2695 for (; q < qend; q++) |
| 2696 { | |
| 2697 unsigned char c = *q; | |
| 2698 | |
| 2699 switch (c) | |
| 2700 { | |
| 2701 case '*': | |
| 2702 case '+': | |
| 2703 if (linestart && c == ct) | |
| 2704 { linestart = 0; | |
| 2705 /* Trim preceding whitespace up to preceding \n | |
| 2706 */ | |
| 2707 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2708 buf.offset--; | |
| 2709 continue; | |
| 2710 } | |
| 2711 break; | |
| 2712 | |
| 2713 case ' ': | |
| 2714 case '\t': | |
| 2715 break; | |
| 2716 | |
| 2717 case '\r': | |
| 2718 if (q[1] == '\n') | |
| 2719 continue; // skip the \r | |
| 2720 goto Lnewline; | |
| 2721 | |
| 2722 default: | |
| 2723 if (c == 226) | |
| 2724 { | |
| 2725 // If LS or PS | |
| 2726 if (q[1] == 128 && | |
| 2727 (q[2] == 168 || q[2] == 169)) | |
| 2728 { | |
| 2729 q += 2; | |
| 2730 goto Lnewline; | |
| 2731 } | |
| 2732 } | |
| 2733 linestart = 0; | |
| 2734 break; | |
| 2735 | |
| 2736 Lnewline: | |
| 2737 c = '\n'; // replace all newlines with \n | |
| 2738 case '\n': | |
| 2739 linestart = 1; | |
| 2740 | |
| 2741 /* Trim trailing whitespace | |
| 2742 */ | |
| 2743 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) | |
| 2744 buf.offset--; | |
| 2745 | |
| 2746 break; | |
| 2747 } | |
| 2748 buf.writeByte(c); | |
| 2749 } | |
| 2750 | |
| 2751 // Always end with a newline | |
| 2752 if (!buf.offset || buf.data[buf.offset - 1] != '\n') | |
| 2753 buf.writeByte('\n'); | |
| 2754 | |
| 2755 buf.writeByte(0); | |
| 2756 | |
| 2757 // It's a line comment if the start of the doc comment comes | |
| 2758 // after other non-whitespace on the same line. | |
| 2759 unsigned char** dc = (lineComment && anyToken) | |
| 2760 ? &t->lineComment | |
| 2761 : &t->blockComment; | |
| 2762 | |
| 2763 // Combine with previous doc comment, if any | |
| 2764 if (*dc) | |
| 2765 *dc = combineComments(*dc, (unsigned char *)buf.data); | |
| 2766 else | |
| 2767 *dc = (unsigned char *)buf.extractData(); | |
| 2768 } | |
| 2769 | |
| 2770 /******************************************** | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2771 * Combine two document comments into one, |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2772 * separated by a newline. |
| 159 | 2773 */ |
| 2774 | |
| 2775 unsigned char *Lexer::combineComments(unsigned char *c1, unsigned char *c2) | |
| 2776 { | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2777 //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
|
2778 |
| 159 | 2779 unsigned char *c = c2; |
| 2780 | |
| 2781 if (c1) | |
| 2782 { c = c1; | |
| 2783 if (c2) | |
| 2784 { size_t len1 = strlen((char *)c1); | |
| 2785 size_t len2 = strlen((char *)c2); | |
| 2786 | |
| 2787 c = (unsigned char *)mem.malloc(len1 + 1 + len2 + 1); | |
