Mercurial > projects > ddmd
annotate dmd/Lexer.d @ 130:60bb0fe4563e
dmdfe 2.037 first main iteration
| author | Eldar Insafutdinov <e.insafutdinov@gmail.com> |
|---|---|
| date | Thu, 09 Sep 2010 22:51:44 +0100 |
| parents | e28b18c23469 |
| children | af1bebfd96a4 |
| rev | line source |
|---|---|
| 0 | 1 module dmd.Lexer; |
| 2 | |
| 114 | 3 import dmd.common; |
| 0 | 4 import dmd.StringTable; |
| 5 import dmd.OutBuffer; | |
| 6 import dmd.Token; | |
| 7 import dmd.Loc; | |
| 8 import dmd.Module; | |
| 9 import dmd.Identifier; | |
| 10 import dmd.TOK; | |
| 11 import dmd.Keyword; | |
| 12 import dmd.StringValue; | |
| 13 import dmd.Global; | |
| 14 import dmd.Util; | |
| 15 import dmd.Id; | |
| 16 import dmd.Dchar; | |
| 17 import dmd.Utf; | |
| 18 | |
| 19 import std.stdio : writeln; | |
| 20 | |
| 4 | 21 import core.memory; |
| 2 | 22 |
| 0 | 23 import core.stdc.ctype; |
| 24 import core.stdc.stdlib; | |
| 25 import core.stdc.string; | |
| 26 import core.stdc.stdio; | |
| 27 import core.stdc.time; | |
| 28 import core.stdc.errno; | |
| 29 | |
| 30 enum LS = 0x2028; // UTF line separator | |
| 31 enum PS = 0x2029; // UTF paragraph separator | |
| 32 | |
| 33 extern (C) extern | |
| 34 { | |
| 35 __gshared char* __locale_decpoint; | |
| 36 } | |
| 37 | |
| 38 int isUniAlpha(uint u) | |
| 39 { | |
| 40 assert(false); | |
| 41 } | |
| 42 | |
| 43 class Lexer | |
| 44 { | |
| 45 static StringTable stringtable; | |
| 46 static OutBuffer stringbuffer; | |
| 47 static Token* freelist; | |
| 48 | |
| 49 Loc loc; // for error messages | |
| 50 | |
| 51 ubyte* base; // pointer to start of buffer | |
| 52 ubyte* end; // past end of buffer | |
| 53 ubyte* p; // current character | |
| 54 Token token; | |
| 55 Module mod; | |
| 56 int doDocComment; // collect doc comment information | |
| 57 int anyToken; // !=0 means seen at least one token | |
| 58 int commentToken; // !=0 means comments are TOKcomment's | |
| 59 | |
| 60 static this() | |
| 61 { | |
| 62 stringtable = new StringTable(); | |
| 63 stringbuffer = new OutBuffer(); | |
| 64 } | |
| 65 | |
| 66 static ~this() | |
| 67 { | |
| 34 | 68 //delete stringtable; |
| 0 | 69 } |
| 70 | |
| 71 this(Module mod, ubyte* base, uint begoffset, uint endoffset, int doDocComment, int commentToken) | |
| 72 { | |
| 73 loc = Loc(mod, 1); | |
| 74 | |
| 75 memset(&token,0,token.sizeof); | |
| 76 this.base = base; | |
| 77 this.end = base + endoffset; | |
| 78 p = base + begoffset; | |
| 79 this.mod = mod; | |
| 80 this.doDocComment = doDocComment; | |
| 81 this.anyToken = 0; | |
| 82 this.commentToken = commentToken; | |
| 83 //initKeywords(); | |
| 84 | |
| 85 /* If first line starts with '#!', ignore the line | |
| 86 */ | |
| 87 | |
| 88 if (p[0] == '#' && p[1] =='!') | |
| 89 { | |
| 90 p += 2; | |
| 91 while (1) | |
| 92 { | |
| 93 ubyte c = *p; | |
| 94 switch (c) | |
| 95 { | |
| 96 case '\n': | |
| 97 p++; | |
| 98 break; | |
| 99 | |
| 100 case '\r': | |
| 101 p++; | |
| 102 if (*p == '\n') | |
| 103 p++; | |
| 104 break; | |
| 105 | |
| 106 case 0: | |
| 107 case 0x1A: | |
| 108 break; | |
| 109 | |
| 110 default: | |
| 111 if (c & 0x80) | |
| 112 { | |
| 113 uint u = decodeUTF(); | |
| 114 if (u == PS || u == LS) | |
| 115 break; | |
| 116 } | |
| 117 p++; | |
| 118 continue; | |
| 119 } | |
| 120 break; | |
| 121 } | |
| 122 loc.linnum = 2; | |
| 123 } | |
| 124 } | |
| 125 | |
| 126 version (DMDV2) { | |
| 127 static Keyword[] keywords = | |
| 128 [ | |
| 129 // { "", TOK }, | |
| 130 | |
| 131 { "this", TOK.TOKthis }, | |
| 132 { "super", TOK.TOKsuper }, | |
| 133 { "assert", TOK.TOKassert }, | |
| 134 { "null", TOK.TOKnull }, | |
| 135 { "true", TOK.TOKtrue }, | |
| 136 { "false", TOK.TOKfalse }, | |
| 137 { "cast", TOK.TOKcast }, | |
| 138 { "new", TOK.TOKnew }, | |
| 139 { "delete", TOK.TOKdelete }, | |
| 140 { "throw", TOK.TOKthrow }, | |
| 141 { "module", TOK.TOKmodule }, | |
| 142 { "pragma", TOK.TOKpragma }, | |
| 143 { "typeof", TOK.TOKtypeof }, | |
| 144 { "typeid", TOK.TOKtypeid }, | |
| 145 | |
| 146 { "template", TOK.TOKtemplate }, | |
| 147 | |
| 148 { "void", TOK.TOKvoid }, | |
| 149 { "byte", TOK.TOKint8 }, | |
| 150 { "ubyte", TOK.TOKuns8 }, | |
| 151 { "short", TOK.TOKint16 }, | |
| 152 { "ushort", TOK.TOKuns16 }, | |
| 153 { "int", TOK.TOKint32 }, | |
| 154 { "uint", TOK.TOKuns32 }, | |
| 155 { "long", TOK.TOKint64 }, | |
| 156 { "ulong", TOK.TOKuns64 }, | |
| 157 { "cent", TOK.TOKcent, }, | |
| 158 { "ucent", TOK.TOKucent, }, | |
| 159 { "float", TOK.TOKfloat32 }, | |
| 160 { "double", TOK.TOKfloat64 }, | |
| 161 { "real", TOK.TOKfloat80 }, | |
| 162 | |
| 163 { "bool", TOK.TOKbool }, | |
| 164 { "char", TOK.TOKchar }, | |
| 165 { "wchar", TOK.TOKwchar }, | |
| 166 { "dchar", TOK.TOKdchar }, | |
| 167 | |
| 168 { "ifloat", TOK.TOKimaginary32 }, | |
| 169 { "idouble", TOK.TOKimaginary64 }, | |
| 170 { "ireal", TOK.TOKimaginary80 }, | |
| 171 | |
| 172 { "cfloat", TOK.TOKcomplex32 }, | |
| 173 { "cdouble", TOK.TOKcomplex64 }, | |
| 174 { "creal", TOK.TOKcomplex80 }, | |
| 175 | |
| 176 { "delegate", TOK.TOKdelegate }, | |
| 177 { "function", TOK.TOKfunction }, | |
| 178 | |
| 179 { "is", TOK.TOKis }, | |
| 180 { "if", TOK.TOKif }, | |
| 181 { "else", TOK.TOKelse }, | |
| 182 { "while", TOK.TOKwhile }, | |
| 183 { "for", TOK.TOKfor }, | |
| 184 { "do", TOK.TOKdo }, | |
| 185 { "switch", TOK.TOKswitch }, | |
| 186 { "case", TOK.TOKcase }, | |
| 187 { "default", TOK.TOKdefault }, | |
| 188 { "break", TOK.TOKbreak }, | |
| 189 { "continue", TOK.TOKcontinue }, | |
| 190 { "synchronized", TOK.TOKsynchronized }, | |
| 191 { "return", TOK.TOKreturn }, | |
| 192 { "goto", TOK.TOKgoto }, | |
| 193 { "try", TOK.TOKtry }, | |
| 194 { "catch", TOK.TOKcatch }, | |
| 195 { "finally", TOK.TOKfinally }, | |
| 196 { "with", TOK.TOKwith }, | |
| 197 { "asm", TOK.TOKasm }, | |
| 198 { "foreach", TOK.TOKforeach }, | |
| 199 { "foreach_reverse", TOK.TOKforeach_reverse }, | |
| 200 { "scope", TOK.TOKscope }, | |
| 201 | |
| 202 { "struct", TOK.TOKstruct }, | |
| 203 { "class", TOK.TOKclass }, | |
| 204 { "interface", TOK.TOKinterface }, | |
| 205 { "union", TOK.TOKunion }, | |
| 206 { "enum", TOK.TOKenum }, | |
| 207 { "import", TOK.TOKimport }, | |
| 208 { "mixin", TOK.TOKmixin }, | |
| 209 { "static", TOK.TOKstatic }, | |
| 210 { "final", TOK.TOKfinal }, | |
| 211 { "const", TOK.TOKconst }, | |
| 212 { "typedef", TOK.TOKtypedef }, | |
| 213 { "alias", TOK.TOKalias }, | |
| 214 { "override", TOK.TOKoverride }, | |
| 215 { "abstract", TOK.TOKabstract }, | |
| 216 { "volatile", TOK.TOKvolatile }, | |
| 217 { "debug", TOK.TOKdebug }, | |
| 218 { "deprecated", TOK.TOKdeprecated }, | |
| 219 { "in", TOK.TOKin }, | |
| 220 { "out", TOK.TOKout }, | |
| 221 { "inout", TOK.TOKinout }, | |
| 222 { "lazy", TOK.TOKlazy }, | |
| 223 { "auto", TOK.TOKauto }, | |
| 224 | |
| 225 { "align", TOK.TOKalign }, | |
| 226 { "extern", TOK.TOKextern }, | |
| 227 { "private", TOK.TOKprivate }, | |
| 228 { "package", TOK.TOKpackage }, | |
| 229 { "protected", TOK.TOKprotected }, | |
| 230 { "public", TOK.TOKpublic }, | |
| 231 { "export", TOK.TOKexport }, | |
| 232 | |
| 233 { "body", TOK.TOKbody }, | |
| 234 { "invariant", TOK.TOKinvariant }, | |
| 235 { "unittest", TOK.TOKunittest }, | |
| 236 { "version", TOK.TOKversion }, | |
| 237 //{ "manifest", TOK.TOKmanifest }, | |
| 238 | |
| 239 // Added after 1.0 | |
| 240 { "ref", TOK.TOKref }, | |
| 241 { "macro", TOK.TOKmacro }, | |
| 242 { "pure", TOK.TOKpure }, | |
| 243 { "nothrow", TOK.TOKnothrow }, | |
| 244 { "__thread", TOK.TOKtls }, | |
| 245 { "__gshared", TOK.TOKgshared }, | |
| 246 { "__traits", TOK.TOKtraits }, | |
| 247 { "__overloadset", TOK.TOKoverloadset }, | |
| 248 { "__FILE__", TOK.TOKfile }, | |
| 249 { "__LINE__", TOK.TOKline }, | |
| 250 { "shared", TOK.TOKshared }, | |
| 251 { "immutable", TOK.TOKimmutable }, | |
| 252 ]; | |
| 253 } else { | |
| 254 static Keyword[] keywords = | |
| 255 [ | |
| 256 // { "", TOK }, | |
| 257 | |
| 258 { "this", TOK.TOKthis }, | |
| 259 { "super", TOK.TOKsuper }, | |
| 260 { "assert", TOK.TOKassert }, | |
| 261 { "null", TOK.TOKnull }, | |
| 262 { "true", TOK.TOKtrue }, | |
| 263 { "false", TOK.TOKfalse }, | |
| 264 { "cast", TOK.TOKcast }, | |
| 265 { "new", TOK.TOKnew }, | |
| 266 { "delete", TOK.TOKdelete }, | |
| 267 { "throw", TOK.TOKthrow }, | |
| 268 { "module", TOK.TOKmodule }, | |
| 269 { "pragma", TOK.TOKpragma }, | |
| 270 { "typeof", TOK.TOKtypeof }, | |
| 271 { "typeid", TOK.TOKtypeid }, | |
| 272 | |
| 273 { "template", TOK.TOKtemplate }, | |
| 274 | |
| 275 { "void", TOK.TOKvoid }, | |
| 276 { "byte", TOK.TOKint8 }, | |
| 277 { "ubyte", TOK.TOKuns8 }, | |
| 278 { "short", TOK.TOKint16 }, | |
| 279 { "ushort", TOK.TOKuns16 }, | |
| 280 { "int", TOK.TOKint32 }, | |
| 281 { "uint", TOK.TOKuns32 }, | |
| 282 { "long", TOK.TOKint64 }, | |
| 283 { "ulong", TOK.TOKuns64 }, | |
| 284 { "cent", TOK.TOKcent, }, | |
| 285 { "ucent", TOK.TOKucent, }, | |
| 286 { "float", TOK.TOKfloat32 }, | |
| 287 { "double", TOK.TOKfloat64 }, | |
| 288 { "real", TOK.TOKfloat80 }, | |
| 289 | |
| 290 { "bool", TOK.TOKbool }, | |
| 291 { "char", TOK.TOKchar }, | |
| 292 { "wchar", TOK.TOKwchar }, | |
| 293 { "dchar", TOK.TOKdchar }, | |
| 294 | |
| 295 { "ifloat", TOK.TOKimaginary32 }, | |
| 296 { "idouble", TOK.TOKimaginary64 }, | |
| 297 { "ireal", TOK.TOKimaginary80 }, | |
| 298 | |
| 299 { "cfloat", TOK.TOKcomplex32 }, | |
| 300 { "cdouble", TOK.TOKcomplex64 }, | |
| 301 { "creal", TOK.TOKcomplex80 }, | |
| 302 | |
| 303 { "delegate", TOK.TOKdelegate }, | |
| 304 { "function", TOK.TOKfunction }, | |
| 305 | |
| 306 { "is", TOK.TOKis }, | |
| 307 { "if", TOK.TOKif }, | |
| 308 { "else", TOK.TOKelse }, | |
| 309 { "while", TOK.TOKwhile }, | |
| 310 { "for", TOK.TOKfor }, | |
| 311 { "do", TOK.TOKdo }, | |
| 312 { "switch", TOK.TOKswitch }, | |
| 313 { "case", TOK.TOKcase }, | |
| 314 { "default", TOK.TOKdefault }, | |
| 315 { "break", TOK.TOKbreak }, | |
| 316 { "continue", TOK.TOKcontinue }, | |
| 317 { "synchronized", TOK.TOKsynchronized }, | |
| 318 { "return", TOK.TOKreturn }, | |
| 319 { "goto", TOK.TOKgoto }, | |
| 320 { "try", TOK.TOKtry }, | |
| 321 { "catch", TOK.TOKcatch }, | |
| 322 { "finally", TOK.TOKfinally }, | |
| 323 { "with", TOK.TOKwith }, | |
| 324 { "asm", TOK.TOKasm }, | |
| 325 { "foreach", TOK.TOKforeach }, | |
| 326 { "foreach_reverse", TOK.TOKforeach_reverse }, | |
| 327 { "scope", TOK.TOKscope }, | |
| 328 | |
| 329 { "struct", TOK.TOKstruct }, | |
