Mercurial > projects > ddmd
annotate dmd/Lexer.d @ 163:fe932c1a9563
*.interpret functions implemenation
| author | korDen |
|---|---|
| date | Thu, 23 Sep 2010 13:55:20 +0400 |
| parents | 438eaa11eed4 |
| children | ceed63f310fb |
| 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 | |
| 162 | 38 bool isUniAlpha(uint u) |
| 0 | 39 { |
| 162 | 40 static ushort table[][2] = |
| 41 [ | |
| 42 [ 0x00AA, 0x00AA ], | |
| 43 [ 0x00B5, 0x00B5 ], | |
| 44 [ 0x00B7, 0x00B7 ], | |
| 45 [ 0x00BA, 0x00BA ], | |
| 46 [ 0x00C0, 0x00D6 ], | |
| 47 [ 0x00D8, 0x00F6 ], | |
| 48 [ 0x00F8, 0x01F5 ], | |
| 49 [ 0x01FA, 0x0217 ], | |
| 50 [ 0x0250, 0x02A8 ], | |
| 51 [ 0x02B0, 0x02B8 ], | |
| 52 [ 0x02BB, 0x02BB ], | |
| 53 [ 0x02BD, 0x02C1 ], | |
| 54 [ 0x02D0, 0x02D1 ], | |
| 55 [ 0x02E0, 0x02E4 ], | |
| 56 [ 0x037A, 0x037A ], | |
| 57 [ 0x0386, 0x0386 ], | |
| 58 [ 0x0388, 0x038A ], | |
| 59 [ 0x038C, 0x038C ], | |
| 60 [ 0x038E, 0x03A1 ], | |
| 61 [ 0x03A3, 0x03CE ], | |
| 62 [ 0x03D0, 0x03D6 ], | |
| 63 [ 0x03DA, 0x03DA ], | |
| 64 [ 0x03DC, 0x03DC ], | |
| 65 [ 0x03DE, 0x03DE ], | |
| 66 [ 0x03E0, 0x03E0 ], | |
| 67 [ 0x03E2, 0x03F3 ], | |
| 68 [ 0x0401, 0x040C ], | |
| 69 [ 0x040E, 0x044F ], | |
| 70 [ 0x0451, 0x045C ], | |
| 71 [ 0x045E, 0x0481 ], | |
| 72 [ 0x0490, 0x04C4 ], | |
| 73 [ 0x04C7, 0x04C8 ], | |
| 74 [ 0x04CB, 0x04CC ], | |
| 75 [ 0x04D0, 0x04EB ], | |
| 76 [ 0x04EE, 0x04F5 ], | |
| 77 [ 0x04F8, 0x04F9 ], | |
| 78 [ 0x0531, 0x0556 ], | |
| 79 [ 0x0559, 0x0559 ], | |
| 80 [ 0x0561, 0x0587 ], | |
| 81 [ 0x05B0, 0x05B9 ], | |
| 82 [ 0x05BB, 0x05BD ], | |
| 83 [ 0x05BF, 0x05BF ], | |
| 84 [ 0x05C1, 0x05C2 ], | |
| 85 [ 0x05D0, 0x05EA ], | |
| 86 [ 0x05F0, 0x05F2 ], | |
| 87 [ 0x0621, 0x063A ], | |
| 88 [ 0x0640, 0x0652 ], | |
| 89 [ 0x0660, 0x0669 ], | |
| 90 [ 0x0670, 0x06B7 ], | |
| 91 [ 0x06BA, 0x06BE ], | |
| 92 [ 0x06C0, 0x06CE ], | |
| 93 [ 0x06D0, 0x06DC ], | |
| 94 [ 0x06E5, 0x06E8 ], | |
| 95 [ 0x06EA, 0x06ED ], | |
| 96 [ 0x06F0, 0x06F9 ], | |
| 97 [ 0x0901, 0x0903 ], | |
| 98 [ 0x0905, 0x0939 ], | |
| 99 [ 0x093D, 0x093D ], | |
| 100 [ 0x093E, 0x094D ], | |
| 101 [ 0x0950, 0x0952 ], | |
| 102 [ 0x0958, 0x0963 ], | |
| 103 [ 0x0966, 0x096F ], | |
| 104 [ 0x0981, 0x0983 ], | |
| 105 [ 0x0985, 0x098C ], | |
| 106 [ 0x098F, 0x0990 ], | |
| 107 [ 0x0993, 0x09A8 ], | |
| 108 [ 0x09AA, 0x09B0 ], | |
| 109 [ 0x09B2, 0x09B2 ], | |
| 110 [ 0x09B6, 0x09B9 ], | |
| 111 [ 0x09BE, 0x09C4 ], | |
| 112 [ 0x09C7, 0x09C8 ], | |
| 113 [ 0x09CB, 0x09CD ], | |
| 114 [ 0x09DC, 0x09DD ], | |
| 115 [ 0x09DF, 0x09E3 ], | |
| 116 [ 0x09E6, 0x09EF ], | |
| 117 [ 0x09F0, 0x09F1 ], | |
| 118 [ 0x0A02, 0x0A02 ], | |
| 119 [ 0x0A05, 0x0A0A ], | |
| 120 [ 0x0A0F, 0x0A10 ], | |
| 121 [ 0x0A13, 0x0A28 ], | |
| 122 [ 0x0A2A, 0x0A30 ], | |
| 123 [ 0x0A32, 0x0A33 ], | |
| 124 [ 0x0A35, 0x0A36 ], | |
| 125 [ 0x0A38, 0x0A39 ], | |
| 126 [ 0x0A3E, 0x0A42 ], | |
| 127 [ 0x0A47, 0x0A48 ], | |
| 128 [ 0x0A4B, 0x0A4D ], | |
| 129 [ 0x0A59, 0x0A5C ], | |
| 130 [ 0x0A5E, 0x0A5E ], | |
| 131 [ 0x0A66, 0x0A6F ], | |
| 132 [ 0x0A74, 0x0A74 ], | |
| 133 [ 0x0A81, 0x0A83 ], | |
| 134 [ 0x0A85, 0x0A8B ], | |
| 135 [ 0x0A8D, 0x0A8D ], | |
| 136 [ 0x0A8F, 0x0A91 ], | |
| 137 [ 0x0A93, 0x0AA8 ], | |
| 138 [ 0x0AAA, 0x0AB0 ], | |
| 139 [ 0x0AB2, 0x0AB3 ], | |
| 140 [ 0x0AB5, 0x0AB9 ], | |
| 141 [ 0x0ABD, 0x0AC5 ], | |
| 142 [ 0x0AC7, 0x0AC9 ], | |
| 143 [ 0x0ACB, 0x0ACD ], | |
| 144 [ 0x0AD0, 0x0AD0 ], | |
| 145 [ 0x0AE0, 0x0AE0 ], | |
| 146 [ 0x0AE6, 0x0AEF ], | |
| 147 [ 0x0B01, 0x0B03 ], | |
| 148 [ 0x0B05, 0x0B0C ], | |
| 149 [ 0x0B0F, 0x0B10 ], | |
| 150 [ 0x0B13, 0x0B28 ], | |
| 151 [ 0x0B2A, 0x0B30 ], | |
| 152 [ 0x0B32, 0x0B33 ], | |
| 153 [ 0x0B36, 0x0B39 ], | |
| 154 [ 0x0B3D, 0x0B3D ], | |
| 155 [ 0x0B3E, 0x0B43 ], | |
| 156 [ 0x0B47, 0x0B48 ], | |
| 157 [ 0x0B4B, 0x0B4D ], | |
| 158 [ 0x0B5C, 0x0B5D ], | |
| 159 [ 0x0B5F, 0x0B61 ], | |
| 160 [ 0x0B66, 0x0B6F ], | |
| 161 [ 0x0B82, 0x0B83 ], | |
| 162 [ 0x0B85, 0x0B8A ], | |
| 163 [ 0x0B8E, 0x0B90 ], | |
| 164 [ 0x0B92, 0x0B95 ], | |
| 165 [ 0x0B99, 0x0B9A ], | |
| 166 [ 0x0B9C, 0x0B9C ], | |
| 167 [ 0x0B9E, 0x0B9F ], | |
| 168 [ 0x0BA3, 0x0BA4 ], | |
| 169 [ 0x0BA8, 0x0BAA ], | |
| 170 [ 0x0BAE, 0x0BB5 ], | |
| 171 [ 0x0BB7, 0x0BB9 ], | |
| 172 [ 0x0BBE, 0x0BC2 ], | |
| 173 [ 0x0BC6, 0x0BC8 ], | |
| 174 [ 0x0BCA, 0x0BCD ], | |
| 175 [ 0x0BE7, 0x0BEF ], | |
| 176 [ 0x0C01, 0x0C03 ], | |
| 177 [ 0x0C05, 0x0C0C ], | |
| 178 [ 0x0C0E, 0x0C10 ], | |
| 179 [ 0x0C12, 0x0C28 ], | |
| 180 [ 0x0C2A, 0x0C33 ], | |
| 181 [ 0x0C35, 0x0C39 ], | |
| 182 [ 0x0C3E, 0x0C44 ], | |
| 183 [ 0x0C46, 0x0C48 ], | |
| 184 [ 0x0C4A, 0x0C4D ], | |
| 185 [ 0x0C60, 0x0C61 ], | |
| 186 [ 0x0C66, 0x0C6F ], | |
| 187 [ 0x0C82, 0x0C83 ], | |
| 188 [ 0x0C85, 0x0C8C ], | |
| 189 [ 0x0C8E, 0x0C90 ], | |
| 190 [ 0x0C92, 0x0CA8 ], | |
| 191 [ 0x0CAA, 0x0CB3 ], | |
| 192 [ 0x0CB5, 0x0CB9 ], | |
| 193 [ 0x0CBE, 0x0CC4 ], | |
| 194 [ 0x0CC6, 0x0CC8 ], | |
| 195 [ 0x0CCA, 0x0CCD ], | |
| 196 [ 0x0CDE, 0x0CDE ], | |
| 197 [ 0x0CE0, 0x0CE1 ], | |
| 198 [ 0x0CE6, 0x0CEF ], | |
| 199 [ 0x0D02, 0x0D03 ], | |
| 200 [ 0x0D05, 0x0D0C ], | |
| 201 [ 0x0D0E, 0x0D10 ], | |
| 202 [ 0x0D12, 0x0D28 ], | |
| 203 [ 0x0D2A, 0x0D39 ], | |
| 204 [ 0x0D3E, 0x0D43 ], | |
| 205 [ 0x0D46, 0x0D48 ], | |
| 206 [ 0x0D4A, 0x0D4D ], | |
| 207 [ 0x0D60, 0x0D61 ], | |
| 208 [ 0x0D66, 0x0D6F ], | |
| 209 [ 0x0E01, 0x0E3A ], | |
| 210 [ 0x0E40, 0x0E5B ], | |
| 211 // { 0x0E50, 0x0E59 }, | |
| 212 [ 0x0E81, 0x0E82 ], | |
| 213 [ 0x0E84, 0x0E84 ], | |
| 214 [ 0x0E87, 0x0E88 ], | |
| 215 [ 0x0E8A, 0x0E8A ], | |
| 216 [ 0x0E8D, 0x0E8D ], | |
| 217 [ 0x0E94, 0x0E97 ], | |
| 218 [ 0x0E99, 0x0E9F ], | |
| 219 [ 0x0EA1, 0x0EA3 ], | |
| 220 [ 0x0EA5, 0x0EA5 ], | |
| 221 [ 0x0EA7, 0x0EA7 ], | |
| 222 [ 0x0EAA, 0x0EAB ], | |
| 223 [ 0x0EAD, 0x0EAE ], | |
| 224 [ 0x0EB0, 0x0EB9 ], | |
| 225 [ 0x0EBB, 0x0EBD ], | |
| 226 [ 0x0EC0, 0x0EC4 ], | |
| 227 [ 0x0EC6, 0x0EC6 ], | |
| 228 [ 0x0EC8, 0x0ECD ], | |
| 229 [ 0x0ED0, 0x0ED9 ], | |
| 230 [ 0x0EDC, 0x0EDD ], | |
| 231 [ 0x0F00, 0x0F00 ], | |
| 232 [ 0x0F18, 0x0F19 ], | |
| 233 [ 0x0F20, 0x0F33 ], | |
| 234 [ 0x0F35, 0x0F35 ], | |
| 235 [ 0x0F37, 0x0F37 ], | |
| 236 [ 0x0F39, 0x0F39 ], | |
| 237 [ 0x0F3E, 0x0F47 ], | |
| 238 [ 0x0F49, 0x0F69 ], | |
| 239 [ 0x0F71, 0x0F84 ], | |
| 240 [ 0x0F86, 0x0F8B ], | |
| 241 [ 0x0F90, 0x0F95 ], | |
| 242 [ 0x0F97, 0x0F97 ], | |
| 243 [ 0x0F99, 0x0FAD ], | |
| 244 [ 0x0FB1, 0x0FB7 ], | |
| 245 [ 0x0FB9, 0x0FB9 ], | |
| 246 [ 0x10A0, 0x10C5 ], | |
| 247 [ 0x10D0, 0x10F6 ], | |
| 248 [ 0x1E00, 0x1E9B ], | |
| 249 [ 0x1EA0, 0x1EF9 ], | |
| 250 [ 0x1F00, 0x1F15 ], | |
| 251 [ 0x1F18, 0x1F1D ], | |
| 252 [ 0x1F20, 0x1F45 ], | |
| 253 [ 0x1F48, 0x1F4D ], | |
| 254 [ 0x1F50, 0x1F57 ], | |
| 255 [ 0x1F59, 0x1F59 ], | |
| 256 [ 0x1F5B, 0x1F5B ], | |
| 257 [ 0x1F5D, 0x1F5D ], | |
| 258 [ 0x1F5F, 0x1F7D ], | |
| 259 [ 0x1F80, 0x1FB4 ], | |
| 260 [ 0x1FB6, 0x1FBC ], | |
| 261 [ 0x1FBE, 0x1FBE ], | |
| 262 [ 0x1FC2, 0x1FC4 ], | |
| 263 [ 0x1FC6, 0x1FCC ], | |
| 264 [ 0x1FD0, 0x1FD3 ], | |
| 265 [ 0x1FD6, 0x1FDB ], | |
| 266 [ 0x1FE0, 0x1FEC ], | |
| 267 [ 0x1FF2, 0x1FF4 ], | |
| 268 [ 0x1FF6, 0x1FFC ], | |
| 269 [ 0x203F, 0x2040 ], | |
| 270 [ 0x207F, 0x207F ], | |
| 271 [ 0x2102, 0x2102 ], | |
| 272 [ 0x2107, 0x2107 ], | |
| 273 [ 0x210A, 0x2113 ], | |
| 274 [ 0x2115, 0x2115 ], | |
| 275 [ 0x2118, 0x211D ], | |
| 276 [ 0x2124, 0x2124 ], | |
| 277 [ 0x2126, 0x2126 ], | |
| 278 [ 0x2128, 0x2128 ], | |
| 279 [ 0x212A, 0x2131 ], | |
| 280 [ 0x2133, 0x2138 ], | |
| 281 [ 0x2160, 0x2182 ], | |
| 282 [ 0x3005, 0x3007 ], | |
| 283 [ 0x3021, 0x3029 ], | |
| 284 [ 0x3041, 0x3093 ], | |
| 285 [ 0x309B, 0x309C ], | |
| 286 [ 0x30A1, 0x30F6 ], | |
| 287 [ 0x30FB, 0x30FC ], | |
| 288 [ 0x3105, 0x312C ], | |
| 289 [ 0x4E00, 0x9FA5 ], | |
| 290 [ 0xAC00, 0xD7A3 ], | |
| 291 ]; | |
| 292 | |