| 2788 memcpy(c, c1, len1); | |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2789 if (len1 && c1[len1 - 1] != '\n') |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2790 { c[len1] = '\n'; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2791 len1++; |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2792 } |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2793 memcpy(c + len1, c2, len2); |
|
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2794 c[len1 + len2] = 0; |
| 159 | 2795 } |
| 2796 } | |
| 2797 return c; | |
| 2798 } | |
| 2799 | |
| 2800 /******************************************** | |
| 2801 * Create an identifier in the string table. | |
| 2802 */ | |
| 2803 | |
| 2804 Identifier *Lexer::idPool(const char *s) | |
| 2805 { | |
| 2806 size_t len = strlen(s); | |
| 2807 StringValue *sv = stringtable.update(s, len); | |
| 2808 Identifier *id = (Identifier *) sv->ptrvalue; | |
| 2809 if (!id) | |
| 2810 { | |
| 2811 id = new Identifier(sv->lstring.string, TOKidentifier); | |
| 2812 sv->ptrvalue = id; | |
| 2813 } | |
| 2814 return id; | |
| 2815 } | |
| 2816 | |
| 2817 /********************************************* | |
| 2818 * Create a unique identifier using the prefix s. | |
| 2819 */ | |
| 2820 | |
| 2821 Identifier *Lexer::uniqueId(const char *s, int num) | |
| 2822 { char buffer[32]; | |
| 2823 size_t slen = strlen(s); | |
| 2824 | |
| 2825 assert(slen + sizeof(num) * 3 + 1 <= sizeof(buffer)); | |
| 2826 sprintf(buffer, "%s%d", s, num); | |
| 2827 return idPool(buffer); | |
| 2828 } | |
| 2829 | |
| 2830 Identifier *Lexer::uniqueId(const char *s) | |
| 2831 { | |
| 2832 static int num; | |
| 2833 return uniqueId(s, ++num); | |
| 2834 } | |
| 2835 | |
| 2836 /**************************************** | |
| 2837 */ | |
| 2838 | |
| 2839 struct Keyword | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2840 { const char *name; |
| 159 | 2841 enum TOK value; |
| 2842 }; | |
| 2843 | |
| 2844 static Keyword keywords[] = | |
| 2845 { | |
| 2846 // { "", TOK }, | |
| 2847 | |
| 2848 { "this", TOKthis }, | |
| 2849 { "super", TOKsuper }, | |
| 2850 { "assert", TOKassert }, | |
| 2851 { "null", TOKnull }, | |
| 2852 { "true", TOKtrue }, | |
| 2853 { "false", TOKfalse }, | |
| 2854 { "cast", TOKcast }, | |
| 2855 { "new", TOKnew }, | |
| 2856 { "delete", TOKdelete }, | |
| 2857 { "throw", TOKthrow }, | |
| 2858 { "module", TOKmodule }, | |
| 2859 { "pragma", TOKpragma }, | |
| 2860 { "typeof", TOKtypeof }, | |
| 2861 { "typeid", TOKtypeid }, | |
| 2862 | |
| 2863 { "template", TOKtemplate }, | |
| 2864 | |
| 2865 { "void", TOKvoid }, | |
| 2866 { "byte", TOKint8 }, | |
| 2867 { "ubyte", TOKuns8 }, | |
| 2868 { "short", TOKint16 }, | |
| 2869 { "ushort", TOKuns16 }, | |
| 2870 { "int", TOKint32 }, | |
| 2871 { "uint", TOKuns32 }, | |
| 2872 { "long", TOKint64 }, | |
| 2873 { "ulong", TOKuns64 }, | |
| 2874 { "cent", TOKcent, }, | |
| 2875 { "ucent", TOKucent, }, | |
| 2876 { "float", TOKfloat32 }, | |
| 2877 { "double", TOKfloat64 }, | |
| 2878 { "real", TOKfloat80 }, | |
| 2879 | |
| 2880 { "bool", TOKbool }, | |