| 330 { "class", TOK.TOKclass }, | |
| 331 { "interface", TOK.TOKinterface }, | |
| 332 { "union", TOK.TOKunion }, | |
| 333 { "enum", TOK.TOKenum }, | |
| 334 { "import", TOK.TOKimport }, | |
| 335 { "mixin", TOK.TOKmixin }, | |
| 336 { "static", TOK.TOKstatic }, | |
| 337 { "final", TOK.TOKfinal }, | |
| 338 { "const", TOK.TOKconst }, | |
| 339 { "typedef", TOK.TOKtypedef }, | |
| 340 { "alias", TOK.TOKalias }, | |
| 341 { "override", TOK.TOKoverride }, | |
| 342 { "abstract", TOK.TOKabstract }, | |
| 343 { "volatile", TOK.TOKvolatile }, | |
| 344 { "debug", TOK.TOKdebug }, | |
| 345 { "deprecated", TOK.TOKdeprecated }, | |
| 346 { "in", TOK.TOKin }, | |
| 347 { "out", TOK.TOKout }, | |
| 348 { "inout", TOK.TOKinout }, | |
| 349 { "lazy", TOK.TOKlazy }, | |
| 350 { "auto", TOK.TOKauto }, | |
| 351 | |
| 352 { "align", TOK.TOKalign }, | |
| 353 { "extern", TOK.TOKextern }, | |
| 354 { "private", TOK.TOKprivate }, | |
| 355 { "package", TOK.TOKpackage }, | |
| 356 { "protected", TOK.TOKprotected }, | |
| 357 { "public", TOK.TOKpublic }, | |
| 358 { "export", TOK.TOKexport }, | |
| 359 | |
| 360 { "body", TOK.TOKbody }, | |
| 361 { "invariant", TOK.TOKinvariant }, | |
| 362 { "unittest", TOK.TOKunittest }, | |
| 363 { "version", TOK.TOKversion }, | |
| 364 //{ "manifest", TOK.TOKmanifest }, | |
| 365 | |
| 366 // Added after 1.0 | |
| 367 { "ref", TOK.TOKref }, | |
| 368 { "macro", TOK.TOKmacro }, | |
| 369 ]; | |
| 370 } | |
| 371 | |
| 372 static ubyte cmtable[256]; | |
| 373 enum CMoctal = 0x1; | |
| 374 enum CMhex = 0x2; | |
| 375 enum CMidchar = 0x4; | |
| 376 | |
| 377 ubyte isoctal (ubyte c) { return cmtable[c] & CMoctal; } | |
| 378 ubyte ishex (ubyte c) { return cmtable[c] & CMhex; } | |
| 379 ubyte isidchar(ubyte c) { return cmtable[c] & CMidchar; } | |
| 380 | |
| 381 static void cmtable_init() | |
| 382 { | |
| 383 for (uint c = 0; c < cmtable.length; c++) | |
| 384 { | |
| 385 if ('0' <= c && c <= '7') | |
| 386 cmtable[c] |= CMoctal; | |
| 387 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 388 cmtable[c] |= CMhex; | |
| 389 if (isalnum(c) || c == '_') | |
| 390 cmtable[c] |= CMidchar; | |
| 391 } | |
| 392 } | |
| 393 | |
| 394 static void initKeywords() | |
| 395 { | |
| 396 uint nkeywords = keywords.length; | |
| 397 | |
| 398 if (global.params.Dversion == 1) | |
| 399 nkeywords -= 2; | |
| 400 | |
| 401 cmtable_init(); | |
| 402 | |
| 403 for (uint u = 0; u < nkeywords; u++) | |
| 404 { | |
| 34 | 405 //printf("keyword[%d] = '%.*s'\n",u, keywords[u].name); |
| 0 | 406 string s = keywords[u].name; |
| 407 TOK v = keywords[u].value; | |
| 408 StringValue* sv = stringtable.insert(s); | |
| 409 sv.ptrvalue = cast(void*) new Identifier(sv.lstring.string_, v); | |
| 410 | |
| 411 //printf("tochars[%d] = '%s'\n",v, s); | |
| 412 Token.tochars[v] = s; | |
| 413 } | |
| 414 | |
| 415 Token.tochars[TOK.TOKeof] = "EOF"; | |
| 416 Token.tochars[TOK.TOKlcurly] = "{"; | |
| 417 Token.tochars[TOK.TOKrcurly] = "}"; | |
| 418 Token.tochars[TOK.TOKlparen] = "("; | |
| 419 Token.tochars[TOK.TOKrparen] = ")"; | |
| 420 Token.tochars[TOK.TOKlbracket] = "["; | |
| 421 Token.tochars[TOK.TOKrbracket] = "]"; | |
| 422 Token.tochars[TOK.TOKsemicolon] = ";"; | |
| 423 Token.tochars[TOK.TOKcolon] = ":"; | |
| 424 Token.tochars[TOK.TOKcomma] = ","; | |
| 425 Token.tochars[TOK.TOKdot] = "."; | |
| 426 Token.tochars[TOK.TOKxor] = "^"; | |
| 427 Token.tochars[TOK.TOKxorass] = "^="; | |
| 428 Token.tochars[TOK.TOKassign] = "="; | |
| 429 Token.tochars[TOK.TOKconstruct] = "="; | |
| 430 version (DMDV2) { | |
| 431 Token.tochars[TOK.TOKblit] = "="; | |
| 432 } | |
| 433 Token.tochars[TOK.TOKlt] = "<"; | |
| 434 Token.tochars[TOK.TOKgt] = ">"; | |
| 435 Token.tochars[TOK.TOKle] = "<="; | |
| 436 Token.tochars[TOK.TOKge] = ">="; | |
| 437 Token.tochars[TOK.TOKequal] = "=="; | |
| 438 Token.tochars[TOK.TOKnotequal] = "!="; | |
| 439 Token.tochars[TOK.TOKnotidentity] = "!is"; | |
| 440 Token.tochars[TOK.TOKtobool] = "!!"; | |
| 441 | |
| 442 Token.tochars[TOK.TOKunord] = "!<>="; | |
| 443 Token.tochars[TOK.TOKue] = "!<>"; | |
| 444 Token.tochars[TOK.TOKlg] = "<>"; | |
| 445 Token.tochars[TOK.TOKleg] = "<>="; | |
| 446 Token.tochars[TOK.TOKule] = "!>"; | |
| 447 Token.tochars[TOK.TOKul] = "!>="; | |
| 448 Token.tochars[TOK.TOKuge] = "!<"; | |
| 449 Token.tochars[TOK.TOKug] = "!<="; | |
| 450 | |
| 451 Token.tochars[TOK.TOKnot] = "!"; | |
| 452 Token.tochars[TOK.TOKtobool] = "!!"; | |
| 453 Token.tochars[TOK.TOKshl] = "<<"; | |
| 454 Token.tochars[TOK.TOKshr] = ">>"; | |
| 455 Token.tochars[TOK.TOKushr] = ">>>"; | |
| 456 Token.tochars[TOK.TOKadd] = "+"; | |
| 457 Token.tochars[TOK.TOKmin] = "-"; | |
| 458 Token.tochars[TOK.TOKmul] = "*"; | |
| 459 Token.tochars[TOK.TOKdiv] = "/"; | |
| 460 Token.tochars[TOK.TOKmod] = "%"; | |
| 461 Token.tochars[TOK.TOKslice] = ".."; | |
| 462 Token.tochars[TOK.TOKdotdotdot] = "..."; | |
| 463 Token.tochars[TOK.TOKand] = "&"; | |
| 464 Token.tochars[TOK.TOKandand] = "&&"; | |
| 465 Token.tochars[TOK.TOKor] = "|"; | |
| 466 Token.tochars[TOK.TOKoror] = "||"; | |
| 467 Token.tochars[TOK.TOKarray] = "[]"; | |
| 468 Token.tochars[TOK.TOKindex] = "[i]"; | |
| 469 Token.tochars[TOK.TOKaddress] = "&"; | |
| 470 Token.tochars[TOK.TOKstar] = "*"; | |
| 471 Token.tochars[TOK.TOKtilde] = "~"; | |
| 472 Token.tochars[TOK.TOKdollar] = "$"; | |
| 473 Token.tochars[TOK.TOKcast] = "cast"; | |
| 474 Token.tochars[TOK.TOKplusplus] = "++"; | |
| 475 Token.tochars[TOK.TOKminusminus] = "--"; | |
| 476 Token.tochars[TOK.TOKtype] = "type"; | |
| 477 Token.tochars[TOK.TOKquestion] = "?"; | |
| 478 Token.tochars[TOK.TOKneg] = "-"; | |
| 479 Token.tochars[TOK.TOKuadd] = "+"; | |
| 480 Token.tochars[TOK.TOKvar] = "var"; | |
| 481 Token.tochars[TOK.TOKaddass] = "+="; | |
| 482 Token.tochars[TOK.TOKminass] = "-="; | |
| 483 Token.tochars[TOK.TOKmulass] = "*="; | |
| 484 Token.tochars[TOK.TOKdivass] = "/="; | |
| 485 Token.tochars[TOK.TOKmodass] = "%="; | |
| 486 Token.tochars[TOK.TOKshlass] = "<<="; | |
| 487 Token.tochars[TOK.TOKshrass] = ">>="; | |
| 488 Token.tochars[TOK.TOKushrass] = ">>>="; | |
| 489 Token.tochars[TOK.TOKandass] = "&="; | |
| 490 Token.tochars[TOK.TOKorass] = "|="; | |
| 491 Token.tochars[TOK.TOKcatass] = "~="; | |
| 492 Token.tochars[TOK.TOKcat] = "~"; | |
| 493 Token.tochars[TOK.TOKcall] = "call"; | |
| 494 Token.tochars[TOK.TOKidentity] = "is"; | |
| 495 Token.tochars[TOK.TOKnotidentity] = "!is"; | |
| 496 | |
| 497 Token.tochars[TOK.TOKorass] = "|="; | |
| 498 Token.tochars[TOK.TOKidentifier] = "identifier"; | |
| 499 Token.tochars[TOK.TOKat] = "@"; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
500 Token.tochars[TOK.TOKpow] = "^^"; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
501 //Token.tochars[TOK.TOKpowass] = "^^="; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
502 |
| 0 | 503 // For debugging |
| 73 | 504 Token.tochars[TOKerror] = "error"; |
| 0 | 505 Token.tochars[TOK.TOKdotexp] = "dotexp"; |
| 506 Token.tochars[TOK.TOKdotti] = "dotti"; | |
| 507 Token.tochars[TOK.TOKdotvar] = "dotvar"; | |
| 508 Token.tochars[TOK.TOKdottype] = "dottype"; | |
| 509 Token.tochars[TOK.TOKsymoff] = "symoff"; | |
| 510 Token.tochars[TOK.TOKarraylength] = "arraylength"; | |
| 511 Token.tochars[TOK.TOKarrayliteral] = "arrayliteral"; | |
| 512 Token.tochars[TOK.TOKassocarrayliteral] = "assocarrayliteral"; | |
| 513 Token.tochars[TOK.TOKstructliteral] = "structliteral"; | |
| 514 Token.tochars[TOK.TOKstring] = "string"; | |
| 515 Token.tochars[TOK.TOKdsymbol] = "symbol"; | |
| 516 Token.tochars[TOK.TOKtuple] = "tuple"; | |
| 517 Token.tochars[TOK.TOKdeclaration] = "declaration"; | |
| 518 Token.tochars[TOK.TOKdottd] = "dottd"; | |
| 519 Token.tochars[TOK.TOKon_scope_exit] = "scope(exit)"; | |
| 520 Token.tochars[TOK.TOKon_scope_success] = "scope(success)"; | |
| 521 Token.tochars[TOK.TOKon_scope_failure] = "scope(failure)"; | |
| 522 } | |
| 523 | |
| 524 static Identifier idPool(string s) | |
| 525 { | |
| 526 StringValue* sv = stringtable.update(s); | |
| 527 Identifier id = cast(Identifier) sv.ptrvalue; | |
| 528 if (id is null) | |
| 529 { | |
| 530 id = new Identifier(sv.lstring.string_, TOK.TOKidentifier); | |
| 531 sv.ptrvalue = cast(void*)id; | |
| 532 } | |
| 533 | |
| 534 return id; | |
| 535 } | |
| 536 | |
| 537 static Identifier uniqueId(string s) | |
| 538 { | |
| 539 static int num; | |
| 540 return uniqueId(s, ++num); | |
| 541 } | |
| 542 | |
| 543 /********************************************* | |
| 544 * Create a unique identifier using the prefix s. | |
| 545 */ | |
| 546 static Identifier uniqueId(string s, int num) | |
| 547 { | |
| 548 char buffer[32]; | |
| 549 size_t slen = s.length; | |
| 550 | |
| 551 assert(slen + num.sizeof * 3 + 1 <= buffer.sizeof); | |
| 552 int len = sprintf(buffer.ptr, "%.*s%d", s, num); | |
| 553 | |
| 554 return idPool(buffer[0..len].idup); | |
| 555 } | |
| 556 | |
| 557 TOK nextToken() | |
| 558 { | |
| 559 Token *t; | |
| 560 | |
| 561 if (token.next) | |
| 562 { | |
| 563 t = token.next; | |
| 564 memcpy(&token, t, Token.sizeof); | |
| 565 t.next = freelist; | |
| 566 freelist = t; | |
| 567 } | |
| 568 else | |
| 569 { | |
| 570 scan(&token); | |
| 571 } | |
| 572 | |
| 573 //token.print(); | |
| 574 return token.value; | |
| 575 } | |
| 576 | |
| 577 /*********************** | |
| 578 * Look ahead at next token's value. | |
| 579 */ | |
| 580 TOK peekNext() | |
| 581 { | |
| 582 return peek(&token).value; | |
| 583 } | |
| 584 | |
| 585 TOK peekNext2() | |
| 586 { | |
| 587 assert(false); | |
| 588 } | |
| 589 | |
| 590 void scan(Token* t) | |
| 591 { | |
| 592 uint lastLine = loc.linnum; | |
| 593 uint linnum; | |
| 594 | |
| 595 t.blockComment = null; | |
| 596 t.lineComment = null; | |
| 597 while (1) | |
| 598 { | |
| 599 t.ptr = p; | |
| 600 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 601 switch (*p) | |
| 602 { | |
| 603 case 0: | |
| 604 case 0x1A: | |
| 605 t.value = TOK.TOKeof; // end of file | |
| 606 return; | |
| 607 | |
| 608 case ' ': | |
| 609 case '\t': | |
| 610 case '\v': | |
| 611 case '\f': | |
| 612 p++; | |
| 613 continue; // skip white space | |
| 614 | |
| 615 case '\r': | |
| 616 p++; | |
| 617 if (*p != '\n') // if CR stands by itself | |
| 618 loc.linnum++; | |
| 619 continue; // skip white space | |
| 620 | |
| 621 case '\n': | |
| 622 p++; | |
| 623 loc.linnum++; | |
| 624 continue; // skip white space | |
| 625 | |