| 293 debug { | |
| 294 for (int i = 0; i < table.length; i++) | |
| 295 { | |
| 296 //printf("%x\n", table[i][0]); | |
| 297 assert(table[i][0] <= table[i][1]); | |
| 298 if (i < table.length - 1) | |
| 299 assert(table[i][1] < table[i + 1][0]); | |
| 300 } | |
| 301 } | |
| 302 | |
| 303 if (u > 0xD7A3) | |
| 304 goto Lisnot; | |
| 305 | |
| 306 // Binary search | |
| 307 int mid; | |
| 308 int low; | |
| 309 int high; | |
| 310 | |
| 311 low = 0; | |
| 312 high = table.length - 1; | |
| 313 while (low <= high) | |
| 314 { | |
| 315 mid = (low + high) >> 1; | |
| 316 if (u < table[mid][0]) | |
| 317 high = mid - 1; | |
| 318 else if (u > table[mid][1]) | |
| 319 low = mid + 1; | |
| 320 else | |
| 321 goto Lis; | |
| 322 } | |
| 323 | |
| 324 Lisnot: | |
| 325 debug { | |
| 326 for (int i = 0; i < table.length; i++) | |
| 327 { | |
| 328 assert(u < table[i][0] || u > table[i][1]); | |
| 329 } | |
| 330 } | |
| 331 return false; | |
| 332 | |
| 333 Lis: | |
| 334 debug { | |
| 335 for (int i = 0; i < table.length; i++) | |
| 336 { | |
| 337 if (u >= table[i][0] && u <= table[i][1]) | |
| 338 return 1; | |
| 339 } | |
| 340 assert(0); // should have been in table | |
| 341 } | |
| 342 return true; | |
| 0 | 343 } |
| 344 | |
| 345 class Lexer | |
| 346 { | |
| 347 static StringTable stringtable; | |
| 348 static OutBuffer stringbuffer; | |
| 349 static Token* freelist; | |
| 350 | |
| 351 Loc loc; // for error messages | |
| 352 | |
| 353 ubyte* base; // pointer to start of buffer | |
| 354 ubyte* end; // past end of buffer | |
| 355 ubyte* p; // current character | |
| 356 Token token; | |
| 357 Module mod; | |
| 358 int doDocComment; // collect doc comment information | |
| 359 int anyToken; // !=0 means seen at least one token | |
| 360 int commentToken; // !=0 means comments are TOKcomment's | |
| 361 | |
| 362 static this() | |
| 363 { | |
| 364 stringtable = new StringTable(); | |
| 365 stringbuffer = new OutBuffer(); | |
| 366 } | |
| 367 | |
| 368 static ~this() | |
| 369 { | |
| 34 | 370 //delete stringtable; |
| 0 | 371 } |
| 372 | |
| 373 this(Module mod, ubyte* base, uint begoffset, uint endoffset, int doDocComment, int commentToken) | |
| 374 { | |
| 375 loc = Loc(mod, 1); | |
| 376 | |
| 377 memset(&token,0,token.sizeof); | |
| 378 this.base = base; | |
| 379 this.end = base + endoffset; | |
| 380 p = base + begoffset; | |
| 381 this.mod = mod; | |
| 382 this.doDocComment = doDocComment; | |
| 383 this.anyToken = 0; | |
| 384 this.commentToken = commentToken; | |
| 385 //initKeywords(); | |
| 386 | |
| 387 /* If first line starts with '#!', ignore the line | |
| 388 */ | |
| 389 | |
| 390 if (p[0] == '#' && p[1] =='!') | |
| 391 { | |
| 392 p += 2; | |
| 393 while (1) | |
| 394 { | |
| 395 ubyte c = *p; | |
| 396 switch (c) | |
| 397 { | |
| 398 case '\n': | |
| 399 p++; | |
| 400 break; | |
| 401 | |
| 402 case '\r': | |
| 403 p++; | |
| 404 if (*p == '\n') | |
| 405 p++; | |
| 406 break; | |
| 407 | |
| 408 case 0: | |
| 409 case 0x1A: | |
| 410 break; | |
| 411 | |
| 412 default: | |
| 413 if (c & 0x80) | |
| 414 { | |
| 415 uint u = decodeUTF(); | |
| 416 if (u == PS || u == LS) | |
| 417 break; | |
| 418 } | |
| 419 p++; | |
| 420 continue; | |
| 421 } | |
| 422 break; | |
| 423 } | |
| 424 loc.linnum = 2; | |
| 425 } | |
| 426 } | |
| 427 | |
| 428 version (DMDV2) { | |
| 429 static Keyword[] keywords = | |
| 430 [ | |
| 431 // { "", TOK }, | |
| 432 | |
| 433 { "this", TOK.TOKthis }, | |
| 434 { "super", TOK.TOKsuper }, | |
| 435 { "assert", TOK.TOKassert }, | |
| 436 { "null", TOK.TOKnull }, | |
| 437 { "true", TOK.TOKtrue }, | |
| 438 { "false", TOK.TOKfalse }, | |
| 439 { "cast", TOK.TOKcast }, | |
| 440 { "new", TOK.TOKnew }, | |
| 441 { "delete", TOK.TOKdelete }, | |
| 442 { "throw", TOK.TOKthrow }, | |
| 443 { "module", TOK.TOKmodule }, | |
| 444 { "pragma", TOK.TOKpragma }, | |
| 445 { "typeof", TOK.TOKtypeof }, | |
| 446 { "typeid", TOK.TOKtypeid }, | |
| 447 | |
| 448 { "template", TOK.TOKtemplate }, | |
| 449 | |
| 450 { "void", TOK.TOKvoid }, | |
| 451 { "byte", TOK.TOKint8 }, | |
| 452 { "ubyte", TOK.TOKuns8 }, | |
| 453 { "short", TOK.TOKint16 }, | |
| 454 { "ushort", TOK.TOKuns16 }, | |
| 455 { "int", TOK.TOKint32 }, | |
| 456 { "uint", TOK.TOKuns32 }, | |
| 457 { "long", TOK.TOKint64 }, | |
| 458 { "ulong", TOK.TOKuns64 }, | |
| 459 { "cent", TOK.TOKcent, }, | |
| 460 { "ucent", TOK.TOKucent, }, | |
| 461 { "float", TOK.TOKfloat32 }, | |
| 462 { "double", TOK.TOKfloat64 }, | |
| 463 { "real", TOK.TOKfloat80 }, | |
| 464 | |
| 465 { "bool", TOK.TOKbool }, | |
| 466 { "char", TOK.TOKchar }, | |
| 467 { "wchar", TOK.TOKwchar }, | |
| 468 { "dchar", TOK.TOKdchar }, | |
| 469 | |
| 470 { "ifloat", TOK.TOKimaginary32 }, | |
| 471 { "idouble", TOK.TOKimaginary64 }, | |
| 472 { "ireal", TOK.TOKimaginary80 }, | |
| 473 | |
| 474 { "cfloat", TOK.TOKcomplex32 }, | |
| 475 { "cdouble", TOK.TOKcomplex64 }, | |
| 476 { "creal", TOK.TOKcomplex80 }, | |
| 477 | |
| 478 { "delegate", TOK.TOKdelegate }, | |
| 479 { "function", TOK.TOKfunction }, | |
| 480 | |
| 481 { "is", TOK.TOKis }, | |
| 482 { "if", TOK.TOKif }, | |
| 483 { "else", TOK.TOKelse }, | |
| 484 { "while", TOK.TOKwhile }, | |
| 485 { "for", TOK.TOKfor }, | |
| 486 { "do", TOK.TOKdo }, | |
| 487 { "switch", TOK.TOKswitch }, | |
| 488 { "case", TOK.TOKcase }, | |
| 489 { "default", TOK.TOKdefault }, | |
| 490 { "break", TOK.TOKbreak }, | |
| 491 { "continue", TOK.TOKcontinue }, | |
| 492 { "synchronized", TOK.TOKsynchronized }, | |
| 493 { "return", TOK.TOKreturn }, | |
| 494 { "goto", TOK.TOKgoto }, | |
| 495 { "try", TOK.TOKtry }, | |
| 496 { "catch", TOK.TOKcatch }, | |
| 497 { "finally", TOK.TOKfinally }, | |
| 498 { "with", TOK.TOKwith }, | |
| 499 { "asm", TOK.TOKasm }, | |
| 500 { "foreach", TOK.TOKforeach }, | |
| 501 { "foreach_reverse", TOK.TOKforeach_reverse }, | |
| 502 { "scope", TOK.TOKscope }, | |
| 503 | |
| 504 { "struct", TOK.TOKstruct }, | |
| 505 { "class", TOK.TOKclass }, | |
| 506 { "interface", TOK.TOKinterface }, | |
| 507 { "union", TOK.TOKunion }, | |
| 508 { "enum", TOK.TOKenum }, | |
| 509 { "import", TOK.TOKimport }, | |
| 510 { "mixin", TOK.TOKmixin }, | |
| 511 { "static", TOK.TOKstatic }, | |
| 512 { "final", TOK.TOKfinal }, | |
| 513 { "const", TOK.TOKconst }, | |
| 514 { "typedef", TOK.TOKtypedef }, | |
| 515 { "alias", TOK.TOKalias }, | |
| 516 { "override", TOK.TOKoverride }, | |
| 517 { "abstract", TOK.TOKabstract }, | |
| 518 { "volatile", TOK.TOKvolatile }, | |
| 519 { "debug", TOK.TOKdebug }, | |
| 520 { "deprecated", TOK.TOKdeprecated }, | |
| 521 { "in", TOK.TOKin }, | |
| 522 { "out", TOK.TOKout }, | |
| 523 { "inout", TOK.TOKinout }, | |
| 524 { "lazy", TOK.TOKlazy }, | |
| 525 { "auto", TOK.TOKauto }, | |
| 526 | |
| 527 { "align", TOK.TOKalign }, | |
| 528 { "extern", TOK.TOKextern }, | |
| 529 { "private", TOK.TOKprivate }, | |
| 530 { "package", TOK.TOKpackage }, | |
| 531 { "protected", TOK.TOKprotected }, | |
| 532 { "public", TOK.TOKpublic }, | |
| 533 { "export", TOK.TOKexport }, | |
| 534 | |
| 535 { "body", TOK.TOKbody }, | |
| 536 { "invariant", TOK.TOKinvariant }, | |
| 537 { "unittest", TOK.TOKunittest }, | |
| 538 { "version", TOK.TOKversion }, | |
| 539 //{ "manifest", TOK.TOKmanifest }, | |
| 540 | |
| 541 // Added after 1.0 | |
| 542 { "ref", TOK.TOKref }, | |
| 543 { "macro", TOK.TOKmacro }, | |
| 544 { "pure", TOK.TOKpure }, | |
| 545 { "nothrow", TOK.TOKnothrow }, | |
| 546 { "__thread", TOK.TOKtls }, | |
| 547 { "__gshared", TOK.TOKgshared }, | |
| 548 { "__traits", TOK.TOKtraits }, | |
| 549 { "__overloadset", TOK.TOKoverloadset }, | |
| 550 { "__FILE__", TOK.TOKfile }, | |
| 551 { "__LINE__", TOK.TOKline }, | |
| 552 { "shared", TOK.TOKshared }, | |
| 553 { "immutable", TOK.TOKimmutable }, | |
| 554 ]; | |
| 555 } else { | |