| 2881 { "char", TOKchar }, | |
| 2882 { "wchar", TOKwchar }, | |
| 2883 { "dchar", TOKdchar }, | |
| 2884 | |
| 2885 { "ifloat", TOKimaginary32 }, | |
| 2886 { "idouble", TOKimaginary64 }, | |
| 2887 { "ireal", TOKimaginary80 }, | |
| 2888 | |
| 2889 { "cfloat", TOKcomplex32 }, | |
| 2890 { "cdouble", TOKcomplex64 }, | |
| 2891 { "creal", TOKcomplex80 }, | |
| 2892 | |
| 2893 { "delegate", TOKdelegate }, | |
| 2894 { "function", TOKfunction }, | |
| 2895 | |
| 2896 { "is", TOKis }, | |
| 2897 { "if", TOKif }, | |
| 2898 { "else", TOKelse }, | |
| 2899 { "while", TOKwhile }, | |
| 2900 { "for", TOKfor }, | |
| 2901 { "do", TOKdo }, | |
| 2902 { "switch", TOKswitch }, | |
| 2903 { "case", TOKcase }, | |
| 2904 { "default", TOKdefault }, | |
| 2905 { "break", TOKbreak }, | |
| 2906 { "continue", TOKcontinue }, | |
| 2907 { "synchronized", TOKsynchronized }, | |
| 2908 { "return", TOKreturn }, | |
| 2909 { "goto", TOKgoto }, | |
| 2910 { "try", TOKtry }, | |
| 2911 { "catch", TOKcatch }, | |
| 2912 { "finally", TOKfinally }, | |
| 2913 { "with", TOKwith }, | |
| 2914 { "asm", TOKasm }, | |
| 2915 { "foreach", TOKforeach }, | |
| 2916 { "foreach_reverse", TOKforeach_reverse }, | |
| 2917 { "scope", TOKscope }, | |
| 2918 | |
| 2919 { "struct", TOKstruct }, | |
| 2920 { "class", TOKclass }, | |
| 2921 { "interface", TOKinterface }, | |
| 2922 { "union", TOKunion }, | |
| 2923 { "enum", TOKenum }, | |
| 2924 { "import", TOKimport }, | |
| 2925 { "mixin", TOKmixin }, | |
| 2926 { "static", TOKstatic }, | |
| 2927 { "final", TOKfinal }, | |
| 2928 { "const", TOKconst }, | |
| 2929 { "typedef", TOKtypedef }, | |
| 2930 { "alias", TOKalias }, | |
| 2931 { "override", TOKoverride }, | |
| 2932 { "abstract", TOKabstract }, | |
| 2933 { "volatile", TOKvolatile }, | |
| 2934 { "debug", TOKdebug }, | |
| 2935 { "deprecated", TOKdeprecated }, | |
| 2936 { "in", TOKin }, | |
| 2937 { "out", TOKout }, | |
| 2938 { "inout", TOKinout }, | |
| 2939 { "lazy", TOKlazy }, | |
| 2940 { "auto", TOKauto }, | |
| 2941 | |
| 2942 { "align", TOKalign }, | |
| 2943 { "extern", TOKextern }, | |
| 2944 { "private", TOKprivate }, | |
| 2945 { "package", TOKpackage }, | |
| 2946 { "protected", TOKprotected }, | |
| 2947 { "public", TOKpublic }, | |
| 2948 { "export", TOKexport }, | |
| 2949 | |
| 2950 { "body", TOKbody }, | |
| 2951 { "invariant", TOKinvariant }, | |
| 2952 { "unittest", TOKunittest }, | |
| 2953 { "version", TOKversion }, | |
| 2954 //{ "manifest", TOKmanifest }, | |
| 2955 | |
| 2956 // Added after 1.0 | |
| 2957 { "ref", TOKref }, | |
| 2958 { "macro", TOKmacro }, | |
| 336 | 2959 #if DMDV2 |
| 159 | 2960 { "pure", TOKpure }, |
| 2961 { "nothrow", TOKnothrow }, | |
| 336 | 2962 { "__thread", TOKtls }, |
| 159 | 2963 { "__traits", TOKtraits }, |
| 2964 { "__overloadset", TOKoverloadset }, | |
| 336 | 2965 { "__FILE__", TOKfile }, |
| 2966 { "__LINE__", TOKline }, | |
|
658
50383e476c7e