| 626 case '0': case '1': case '2': case '3': case '4': | |
| 627 case '5': case '6': case '7': case '8': case '9': | |
| 628 t.value = number(t); | |
| 629 return; | |
| 630 | |
| 631 version (CSTRINGS) { | |
| 632 case '\'': | |
| 633 t.value = charConstant(t, 0); | |
| 634 return; | |
| 635 | |
| 636 case '"': | |
| 637 t.value = stringConstant(t,0); | |
| 638 return; | |
| 639 | |
| 640 case 'l': | |
| 641 case 'L': | |
| 642 if (p[1] == '\'') | |
| 643 { | |
| 644 p++; | |
| 645 t.value = charConstant(t, 1); | |
| 646 return; | |
| 647 } | |
| 648 else if (p[1] == '"') | |
| 649 { | |
| 650 p++; | |
| 651 t.value = stringConstant(t, 1); | |
| 652 return; | |
| 653 } | |
| 654 } else { | |
| 655 case '\'': | |
| 656 t.value = charConstant(t,0); | |
| 657 return; | |
| 658 | |
| 659 case 'r': | |
| 660 if (p[1] != '"') | |
| 661 goto case_ident; | |
| 662 p++; | |
| 663 case '`': | |
| 664 t.value = wysiwygStringConstant(t, *p); | |
| 665 return; | |
| 666 | |
| 667 case 'x': | |
| 668 if (p[1] != '"') | |
| 669 goto case_ident; | |
| 670 p++; | |
| 671 t.value = hexStringConstant(t); | |
| 672 return; | |
| 673 | |
| 674 version (DMDV2) { | |
| 675 case 'q': | |
| 676 if (p[1] == '"') | |
| 677 { | |
| 678 p++; | |
| 679 t.value = delimitedStringConstant(t); | |
| 680 return; | |
| 681 } | |
| 682 else if (p[1] == '{') | |
| 683 { | |
| 684 p++; | |
| 685 t.value = tokenStringConstant(t); | |
| 686 return; | |
| 687 } | |
| 688 else | |
| 689 goto case_ident; | |
| 690 } | |
| 691 | |
| 692 case '"': | |
| 693 t.value = escapeStringConstant(t,0); | |
| 694 return; | |
| 695 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 696 } else { | |
| 697 case '\\': // escaped string literal | |
| 698 { uint c; | |
| 699 ubyte* pstart = p; | |
| 700 | |
| 701 stringbuffer.reset(); | |
| 702 do | |
| 703 { | |
| 704 p++; | |
| 705 switch (*p) | |
| 706 { | |
| 707 case 'u': | |
| 708 case 'U': | |
| 709 case '&': | |
| 710 c = escapeSequence(); | |
| 711 stringbuffer.writeUTF8(c); | |
| 712 break; | |
| 713 | |
| 714 default: | |
| 715 c = escapeSequence(); | |
| 716 stringbuffer.writeByte(c); | |
| 717 break; | |
| 718 } | |
| 719 } while (*p == '\\'); | |
| 720 t.len = stringbuffer.offset; | |
| 721 stringbuffer.writeByte(0); | |
| 2 | 722 char* cc = cast(char*)GC.malloc(stringbuffer.offset); |
| 0 | 723 memcpy(cc, stringbuffer.data, stringbuffer.offset); |
| 724 t.ustring = cc; | |
| 725 t.postfix = 0; | |
| 726 t.value = TOK.TOKstring; | |
| 727 if (!global.params.useDeprecated) | |
| 728 error("Escape String literal %.*s is deprecated, use double quoted string literal \"%.*s\" instead", p - pstart, pstart, p - pstart, pstart); | |
| 729 return; | |
| 730 } | |
| 731 } | |
| 732 case 'l': | |
| 733 case 'L': | |
| 734 } | |
| 735 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 736 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 737 case 'k': case 'm': case 'n': case 'o': | |
| 738 version (DMDV2) { | |
| 739 case 'p': /*case 'q': case 'r':*/ case 's': case 't': | |
| 740 } else { | |
| 741 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 742 } | |
| 743 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 744 case 'z': | |
| 745 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 746 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 747 case 'K': case 'M': case 'N': case 'O': | |
| 748 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 749 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 750 case 'Z': | |
| 751 case '_': | |
| 752 case_ident: | |
| 753 { ubyte c; | |
| 754 StringValue *sv; | |
| 755 Identifier id; | |
| 756 | |
| 757 do | |
| 758 { | |
| 759 c = *++p; | |
| 760 } while (isidchar(c) || (c & 0x80 && isUniAlpha(decodeUTF()))); | |
| 761 sv = stringtable.update((cast(immutable(char)*)t.ptr)[0.. p - t.ptr]); /// | |
| 762 id = cast(Identifier) sv.ptrvalue; | |
| 763 if (id is null) | |
| 764 { id = new Identifier(sv.lstring.string_, TOK.TOKidentifier); | |
| 765 sv.ptrvalue = cast(void*)id; | |
| 766 } | |
| 767 t.ident = id; | |
| 768 t.value = cast(TOK) id.value; | |
| 769 anyToken = 1; | |
| 770 if (*t.ptr == '_') // if special identifier token | |
| 771 { | |
| 772 static char date[11+1]; | |
| 773 static char time[8+1]; | |
| 774 static char timestamp[24+1]; | |
| 775 | |
| 776 if (!date[0]) // lazy evaluation | |
| 777 { time_t tm; | |
| 778 char *p; | |
| 779 | |
| 780 .time(&tm); | |
| 781 p = ctime(&tm); | |
| 782 assert(p); | |
| 783 sprintf(date.ptr, "%.6s %.4s", p + 4, p + 20); | |
| 784 sprintf(time.ptr, "%.8s", p + 11); | |
| 785 sprintf(timestamp.ptr, "%.24s", p); | |
| 786 } | |
| 787 | |
| 788 ///version (DMDV1) { | |
| 789 /// if (mod && id == Id.FILE) | |
| 790 /// { | |
| 791 /// t.ustring = cast(ubyte*)(loc.filename ? loc.filename : mod.ident.toChars()); | |
| 792 /// goto Lstr; | |
| 793 /// } | |
| 794 /// else if (mod && id == Id.LINE) | |
| 795 /// { | |
| 796 /// t.value = TOK.TOKint64v; | |
| 797 /// t.uns64value = loc.linnum; | |
| 798 /// } | |
| 799 /// else | |
| 800 ///} | |
| 801 if (id == Id.DATE) | |
| 802 { | |
| 803 t.ustring = date.ptr; | |
| 804 goto Lstr; | |
| 805 } | |
| 806 else if (id == Id.TIME) | |
| 807 { | |
| 808 t.ustring = time.ptr; | |
| 809 goto Lstr; | |
| 810 } | |
| 811 else if (id == Id.VENDOR) | |
| 812 { | |
| 813 t.ustring = "Digital Mars D".ptr; | |
| 814 goto Lstr; | |
| 815 } | |
| 816 else if (id == Id.TIMESTAMP) | |
| 817 { | |
| 818 t.ustring = timestamp.ptr; | |
| 819 Lstr: | |
| 820 t.value = TOK.TOKstring; | |
| 821 Llen: | |
| 822 t.postfix = 0; | |
| 823 t.len = strlen(cast(char*)t.ustring); | |
| 824 } | |
| 825 else if (id == Id.VERSIONX) | |
| 826 { | |
| 827 uint major = 0; | |
| 828 uint minor = 0; | |
| 829 | |
| 830 foreach (char cc; global.version_[1..$]) | |
| 831 { | |
| 832 if (isdigit(cc)) | |
| 833 minor = minor * 10 + cc - '0'; | |
| 834 else if (cc == '.') | |
| 835 { | |
| 836 major = minor; | |
| 837 minor = 0; | |
| 838 } | |
| 839 else | |
| 840 break; | |
| 841 } | |
| 842 t.value = TOK.TOKint64v; | |
| 843 t.uns64value = major * 1000 + minor; | |
| 844 } | |
| 845 ///version (DMDV2) { | |
| 846 else if (id == Id.EOFX) | |
| 847 { | |
| 848 t.value = TOK.TOKeof; | |
| 849 // Advance scanner to end of file | |
| 850 while (!(*p == 0 || *p == 0x1A)) | |
| 851 p++; | |
| 852 } | |
| 853 ///} | |
| 854 } | |
| 855 //printf("t.value = %d\n",t.value); | |
| 856 return; | |
| 857 } | |
| 858 | |
| 859 case '/': | |
| 860 p++; | |
| 861 switch (*p) | |
| 862 { | |
| 863 case '=': | |
| 864 p++; | |
| 865 t.value = TOK.TOKdivass; | |
| 866 return; | |
| 867 | |
| 868 case '*': | |
| 869 p++; | |
| 870 linnum = loc.linnum; | |
| 871 while (1) | |
| 872 { | |
| 873 while (1) | |
| 874 { | |
| 875 ubyte c = *p; | |
| 876 switch (c) | |
| 877 { | |
| 878 case '/': | |
| 879 break; | |
| 880 | |
| 881 case '\n': | |
| 882 loc.linnum++; | |
| 883 p++; | |
| 884 continue; | |
| 885 | |
| 886 case '\r': | |
| 887 p++; | |
| 888 if (*p != '\n') | |
| 889 loc.linnum++; | |
| 890 continue; | |
| 891 | |
| 892 case 0: | |
| 893 case 0x1A: | |
| 894 error("unterminated /* */ comment"); | |
| 895 p = end; | |
| 896 t.value = TOK.TOKeof; | |
| 897 return; | |
| 898 | |
| 899 default: | |
| 900 if (c & 0x80) | |
| 901 { uint u = decodeUTF(); | |
| 902 if (u == PS || u == LS) | |
| 903 loc.linnum++; | |
| 904 } | |
| 905 p++; | |
| 906 continue; | |
| 907 } | |
| 908 break; | |
| 909 } | |
| 910 p++; | |
| 911 if (p[-2] == '*' && p - 3 != t.ptr) | |
| 912 break; | |
| 913 } | |
| 914 if (commentToken) | |
| 915 { | |
| 916 t.value = TOK.TOKcomment; | |
| 917 return; | |
| 918 } | |
| 919 else if (doDocComment && t.ptr[2] == '*' && p - 4 != t.ptr) | |
| 920 { // if /** but not /**/ | |
| 921 getDocComment(t, lastLine == linnum); | |
| 922 } | |
| 923 continue; | |
| 924 | |
| 925 case '/': // do // style comments | |
| 926 linnum = loc.linnum; | |
| 927 while (1) | |
| 928 { ubyte c = *++p; | |
| 929 switch (c) | |
| 930 { | |
| 931 case '\n': | |
| 932 break; | |
| 933 | |
| 934 case '\r': | |
| 935 if (p[1] == '\n') | |
| 936 p++; | |
| 937 break; | |
| 938 | |
| 939 case 0: | |
| 940 case 0x1A: | |
| 941 if (commentToken) | |
| 942 { | |
| 943 p = end; | |
| 944 t.value = TOK.TOKcomment; | |
| 945 return; | |
| 946 } | |
| 947 if (doDocComment && t.ptr[2] == '/') | |
| 948 getDocComment(t, lastLine == linnum); | |
| 949 p = end; | |
| 950 t.value = TOK.TOKeof; | |
| 951 return; | |
| 952 | |
| 953 default: | |
| 954 if (c & 0x80) | |
| 955 { uint u = decodeUTF(); | |
| 956 if (u == PS || u == LS) | |
| 957 break; | |
| 958 } | |
| 959 continue; | |
| 960 } | |
| 961 break; | |
| 962 } | |
| 963 | |
| 964 if (commentToken) | |
| 965 { | |
| 966 p++; | |
| 967 loc.linnum++; | |
| 968 t.value = TOK.TOKcomment; | |
| 969 return; | |
| 970 } | |
| 971 if (doDocComment && t.ptr[2] == '/') | |
| 972 getDocComment(t, lastLine == linnum); | |
| 973 | |
| 974 p++; | |
| 975 loc.linnum++; | |
| 976 continue; | |
| 977 | |
| 978 case '+': | |
| 979 { | |
| 980 int nest; | |
| 981 | |
| 982 linnum = loc.linnum; | |
| 983 p++; | |
| 984 nest = 1; | |
| 985 while (1) | |
| 986 { ubyte c = *p; | |
| 987 switch (c) | |
| 988 { | |
| 989 case '/': | |
| 990 p++; | |
| 991 if (*p == '+') | |
| 992 { | |
| 993 p++; | |
| 994 nest++; | |
| 995 } | |
| 996 continue; | |
| 997 | |
| 998 case '+': | |
| 999 p++; | |
| 1000 if (*p == '/') | |
| 1001 { | |
| 1002 p++; | |
| 1003 if (--nest == 0) | |
| 1004 break; | |
| 1005 } | |
| 1006 continue; | |
| 1007 | |
| 1008 case '\r': | |
| 1009 p++; | |
| 1010 if (*p != '\n') | |
| 1011 loc.linnum++; | |
| 1012 continue; | |
| 1013 | |
| 1014 case '\n': | |
| 1015 loc.linnum++; | |
| 1016 p++; | |
| 1017 continue; | |
| 1018 | |
| 1019 case 0: | |
| 1020 case 0x1A: | |
| 1021 error("unterminated /+ +/ comment"); | |
| 1022 p = end; | |
| 1023 t.value = TOK.TOKeof; | |
| 1024 return; | |
| 1025 | |
| 1026 default: | |
| 1027 if (c & 0x80) | |
| 1028 { uint u = decodeUTF(); | |
| 1029 if (u == PS || u == LS) | |
| 1030 loc.linnum++; | |
| 1031 } | |
| 1032 p++; | |
| 1033 continue; | |
| 1034 } | |
| 1035 break; | |
| 1036 } | |
| 1037 if (commentToken) | |
| 1038 { | |
| 1039 t.value = TOK.TOKcomment; | |
| 1040 return; | |
| 1041 } | |
| 1042 if (doDocComment && t.ptr[2] == '+' && p - 4 != t.ptr) | |
| 1043 { // if /++ but not /++/ | |
| 1044 getDocComment(t, lastLine == linnum); | |
| 1045 } | |
| 1046 continue; | |
| 1047 } | |
| 1048 | |
| 1049 default: | |