| 556 static Keyword[] keywords = | |
| 557 [ | |
| 558 // { "", TOK }, | |
| 559 | |
| 560 { "this", TOK.TOKthis }, | |
| 561 { "super", TOK.TOKsuper }, | |
| 562 { "assert", TOK.TOKassert }, | |
| 563 { "null", TOK.TOKnull }, | |
| 564 { "true", TOK.TOKtrue }, | |
| 565 { "false", TOK.TOKfalse }, | |
| 566 { "cast", TOK.TOKcast }, | |
| 567 { "new", TOK.TOKnew }, | |
| 568 { "delete", TOK.TOKdelete }, | |
| 569 { "throw", TOK.TOKthrow }, | |
| 570 { "module", TOK.TOKmodule }, | |
| 571 { "pragma", TOK.TOKpragma }, | |
| 572 { "typeof", TOK.TOKtypeof }, | |
| 573 { "typeid", TOK.TOKtypeid }, | |
| 574 | |
| 575 { "template", TOK.TOKtemplate }, | |
| 576 | |
| 577 { "void", TOK.TOKvoid }, | |
| 578 { "byte", TOK.TOKint8 }, | |
| 579 { "ubyte", TOK.TOKuns8 }, | |
| 580 { "short", TOK.TOKint16 }, | |
| 581 { "ushort", TOK.TOKuns16 }, | |
| 582 { "int", TOK.TOKint32 }, | |
| 583 { "uint", TOK.TOKuns32 }, | |
| 584 { "long", TOK.TOKint64 }, | |
| 585 { "ulong", TOK.TOKuns64 }, | |
| 586 { "cent", TOK.TOKcent, }, | |
| 587 { "ucent", TOK.TOKucent, }, | |
| 588 { "float", TOK.TOKfloat32 }, | |
| 589 { "double", TOK.TOKfloat64 }, | |
| 590 { "real", TOK.TOKfloat80 }, | |
| 591 | |
| 592 { "bool", TOK.TOKbool }, | |
| 593 { "char", TOK.TOKchar }, | |
| 594 { "wchar", TOK.TOKwchar }, | |
| 595 { "dchar", TOK.TOKdchar }, | |
| 596 | |
| 597 { "ifloat", TOK.TOKimaginary32 }, | |
| 598 { "idouble", TOK.TOKimaginary64 }, | |
| 599 { "ireal", TOK.TOKimaginary80 }, | |
| 600 | |
| 601 { "cfloat", TOK.TOKcomplex32 }, | |
| 602 { "cdouble", TOK.TOKcomplex64 }, | |
| 603 { "creal", TOK.TOKcomplex80 }, | |
| 604 | |
| 605 { "delegate", TOK.TOKdelegate }, | |
| 606 { "function", TOK.TOKfunction }, | |
| 607 | |
| 608 { "is", TOK.TOKis }, | |
| 609 { "if", TOK.TOKif }, | |
| 610 { "else", TOK.TOKelse }, | |
| 611 { "while", TOK.TOKwhile }, | |
| 612 { "for", TOK.TOKfor }, | |
| 613 { "do", TOK.TOKdo }, | |
| 614 { "switch", TOK.TOKswitch }, | |
| 615 { "case", TOK.TOKcase }, | |
| 616 { "default", TOK.TOKdefault }, | |
| 617 { "break", TOK.TOKbreak }, | |
| 618 { "continue", TOK.TOKcontinue }, | |
| 619 { "synchronized", TOK.TOKsynchronized }, | |
| 620 { "return", TOK.TOKreturn }, | |
| 621 { "goto", TOK.TOKgoto }, | |
| 622 { "try", TOK.TOKtry }, | |
| 623 { "catch", TOK.TOKcatch }, | |
| 624 { "finally", TOK.TOKfinally }, | |
| 625 { "with", TOK.TOKwith }, | |
| 626 { "asm", TOK.TOKasm }, | |
| 627 { "foreach", TOK.TOKforeach }, | |
| 628 { "foreach_reverse", TOK.TOKforeach_reverse }, | |
| 629 { "scope", TOK.TOKscope }, | |
| 630 | |
| 631 { "struct", TOK.TOKstruct }, | |
| 632 { "class", TOK.TOKclass }, | |
| 633 { "interface", TOK.TOKinterface }, | |
| 634 { "union", TOK.TOKunion }, | |
| 635 { "enum", TOK.TOKenum }, | |
| 636 { "import", TOK.TOKimport }, | |
| 637 { "mixin", TOK.TOKmixin }, | |
| 638 { "static", TOK.TOKstatic }, | |
| 639 { "final", TOK.TOKfinal }, | |
| 640 { "const", TOK.TOKconst }, | |
| 641 { "typedef", TOK.TOKtypedef }, | |
| 642 { "alias", TOK.TOKalias }, | |
| 643 { "override", TOK.TOKoverride }, | |
| 644 { "abstract", TOK.TOKabstract }, | |
| 645 { "volatile", TOK.TOKvolatile }, | |
| 646 { "debug", TOK.TOKdebug }, | |
| 647 { "deprecated", TOK.TOKdeprecated }, | |
| 648 { "in", TOK.TOKin }, | |
| 649 { "out", TOK.TOKout }, | |
| 650 { "inout", TOK.TOKinout }, | |
| 651 { "lazy", TOK.TOKlazy }, | |
| 652 { "auto", TOK.TOKauto }, | |
| 653 | |
| 654 { "align", TOK.TOKalign }, | |
| 655 { "extern", TOK.TOKextern }, | |
| 656 { "private", TOK.TOKprivate }, | |
| 657 { "package", TOK.TOKpackage }, | |
| 658 { "protected", TOK.TOKprotected }, | |
| 659 { "public", TOK.TOKpublic }, | |
| 660 { "export", TOK.TOKexport }, | |
| 661 | |
| 662 { "body", TOK.TOKbody }, | |
| 663 { "invariant", TOK.TOKinvariant }, | |
| 664 { "unittest", TOK.TOKunittest }, | |
| 665 { "version", TOK.TOKversion }, | |
| 666 //{ "manifest", TOK.TOKmanifest }, | |
| 667 | |
| 668 // Added after 1.0 | |
| 669 { "ref", TOK.TOKref }, | |
| 670 { "macro", TOK.TOKmacro }, | |
| 671 ]; | |
| 672 } | |
| 673 | |
| 674 static ubyte cmtable[256]; | |
| 675 enum CMoctal = 0x1; | |
| 676 enum CMhex = 0x2; | |
| 677 enum CMidchar = 0x4; | |
| 678 | |
| 679 ubyte isoctal (ubyte c) { return cmtable[c] & CMoctal; } | |
| 680 ubyte ishex (ubyte c) { return cmtable[c] & CMhex; } | |
| 681 ubyte isidchar(ubyte c) { return cmtable[c] & CMidchar; } | |
| 682 | |
| 683 static void cmtable_init() | |
| 684 { | |
| 685 for (uint c = 0; c < cmtable.length; c++) | |
| 686 { | |
| 687 if ('0' <= c && c <= '7') | |
| 688 cmtable[c] |= CMoctal; | |
| 689 if (isdigit(c) || ('a' <= c && c <= 'f') || ('A' <= c && c <= 'F')) | |
| 690 cmtable[c] |= CMhex; | |
| 691 if (isalnum(c) || c == '_') | |
| 692 cmtable[c] |= CMidchar; | |
| 693 } | |
| 694 } | |
| 695 | |
| 696 static void initKeywords() | |
| 697 { | |
| 698 uint nkeywords = keywords.length; | |
| 699 | |
| 700 if (global.params.Dversion == 1) | |
| 701 nkeywords -= 2; | |
| 702 | |
| 703 cmtable_init(); | |
| 704 | |
| 705 for (uint u = 0; u < nkeywords; u++) | |
| 706 { | |
| 34 | 707 //printf("keyword[%d] = '%.*s'\n",u, keywords[u].name); |
| 0 | 708 string s = keywords[u].name; |
| 709 TOK v = keywords[u].value; | |
| 710 StringValue* sv = stringtable.insert(s); | |
| 711 sv.ptrvalue = cast(void*) new Identifier(sv.lstring.string_, v); | |
| 712 | |
| 713 //printf("tochars[%d] = '%s'\n",v, s); | |
| 714 Token.tochars[v] = s; | |
| 715 } | |
| 716 | |
| 717 Token.tochars[TOK.TOKeof] = "EOF"; | |
| 718 Token.tochars[TOK.TOKlcurly] = "{"; | |
| 719 Token.tochars[TOK.TOKrcurly] = "}"; | |
| 720 Token.tochars[TOK.TOKlparen] = "("; | |
| 721 Token.tochars[TOK.TOKrparen] = ")"; | |
| 722 Token.tochars[TOK.TOKlbracket] = "["; | |
| 723 Token.tochars[TOK.TOKrbracket] = "]"; | |
| 724 Token.tochars[TOK.TOKsemicolon] = ";"; | |
| 725 Token.tochars[TOK.TOKcolon] = ":"; | |
| 726 Token.tochars[TOK.TOKcomma] = ","; | |
| 727 Token.tochars[TOK.TOKdot] = "."; | |
| 728 Token.tochars[TOK.TOKxor] = "^"; | |
| 729 Token.tochars[TOK.TOKxorass] = "^="; | |
| 730 Token.tochars[TOK.TOKassign] = "="; | |
| 731 Token.tochars[TOK.TOKconstruct] = "="; | |
| 732 version (DMDV2) { | |
| 733 Token.tochars[TOK.TOKblit] = "="; | |
| 734 } | |
| 735 Token.tochars[TOK.TOKlt] = "<"; | |
| 736 Token.tochars[TOK.TOKgt] = ">"; | |
| 737 Token.tochars[TOK.TOKle] = "<="; | |
| 738 Token.tochars[TOK.TOKge] = ">="; | |
| 739 Token.tochars[TOK.TOKequal] = "=="; | |
| 740 Token.tochars[TOK.TOKnotequal] = "!="; | |
| 741 Token.tochars[TOK.TOKnotidentity] = "!is"; | |
| 742 Token.tochars[TOK.TOKtobool] = "!!"; | |
| 743 | |
| 744 Token.tochars[TOK.TOKunord] = "!<>="; | |
| 745 Token.tochars[TOK.TOKue] = "!<>"; | |
| 746 Token.tochars[TOK.TOKlg] = "<>"; | |
| 747 Token.tochars[TOK.TOKleg] = "<>="; | |
| 748 Token.tochars[TOK.TOKule] = "!>"; | |
| 749 Token.tochars[TOK.TOKul] = "!>="; | |
| 750 Token.tochars[TOK.TOKuge] = "!<"; | |
| 751 Token.tochars[TOK.TOKug] = "!<="; | |
| 752 | |
| 753 Token.tochars[TOK.TOKnot] = "!"; | |
| 754 Token.tochars[TOK.TOKtobool] = "!!"; | |
| 755 Token.tochars[TOK.TOKshl] = "<<"; | |
| 756 Token.tochars[TOK.TOKshr] = ">>"; | |
| 757 Token.tochars[TOK.TOKushr] = ">>>"; | |
| 758 Token.tochars[TOK.TOKadd] = "+"; | |
| 759 Token.tochars[TOK.TOKmin] = "-"; | |
| 760 Token.tochars[TOK.TOKmul] = "*"; | |
| 761 Token.tochars[TOK.TOKdiv] = "/"; | |
| 762 Token.tochars[TOK.TOKmod] = "%"; | |
| 763 Token.tochars[TOK.TOKslice] = ".."; | |
| 764 Token.tochars[TOK.TOKdotdotdot] = "..."; | |
| 765 Token.tochars[TOK.TOKand] = "&"; | |
| 766 Token.tochars[TOK.TOKandand] = "&&"; | |
| 767 Token.tochars[TOK.TOKor] = "|"; | |
| 768 Token.tochars[TOK.TOKoror] = "||"; | |
| 769 Token.tochars[TOK.TOKarray] = "[]"; | |
| 770 Token.tochars[TOK.TOKindex] = "[i]"; | |
| 771 Token.tochars[TOK.TOKaddress] = "&"; | |