Upgraded frontend to DMD 1.035
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
366
diff
changeset
|
2967 { "shared", TOKshared }, |
|
846
bc982f1ad106
Merged DMD 1.037 frontend
Tomas Lindquist Olsen <tomas.l.olsen@gmail.com>
parents:
794
diff
changeset
|
2968 { "immutable", TOKimmutable }, |
| 159 | 2969 #endif |
| 2970 }; | |
| 2971 | |
| 2972 int Token::isKeyword() | |
| 2973 { | |
| 2974 for (unsigned u = 0; u < sizeof(keywords) / sizeof(keywords[0]); u++) | |
| 2975 { | |
| 2976 if (keywords[u].value == value) | |
| 2977 return 1; | |
| 2978 } | |
| 2979 return 0; | |
| 2980 } | |
| 2981 | |
| 2982 void Lexer::initKeywords() | |
| 2983 { StringValue *sv; | |
| 2984 unsigned u; | |
| 2985 enum TOK v; | |
| 2986 unsigned nkeywords = sizeof(keywords) / sizeof(keywords[0]); | |
| 2987 | |
| 2988 if (global.params.Dversion == 1) | |
| 2989 nkeywords -= 2; | |
| 2990 | |
| 2991 cmtable_init(); | |
| 2992 | |
| 2993 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
|
2994 { const char *s; |
| 159 | 2995 |
| 2996 //printf("keyword[%d] = '%s'\n",u, keywords[u].name); | |
| 2997 s = keywords[u].name; | |
| 2998 v = keywords[u].value; | |
| 2999 sv = stringtable.insert(s, strlen(s)); | |
| 3000 sv->ptrvalue = (void *) new Identifier(sv->lstring.string,v); | |
| 3001 | |
| 3002 //printf("tochars[%d] = '%s'\n",v, s); | |
| 3003 Token::tochars[v] = s; | |
| 3004 } | |
| 3005 | |
| 3006 Token::tochars[TOKeof] = "EOF"; | |
| 3007 Token::tochars[TOKlcurly] = "{"; | |
| 3008 Token::tochars[TOKrcurly] = "}"; | |
| 3009 Token::tochars[TOKlparen] = "("; | |
| 3010 Token::tochars[TOKrparen] = ")"; | |
| 3011 Token::tochars[TOKlbracket] = "["; | |
| 3012 Token::tochars[TOKrbracket] = "]"; | |
| 3013 Token::tochars[TOKsemicolon] = ";"; | |
| 3014 Token::tochars[TOKcolon] = ":"; | |
| 3015 Token::tochars[TOKcomma] = ","; | |
| 3016 Token::tochars[TOKdot] = "."; | |
| 3017 Token::tochars[TOKxor] = "^"; | |
| 3018 Token::tochars[TOKxorass] = "^="; | |
| 3019 Token::tochars[TOKassign] = "="; | |
| 3020 Token::tochars[TOKconstruct] = "="; | |
| 336 | 3021 #if DMDV2 |
| 159 | 3022 Token::tochars[TOKblit] = "="; |
| 3023 #endif | |
| 3024 Token::tochars[TOKlt] = "<"; | |
| 3025 Token::tochars[TOKgt] = ">"; | |
| 3026 Token::tochars[TOKle] = "<="; | |
| 3027 Token::tochars[TOKge] = ">="; | |
| 3028 Token::tochars[TOKequal] = "=="; | |
| 3029 Token::tochars[TOKnotequal] = "!="; | |
| 3030 Token::tochars[TOKnotidentity] = "!is"; | |
| 3031 Token::tochars[TOKtobool] = "!!"; | |
| 3032 | |
| 3033 Token::tochars[TOKunord] = "!<>="; | |
| 3034 Token::tochars[TOKue] = "!<>"; | |
| 3035 Token::tochars[TOKlg] = "<>"; | |
| 3036 Token::tochars[TOKleg] = "<>="; | |
| 3037 Token::tochars[TOKule] = "!>"; | |
| 3038 Token::tochars[TOKul] = "!>="; | |
| 3039 Token::tochars[TOKuge] = "!<"; | |
| 3040 Token::tochars[TOKug] = "!<="; | |
| 3041 | |
| 3042 Token::tochars[TOKnot] = "!"; | |
| 3043 Token::tochars[TOKtobool] = "!!"; | |