| 1050 break; /// | |
| 1051 } | |
| 1052 t.value = TOK.TOKdiv; | |
| 1053 return; | |
| 1054 | |
| 1055 case '.': | |
| 1056 p++; | |
| 1057 if (isdigit(*p)) | |
| 1058 { /* Note that we don't allow ._1 and ._ as being | |
| 1059 * valid floating point numbers. | |
| 1060 */ | |
| 1061 p--; | |
| 1062 t.value = inreal(t); | |
| 1063 } | |
| 1064 else if (p[0] == '.') | |
| 1065 { | |
| 1066 if (p[1] == '.') | |
| 1067 { p += 2; | |
| 1068 t.value = TOK.TOKdotdotdot; | |
| 1069 } | |
| 1070 else | |
| 1071 { p++; | |
| 1072 t.value = TOK.TOKslice; | |
| 1073 } | |
| 1074 } | |
| 1075 else | |
| 1076 t.value = TOK.TOKdot; | |
| 1077 return; | |
| 1078 | |
| 1079 case '&': | |
| 1080 p++; | |
| 1081 if (*p == '=') | |
| 1082 { p++; | |
| 1083 t.value = TOK.TOKandass; | |
| 1084 } | |
| 1085 else if (*p == '&') | |
| 1086 { p++; | |
| 1087 t.value = TOK.TOKandand; | |
| 1088 } | |
| 1089 else | |
| 1090 t.value = TOK.TOKand; | |
| 1091 return; | |
| 1092 | |
| 1093 case '|': | |
| 1094 p++; | |
| 1095 if (*p == '=') | |
| 1096 { p++; | |
| 1097 t.value = TOK.TOKorass; | |
| 1098 } | |
| 1099 else if (*p == '|') | |
| 1100 { p++; | |
| 1101 t.value = TOK.TOKoror; | |
| 1102 } | |
| 1103 else | |
| 1104 t.value = TOK.TOKor; | |
| 1105 return; | |
| 1106 | |
| 1107 case '-': | |
| 1108 p++; | |
| 1109 if (*p == '=') | |
| 1110 { p++; | |
| 1111 t.value = TOK.TOKminass; | |
| 1112 } | |
| 1113 /// #if 0 | |
| 1114 /// else if (*p == '>') | |
| 1115 /// { p++; | |
| 1116 /// t.value = TOK.TOKarrow; | |
| 1117 /// } | |
| 1118 /// #endif | |
| 1119 else if (*p == '-') | |
| 1120 { p++; | |
| 1121 t.value = TOK.TOKminusminus; | |
| 1122 } | |
| 1123 else | |
| 1124 t.value = TOK.TOKmin; | |
| 1125 return; | |
| 1126 | |
| 1127 case '+': | |
| 1128 p++; | |
| 1129 if (*p == '=') | |
| 1130 { p++; | |
| 1131 t.value = TOK.TOKaddass; | |
| 1132 } | |
| 1133 else if (*p == '+') | |
| 1134 { p++; | |
| 1135 t.value = TOK.TOKplusplus; | |
| 1136 } | |
| 1137 else | |
| 1138 t.value = TOK.TOKadd; | |
| 1139 return; | |
| 1140 | |
| 1141 case '<': | |
| 1142 p++; | |
| 1143 if (*p == '=') | |
| 1144 { p++; | |
| 1145 t.value = TOK.TOKle; // <= | |
| 1146 } | |
| 1147 else if (*p == '<') | |
| 1148 { p++; | |
| 1149 if (*p == '=') | |
| 1150 { p++; | |
| 1151 t.value = TOK.TOKshlass; // <<= | |
| 1152 } | |
| 1153 else | |
| 1154 t.value = TOK.TOKshl; // << | |
| 1155 } | |
| 1156 else if (*p == '>') | |
| 1157 { p++; | |
| 1158 if (*p == '=') | |
| 1159 { p++; | |
| 1160 t.value = TOK.TOKleg; // <>= | |
| 1161 } | |
| 1162 else | |
| 1163 t.value = TOK.TOKlg; // <> | |
| 1164 } | |
| 1165 else | |
| 1166 t.value = TOK.TOKlt; // < | |
| 1167 return; | |
| 1168 | |
| 1169 case '>': | |
| 1170 p++; | |
| 1171 if (*p == '=') | |
| 1172 { p++; | |
| 1173 t.value = TOK.TOKge; // >= | |
| 1174 } | |
| 1175 else if (*p == '>') | |
| 1176 { p++; | |
| 1177 if (*p == '=') | |
| 1178 { p++; | |
| 1179 t.value = TOK.TOKshrass; // >>= | |
| 1180 } | |
| 1181 else if (*p == '>') | |
| 1182 { p++; | |
| 1183 if (*p == '=') | |
| 1184 { p++; | |
| 1185 t.value = TOK.TOKushrass; // >>>= | |
| 1186 } | |
| 1187 else | |
| 1188 t.value = TOK.TOKushr; // >>> | |
| 1189 } | |
| 1190 else | |
| 1191 t.value = TOK.TOKshr; // >> | |
| 1192 } | |
| 1193 else | |
| 1194 t.value = TOK.TOKgt; // > | |
| 1195 return; | |
| 1196 | |
| 1197 case '!': | |
| 1198 p++; | |
| 1199 if (*p == '=') | |
| 1200 { p++; | |
| 1201 if (*p == '=' && global.params.Dversion == 1) | |
| 1202 { p++; | |
| 1203 t.value = TOK.TOKnotidentity; // !== | |
| 1204 } | |
| 1205 else | |
| 1206 t.value = TOK.TOKnotequal; // != | |
| 1207 } | |
| 1208 else if (*p == '<') | |
| 1209 { p++; | |
| 1210 if (*p == '>') | |
| 1211 { p++; | |
| 1212 if (*p == '=') | |
| 1213 { p++; | |
| 1214 t.value = TOK.TOKunord; // !<>= | |
| 1215 } | |
| 1216 else | |
| 1217 t.value = TOK.TOKue; // !<> | |
| 1218 } | |
| 1219 else if (*p == '=') | |
| 1220 { p++; | |
| 1221 t.value = TOK.TOKug; // !<= | |
| 1222 } | |
| 1223 else | |
| 1224 t.value = TOK.TOKuge; // !< | |
| 1225 } | |
| 1226 else if (*p == '>') | |
| 1227 { p++; | |
| 1228 if (*p == '=') | |
| 1229 { p++; | |
| 1230 t.value = TOK.TOKul; // !>= | |
| 1231 } | |
| 1232 else | |
| 1233 t.value = TOK.TOKule; // !> | |
| 1234 } | |
| 1235 else | |
| 1236 t.value = TOK.TOKnot; // ! | |
| 1237 return; | |
| 1238 | |
| 1239 case '=': | |
| 1240 p++; | |
| 1241 if (*p == '=') | |
| 1242 { p++; | |
| 1243 if (*p == '=' && global.params.Dversion == 1) | |
| 1244 { p++; | |
| 1245 t.value = TOK.TOKidentity; // === | |
| 1246 } | |
| 1247 else | |
| 1248 t.value = TOK.TOKequal; // == | |
| 1249 } | |
| 1250 else | |
| 1251 t.value = TOK.TOKassign; // = | |
| 1252 return; | |
| 1253 | |
| 1254 case '~': | |
| 1255 p++; | |
| 1256 if (*p == '=') | |
| 1257 { p++; | |
| 1258 t.value = TOK.TOKcatass; // ~= | |
| 1259 } | |
| 1260 else | |
| 1261 t.value = TOK.TOKtilde; // ~ | |
| 1262 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1263 |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1264 version(DMDV2) { |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1265 case '^': |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1266 p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1267 if (*p == '^') |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1268 { p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1269 //static if (false) { |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1270 // if (*p == '=') |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1271 // { p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1272 // t.value = TOKpowass; // ^^= |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1273 // } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1274 // else |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1275 //} |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1276 t.value = TOKpow; // ^^ |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1277 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1278 else if (*p == '=') |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1279 { p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1280 t.value = TOKxorass; // ^= |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1281 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1282 else |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1283 t.value = TOKxor; // ^ |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1284 return; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1285 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1286 |
| 0 | 1287 /* |
| 1288 #define SINGLE(c,tok) case c: p++; t.value = tok; return; | |
| 1289 | |
| 1290 SINGLE('(', TOKlparen) | |
| 1291 SINGLE(')', TOKrparen) | |
| 1292 SINGLE('[', TOKlbracket) | |
| 1293 SINGLE(']', TOKrbracket) | |
| 1294 SINGLE('{', TOKlcurly) | |
| 1295 SINGLE('}', TOKrcurly) | |
| 1296 SINGLE('?', TOKquestion) | |
| 1297 SINGLE(',', TOKcomma) | |
| 1298 SINGLE(';', TOKsemicolon) | |
| 1299 SINGLE(':', TOKcolon) | |
| 1300 SINGLE('$', TOKdollar) | |
| 1301 SINGLE('@', TOKat) | |
| 1302 | |
| 1303 #undef SINGLE | |
| 1304 | |
| 1305 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1306 case c1: \ | |
| 1307 p++; \ | |
| 1308 if (*p == c2) \ | |
| 1309 { p++; \ | |
| 1310 t.value = tok2; \ | |
| 1311 } \ | |
| 1312 else \ | |
| 1313 t.value = tok1; \ | |
| 1314 return; | |
| 1315 | |
| 1316 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1317 DOUBLE('%', TOKmod, '=', TOKmodass) | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1318 #if DMDV1 |
| 0 | 1319 DOUBLE('^', TOKxor, '=', TOKxorass) |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1320 #endif |
| 0 | 1321 #undef DOUBLE |
| 1322 */ | |
| 1323 | |
| 1324 case '(': p++; t.value = TOK.TOKlparen; return; | |
| 1325 case ')': p++; t.value = TOK.TOKrparen; return; | |
| 1326 case '[': p++; t.value = TOK.TOKlbracket; return; | |
| 1327 case ']': p++; t.value = TOK.TOKrbracket; return; | |
| 1328 case '{': p++; t.value = TOK.TOKlcurly; return; | |
| 1329 case '}': p++; t.value = TOK.TOKrcurly; return; | |
| 1330 case '?': p++; t.value = TOK.TOKquestion; return; | |
| 1331 case ',': p++; t.value = TOK.TOKcomma; return; | |
| 1332 case ';': p++; t.value = TOK.TOKsemicolon; return; | |
| 1333 case ':': p++; t.value = TOK.TOKcolon; return; | |
| 1334 case '$': p++; t.value = TOK.TOKdollar; return; | |
| 1335 case '@': p++; t.value = TOK.TOKat; return; | |
| 1336 | |
| 1337 case '*': | |
| 1338 p++; | |
| 1339 if (*p == '=') { | |
| 1340 p++; | |
| 1341 t.value = TOK.TOKmulass; | |
| 1342 } else { | |
| 1343 t.value = TOK.TOKmul; | |
| 1344 } | |
| 1345 return; | |
| 1346 | |
| 1347 case '%': | |
| 1348 p++; | |
| 1349 if (*p == '=') { | |
| 1350 p++; | |
| 1351 t.value = TOK.TOKmodass; | |
| 1352 } else { | |
| 1353 t.value = TOK.TOKmod; | |
| 1354 } | |
| 1355 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1356 version(DMDV1) { |
| 0 | 1357 case '^': |
| 1358 p++; | |
| 1359 if (*p == '=') { | |
| 1360 p++; | |
| 1361 t.value = TOK.TOKxorass; | |
| 1362 } else { | |
| 1363 t.value = TOK.TOKxor; | |
| 1364 } | |
| 1365 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1366 } |
| 0 | 1367 case '#': |
| 1368 p++; | |
| 1369 pragma_(); | |
| 1370 continue; | |
| 1371 | |
| 1372 default: | |
| 1373 { ubyte c = *p; | |
| 1374 | |
| 1375 if (c & 0x80) | |
| 1376 { uint u = decodeUTF(); | |
| 1377 | |
| 1378 // Check for start of unicode identifier | |
| 1379 if (isUniAlpha(u)) | |
| 1380 goto case_ident; | |
| 1381 | |
| 1382 if (u == PS || u == LS) | |
| 1383 { | |
| 1384 loc.linnum++; | |
| 1385 p++; | |
| 1386 continue; | |
| 1387 } | |
| 1388 } | |
| 1389 if (isprint(c)) | |
| 1390 error("unsupported char '%c'", c); | |
| 1391 else | |
| 1392 error("unsupported char 0x%02x", c); | |
| 1393 p++; | |
| 1394 continue; | |
| 1395 } | |
| 1396 } | |
| 1397 } | |
| 1398 } | |
| 1399 | |
| 1400 Token* peek(Token* ct) | |
| 1401 { | |
| 1402 Token* t; | |
| 1403 | |
| 1404 if (ct.next) | |
| 1405 t = ct.next; | |
| 1406 else | |
| 1407 { | |
| 1408 t = new Token(); | |
| 1409 scan(t); | |
| 1410 t.next = null; | |
| 1411 ct.next = t; | |
| 1412 } | |
| 1413 return t; | |
| 1414 } | |
| 1415 | |
| 1416 Token* peekPastParen(Token* tk) | |
| 1417 { | |
| 1418 //printf("peekPastParen()\n"); | |
| 1419 int parens = 1; | |
| 1420 int curlynest = 0; | |
| 1421 while (1) | |
| 1422 { | |
| 1423 tk = peek(tk); | |
| 1424 //tk.print(); | |