| 772 Token.tochars[TOK.TOKstar] = "*"; | |
| 773 Token.tochars[TOK.TOKtilde] = "~"; | |
| 774 Token.tochars[TOK.TOKdollar] = "$"; | |
| 775 Token.tochars[TOK.TOKcast] = "cast"; | |
| 776 Token.tochars[TOK.TOKplusplus] = "++"; | |
| 777 Token.tochars[TOK.TOKminusminus] = "--"; | |
| 778 Token.tochars[TOK.TOKtype] = "type"; | |
| 779 Token.tochars[TOK.TOKquestion] = "?"; | |
| 780 Token.tochars[TOK.TOKneg] = "-"; | |
| 781 Token.tochars[TOK.TOKuadd] = "+"; | |
| 782 Token.tochars[TOK.TOKvar] = "var"; | |
| 783 Token.tochars[TOK.TOKaddass] = "+="; | |
| 784 Token.tochars[TOK.TOKminass] = "-="; | |
| 785 Token.tochars[TOK.TOKmulass] = "*="; | |
| 786 Token.tochars[TOK.TOKdivass] = "/="; | |
| 787 Token.tochars[TOK.TOKmodass] = "%="; | |
| 788 Token.tochars[TOK.TOKshlass] = "<<="; | |
| 789 Token.tochars[TOK.TOKshrass] = ">>="; | |
| 790 Token.tochars[TOK.TOKushrass] = ">>>="; | |
| 791 Token.tochars[TOK.TOKandass] = "&="; | |
| 792 Token.tochars[TOK.TOKorass] = "|="; | |
| 793 Token.tochars[TOK.TOKcatass] = "~="; | |
| 794 Token.tochars[TOK.TOKcat] = "~"; | |
| 795 Token.tochars[TOK.TOKcall] = "call"; | |
| 796 Token.tochars[TOK.TOKidentity] = "is"; | |
| 797 Token.tochars[TOK.TOKnotidentity] = "!is"; | |
| 798 | |
| 799 Token.tochars[TOK.TOKorass] = "|="; | |
| 800 Token.tochars[TOK.TOKidentifier] = "identifier"; | |
| 801 Token.tochars[TOK.TOKat] = "@"; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
802 Token.tochars[TOK.TOKpow] = "^^"; |
| 135 | 803 Token.tochars[TOK.TOKpowass] = "^^="; |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
804 |
| 0 | 805 // For debugging |
| 73 | 806 Token.tochars[TOKerror] = "error"; |
| 0 | 807 Token.tochars[TOK.TOKdotexp] = "dotexp"; |
| 808 Token.tochars[TOK.TOKdotti] = "dotti"; | |
| 809 Token.tochars[TOK.TOKdotvar] = "dotvar"; | |
| 810 Token.tochars[TOK.TOKdottype] = "dottype"; | |
| 811 Token.tochars[TOK.TOKsymoff] = "symoff"; | |
| 812 Token.tochars[TOK.TOKarraylength] = "arraylength"; | |
| 813 Token.tochars[TOK.TOKarrayliteral] = "arrayliteral"; | |
| 814 Token.tochars[TOK.TOKassocarrayliteral] = "assocarrayliteral"; | |
| 815 Token.tochars[TOK.TOKstructliteral] = "structliteral"; | |
| 816 Token.tochars[TOK.TOKstring] = "string"; | |
| 817 Token.tochars[TOK.TOKdsymbol] = "symbol"; | |
| 818 Token.tochars[TOK.TOKtuple] = "tuple"; | |
| 819 Token.tochars[TOK.TOKdeclaration] = "declaration"; | |
| 820 Token.tochars[TOK.TOKdottd] = "dottd"; | |
| 821 Token.tochars[TOK.TOKon_scope_exit] = "scope(exit)"; | |
| 822 Token.tochars[TOK.TOKon_scope_success] = "scope(success)"; | |
| 823 Token.tochars[TOK.TOKon_scope_failure] = "scope(failure)"; | |
| 824 } | |
| 825 | |
| 826 static Identifier idPool(string s) | |
| 827 { | |
| 828 StringValue* sv = stringtable.update(s); | |
| 829 Identifier id = cast(Identifier) sv.ptrvalue; | |
| 830 if (id is null) | |
| 831 { | |
| 832 id = new Identifier(sv.lstring.string_, TOK.TOKidentifier); | |
| 833 sv.ptrvalue = cast(void*)id; | |
| 834 } | |
| 835 | |
| 836 return id; | |
| 837 } | |
| 838 | |
| 839 static Identifier uniqueId(string s) | |
| 840 { | |
| 841 static int num; | |
| 842 return uniqueId(s, ++num); | |
| 843 } | |
| 844 | |
| 845 /********************************************* | |
| 846 * Create a unique identifier using the prefix s. | |
| 847 */ | |
| 848 static Identifier uniqueId(string s, int num) | |
| 849 { | |
| 850 char buffer[32]; | |
| 851 size_t slen = s.length; | |
| 852 | |
| 853 assert(slen + num.sizeof * 3 + 1 <= buffer.sizeof); | |
| 854 int len = sprintf(buffer.ptr, "%.*s%d", s, num); | |
| 855 | |
| 856 return idPool(buffer[0..len].idup); | |
| 857 } | |
| 858 | |
| 859 TOK nextToken() | |
| 860 { | |
| 861 Token *t; | |
| 862 | |
| 863 if (token.next) | |
| 864 { | |
| 865 t = token.next; | |
| 866 memcpy(&token, t, Token.sizeof); | |
| 867 t.next = freelist; | |
| 868 freelist = t; | |
| 869 } | |
| 870 else | |
| 871 { | |
| 872 scan(&token); | |
| 873 } | |
| 874 | |
| 875 //token.print(); | |
| 876 return token.value; | |
| 877 } | |
| 878 | |
| 879 /*********************** | |
| 880 * Look ahead at next token's value. | |
| 881 */ | |
| 882 TOK peekNext() | |
| 883 { | |
| 884 return peek(&token).value; | |
| 885 } | |
| 886 | |
| 163 | 887 /*********************** |
| 888 * Look 2 tokens ahead at value. | |
| 889 */ | |
| 0 | 890 TOK peekNext2() |
| 891 { | |
| 163 | 892 Token* t = peek(&token); |
| 893 return peek(t).value; | |
| 0 | 894 } |
| 895 | |
| 896 void scan(Token* t) | |
| 897 { | |
| 898 uint lastLine = loc.linnum; | |
| 899 uint linnum; | |
| 900 | |
| 901 t.blockComment = null; | |
| 902 t.lineComment = null; | |
| 903 while (1) | |
| 904 { | |
| 905 t.ptr = p; | |
| 906 //printf("p = %p, *p = '%c'\n",p,*p); | |
| 907 switch (*p) | |
| 908 { | |
| 909 case 0: | |
| 910 case 0x1A: | |
| 911 t.value = TOK.TOKeof; // end of file | |
| 912 return; | |
| 913 | |
| 914 case ' ': | |
| 915 case '\t': | |
| 916 case '\v': | |
| 917 case '\f': | |
| 918 p++; | |
| 919 continue; // skip white space | |
| 920 | |
| 921 case '\r': | |
| 922 p++; | |
| 923 if (*p != '\n') // if CR stands by itself | |
| 924 loc.linnum++; | |
| 925 continue; // skip white space | |
| 926 | |
| 927 case '\n': | |
| 928 p++; | |
| 929 loc.linnum++; | |
| 930 continue; // skip white space | |
| 931 | |
| 932 case '0': case '1': case '2': case '3': case '4': | |
| 933 case '5': case '6': case '7': case '8': case '9': | |
| 934 t.value = number(t); | |
| 935 return; | |
| 936 | |
| 937 version (CSTRINGS) { | |
| 938 case '\'': | |
| 939 t.value = charConstant(t, 0); | |
| 940 return; | |
| 941 | |
| 942 case '"': | |
| 943 t.value = stringConstant(t,0); | |
| 944 return; | |
| 945 | |
| 946 case 'l': | |
| 947 case 'L': | |
| 948 if (p[1] == '\'') | |
| 949 { | |
| 950 p++; | |
| 951 t.value = charConstant(t, 1); | |
| 952 return; | |
| 953 } | |
| 954 else if (p[1] == '"') | |
| 955 { | |
| 956 p++; | |
| 957 t.value = stringConstant(t, 1); | |
| 958 return; | |
| 959 } | |
| 960 } else { | |
| 961 case '\'': | |
| 962 t.value = charConstant(t,0); | |
| 963 return; | |
| 964 | |
| 965 case 'r': | |
| 966 if (p[1] != '"') | |
| 967 goto case_ident; | |
| 968 p++; | |
| 969 case '`': | |
| 970 t.value = wysiwygStringConstant(t, *p); | |
| 971 return; | |
| 972 | |
| 973 case 'x': | |
| 974 if (p[1] != '"') | |
| 975 goto case_ident; | |
| 976 p++; | |
| 977 t.value = hexStringConstant(t); | |
| 978 return; | |
| 979 | |
| 980 version (DMDV2) { | |
| 981 case 'q': | |
| 982 if (p[1] == '"') | |
| 983 { | |
| 984 p++; | |
| 985 t.value = delimitedStringConstant(t); | |
| 986 return; | |
| 987 } | |
| 988 else if (p[1] == '{') | |
| 989 { | |
| 990 p++; | |
| 991 t.value = tokenStringConstant(t); | |
| 992 return; | |
| 993 } | |
| 994 else | |
| 995 goto case_ident; | |
| 996 } | |
| 997 | |
| 998 case '"': | |
| 999 t.value = escapeStringConstant(t,0); | |
| 1000 return; | |
| 1001 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 1002 } else { | |
| 1003 case '\\': // escaped string literal | |
| 1004 { uint c; | |
| 1005 ubyte* pstart = p; | |
| 1006 | |
| 1007 stringbuffer.reset(); | |
| 1008 do | |
| 1009 { | |
| 1010 p++; | |
| 1011 switch (*p) | |
| 1012 { | |
| 1013 case 'u': | |
| 1014 case 'U': | |
| 1015 case '&': | |
| 1016 c = escapeSequence(); | |
| 1017 stringbuffer.writeUTF8(c); | |
| 1018 break; | |
| 1019 | |
| 1020 default: | |
| 1021 c = escapeSequence(); | |
| 1022 stringbuffer.writeByte(c); | |
| 1023 break; | |
| 1024 } | |
| 1025 } while (*p == '\\'); | |
| 1026 t.len = stringbuffer.offset; | |
| 1027 stringbuffer.writeByte(0); | |
| 2 | 1028 char* cc = cast(char*)GC.malloc(stringbuffer.offset); |
| 0 | 1029 memcpy(cc, stringbuffer.data, stringbuffer.offset); |
| 1030 t.ustring = cc; | |
| 1031 t.postfix = 0; | |
| 1032 t.value = TOK.TOKstring; | |