| 3044 Token::tochars[TOKshl] = "<<"; | |
| 3045 Token::tochars[TOKshr] = ">>"; | |
| 3046 Token::tochars[TOKushr] = ">>>"; | |
| 3047 Token::tochars[TOKadd] = "+"; | |
| 3048 Token::tochars[TOKmin] = "-"; | |
| 3049 Token::tochars[TOKmul] = "*"; | |
| 3050 Token::tochars[TOKdiv] = "/"; | |
| 3051 Token::tochars[TOKmod] = "%"; | |
| 3052 Token::tochars[TOKslice] = ".."; | |
| 3053 Token::tochars[TOKdotdotdot] = "..."; | |
| 3054 Token::tochars[TOKand] = "&"; | |
| 3055 Token::tochars[TOKandand] = "&&"; | |
| 3056 Token::tochars[TOKor] = "|"; | |
| 3057 Token::tochars[TOKoror] = "||"; | |
| 3058 Token::tochars[TOKarray] = "[]"; | |
| 3059 Token::tochars[TOKindex] = "[i]"; | |
| 3060 Token::tochars[TOKaddress] = "&"; | |
| 3061 Token::tochars[TOKstar] = "*"; | |
| 3062 Token::tochars[TOKtilde] = "~"; | |
| 3063 Token::tochars[TOKdollar] = "$"; | |
| 3064 Token::tochars[TOKcast] = "cast"; | |
| 3065 Token::tochars[TOKplusplus] = "++"; | |
| 3066 Token::tochars[TOKminusminus] = "--"; | |
| 3067 Token::tochars[TOKtype] = "type"; | |
| 3068 Token::tochars[TOKquestion] = "?"; | |
| 3069 Token::tochars[TOKneg] = "-"; | |
| 3070 Token::tochars[TOKuadd] = "+"; | |
| 3071 Token::tochars[TOKvar] = "var"; | |
| 3072 Token::tochars[TOKaddass] = "+="; | |
| 3073 Token::tochars[TOKminass] = "-="; | |
| 3074 Token::tochars[TOKmulass] = "*="; | |
| 3075 Token::tochars[TOKdivass] = "/="; | |
| 3076 Token::tochars[TOKmodass] = "%="; | |
| 3077 Token::tochars[TOKshlass] = "<<="; | |
| 3078 Token::tochars[TOKshrass] = ">>="; | |
| 3079 Token::tochars[TOKushrass] = ">>>="; | |
| 3080 Token::tochars[TOKandass] = "&="; | |
| 3081 Token::tochars[TOKorass] = "|="; | |
| 3082 Token::tochars[TOKcatass] = "~="; | |
| 3083 Token::tochars[TOKcat] = "~"; | |
| 3084 Token::tochars[TOKcall] = "call"; | |
| 3085 Token::tochars[TOKidentity] = "is"; | |
| 3086 Token::tochars[TOKnotidentity] = "!is"; | |
| 3087 | |
| 3088 Token::tochars[TOKorass] = "|="; | |
| 3089 Token::tochars[TOKidentifier] = "identifier"; | |
| 3090 | |
| 3091 // For debugging | |
| 3092 Token::tochars[TOKdotexp] = "dotexp"; | |
| 3093 Token::tochars[TOKdotti] = "dotti"; | |
| 3094 Token::tochars[TOKdotvar] = "dotvar"; | |
| 3095 Token::tochars[TOKdottype] = "dottype"; | |
| 3096 Token::tochars[TOKsymoff] = "symoff"; | |
| 3097 Token::tochars[TOKtypedot] = "typedot"; | |
| 3098 Token::tochars[TOKarraylength] = "arraylength"; | |
| 3099 Token::tochars[TOKarrayliteral] = "arrayliteral"; | |
| 3100 Token::tochars[TOKassocarrayliteral] = "assocarrayliteral"; | |
| 3101 Token::tochars[TOKstructliteral] = "structliteral"; | |
| 3102 Token::tochars[TOKstring] = "string"; | |
| 3103 Token::tochars[TOKdsymbol] = "symbol"; | |
| 3104 Token::tochars[TOKtuple] = "tuple"; | |
| 3105 Token::tochars[TOKdeclaration] = "declaration"; | |
| 3106 Token::tochars[TOKdottd] = "dottd"; | |
| 3107 Token::tochars[TOKon_scope_exit] = "scope(exit)"; | |
| 3108 } |