| 1425 switch (tk.value) | |
| 1426 { | |
| 1427 case TOK.TOKlparen: | |
| 1428 parens++; | |
| 1429 continue; | |
| 1430 | |
| 1431 case TOK.TOKrparen: | |
| 1432 --parens; | |
| 1433 if (parens) | |
| 1434 continue; | |
| 1435 tk = peek(tk); | |
| 1436 break; | |
| 1437 | |
| 1438 case TOK.TOKlcurly: | |
| 1439 curlynest++; | |
| 1440 continue; | |
| 1441 | |
| 1442 case TOK.TOKrcurly: | |
| 1443 if (--curlynest >= 0) | |
| 1444 continue; | |
| 1445 break; | |
| 1446 | |
| 1447 case TOK.TOKsemicolon: | |
| 1448 if (curlynest) | |
| 1449 continue; | |
| 1450 break; | |
| 1451 | |
| 1452 case TOK.TOKeof: | |
| 1453 break; | |
| 1454 | |
| 1455 default: | |
| 1456 continue; | |
| 1457 } | |
| 1458 return tk; | |
| 1459 } | |
| 1460 } | |
| 1461 | |
| 1462 /******************************************* | |
| 1463 * Parse escape sequence. | |
| 1464 */ | |
| 1465 uint escapeSequence() | |
| 1466 { | |
| 1467 uint c = *p; | |
| 1468 | |
| 1469 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 1470 return c; | |
| 1471 } | |
| 1472 int n; | |
| 1473 int ndigits; | |
| 1474 | |
| 1475 switch (c) | |
| 1476 { | |
| 1477 case '\'': | |
| 1478 case '"': | |
| 1479 case '?': | |
| 1480 case '\\': | |
| 1481 Lconsume: | |
| 1482 p++; | |
| 1483 break; | |
| 1484 | |
| 1485 case 'a': c = 7; goto Lconsume; | |
| 1486 case 'b': c = 8; goto Lconsume; | |
| 1487 case 'f': c = 12; goto Lconsume; | |
| 1488 case 'n': c = 10; goto Lconsume; | |
| 1489 case 'r': c = 13; goto Lconsume; | |
| 1490 case 't': c = 9; goto Lconsume; | |
| 1491 case 'v': c = 11; goto Lconsume; | |
| 1492 | |
| 1493 case 'u': | |
| 1494 ndigits = 4; | |
| 1495 goto Lhex; | |
| 1496 case 'U': | |
| 1497 ndigits = 8; | |
| 1498 goto Lhex; | |
| 1499 case 'x': | |
| 1500 ndigits = 2; | |
| 1501 Lhex: | |
| 1502 p++; | |
| 1503 c = *p; | |
| 1504 if (ishex(cast(ubyte)c)) | |
| 1505 { | |
| 1506 uint v; | |
| 1507 | |
| 1508 n = 0; | |
| 1509 v = 0; | |
| 1510 while (1) | |
| 1511 { | |
| 1512 if (isdigit(c)) | |
| 1513 c -= '0'; | |
| 1514 else if (islower(c)) | |
| 1515 c -= 'a' - 10; | |
| 1516 else | |
| 1517 c -= 'A' - 10; | |
| 1518 v = v * 16 + c; | |
| 1519 c = *++p; | |
| 1520 if (++n == ndigits) | |
| 1521 break; | |
| 1522 if (!ishex(cast(ubyte)c)) | |
| 1523 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1524 break; | |
| 1525 } | |
| 1526 } | |
| 1527 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1528 { error("invalid UTF character \\U%08x", v); | |
| 1529 v = '?'; // recover with valid UTF character | |
| 1530 } | |
| 1531 c = v; | |
| 1532 } | |
| 1533 else | |
| 1534 error("undefined escape hex sequence \\%c\n",c); | |
| 1535 break; | |
| 1536 | |
| 1537 case '&': // named character entity | |
| 1538 for (ubyte* idstart = ++p; true; p++) | |
| 1539 { | |
| 1540 switch (*p) | |
| 1541 { | |
| 1542 case ';': | |
| 1543 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1544 if (c == ~0) | |
| 1545 { | |
| 1546 error("unnamed character entity &%s;", idstart[0..(p - idstart)]); | |
| 1547 c = ' '; | |
| 1548 } | |
| 1549 p++; | |
| 1550 break; | |
| 1551 | |
| 1552 default: | |
| 1553 if (isalpha(*p) || | |
| 1554 (p != idstart + 1 && isdigit(*p))) | |
| 1555 continue; | |
| 1556 error("unterminated named entity"); | |
| 1557 break; | |
| 1558 } | |
| 1559 break; | |
| 1560 } | |
| 1561 break; | |
| 1562 | |
| 1563 case 0: | |
| 1564 case 0x1A: // end of file | |
| 1565 c = '\\'; | |
| 1566 break; | |
| 1567 | |
| 1568 default: | |
| 1569 if (isoctal(cast(ubyte)c)) | |
| 1570 { | |
| 1571 uint v; | |
| 1572 | |
| 1573 n = 0; | |
| 1574 v = 0; | |
| 1575 do | |
| 1576 { | |
| 1577 v = v * 8 + (c - '0'); | |
| 1578 c = *++p; | |
| 1579 } while (++n < 3 && isoctal(cast(ubyte)c)); | |
| 1580 c = v; | |
| 1581 if (c > 0xFF) | |
| 1582 error("0%03o is larger than a byte", c); | |
| 1583 } | |
| 1584 else | |
| 1585 error("undefined escape sequence \\%c\n",c); | |
| 1586 break; | |
| 1587 } | |
| 1588 return c; | |
| 1589 } | |
| 1590 | |
| 1591 TOK wysiwygStringConstant(Token* t, int tc) | |
| 1592 { | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1593 uint c; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1594 Loc start = loc; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1595 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1596 p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1597 stringbuffer.reset(); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1598 while (true) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1599 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1600 c = *p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1601 switch (c) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1602 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1603 case '\n': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1604 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1605 break; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1606 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1607 case '\r': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1608 if (*p == '\n') |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1609 continue; // ignore |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1610 c = '\n'; // treat EndOfLine as \n character |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1611 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1612 break; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1613 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1614 case 0: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1615 case 0x1A: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1616 error("unterminated string constant starting at %s", start.toChars()); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1617 t.ustring = "".ptr; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1618 t.len = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1619 t.postfix = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1620 return TOKstring; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1621 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1622 case '"': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1623 case '`': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1624 if (c == tc) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1625 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1626 t.len = stringbuffer.offset; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1627 stringbuffer.writeByte(0); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1628 char* tmp = cast(char*)GC.malloc(stringbuffer.offset); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1629 memcpy(tmp, stringbuffer.data, stringbuffer.offset); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1630 t.ustring = tmp; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1631 stringPostfix(t); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1632 return TOKstring; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1633 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1634 break; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1635 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1636 default: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1637 if (c & 0x80) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1638 { p--; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1639 uint u = decodeUTF(); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1640 p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1641 if (u == PS || u == LS) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1642 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1643 stringbuffer.writeUTF8(u); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1644 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1645 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1646 break; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1647 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1648 stringbuffer.writeByte(c); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1649 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1650 |
| 0 | 1651 assert(false); |
| 1652 } | |
| 1653 | |
| 51 | 1654 /************************************** |
| 1655 * Lex hex strings: | |
| 1656 * x"0A ae 34FE BD" | |
| 1657 */ | |
| 0 | 1658 TOK hexStringConstant(Token* t) |
| 1659 { | |
| 51 | 1660 uint c; |
| 1661 Loc start = loc; | |
| 1662 uint n = 0; | |
| 1663 uint v; | |
| 1664 | |
| 1665 p++; | |
| 1666 stringbuffer.reset(); | |
| 1667 while (1) | |
| 1668 { | |
| 1669 c = *p++; | |
| 1670 switch (c) | |
| 1671 { | |
| 1672 case ' ': | |
| 1673 case '\t': | |
| 1674 case '\v': | |
| 1675 case '\f': | |
| 1676 continue; // skip white space | |
| 1677 | |
| 1678 case '\r': | |
| 1679 if (*p == '\n') | |
| 1680 continue; // ignore | |
| 1681 // Treat isolated '\r' as if it were a '\n' | |
| 1682 case '\n': | |
| 1683 loc.linnum++; | |
| 1684 continue; | |
| 1685 | |
| 1686 case 0: | |
| 1687 case 0x1A: | |
| 1688 error("unterminated string constant starting at %s", start.toChars()); | |
| 1689 t.ustring = "".ptr; | |
| 1690 t.len = 0; | |
| 1691 t.postfix = 0; | |
| 1692 return TOKstring; | |
| 1693 | |
| 1694 case '"': | |
| 1695 if (n & 1) | |
| 1696 { | |
| 1697 error("odd number (%d) of hex characters in hex string", n); | |
| 1698 stringbuffer.writeByte(v); | |
| 1699 } | |
| 1700 t.len = stringbuffer.offset; | |
| 1701 stringbuffer.writeByte(0); | |
| 1702 void* mem = malloc(stringbuffer.offset); | |
| 1703 memcpy(mem, stringbuffer.data, stringbuffer.offset); | |
| 1704 t.ustring = cast(const(char)*)mem; | |
| 1705 stringPostfix(t); | |
| 1706 return TOKstring; | |
| 1707 | |
| 1708 default: | |
| 1709 if (c >= '0' && c <= '9') | |
| 1710 c -= '0'; | |
| 1711 else if (c >= 'a' && c <= 'f') | |
| 1712 c -= 'a' - 10; | |
| 1713 else if (c >= 'A' && c <= 'F') | |
| 1714 c -= 'A' - 10; | |
| 1715 else if (c & 0x80) | |
| 1716 { p--; | |
| 1717 uint u = decodeUTF(); | |
| 1718 p++; | |
| 1719 if (u == PS || u == LS) | |
| 1720 loc.linnum++; | |
| 1721 else | |
| 1722 error("non-hex character \\u%x", u); | |
| 1723 } | |
| 1724 else | |
| 1725 error("non-hex character '%c'", c); | |
| 1726 if (n & 1) | |
| 1727 { v = (v << 4) | c; | |
| 1728 stringbuffer.writeByte(v); | |
| 1729 } | |
| 1730 else | |
| 1731 v = c; | |
| 1732 n++; | |
| 1733 break; | |
| 1734 } | |
| 1735 } | |
| 0 | 1736 } |
| 1737 | |
| 1738 version (DMDV2) { | |