| 1033 if (!global.params.useDeprecated) | |
| 1034 error("Escape String literal %.*s is deprecated, use double quoted string literal \"%.*s\" instead", p - pstart, pstart, p - pstart, pstart); | |
| 1035 return; | |
| 1036 } | |
| 1037 } | |
| 1038 case 'l': | |
| 1039 case 'L': | |
| 1040 } | |
| 1041 case 'a': case 'b': case 'c': case 'd': case 'e': | |
| 1042 case 'f': case 'g': case 'h': case 'i': case 'j': | |
| 1043 case 'k': case 'm': case 'n': case 'o': | |
| 1044 version (DMDV2) { | |
| 1045 case 'p': /*case 'q': case 'r':*/ case 's': case 't': | |
| 1046 } else { | |
| 1047 case 'p': case 'q': /*case 'r':*/ case 's': case 't': | |
| 1048 } | |
| 1049 case 'u': case 'v': case 'w': /*case 'x':*/ case 'y': | |
| 1050 case 'z': | |
| 1051 case 'A': case 'B': case 'C': case 'D': case 'E': | |
| 1052 case 'F': case 'G': case 'H': case 'I': case 'J': | |
| 1053 case 'K': case 'M': case 'N': case 'O': | |
| 1054 case 'P': case 'Q': case 'R': case 'S': case 'T': | |
| 1055 case 'U': case 'V': case 'W': case 'X': case 'Y': | |
| 1056 case 'Z': | |
| 1057 case '_': | |
| 1058 case_ident: | |
| 135 | 1059 { |
| 1060 ubyte c; | |
| 0 | 1061 |
| 135 | 1062 while (1) |
| 1063 { | |
| 1064 c = *++p; | |
| 1065 if (isidchar(c)) | |
| 1066 continue; | |
| 1067 else if (c & 0x80) | |
| 1068 { | |
| 1069 ubyte *s = p; | |
| 1070 uint u = decodeUTF(); | |
| 1071 if (isUniAlpha(u)) | |
| 1072 continue; | |
| 1073 error("char 0x%04x not allowed in identifier", u); | |
| 1074 p = s; | |
| 1075 } | |
| 1076 break; | |
| 1077 } | |
| 1078 | |
| 1079 StringValue *sv = stringtable.update((cast(immutable(char)*)t.ptr)[0.. p - t.ptr]); | |
| 1080 Identifier id = cast(Identifier) sv.ptrvalue; | |
| 1081 | |
| 0 | 1082 if (id is null) |
| 1083 { id = new Identifier(sv.lstring.string_, TOK.TOKidentifier); | |
| 1084 sv.ptrvalue = cast(void*)id; | |
| 1085 } | |
| 1086 t.ident = id; | |
| 1087 t.value = cast(TOK) id.value; | |
| 1088 anyToken = 1; | |
| 1089 if (*t.ptr == '_') // if special identifier token | |
| 1090 { | |
| 1091 static char date[11+1]; | |
| 1092 static char time[8+1]; | |
| 1093 static char timestamp[24+1]; | |
| 1094 | |
| 1095 if (!date[0]) // lazy evaluation | |
| 1096 { time_t tm; | |
| 1097 char *p; | |
| 1098 | |
| 1099 .time(&tm); | |
| 1100 p = ctime(&tm); | |
| 1101 assert(p); | |
| 1102 sprintf(date.ptr, "%.6s %.4s", p + 4, p + 20); | |
| 1103 sprintf(time.ptr, "%.8s", p + 11); | |
| 1104 sprintf(timestamp.ptr, "%.24s", p); | |
| 1105 } | |
| 1106 | |
| 1107 ///version (DMDV1) { | |
| 1108 /// if (mod && id == Id.FILE) | |
| 1109 /// { | |
| 1110 /// t.ustring = cast(ubyte*)(loc.filename ? loc.filename : mod.ident.toChars()); | |
| 1111 /// goto Lstr; | |
| 1112 /// } | |
| 1113 /// else if (mod && id == Id.LINE) | |
| 1114 /// { | |
| 1115 /// t.value = TOK.TOKint64v; | |
| 1116 /// t.uns64value = loc.linnum; | |
| 1117 /// } | |
| 1118 /// else | |
| 1119 ///} | |
| 1120 if (id == Id.DATE) | |
| 1121 { | |
| 1122 t.ustring = date.ptr; | |
| 1123 goto Lstr; | |
| 1124 } | |
| 1125 else if (id == Id.TIME) | |
| 1126 { | |
| 1127 t.ustring = time.ptr; | |
| 1128 goto Lstr; | |
| 1129 } | |
| 1130 else if (id == Id.VENDOR) | |
| 1131 { | |
| 1132 t.ustring = "Digital Mars D".ptr; | |
| 1133 goto Lstr; | |
| 1134 } | |
| 1135 else if (id == Id.TIMESTAMP) | |
| 1136 { | |
| 1137 t.ustring = timestamp.ptr; | |
| 1138 Lstr: | |
| 1139 t.value = TOK.TOKstring; | |
| 1140 Llen: | |
| 1141 t.postfix = 0; | |
| 1142 t.len = strlen(cast(char*)t.ustring); | |
| 1143 } | |
| 1144 else if (id == Id.VERSIONX) | |
| 1145 { | |
| 1146 uint major = 0; | |
| 1147 uint minor = 0; | |
| 1148 | |
| 1149 foreach (char cc; global.version_[1..$]) | |
| 1150 { | |
| 1151 if (isdigit(cc)) | |
| 1152 minor = minor * 10 + cc - '0'; | |
| 1153 else if (cc == '.') | |
| 1154 { | |
| 1155 major = minor; | |
| 1156 minor = 0; | |
| 1157 } | |
| 1158 else | |
| 1159 break; | |
| 1160 } | |
| 1161 t.value = TOK.TOKint64v; | |
| 1162 t.uns64value = major * 1000 + minor; | |
| 1163 } | |
| 1164 ///version (DMDV2) { | |
| 1165 else if (id == Id.EOFX) | |
| 1166 { | |
| 1167 t.value = TOK.TOKeof; | |
| 1168 // Advance scanner to end of file | |
| 1169 while (!(*p == 0 || *p == 0x1A)) | |
| 1170 p++; | |
| 1171 } | |
| 1172 ///} | |
| 1173 } | |
| 1174 //printf("t.value = %d\n",t.value); | |
| 1175 return; | |
| 1176 } | |
| 1177 | |
| 1178 case '/': | |
| 1179 p++; | |
| 1180 switch (*p) | |
| 1181 { | |
| 1182 case '=': | |
| 1183 p++; | |
| 1184 t.value = TOK.TOKdivass; | |
| 1185 return; | |
| 1186 | |
| 1187 case '*': | |
| 1188 p++; | |
| 1189 linnum = loc.linnum; | |
| 1190 while (1) | |
| 1191 { | |
| 1192 while (1) | |
| 1193 { | |
| 1194 ubyte c = *p; | |
| 1195 switch (c) | |
| 1196 { | |
| 1197 case '/': | |
| 1198 break; | |
| 1199 | |
| 1200 case '\n': | |
| 1201 loc.linnum++; | |
| 1202 p++; | |
| 1203 continue; | |
| 1204 | |
| 1205 case '\r': | |
| 1206 p++; | |
| 1207 if (*p != '\n') | |
| 1208 loc.linnum++; | |
| 1209 continue; | |
| 1210 | |
| 1211 case 0: | |
| 1212 case 0x1A: | |
| 1213 error("unterminated /* */ comment"); | |
| 1214 p = end; | |
| 1215 t.value = TOK.TOKeof; | |
| 1216 return; | |
| 1217 | |
| 1218 default: | |
| 1219 if (c & 0x80) | |
| 1220 { uint u = decodeUTF(); | |
| 1221 if (u == PS || u == LS) | |
| 1222 loc.linnum++; | |
| 1223 } | |
| 1224 p++; | |
| 1225 continue; | |
| 1226 } | |
| 1227 break; | |
| 1228 } | |
| 1229 p++; | |
| 1230 if (p[-2] == '*' && p - 3 != t.ptr) | |
| 1231 break; | |
| 1232 } | |
| 1233 if (commentToken) | |
| 1234 { | |
| 1235 t.value = TOK.TOKcomment; | |
| 1236 return; | |
| 1237 } | |
| 1238 else if (doDocComment && t.ptr[2] == '*' && p - 4 != t.ptr) | |
| 1239 { // if /** but not /**/ | |
| 1240 getDocComment(t, lastLine == linnum); | |
| 1241 } | |
| 1242 continue; | |
| 1243 | |
| 1244 case '/': // do // style comments | |
| 1245 linnum = loc.linnum; | |
| 1246 while (1) | |
| 1247 { ubyte c = *++p; | |
| 1248 switch (c) | |
| 1249 { | |
| 1250 case '\n': | |
| 1251 break; | |
| 1252 | |
| 1253 case '\r': | |
| 1254 if (p[1] == '\n') | |
| 1255 p++; | |
| 1256 break; | |
| 1257 | |
| 1258 case 0: | |
| 1259 case 0x1A: | |
| 1260 if (commentToken) | |
| 1261 { | |
| 1262 p = end; | |
| 1263 t.value = TOK.TOKcomment; | |
| 1264 return; | |
| 1265 } | |
|
157
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
1266 if (doDocComment && t.ptr[2] == '/' || t.ptr[2] == '!') // '///' or '//!' |
| 0 | 1267 getDocComment(t, lastLine == linnum); |
| 1268 p = end; | |
| 1269 t.value = TOK.TOKeof; | |
| 1270 return; | |
| 1271 | |
| 1272 default: | |
| 1273 if (c & 0x80) | |
| 1274 { uint u = decodeUTF(); | |
| 1275 if (u == PS || u == LS) | |
| 1276 break; | |
| 1277 } | |
| 1278 continue; | |
| 1279 } | |
| 1280 break; | |
| 1281 } | |
| 1282 | |
| 1283 if (commentToken) | |
| 1284 { | |
| 1285 p++; | |
| 1286 loc.linnum++; | |
| 1287 t.value = TOK.TOKcomment; | |
| 1288 return; | |
| 1289 } | |
|
157
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
1290 if (doDocComment && t.ptr[2] == '/' || t.ptr[2] == '!') // '///' or '//!' |
| 0 | 1291 getDocComment(t, lastLine == linnum); |
| 1292 | |
| 1293 p++; | |
| 1294 loc.linnum++; | |
| 1295 continue; | |
| 1296 | |
| 1297 case '+': | |
| 1298 { | |
| 1299 int nest; | |
| 1300 | |
| 1301 linnum = loc.linnum; | |
| 1302 p++; | |
| 1303 nest = 1; | |
| 1304 while (1) | |
| 1305 { ubyte c = *p; | |
| 1306 switch (c) | |
| 1307 { | |
| 1308 case '/': | |
| 1309 p++; | |
| 1310 if (*p == '+') | |
| 1311 { | |
| 1312 p++; | |
| 1313 nest++; | |
| 1314 } | |
| 1315 continue; | |
| 1316 | |
| 1317 case '+': | |
| 1318 p++; | |
| 1319 if (*p == '/') | |
| 1320 { | |
| 1321 p++; | |
| 1322 if (--nest == 0) | |
| 1323 break; | |
| 1324 } | |
| 1325 continue; | |
| 1326 | |
| 1327 case '\r': | |
| 1328 p++; | |