| 51 | 1739 /************************************** |
| 1740 * Lex delimited strings: | |
| 1741 * q"(foo(xxx))" // "foo(xxx)" | |
| 1742 * q"[foo(]" // "foo(" | |
| 1743 * q"/foo]/" // "foo]" | |
| 1744 * q"HERE | |
| 1745 * foo | |
| 1746 * HERE" // "foo\n" | |
| 1747 * Input: | |
| 1748 * p is on the " | |
| 1749 */ | |
| 0 | 1750 TOK delimitedStringConstant(Token* t) |
| 1751 { | |
| 51 | 1752 uint c; |
| 1753 Loc start = loc; | |
| 1754 uint delimleft = 0; | |
| 1755 uint delimright = 0; | |
| 1756 uint nest = 1; | |
| 1757 uint nestcount; | |
| 1758 Identifier hereid = null; | |
| 1759 uint blankrol = 0; | |
| 1760 uint startline = 0; | |
| 1761 | |
| 1762 p++; | |
| 1763 stringbuffer.reset(); | |
| 1764 while (1) | |
| 1765 { | |
| 1766 c = *p++; | |
| 1767 //printf("c = '%c'\n", c); | |
| 1768 switch (c) | |
| 1769 { | |
| 1770 case '\n': | |
| 1771 Lnextline: | |
| 1772 loc.linnum++; | |
| 1773 startline = 1; | |
| 1774 if (blankrol) | |
| 1775 { blankrol = 0; | |
| 1776 continue; | |
| 1777 } | |
| 1778 if (hereid) | |
| 1779 { | |
| 1780 stringbuffer.writeUTF8(c); | |
| 1781 continue; | |
| 1782 } | |
| 1783 break; | |
| 1784 | |
| 1785 case '\r': | |
| 1786 if (*p == '\n') | |
| 1787 continue; // ignore | |
| 1788 c = '\n'; // treat EndOfLine as \n character | |
| 1789 goto Lnextline; | |
| 1790 | |
| 1791 case 0: | |
| 1792 case 0x1A: | |
| 1793 goto Lerror; | |
| 1794 | |
| 1795 default: | |
| 1796 if (c & 0x80) | |
| 1797 { p--; | |
| 1798 c = decodeUTF(); | |
| 1799 p++; | |
| 1800 if (c == PS || c == LS) | |
| 1801 goto Lnextline; | |
| 1802 } | |
| 1803 break; | |
| 1804 } | |
| 1805 if (delimleft == 0) | |
| 1806 { | |
| 1807 delimleft = c; | |
| 1808 nest = 1; | |
| 1809 nestcount = 1; | |
| 1810 if (c == '(') | |
| 1811 delimright = ')'; | |
| 1812 else if (c == '{') | |
| 1813 delimright = '}'; | |
| 1814 else if (c == '[') | |
| 1815 delimright = ']'; | |
| 1816 else if (c == '<') | |
| 1817 delimright = '>'; | |
| 1818 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 1819 { | |
| 1820 // Start of identifier; must be a heredoc | |
| 1821 Token t2; | |
| 1822 p--; | |
| 1823 scan(&t2); // read in heredoc identifier | |
| 1824 if (t2.value != TOKidentifier) | |
| 1825 { | |
| 1826 error("identifier expected for heredoc, not %s", t2.toChars()); | |
| 1827 delimright = c; | |
| 1828 } | |
| 1829 else | |
| 1830 { | |
| 1831 hereid = t2.ident; | |
| 1832 //printf("hereid = '%s'\n", hereid.toChars()); | |
| 1833 blankrol = 1; | |
| 1834 } | |
| 1835 nest = 0; | |
| 1836 } | |
| 1837 else | |
| 1838 { | |
| 1839 delimright = c; | |
| 1840 nest = 0; | |
| 1841 if (isspace(c)) | |
| 1842 error("delimiter cannot be whitespace"); | |
| 1843 } | |
| 1844 } | |
| 1845 else | |
| 1846 { | |
| 1847 if (blankrol) | |
| 1848 { | |
| 1849 error("heredoc rest of line should be blank"); | |
| 1850 blankrol = 0; | |
| 1851 continue; | |
| 1852 } | |
| 1853 if (nest == 1) | |
| 1854 { | |
| 1855 if (c == delimleft) | |
| 1856 nestcount++; | |
| 1857 else if (c == delimright) | |
| 1858 { nestcount--; | |
| 1859 if (nestcount == 0) | |
| 1860 goto Ldone; | |
| 1861 } | |
| 1862 } | |
| 1863 else if (c == delimright) | |
| 1864 goto Ldone; | |
| 1865 if (startline && isalpha(c) && hereid) | |
| 1866 { | |
| 1867 Token t2; | |
| 1868 ubyte* psave = p; | |
| 1869 p--; | |
| 1870 scan(&t2); // read in possible heredoc identifier | |
| 1871 //printf("endid = '%s'\n", t2.ident.toChars()); | |
| 1872 if (t2.value == TOKidentifier && t2.ident.equals(hereid)) | |
| 1873 { | |
| 1874 /* should check that rest of line is blank | |
| 1875 */ | |
| 1876 goto Ldone; | |
| 1877 } | |
| 1878 p = psave; | |
| 1879 } | |
| 1880 stringbuffer.writeUTF8(c); | |
| 1881 startline = 0; | |
| 1882 } | |
| 1883 } | |
| 1884 | |
| 1885 Ldone: | |
| 1886 if (*p == '"') | |
| 1887 p++; | |
| 1888 else | |
| 1889 error("delimited string must end in %c\"", delimright); | |
| 1890 t.len = stringbuffer.offset; | |
| 1891 stringbuffer.writeByte(0); | |
| 1892 void* mem = malloc(stringbuffer.offset); | |
| 1893 memcpy(mem, stringbuffer.data, stringbuffer.offset); | |
| 1894 t.ustring = cast(const(char)*)mem; | |
| 1895 stringPostfix(t); | |
| 1896 return TOKstring; | |
| 1897 | |
| 1898 Lerror: | |
| 1899 error("unterminated string constant starting at %s", start.toChars()); | |
| 1900 t.ustring = "".ptr; | |
| 1901 t.len = 0; | |
| 1902 t.postfix = 0; | |
| 1903 return TOKstring; | |
| 0 | 1904 } |
| 1905 | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1906 /************************************** |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1907 * Lex delimited strings: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1908 * q{ foo(xxx) } // " foo(xxx) " |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1909 * q{foo(} // "foo(" |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1910 * q{{foo}"}"} // "{foo}"}"" |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1911 * Input: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1912 * p is on the q |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1913 */ |
| 0 | 1914 TOK tokenStringConstant(Token* t) |
| 1915 { | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1916 uint nest = 1; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1917 Loc start = loc; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1918 ubyte* pstart = ++p; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1919 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1920 while (true) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1921 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1922 Token tok; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1923 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1924 scan(&tok); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1925 switch (tok.value) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1926 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1927 case TOKlcurly: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1928 nest++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1929 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1930 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1931 case TOKrcurly: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1932 if (--nest == 0) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1933 goto Ldone; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1934 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1935 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1936 case TOKeof: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1937 goto Lerror; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1938 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1939 default: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1940 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1941 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1942 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1943 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1944 Ldone: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1945 t.len = p - 1 - pstart; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1946 char* tmp = cast(char*)GC.malloc(t.len + 1); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1947 memcpy(tmp, pstart, t.len); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1948 tmp[t.len] = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1949 t.ustring = tmp; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1950 stringPostfix(t); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1951 return TOKstring; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1952 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1953 Lerror: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1954 error("unterminated token string constant starting at %s", start.toChars()); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1955 t.ustring = "".ptr; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1956 t.len = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1957 t.postfix = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1958 return TOKstring; |
| 0 | 1959 } |
| 1960 } | |
| 1961 TOK escapeStringConstant(Token* t, int wide) | |
| 1962 { | |
| 1963 uint c; | |
| 1964 Loc start = loc; | |
| 1965 | |
| 1966 p++; | |
| 1967 stringbuffer.reset(); | |
| 1968 while (true) | |
| 1969 { | |
| 1970 c = *p++; | |
| 1971 switch (c) | |
| 1972 { | |
| 1973 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 1974 } else { | |
| 1975 case '\\': | |
| 1976 switch (*p) | |
| 1977 { | |
| 1978 case 'u': | |
| 1979 case 'U': | |
| 1980 case '&': | |
| 1981 c = escapeSequence(); | |
| 1982 stringbuffer.writeUTF8(c); | |
| 1983 continue; | |
| 1984 | |
| 1985 default: | |
| 1986 c = escapeSequence(); | |
| 1987 break; | |
| 1988 } | |
| 1989 break; | |
| 1990 } | |
| 1991 case '\n': | |
| 1992 loc.linnum++; | |
| 1993 break; | |
| 1994 | |
| 1995 case '\r': | |
| 1996 if (*p == '\n') | |
| 1997 continue; // ignore | |
| 1998 c = '\n'; // treat EndOfLine as \n character | |
| 1999 loc.linnum++; | |
| 2000 break; | |
| 2001 | |
| 2002 case '"': | |
| 2003 t.len = stringbuffer.offset; | |
| 2004 stringbuffer.writeByte(0); | |
| 2 | 2005 char* tmp = cast(char*)GC.malloc(stringbuffer.offset); |
| 0 | 2006 memcpy(tmp, stringbuffer.data, stringbuffer.offset); |
| 2007 t.ustring = tmp; | |
| 2008 stringPostfix(t); | |
| 2009 return TOK.TOKstring; | |
| 2010 | |
| 2011 case 0: | |
| 2012 case 0x1A: | |
| 2013 p--; | |
| 2014 error("unterminated string constant starting at %s", start.toChars()); | |
| 2015 t.ustring = "".ptr; | |
| 2016 t.len = 0; | |
| 2017 t.postfix = 0; | |
| 2018 return TOK.TOKstring; | |
| 2019 | |
| 2020 default: | |
| 2021 if (c & 0x80) | |
| 2022 { | |
| 2023 p--; | |
| 2024 c = decodeUTF(); | |
| 2025 if (c == LS || c == PS) | |
| 2026 { c = '\n'; | |
| 2027 loc.linnum++; | |
| 2028 } | |
| 2029 p++; | |
| 2030 stringbuffer.writeUTF8(c); | |
| 2031 continue; | |
| 2032 } | |
| 2033 break; | |
| 2034 } | |
| 2035 stringbuffer.writeByte(c); | |
| 2036 } | |
| 2037 | |
| 2038 assert(false); | |
| 2039 } | |
| 2040 | |
| 2041 TOK charConstant(Token* t, int wide) | |
| 2042 { | |
| 2043 uint c; | |
| 2044 TOK tk = TOKcharv; | |
| 2045 | |
| 2046 //printf("Lexer.charConstant\n"); | |
| 2047 p++; | |
| 2048 c = *p++; | |
| 2049 switch (c) | |
| 2050 { | |
| 2051 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 2052 } else { | |
| 2053 case '\\': | |
| 2054 switch (*p) | |
| 2055 { | |
| 2056 case 'u': | |
| 2057 t.uns64value = escapeSequence(); | |
| 2058 tk = TOKwcharv; | |
| 2059 break; | |
| 2060 | |
| 2061 case 'U': | |
| 2062 case '&': | |
| 2063 t.uns64value = escapeSequence(); | |
| 2064 tk = TOKdcharv; | |
| 2065 break; | |
| 2066 | |
| 2067 default: | |
| 2068 t.uns64value = escapeSequence(); | |
| 2069 break; | |
| 2070 } | |
| 2071 break; | |
| 2072 } | |
| 2073 case '\n': | |
| 2074 L1: | |
| 2075 loc.linnum++; | |
| 2076 case '\r': | |
| 2077 case 0: | |
| 2078 case 0x1A: | |
| 2079 case '\'': | |
| 2080 error("unterminated character constant"); | |
| 2081 return tk; | |
| 2082 | |
| 2083 default: | |
| 2084 if (c & 0x80) | |
| 2085 { | |
| 2086 p--; | |
| 2087 c = decodeUTF(); | |
| 2088 p++; | |
| 2089 if (c == LS || c == PS) | |
| 2090 goto L1; | |
| 2091 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 2092 tk = TOKwcharv; | |
| 2093 else | |
| 2094 tk = TOKdcharv; | |
| 2095 } | |
| 2096 t.uns64value = c; | |
| 2097 break; | |
| 2098 } | |
| 2099 | |
| 2100 if (*p != '\'') | |
| 2101 { | |
| 2102 error("unterminated character constant"); | |
| 2103 return tk; | |
| 2104 } | |
| 2105 p++; | |
| 2106 return tk; | |
| 2107 } | |
| 2108 | |
| 2109 /*************************************** | |
| 2110 * Get postfix of string literal. | |
| 2111 */ | |
| 2112 void stringPostfix(Token* t) | |
| 2113 { | |
| 2114 switch (*p) | |
| 2115 { | |
| 2116 case 'c': | |
| 2117 case 'w': | |
| 2118 case 'd': | |
| 2119 t.postfix = *p; | |
| 2120 p++; | |
| 2121 break; | |
| 2122 | |
| 2123 default: | |
| 2124 t.postfix = 0; | |
| 2125 break; | |
| 2126 } | |
| 2127 } | |
| 2128 | |
| 2129 uint wchar_(uint u) | |
| 2130 { | |
| 2131 assert(false); | |
| 2132 } | |
| 2133 | |
| 2134 /************************************** | |
| 2135 * Read in a number. | |
| 2136 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 2137 * integers can be decimal, octal or hex | |
| 2138 * Handle the suffixes U, UL, LU, L, etc. | |
| 2139 * If it's double, store it in tok.TKutok.Vdouble. | |
| 2140 * Returns: | |
| 2141 * TKnum | |
| 2142 * TKdouble,... | |
| 2143 */ | |
| 2144 | |
| 2145 TOK number(Token* t) | |
| 2146 { | |
| 2147 // We use a state machine to collect numbers | |
| 2148 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 2149 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 2150 STATE_hexh, STATE_error }; | |
| 2151 STATE state; | |
| 2152 | |
| 2153 enum FLAGS | |
| 2154 { | |
| 2155 FLAGS_undefined = 0, | |
| 2156 FLAGS_decimal = 1, // decimal | |
| 2157 FLAGS_unsigned = 2, // u or U suffix | |
| 2158 FLAGS_long = 4, // l or L suffix | |
| 2159 }; | |
| 2160 | |
| 2161 FLAGS flags = FLAGS.FLAGS_decimal; | |
| 2162 | |
| 2163 int i; | |
| 2164 int base; | |
| 2165 uint c; | |
| 2166 ubyte *start; | |
| 2167 TOK result; | |
| 2168 | |
| 2169 //printf("Lexer.number()\n"); | |
| 2170 state = STATE.STATE_initial; | |
| 2171 base = 0; | |
| 2172 stringbuffer.reset(); | |
| 2173 start = p; | |
| 2174 while (1) | |
| 2175 { | |
| 2176 c = *p; | |
| 2177 switch (state) | |
| 2178 { | |
| 2179 case STATE.STATE_initial: // opening state | |
| 2180 if (c == '0') | |
| 2181 state = STATE.STATE_0; | |
| 2182 else | |
| 2183 state = STATE.STATE_decimal; | |
| 2184 break; | |
| 2185 | |
| 2186 case STATE.STATE_0: | |
| 2187 flags = (flags & ~FLAGS.FLAGS_decimal); | |
| 2188 switch (c) | |
| 2189 { | |
| 2190 version (ZEROH) { | |
| 2191 case 'H': // 0h | |
| 2192 case 'h': | |
| 2193 goto hexh; | |
| 2194 } | |
| 2195 case 'X': | |
| 2196 case 'x': | |
| 2197 state = STATE.STATE_hex0; | |
| 2198 break; | |
| 2199 | |
| 2200 case '.': | |
| 2201 if (p[1] == '.') // .. is a separate token | |
| 2202 goto done; | |
| 2203 case 'i': | |
| 2204 case 'f': | |
| 2205 case 'F': | |
| 2206 goto real_; | |
| 2207 version (ZEROH) { | |
| 2208 case 'E': | |
| 2209 case 'e': | |
| 2210 goto case_hex; | |
| 2211 } | |
| 2212 case 'B': | |
| 2213 case 'b': | |
| 2214 state = STATE.STATE_binary0; | |
| 2215 break; | |
| 2216 | |
| 2217 case '0': case '1': case '2': case '3': | |
| 2218 case '4': case '5': case '6': case '7': | |
| 2219 state = STATE.STATE_octal; | |
| 2220 break; | |
| 2221 | |
| 2222 version (ZEROH) { | |
| 2223 case '8': case '9': case 'A': | |
| 2224 case 'C': case 'D': case 'F': | |
| 2225 case 'a': case 'c': case 'd': case 'f': | |
| 2226 case_hex: | |
| 2227 state = STATE.STATE_hexh; | |
| 2228 break; | |
| 2229 } | |
| 2230 case '_': | |
| 2231 state = STATE.STATE_octal; | |
| 2232 p++; | |
| 2233 continue; | |
| 2234 | |
| 2235 case 'L': | |
| 2236 if (p[1] == 'i') | |
| 2237 goto real_; | |
| 2238 goto done; | |
| 2239 | |
| 2240 default: | |
| 2241 goto done; | |
| 2242 } | |
| 2243 break; | |
| 2244 | |
| 2245 case STATE.STATE_decimal: // reading decimal number | |
| 2246 if (!isdigit(c)) | |
| 2247 { | |
| 2248 version (ZEROH) { | |
| 2249 if (ishex(c) | |
| 2250 || c == 'H' || c == 'h' | |
| 2251 ) | |
| 2252 goto hexh; | |
| 2253 } | |
| 2254 if (c == '_') // ignore embedded _ | |
| 2255 { p++; | |
| 2256 continue; | |
| 2257 } | |
| 2258 if (c == '.' && p[1] != '.') | |
| 2259 goto real_; | |
| 2260 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2261 c == 'e' || c == 'E') | |
| 2262 { | |
| 2263 real_: // It's a real number. Back up and rescan as a real | |
| 2264 p = start; | |
| 2265 return inreal(t); | |
| 2266 } | |
| 2267 else if (c == 'L' && p[1] == 'i') | |
| 2268 goto real_; | |
| 2269 goto done; | |
| 2270 } | |
| 2271 break; | |
| 2272 | |
| 2273 case STATE.STATE_hex0: // reading hex number | |
| 2274 case STATE.STATE_hex: | |
| 2275 if (! ishex(cast(ubyte)c)) | |
| 2276 { | |
| 2277 if (c == '_') // ignore embedded _ | |
| 2278 { p++; | |
| 2279 continue; | |
| 2280 } | |
| 2281 if (c == '.' && p[1] != '.') | |
| 2282 goto real_; | |
| 2283 if (c == 'P' || c == 'p' || c == 'i') | |
| 2284 goto real_; | |
| 2285 if (state == STATE.STATE_hex0) | |
| 2286 error("Hex digit expected, not '%c'", c); | |
| 2287 goto done; | |
| 2288 } | |
| 2289 state = STATE.STATE_hex; | |
| 2290 break; | |
| 2291 | |
| 2292 version (ZEROH) { | |
| 2293 hexh: | |
| 2294 state = STATE.STATE_hexh; | |
| 2295 case STATE.STATE_hexh: // parse numbers like 0FFh | |
| 2296 if (!ishex(c)) | |
| 2297 { | |
| 2298 if (c == 'H' || c == 'h') | |
| 2299 { | |
| 2300 p++; | |
| 2301 base = 16; | |
| 2302 goto done; | |
| 2303 } | |
| 2304 else | |
| 2305 { | |
| 2306 // Check for something like 1E3 or 0E24 | |
| 2307 if (memchr(cast(char*)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2308 memchr(cast(char*)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2309 goto real_; | |
| 2310 error("Hex digit expected, not '%c'", c); | |
| 2311 goto done; | |
| 2312 } | |
| 2313 } | |
| 2314 break; | |
| 2315 } | |
| 2316 | |
| 2317 case STATE.STATE_octal: // reading octal number | |
| 2318 case STATE.STATE_octale: // reading octal number with non-octal digits | |
| 2319 if (!isoctal(cast(ubyte)c)) | |
| 2320 { | |
| 2321 version (ZEROH) { | |
| 2322 if (ishex(c) | |
| 2323 || c == 'H' || c == 'h' | |
| 2324 ) | |
| 2325 goto hexh; | |
| 2326 } | |
| 2327 if (c == '_') // ignore embedded _ | |
| 2328 { p++; | |
| 2329 continue; | |
| 2330 } | |
| 2331 if (c == '.' && p[1] != '.') | |
| 2332 goto real_; | |
| 2333 if (c == 'i') | |
| 2334 goto real_; | |
| 2335 if (isdigit(c)) | |
| 2336 { | |
| 2337 state = STATE.STATE_octale; | |
| 2338 } | |
| 2339 else | |
| 2340 goto done; | |
| 2341 } | |
| 2342 break; | |
| 2343 | |
| 2344 case STATE.STATE_binary0: // starting binary number | |
| 2345 case STATE.STATE_binary: // reading binary number | |
| 2346 if (c != '0' && c != '1') | |
| 2347 { | |
| 2348 version (ZEROH) { | |
| 2349 if (ishex(c) | |
| 2350 || c == 'H' || c == 'h' | |
| 2351 ) | |
| 2352 goto hexh; | |
| 2353 } | |
| 2354 if (c == '_') // ignore embedded _ | |
| 2355 { p++; | |
| 2356 continue; | |
| 2357 } | |
| 2358 if (state == STATE.STATE_binary0) | |
| 2359 { error("binary digit expected"); | |
| 2360 state = STATE.STATE_error; | |
| 2361 break; | |
| 2362 } | |
| 2363 else | |
| 2364 goto done; | |
| 2365 } | |
| 2366 state = STATE.STATE_binary; | |
| 2367 break; | |
| 2368 | |
| 2369 case STATE.STATE_error: // for error recovery | |
| 2370 if (!isdigit(c)) // scan until non-digit | |
| 2371 goto done; | |
| 2372 break; | |
| 2373 | |
| 2374 default: | |
| 2375 assert(0); | |
| 2376 } | |
| 2377 stringbuffer.writeByte(c); | |
| 2378 p++; | |
| 2379 } | |
| 2380 done: | |
| 2381 stringbuffer.writeByte(0); // terminate string | |
| 2382 if (state == STATE.STATE_octale) | |
| 2383 error("Octal digit expected"); | |
| 2384 | |
| 2385 ulong n; // unsigned >=64 bit integer type | |
| 2386 | |
| 2387 if (stringbuffer.offset == 2 && (state == STATE.STATE_decimal || state == STATE.STATE_0)) | |
| 2388 n = stringbuffer.data[0] - '0'; | |
| 2389 else | |
| 2390 { | |
| 2391 // Convert string to integer | |
| 2392 version (__DMC__) { | |
| 2393 errno = 0; | |
| 2394 n = strtoull(cast(char*)stringbuffer.data,null,base); | |
| 2395 if (errno == ERANGE) | |
| 2396 error("integer overflow"); | |
| 2397 } else { | |
| 2398 // Not everybody implements strtoull() | |
| 2399 char* p = cast(char*)stringbuffer.data; | |
| 2400 int r = 10, d; | |
| 2401 | |
| 2402 if (*p == '0') | |
| 2403 { | |
| 2404 if (p[1] == 'x' || p[1] == 'X') | |
| 2405 p += 2, r = 16; | |
| 2406 else if (p[1] == 'b' || p[1] == 'B') | |
| 2407 p += 2, r = 2; | |
| 2408 else if (isdigit(p[1])) | |
| 2409 p += 1, r = 8; | |
| 2410 } | |
| 2411 | |
| 2412 n = 0; | |
| 2413 while (1) | |
| 2414 { | |
| 2415 if (*p >= '0' && *p <= '9') | |
| 2416 d = *p - '0'; | |
| 2417 else if (*p >= 'a' && *p <= 'z') | |
| 2418 d = *p - 'a' + 10; | |
| 2419 else if (*p >= 'A' && *p <= 'Z') | |
| 2420 d = *p - 'A' + 10; | |
| 2421 else | |
| 2422 break; | |
| 2423 if (d >= r) | |
| 2424 break; | |
| 2425 ulong n2 = n * r; | |
| 2426 //printf("n2 / r = %llx, n = %llx\n", n2/r, n); | |
| 2427 if (n2 / r != n || n2 + d < n) | |
| 2428 { | |
| 2429 error ("integer overflow"); | |
| 2430 break; | |
| 2431 } | |
| 2432 | |
| 2433 n = n2 + d; | |
| 2434 p++; | |
| 2435 } | |
| 2436 } | |
| 2437 if (n.sizeof > 8 && | |
| 2438 n > 0xFFFFFFFFFFFFFFFF) // if n needs more than 64 bits | |
| 2439 error("integer overflow"); | |
| 2440 } | |
| 2441 | |
| 2442 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2443 while (1) | |
| 2444 { FLAGS f; | |
| 2445 | |
| 2446 switch (*p) | |