| 1329 if (*p != '\n') | |
| 1330 loc.linnum++; | |
| 1331 continue; | |
| 1332 | |
| 1333 case '\n': | |
| 1334 loc.linnum++; | |
| 1335 p++; | |
| 1336 continue; | |
| 1337 | |
| 1338 case 0: | |
| 1339 case 0x1A: | |
| 1340 error("unterminated /+ +/ comment"); | |
| 1341 p = end; | |
| 1342 t.value = TOK.TOKeof; | |
| 1343 return; | |
| 1344 | |
| 1345 default: | |
| 1346 if (c & 0x80) | |
| 1347 { uint u = decodeUTF(); | |
| 1348 if (u == PS || u == LS) | |
| 1349 loc.linnum++; | |
| 1350 } | |
| 1351 p++; | |
| 1352 continue; | |
| 1353 } | |
| 1354 break; | |
| 1355 } | |
| 1356 if (commentToken) | |
| 1357 { | |
| 1358 t.value = TOK.TOKcomment; | |
| 1359 return; | |
| 1360 } | |
| 1361 if (doDocComment && t.ptr[2] == '+' && p - 4 != t.ptr) | |
| 1362 { // if /++ but not /++/ | |
| 1363 getDocComment(t, lastLine == linnum); | |
| 1364 } | |
| 1365 continue; | |
| 1366 } | |
| 1367 | |
| 1368 default: | |
| 1369 break; /// | |
| 1370 } | |
| 1371 t.value = TOK.TOKdiv; | |
| 1372 return; | |
| 1373 | |
| 1374 case '.': | |
| 1375 p++; | |
| 1376 if (isdigit(*p)) | |
| 1377 { /* Note that we don't allow ._1 and ._ as being | |
| 1378 * valid floating point numbers. | |
| 1379 */ | |
| 1380 p--; | |
| 1381 t.value = inreal(t); | |
| 1382 } | |
| 1383 else if (p[0] == '.') | |
| 1384 { | |
| 1385 if (p[1] == '.') | |
| 1386 { p += 2; | |
| 1387 t.value = TOK.TOKdotdotdot; | |
| 1388 } | |
| 1389 else | |
| 1390 { p++; | |
| 1391 t.value = TOK.TOKslice; | |
| 1392 } | |
| 1393 } | |
| 1394 else | |
| 1395 t.value = TOK.TOKdot; | |
| 1396 return; | |
| 1397 | |
| 1398 case '&': | |
| 1399 p++; | |
| 1400 if (*p == '=') | |
| 1401 { p++; | |
| 1402 t.value = TOK.TOKandass; | |
| 1403 } | |
| 1404 else if (*p == '&') | |
| 1405 { p++; | |
| 1406 t.value = TOK.TOKandand; | |
| 1407 } | |
| 1408 else | |
| 1409 t.value = TOK.TOKand; | |
| 1410 return; | |
| 1411 | |
| 1412 case '|': | |
| 1413 p++; | |
| 1414 if (*p == '=') | |
| 1415 { p++; | |
| 1416 t.value = TOK.TOKorass; | |
| 1417 } | |
| 1418 else if (*p == '|') | |
| 1419 { p++; | |
| 1420 t.value = TOK.TOKoror; | |
| 1421 } | |
| 1422 else | |
| 1423 t.value = TOK.TOKor; | |
| 1424 return; | |
| 1425 | |
| 1426 case '-': | |
| 1427 p++; | |
| 1428 if (*p == '=') | |
| 1429 { p++; | |
| 1430 t.value = TOK.TOKminass; | |
| 1431 } | |
| 1432 /// #if 0 | |
| 1433 /// else if (*p == '>') | |
| 1434 /// { p++; | |
| 1435 /// t.value = TOK.TOKarrow; | |
| 1436 /// } | |
| 1437 /// #endif | |
| 1438 else if (*p == '-') | |
| 1439 { p++; | |
| 1440 t.value = TOK.TOKminusminus; | |
| 1441 } | |
| 1442 else | |
| 1443 t.value = TOK.TOKmin; | |
| 1444 return; | |
| 1445 | |
| 1446 case '+': | |
| 1447 p++; | |
| 1448 if (*p == '=') | |
| 1449 { p++; | |
| 1450 t.value = TOK.TOKaddass; | |
| 1451 } | |
| 1452 else if (*p == '+') | |
| 1453 { p++; | |
| 1454 t.value = TOK.TOKplusplus; | |
| 1455 } | |
| 1456 else | |
| 1457 t.value = TOK.TOKadd; | |
| 1458 return; | |
| 1459 | |
| 1460 case '<': | |
| 1461 p++; | |
| 1462 if (*p == '=') | |
| 1463 { p++; | |
| 1464 t.value = TOK.TOKle; // <= | |
| 1465 } | |
| 1466 else if (*p == '<') | |
| 1467 { p++; | |
| 1468 if (*p == '=') | |
| 1469 { p++; | |
| 1470 t.value = TOK.TOKshlass; // <<= | |
| 1471 } | |
| 1472 else | |
| 1473 t.value = TOK.TOKshl; // << | |
| 1474 } | |
| 1475 else if (*p == '>') | |
| 1476 { p++; | |
| 1477 if (*p == '=') | |
| 1478 { p++; | |
| 1479 t.value = TOK.TOKleg; // <>= | |
| 1480 } | |
| 1481 else | |
| 1482 t.value = TOK.TOKlg; // <> | |
| 1483 } | |
| 1484 else | |
| 1485 t.value = TOK.TOKlt; // < | |
| 1486 return; | |
| 1487 | |
| 1488 case '>': | |
| 1489 p++; | |
| 1490 if (*p == '=') | |
| 1491 { p++; | |
| 1492 t.value = TOK.TOKge; // >= | |
| 1493 } | |
| 1494 else if (*p == '>') | |
| 1495 { p++; | |
| 1496 if (*p == '=') | |
| 1497 { p++; | |
| 1498 t.value = TOK.TOKshrass; // >>= | |
| 1499 } | |
| 1500 else if (*p == '>') | |
| 1501 { p++; | |
| 1502 if (*p == '=') | |
| 1503 { p++; | |
| 1504 t.value = TOK.TOKushrass; // >>>= | |
| 1505 } | |
| 1506 else | |
| 1507 t.value = TOK.TOKushr; // >>> | |
| 1508 } | |
| 1509 else | |
| 1510 t.value = TOK.TOKshr; // >> | |
| 1511 } | |
| 1512 else | |
| 1513 t.value = TOK.TOKgt; // > | |
| 1514 return; | |
| 1515 | |
| 1516 case '!': | |
| 1517 p++; | |
| 1518 if (*p == '=') | |
| 1519 { p++; | |
| 1520 if (*p == '=' && global.params.Dversion == 1) | |
| 1521 { p++; | |
| 1522 t.value = TOK.TOKnotidentity; // !== | |
| 1523 } | |
| 1524 else | |
| 1525 t.value = TOK.TOKnotequal; // != | |
| 1526 } | |
| 1527 else if (*p == '<') | |
| 1528 { p++; | |
| 1529 if (*p == '>') | |
| 1530 { p++; | |
| 1531 if (*p == '=') | |
| 1532 { p++; | |
| 1533 t.value = TOK.TOKunord; // !<>= | |
| 1534 } | |
| 1535 else | |
| 1536 t.value = TOK.TOKue; // !<> | |
| 1537 } | |
| 1538 else if (*p == '=') | |
| 1539 { p++; | |
| 1540 t.value = TOK.TOKug; // !<= | |
| 1541 } | |
| 1542 else | |
| 1543 t.value = TOK.TOKuge; // !< | |
| 1544 } | |
| 1545 else if (*p == '>') | |
| 1546 { p++; | |
| 1547 if (*p == '=') | |
| 1548 { p++; | |
| 1549 t.value = TOK.TOKul; // !>= | |
| 1550 } | |
| 1551 else | |
| 1552 t.value = TOK.TOKule; // !> | |
| 1553 } | |
| 1554 else | |
| 1555 t.value = TOK.TOKnot; // ! | |
| 1556 return; | |
| 1557 | |
| 1558 case '=': | |
| 1559 p++; | |
| 1560 if (*p == '=') | |
| 1561 { p++; | |
| 1562 if (*p == '=' && global.params.Dversion == 1) | |
| 1563 { p++; | |
| 1564 t.value = TOK.TOKidentity; // === | |
| 1565 } | |
| 1566 else | |
| 1567 t.value = TOK.TOKequal; // == | |
| 1568 } | |
| 1569 else | |
| 1570 t.value = TOK.TOKassign; // = | |
| 1571 return; | |
| 1572 | |
| 1573 case '~': | |
| 1574 p++; | |
| 1575 if (*p == '=') | |
| 1576 { p++; | |
| 1577 t.value = TOK.TOKcatass; // ~= | |
| 1578 } | |
| 1579 else | |
| 1580 t.value = TOK.TOKtilde; // ~ | |
| 1581 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1582 |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1583 version(DMDV2) { |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1584 case '^': |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1585 p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1586 if (*p == '^') |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1587 { p++; |
| 135 | 1588 if (*p == '=') |
| 1589 { p++; | |
| 1590 t.value = TOKpowass; // ^^= | |
| 1591 } | |
| 1592 else | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1593 t.value = TOKpow; // ^^ |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1594 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1595 else if (*p == '=') |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1596 { p++; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1597 t.value = TOKxorass; // ^= |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1598 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1599 else |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1600 t.value = TOKxor; // ^ |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1601 return; |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1602 } |
|
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1603 |
| 0 | 1604 /* |
| 1605 #define SINGLE(c,tok) case c: p++; t.value = tok; return; | |
| 1606 | |
| 1607 SINGLE('(', TOKlparen) | |
| 1608 SINGLE(')', TOKrparen) | |
| 1609 SINGLE('[', TOKlbracket) | |
| 1610 SINGLE(']', TOKrbracket) | |
| 1611 SINGLE('{', TOKlcurly) | |
| 1612 SINGLE('}', TOKrcurly) | |
| 1613 SINGLE('?', TOKquestion) | |
| 1614 SINGLE(',', TOKcomma) | |
| 1615 SINGLE(';', TOKsemicolon) | |
| 1616 SINGLE(':', TOKcolon) | |
| 1617 SINGLE('$', TOKdollar) | |
| 1618 SINGLE('@', TOKat) | |
| 1619 | |
| 1620 #undef SINGLE | |
| 1621 | |
| 1622 #define DOUBLE(c1,tok1,c2,tok2) \ | |
| 1623 case c1: \ | |
| 1624 p++; \ | |
| 1625 if (*p == c2) \ | |
| 1626 { p++; \ | |