| 2447 { case 'U': | |
| 2448 case 'u': | |
| 2449 f = FLAGS.FLAGS_unsigned; | |
| 2450 goto L1; | |
| 2451 | |
| 2452 case 'l': | |
| 2453 if (1 || !global.params.useDeprecated) | |
| 2454 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2455 case 'L': | |
| 2456 f = FLAGS.FLAGS_long; | |
| 2457 L1: | |
| 2458 p++; | |
| 2459 if (flags & f) | |
| 2460 error("unrecognized token"); | |
| 2461 flags = (flags | f); | |
| 2462 continue; | |
| 2463 default: | |
| 2464 break; | |
| 2465 } | |
| 2466 break; | |
| 2467 } | |
| 2468 | |
| 2469 switch (flags) | |
| 2470 { | |
| 2471 case FLAGS.FLAGS_undefined: | |
| 2472 /* Octal or Hexadecimal constant. | |
| 2473 * First that fits: int, uint, long, ulong | |
| 2474 */ | |
| 2475 if (n & 0x8000000000000000) | |
| 2476 result = TOK.TOKuns64v; | |
| 2477 else if (n & 0xFFFFFFFF00000000) | |
| 2478 result = TOK.TOKint64v; | |
| 2479 else if (n & 0x80000000) | |
| 2480 result = TOK.TOKuns32v; | |
| 2481 else | |
| 2482 result = TOK.TOKint32v; | |
| 2483 break; | |
| 2484 | |
| 2485 case FLAGS.FLAGS_decimal: | |
| 2486 /* First that fits: int, long, long long | |
| 2487 */ | |
| 2488 if (n & 0x8000000000000000) | |
| 2489 { error("signed integer overflow"); | |
| 2490 result = TOK.TOKuns64v; | |
| 2491 } | |
| 2492 else if (n & 0xFFFFFFFF80000000) | |
| 2493 result = TOK.TOKint64v; | |
| 2494 else | |
| 2495 result = TOK.TOKint32v; | |
| 2496 break; | |
| 2497 | |
| 2498 case FLAGS.FLAGS_unsigned: | |
| 2499 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_unsigned: | |
| 2500 /* First that fits: uint, ulong | |
| 2501 */ | |
| 2502 if (n & 0xFFFFFFFF00000000) | |
| 2503 result = TOK.TOKuns64v; | |
| 2504 else | |
| 2505 result = TOK.TOKuns32v; | |
| 2506 break; | |
| 2507 | |
| 2508 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_long: | |
| 2509 if (n & 0x8000000000000000) | |
| 2510 { error("signed integer overflow"); | |
| 2511 result = TOK.TOKuns64v; | |
| 2512 } | |
| 2513 else | |
| 2514 result = TOK.TOKint64v; | |
| 2515 break; | |
| 2516 | |
| 2517 case FLAGS.FLAGS_long: | |
| 2518 if (n & 0x8000000000000000) | |
| 2519 result = TOK.TOKuns64v; | |
| 2520 else | |
| 2521 result = TOK.TOKint64v; | |
| 2522 break; | |
| 2523 | |
| 2524 case FLAGS.FLAGS_unsigned | FLAGS.FLAGS_long: | |
| 2525 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_unsigned | FLAGS.FLAGS_long: | |
| 2526 result = TOK.TOKuns64v; | |
| 2527 break; | |
| 2528 | |
| 2529 default: | |
| 2530 debug { | |
| 2531 printf("%x\n",flags); | |
| 2532 } | |
| 2533 assert(0); | |
| 2534 } | |
| 2535 t.uns64value = n; | |
| 2536 return result; | |
| 2537 } | |
| 2538 | |
| 2539 /************************************** | |
| 2540 * Read in characters, converting them to real. | |
| 2541 * Bugs: | |
| 2542 * Exponent overflow not detected. | |
| 2543 * Too much requested precision is not detected. | |
| 2544 */ | |
| 2545 TOK inreal(Token* t) | |
| 2546 in | |
| 2547 { | |
| 2548 assert(*p == '.' || isdigit(*p)); | |
| 2549 } | |
| 2550 out (result) | |
| 2551 { | |
| 2552 switch (result) | |
| 2553 { | |
| 2554 case TOKfloat32v: | |
| 2555 case TOKfloat64v: | |
| 2556 case TOKfloat80v: | |
| 2557 case TOKimaginary32v: | |
| 2558 case TOKimaginary64v: | |
| 2559 case TOKimaginary80v: | |
| 2560 break; | |
| 2561 | |
| 2562 default: | |
| 2563 assert(0); | |
| 2564 } | |
| 2565 } | |
| 2566 body | |
| 2567 { | |
| 2568 int dblstate; | |
| 2569 uint c; | |
| 2570 char hex; // is this a hexadecimal-floating-constant? | |
| 2571 TOK result; | |
| 2572 | |
| 2573 //printf("Lexer.inreal()\n"); | |
| 2574 stringbuffer.reset(); | |
| 2575 dblstate = 0; | |
| 2576 hex = 0; | |
| 2577 Lnext: | |
| 2578 while (true) | |
| 2579 { | |
| 2580 // Get next char from input | |
| 2581 c = *p++; | |
| 2582 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2583 while (true) | |
| 2584 { | |
| 2585 switch (dblstate) | |
| 2586 { | |
| 2587 case 0: // opening state | |
| 2588 if (c == '0') | |
| 2589 dblstate = 9; | |
| 2590 else if (c == '.') | |
| 2591 dblstate = 3; | |
| 2592 else | |
| 2593 dblstate = 1; | |
| 2594 break; | |
| 2595 | |
| 2596 case 9: | |
| 2597 dblstate = 1; | |
| 2598 if (c == 'X' || c == 'x') | |
| 2599 { | |
| 2600 hex++; | |
| 2601 break; | |
| 2602 } | |
| 2603 case 1: // digits to left of . | |
| 2604 case 3: // digits to right of . | |
| 2605 case 7: // continuing exponent digits | |
| 2606 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2607 { | |
| 2608 if (c == '_') | |
| 2609 goto Lnext; // ignore embedded '_' | |
| 2610 dblstate++; | |
| 2611 continue; | |
| 2612 } | |
| 2613 break; | |
| 2614 | |
| 2615 case 2: // no more digits to left of . | |
| 2616 if (c == '.') | |
| 2617 { | |
| 2618 dblstate++; | |
| 2619 break; | |
| 2620 } | |
| 2621 case 4: // no more digits to right of . | |
| 2622 if ((c == 'E' || c == 'e') || | |
| 2623 hex && (c == 'P' || c == 'p')) | |
| 2624 { | |
| 2625 dblstate = 5; | |
| 2626 hex = 0; // exponent is always decimal | |
| 2627 break; | |
| 2628 } | |
| 2629 if (hex) | |
| 2630 error("binary-exponent-part required"); | |
| 2631 goto done; | |
| 2632 | |
| 2633 case 5: // looking immediately to right of E | |
| 2634 dblstate++; | |
| 2635 if (c == '-' || c == '+') | |
| 2636 break; | |
| 2637 case 6: // 1st exponent digit expected | |
| 2638 if (!isdigit(c)) | |
| 2639 error("exponent expected"); | |
| 2640 dblstate++; | |
| 2641 break; | |
| 2642 | |
| 2643 case 8: // past end of exponent digits | |
| 2644 goto done; | |
| 79 | 2645 |
| 2646 default: | |
| 2647 assert(0, "inreal.dblstate has unexpected value"); | |
| 0 | 2648 } |
| 2649 break; | |
| 2650 } | |
| 2651 stringbuffer.writeByte(c); | |
| 2652 } | |
| 2653 done: | |
| 2654 p--; | |
| 2655 | |
| 2656 stringbuffer.writeByte(0); | |
| 2657 | |
| 114 | 2658 version (Windows) { /// && __DMC__ |
| 0 | 2659 char* save = __locale_decpoint; |
| 2660 __locale_decpoint = cast(char*)".".ptr; | |
| 2661 } | |
| 2662 t.float80value = strtold(cast(char*)stringbuffer.data, null); | |
| 2663 | |
| 2664 errno = 0; | |
| 2665 switch (*p) | |
| 2666 { | |
| 2667 case 'F': | |
| 2668 case 'f': | |
| 2669 strtof(cast(char*)stringbuffer.data, null); | |
| 2670 result = TOKfloat32v; | |
| 2671 p++; | |
| 2672 break; | |
| 2673 | |
| 2674 default: | |
| 2675 strtod(cast(char*)stringbuffer.data, null); | |
| 2676 result = TOKfloat64v; | |
| 2677 break; | |
| 2678 | |
| 2679 case 'l': | |
| 2680 if (!global.params.useDeprecated) | |
| 2681 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2682 case 'L': | |
| 2683 result = TOKfloat80v; | |
| 2684 p++; | |
| 2685 break; | |
| 2686 } | |
| 2687 if (*p == 'i' || *p == 'I') | |
| 2688 { | |
| 2689 if (!global.params.useDeprecated && *p == 'I') | |
| 2690 error("'I' suffix is deprecated, use 'i' instead"); | |
| 2691 p++; | |
| 2692 switch (result) | |
| 2693 { | |
| 2694 case TOKfloat32v: | |
| 2695 result = TOKimaginary32v; | |
| 2696 break; | |
| 2697 case TOKfloat64v: | |
| 2698 result = TOKimaginary64v; | |
| 2699 break; | |
| 2700 case TOKfloat80v: | |
| 2701 result = TOKimaginary80v; | |
| 2702 break; | |
| 2703 } | |
| 2704 } | |
| 2705 | |
| 114 | 2706 version (Windows) { ///&& __DMC__ |
| 0 | 2707 __locale_decpoint = save; |
| 2708 } | |
| 2709 if (errno == ERANGE) | |
| 2710 error("number is not representable"); | |
| 2711 | |
| 2712 return result; | |
| 2713 } | |
| 2714 | |
| 2715 void error(T...)(string format, T t) | |
| 2716 { | |
| 2717 error(this.loc, format, t); | |
| 2718 } | |
| 2719 | |
| 2720 void error(T...)(Loc loc, string format, T t) | |
| 2721 { | |
| 2722 if (mod && !global.gag) | |
| 2723 { | |
| 2724 string p = loc.toChars(); | |
| 2725 if (p.length != 0) | |
| 2726 writef("%s: ", p); | |
| 2727 | |
| 2728 writefln(format, t); | |
| 2729 | |
| 2730 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 2731 fatal(); | |
| 2732 } | |
| 2733 | |
| 2734 global.errors++; | |
| 2735 } | |
| 2736 | |
| 2737 void pragma_() | |
| 2738 { | |
| 2739 assert(false); | |
| 2740 } | |
| 2741 | |
| 49 | 2742 /******************************************** |
| 2743 * Decode UTF character. | |
| 2744 * Issue error messages for invalid sequences. | |
| 2745 * Return decoded character, advance p to last character in UTF sequence. | |
| 2746 */ | |
| 0 | 2747 uint decodeUTF() |
| 2748 { | |
| 49 | 2749 dchar u; |
| 2750 ubyte c; | |
| 2751 ubyte* s = p; | |
| 2752 size_t len; | |
| 2753 size_t idx; | |
| 2754 string msg; | |
| 2755 | |
| 2756 c = *s; | |
| 2757 assert(c & 0x80); | |
| 2758 | |
| 2759 // Check length of remaining string up to 6 UTF-8 characters | |
| 2760 for (len = 1; len < 6 && s[len]; len++) { | |
| 2761 ; | |
| 2762 } | |
| 2763 | |
| 2764 idx = 0; | |
| 2765 msg = utf_decodeChar(cast(string)s[0..len], &idx, &u); | |
| 2766 p += idx - 1; | |
| 2767 if (msg) | |
| 2768 { | |
| 2769 error("%s", msg); | |
| 2770 } | |
| 2771 return u; | |
| 0 | 2772 } |
| 2773 | |
| 2774 void getDocComment(Token* t, uint lineComment) | |
| 2775 { | |
| 2776 assert(false); | |
| 2777 } | |
| 2778 | |
| 2779 static bool isValidIdentifier(string p) | |
| 2780 { | |
| 2781 if (p.length == 0) { | |
| 2782 return false; | |
| 2783 } | |
| 2784 | |
| 2785 if (p[0] >= '0' && p[0] <= '9') { // beware of isdigit() on signed chars | |
| 2786 return false; | |
| 2787 } | |
| 2788 | |
| 2789 size_t idx = 0; | |
| 2790 while (idx < p.length) | |
| 2791 { | |
| 2792 dchar dc; | |
| 2793 | |
| 2794 if (utf_decodeChar(p, &idx, &dc) !is null) { | |
| 2795 return false; | |
| 2796 } | |
| 2797 | |
| 2798 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) { | |
| 2799 return false; | |
| 2800 } | |
| 2801 } | |
| 2802 | |
| 2803 return true; | |
| 2804 } | |
| 2805 | |
| 79 | 2806 /// TODO: use normal string append when GC works |
| 2807 static string combineComments(const(char)[] c1, const(char)[] c2) | |
| 0 | 2808 { |
| 79 | 2809 //writef("Lexer.combineComments('%s', '%s')\n", c1, c2); |
| 0 | 2810 |
| 79 | 2811 char[] c = cast(char[]) c2; |
| 2812 | |
| 2813 if (c1 !is null) | |
| 0 | 2814 { |
| 79 | 2815 c = cast(char[]) c1; |
| 2816 if (c2 !is null) | |
| 0 | 2817 { |
| 79 | 2818 c = cast(char[]) (GC.malloc(c1.length + 1 + c2.length)[0 .. c1.length + 1 + c2.length]); |
| 2819 size_t len1 = c1.length; | |
| 2820 c[0..len1] = c1[]; | |
| 2821 c[len1++] = '\n'; | |
| 2822 c[len1 .. len1 + c2.length] = c2[]; | |
| 0 | 2823 } |
| 2824 } | |
| 79 | 2825 |
| 2826 return cast(string)c; | |
| 0 | 2827 } |
| 2828 } |