| 1627 t.value = tok2; \ | |
| 1628 } \ | |
| 1629 else \ | |
| 1630 t.value = tok1; \ | |
| 1631 return; | |
| 1632 | |
| 1633 DOUBLE('*', TOKmul, '=', TOKmulass) | |
| 1634 DOUBLE('%', TOKmod, '=', TOKmodass) | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1635 #if DMDV1 |
| 0 | 1636 DOUBLE('^', TOKxor, '=', TOKxorass) |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1637 #endif |
| 0 | 1638 #undef DOUBLE |
| 1639 */ | |
| 1640 | |
| 1641 case '(': p++; t.value = TOK.TOKlparen; return; | |
| 1642 case ')': p++; t.value = TOK.TOKrparen; return; | |
| 1643 case '[': p++; t.value = TOK.TOKlbracket; return; | |
| 1644 case ']': p++; t.value = TOK.TOKrbracket; return; | |
| 1645 case '{': p++; t.value = TOK.TOKlcurly; return; | |
| 1646 case '}': p++; t.value = TOK.TOKrcurly; return; | |
| 1647 case '?': p++; t.value = TOK.TOKquestion; return; | |
| 1648 case ',': p++; t.value = TOK.TOKcomma; return; | |
| 1649 case ';': p++; t.value = TOK.TOKsemicolon; return; | |
| 1650 case ':': p++; t.value = TOK.TOKcolon; return; | |
| 1651 case '$': p++; t.value = TOK.TOKdollar; return; | |
| 1652 case '@': p++; t.value = TOK.TOKat; return; | |
| 1653 | |
| 1654 case '*': | |
| 1655 p++; | |
| 1656 if (*p == '=') { | |
| 1657 p++; | |
| 1658 t.value = TOK.TOKmulass; | |
| 1659 } else { | |
| 1660 t.value = TOK.TOKmul; | |
| 1661 } | |
| 1662 return; | |
| 1663 | |
| 1664 case '%': | |
| 1665 p++; | |
| 1666 if (*p == '=') { | |
| 1667 p++; | |
| 1668 t.value = TOK.TOKmodass; | |
| 1669 } else { | |
| 1670 t.value = TOK.TOKmod; | |
| 1671 } | |
| 1672 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1673 version(DMDV1) { |
| 0 | 1674 case '^': |
| 1675 p++; | |
| 1676 if (*p == '=') { | |
| 1677 p++; | |
| 1678 t.value = TOK.TOKxorass; | |
| 1679 } else { | |
| 1680 t.value = TOK.TOKxor; | |
| 1681 } | |
| 1682 return; | |
|
130
60bb0fe4563e
dmdfe 2.037 first main iteration
Eldar Insafutdinov <e.insafutdinov@gmail.com>
parents:
114
diff
changeset
|
1683 } |
| 0 | 1684 case '#': |
| 1685 p++; | |
| 1686 pragma_(); | |
| 1687 continue; | |
| 1688 | |
| 1689 default: | |
| 135 | 1690 { uint c = *p; |
| 0 | 1691 |
| 1692 if (c & 0x80) | |
| 135 | 1693 { c = decodeUTF(); |
| 0 | 1694 |
| 1695 // Check for start of unicode identifier | |
| 135 | 1696 if (isUniAlpha(c)) |
| 0 | 1697 goto case_ident; |
| 1698 | |
| 135 | 1699 if (c == PS || c == LS) |
| 0 | 1700 { |
| 1701 loc.linnum++; | |
| 1702 p++; | |
| 1703 continue; | |
| 1704 } | |
| 1705 } | |
| 135 | 1706 if (c < 0x80 && isprint(c)) |
| 0 | 1707 error("unsupported char '%c'", c); |
| 1708 else | |
| 1709 error("unsupported char 0x%02x", c); | |
| 1710 p++; | |
| 1711 continue; | |
| 1712 } | |
| 1713 } | |
| 1714 } | |
| 1715 } | |
| 1716 | |
| 1717 Token* peek(Token* ct) | |
| 1718 { | |
| 1719 Token* t; | |
| 1720 | |
| 1721 if (ct.next) | |
| 1722 t = ct.next; | |
| 1723 else | |
| 1724 { | |
| 1725 t = new Token(); | |
| 1726 scan(t); | |
| 1727 t.next = null; | |
| 1728 ct.next = t; | |
| 1729 } | |
| 1730 return t; | |
| 1731 } | |
| 1732 | |
| 1733 Token* peekPastParen(Token* tk) | |
| 1734 { | |
| 1735 //printf("peekPastParen()\n"); | |
| 1736 int parens = 1; | |
| 1737 int curlynest = 0; | |
| 1738 while (1) | |
| 1739 { | |
| 1740 tk = peek(tk); | |
| 1741 //tk.print(); | |
| 1742 switch (tk.value) | |
| 1743 { | |
| 1744 case TOK.TOKlparen: | |
| 1745 parens++; | |
| 1746 continue; | |
| 1747 | |
| 1748 case TOK.TOKrparen: | |
| 1749 --parens; | |
| 1750 if (parens) | |
| 1751 continue; | |
| 1752 tk = peek(tk); | |
| 1753 break; | |
| 1754 | |
| 1755 case TOK.TOKlcurly: | |
| 1756 curlynest++; | |
| 1757 continue; | |
| 1758 | |
| 1759 case TOK.TOKrcurly: | |
| 1760 if (--curlynest >= 0) | |
| 1761 continue; | |
| 1762 break; | |
| 1763 | |
| 1764 case TOK.TOKsemicolon: | |
| 1765 if (curlynest) | |
| 1766 continue; | |
| 1767 break; | |
| 1768 | |
| 1769 case TOK.TOKeof: | |
| 1770 break; | |
| 1771 | |
| 1772 default: | |
| 1773 continue; | |
| 1774 } | |
| 1775 return tk; | |
| 1776 } | |
| 1777 } | |
| 1778 | |
| 1779 /******************************************* | |
| 1780 * Parse escape sequence. | |
| 1781 */ | |
| 1782 uint escapeSequence() | |
| 1783 { | |
| 1784 uint c = *p; | |
| 1785 | |
| 1786 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 1787 return c; | |
| 1788 } | |
| 1789 int n; | |
| 1790 int ndigits; | |
| 1791 | |
| 1792 switch (c) | |
| 1793 { | |
| 1794 case '\'': | |
| 1795 case '"': | |
| 1796 case '?': | |
| 1797 case '\\': | |
| 1798 Lconsume: | |
| 1799 p++; | |
| 1800 break; | |
| 1801 | |
| 1802 case 'a': c = 7; goto Lconsume; | |
| 1803 case 'b': c = 8; goto Lconsume; | |
| 1804 case 'f': c = 12; goto Lconsume; | |
| 1805 case 'n': c = 10; goto Lconsume; | |
| 1806 case 'r': c = 13; goto Lconsume; | |
| 1807 case 't': c = 9; goto Lconsume; | |
| 1808 case 'v': c = 11; goto Lconsume; | |
| 1809 | |
| 1810 case 'u': | |
| 1811 ndigits = 4; | |
| 1812 goto Lhex; | |
| 1813 case 'U': | |
| 1814 ndigits = 8; | |
| 1815 goto Lhex; | |
| 1816 case 'x': | |
| 1817 ndigits = 2; | |
| 1818 Lhex: | |
| 1819 p++; | |
| 1820 c = *p; | |
| 1821 if (ishex(cast(ubyte)c)) | |
| 1822 { | |
| 1823 uint v; | |
| 1824 | |
| 1825 n = 0; | |
| 1826 v = 0; | |
| 1827 while (1) | |
| 1828 { | |
| 1829 if (isdigit(c)) | |
| 1830 c -= '0'; | |
| 1831 else if (islower(c)) | |
| 1832 c -= 'a' - 10; | |
| 1833 else | |
| 1834 c -= 'A' - 10; | |
| 1835 v = v * 16 + c; | |
| 1836 c = *++p; | |
| 1837 if (++n == ndigits) | |
| 1838 break; | |
| 1839 if (!ishex(cast(ubyte)c)) | |
| 1840 { error("escape hex sequence has %d hex digits instead of %d", n, ndigits); | |
| 1841 break; | |
| 1842 } | |
| 1843 } | |
| 1844 if (ndigits != 2 && !utf_isValidDchar(v)) | |
| 1845 { error("invalid UTF character \\U%08x", v); | |
| 1846 v = '?'; // recover with valid UTF character | |
| 1847 } | |
| 1848 c = v; | |
| 1849 } | |
| 1850 else | |
| 1851 error("undefined escape hex sequence \\%c\n",c); | |
| 1852 break; | |
| 1853 | |
| 1854 case '&': // named character entity | |
| 1855 for (ubyte* idstart = ++p; true; p++) | |
| 1856 { | |
| 1857 switch (*p) | |
| 1858 { | |
| 1859 case ';': | |
| 1860 c = HtmlNamedEntity(idstart, p - idstart); | |
| 1861 if (c == ~0) | |
| 1862 { | |
| 1863 error("unnamed character entity &%s;", idstart[0..(p - idstart)]); | |
| 1864 c = ' '; | |
| 1865 } | |
| 1866 p++; | |
| 1867 break; | |
| 1868 | |
| 1869 default: | |
| 1870 if (isalpha(*p) || | |
| 1871 (p != idstart + 1 && isdigit(*p))) | |
| 1872 continue; | |
| 1873 error("unterminated named entity"); | |
| 1874 break; | |
| 1875 } | |
| 1876 break; | |
| 1877 } | |
| 1878 break; | |
| 1879 | |
| 1880 case 0: | |
| 1881 case 0x1A: // end of file | |
| 1882 c = '\\'; | |
| 1883 break; | |
| 1884 | |
| 1885 default: | |
| 1886 if (isoctal(cast(ubyte)c)) | |
| 1887 { | |
| 1888 uint v; | |
| 1889 | |
| 1890 n = 0; | |
| 1891 v = 0; | |
| 1892 do | |
| 1893 { | |
| 1894 v = v * 8 + (c - '0'); | |
| 1895 c = *++p; | |
| 1896 } while (++n < 3 && isoctal(cast(ubyte)c)); | |
| 1897 c = v; | |
| 1898 if (c > 0xFF) | |
| 1899 error("0%03o is larger than a byte", c); | |
| 1900 } | |
| 1901 else | |
| 1902 error("undefined escape sequence \\%c\n",c); | |
| 1903 break; | |
| 1904 } | |
| 1905 return c; | |
| 1906 } | |
| 1907 | |
| 1908 TOK wysiwygStringConstant(Token* t, int tc) | |
| 1909 { | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1910 uint c; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1911 Loc start = loc; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1912 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1913 p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1914 stringbuffer.reset(); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1915 while (true) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1916 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1917 c = *p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1918 switch (c) |
|
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 case '\n': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1921 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1922 break; |
|
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 case '\r': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1925 if (*p == '\n') |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1926 continue; // ignore |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1927 c = '\n'; // treat EndOfLine as \n character |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1928 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1929 break; |
|
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 0: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1932 case 0x1A: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1933 error("unterminated string constant starting at %s", start.toChars()); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1934 t.ustring = "".ptr; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1935 t.len = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1936 t.postfix = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1937 return TOKstring; |
|
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 case '"': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1940 case '`': |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1941 if (c == tc) |
|
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 t.len = stringbuffer.offset; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1944 stringbuffer.writeByte(0); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1945 char* tmp = cast(char*)GC.malloc(stringbuffer.offset); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1946 memcpy(tmp, stringbuffer.data, stringbuffer.offset); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1947 t.ustring = tmp; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1948 stringPostfix(t); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1949 return TOKstring; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1950 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1951 break; |
|
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 default: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1954 if (c & 0x80) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1955 { p--; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1956 uint u = decodeUTF(); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1957 p++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1958 if (u == PS || u == LS) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1959 loc.linnum++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1960 stringbuffer.writeUTF8(u); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1961 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1962 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1963 break; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1964 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1965 stringbuffer.writeByte(c); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1966 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
1967 |
| 0 | 1968 assert(false); |
| 1969 } | |
| 1970 | |
| 51 | 1971 /************************************** |
| 1972 * Lex hex strings: | |
| 1973 * x"0A ae 34FE BD" | |
| 1974 */ | |
| 0 | 1975 TOK hexStringConstant(Token* t) |
| 1976 { | |
| 51 | 1977 uint c; |
| 1978 Loc start = loc; | |
| 1979 uint n = 0; | |
| 1980 uint v; | |
| 1981 | |
| 1982 p++; | |
| 1983 stringbuffer.reset(); | |
| 1984 while (1) | |
| 1985 { | |
| 1986 c = *p++; | |
| 1987 switch (c) | |
| 1988 { | |
| 1989 case ' ': | |
| 1990 case '\t': | |
| 1991 case '\v': | |
| 1992 case '\f': | |
| 1993 continue; // skip white space | |
| 1994 | |
| 1995 case '\r': | |
| 1996 if (*p == '\n') | |
| 1997 continue; // ignore | |
| 1998 // Treat isolated '\r' as if it were a '\n' | |
| 1999 case '\n': | |
| 2000 loc.linnum++; | |
| 2001 continue; | |
| 2002 | |
| 2003 case 0: | |
| 2004 case 0x1A: | |
| 2005 error("unterminated string constant starting at %s", start.toChars()); | |
| 2006 t.ustring = "".ptr; | |
| 2007 t.len = 0; | |
| 2008 t.postfix = 0; | |
| 2009 return TOKstring; | |
| 2010 | |
| 2011 case '"': | |
| 2012 if (n & 1) | |
| 2013 { | |
| 2014 error("odd number (%d) of hex characters in hex string", n); | |
| 2015 stringbuffer.writeByte(v); | |
| 2016 } | |
| 2017 t.len = stringbuffer.offset; | |
| 2018 stringbuffer.writeByte(0); | |
| 2019 void* mem = malloc(stringbuffer.offset); | |
| 2020 memcpy(mem, stringbuffer.data, stringbuffer.offset); | |
| 2021 t.ustring = cast(const(char)*)mem; | |
| 2022 stringPostfix(t); | |
| 2023 return TOKstring; | |
| 2024 | |
| 2025 default: | |
| 2026 if (c >= '0' && c <= '9') | |
| 2027 c -= '0'; | |
| 2028 else if (c >= 'a' && c <= 'f') | |
| 2029 c -= 'a' - 10; | |
| 2030 else if (c >= 'A' && c <= 'F') | |
| 2031 c -= 'A' - 10; | |
| 2032 else if (c & 0x80) | |
| 2033 { p--; | |
| 2034 uint u = decodeUTF(); | |
| 2035 p++; | |
| 2036 if (u == PS || u == LS) | |
| 2037 loc.linnum++; | |
| 2038 else | |
| 135 | 2039 error("non-hex character \\u%04x", u); |
| 51 | 2040 } |
| 2041 else | |
| 2042 error("non-hex character '%c'", c); | |
| 2043 if (n & 1) | |
| 2044 { v = (v << 4) | c; | |
| 2045 stringbuffer.writeByte(v); | |
| 2046 } | |
| 2047 else | |
| 2048 v = c; | |
| 2049 n++; | |
| 2050 break; | |
| 2051 } | |
| 2052 } | |
| 0 | 2053 } |
| 2054 | |
| 2055 version (DMDV2) { | |
| 51 | 2056 /************************************** |
| 2057 * Lex delimited strings: | |
| 2058 * q"(foo(xxx))" // "foo(xxx)" | |
| 2059 * q"[foo(]" // "foo(" | |
| 2060 * q"/foo]/" // "foo]" | |
| 2061 * q"HERE | |
| 2062 * foo | |
| 2063 * HERE" // "foo\n" | |
| 2064 * Input: | |
| 2065 * p is on the " | |
| 2066 */ | |
| 0 | 2067 TOK delimitedStringConstant(Token* t) |
| 2068 { | |
| 51 | 2069 uint c; |
| 2070 Loc start = loc; | |
| 2071 uint delimleft = 0; | |
| 2072 uint delimright = 0; | |
| 2073 uint nest = 1; | |
| 2074 uint nestcount; | |
| 2075 Identifier hereid = null; | |
| 2076 uint blankrol = 0; | |
| 2077 uint startline = 0; | |
| 2078 | |
| 2079 p++; | |
| 2080 stringbuffer.reset(); | |
| 2081 while (1) | |
| 2082 { | |
| 2083 c = *p++; | |
| 2084 //printf("c = '%c'\n", c); | |
| 2085 switch (c) | |
| 2086 { | |
| 2087 case '\n': | |
| 2088 Lnextline: | |
| 2089 loc.linnum++; | |
| 2090 startline = 1; | |
| 2091 if (blankrol) | |
| 2092 { blankrol = 0; | |
| 2093 continue; | |
| 2094 } | |
| 2095 if (hereid) | |
| 2096 { | |
| 2097 stringbuffer.writeUTF8(c); | |
| 2098 continue; | |
| 2099 } | |
| 2100 break; | |
| 2101 | |
| 2102 case '\r': | |
| 2103 if (*p == '\n') | |
| 2104 continue; // ignore | |
| 2105 c = '\n'; // treat EndOfLine as \n character | |
| 2106 goto Lnextline; | |
| 2107 | |
| 2108 case 0: | |
| 2109 case 0x1A: | |
| 2110 goto Lerror; | |
| 2111 | |
| 2112 default: | |
| 2113 if (c & 0x80) | |
| 2114 { p--; | |
| 2115 c = decodeUTF(); | |
| 2116 p++; | |
| 2117 if (c == PS || c == LS) | |
| 2118 goto Lnextline; | |
| 2119 } | |
| 2120 break; | |
| 2121 } | |
| 2122 if (delimleft == 0) | |
| 2123 { | |
| 2124 delimleft = c; | |
| 2125 nest = 1; | |
| 2126 nestcount = 1; | |
| 2127 if (c == '(') | |
| 2128 delimright = ')'; | |
| 2129 else if (c == '{') | |
| 2130 delimright = '}'; | |
| 2131 else if (c == '[') | |
| 2132 delimright = ']'; | |
| 2133 else if (c == '<') | |
| 2134 delimright = '>'; | |
| 2135 else if (isalpha(c) || c == '_' || (c >= 0x80 && isUniAlpha(c))) | |
| 2136 { | |
| 2137 // Start of identifier; must be a heredoc | |
| 2138 Token t2; | |
| 2139 p--; | |
| 2140 scan(&t2); // read in heredoc identifier | |
| 2141 if (t2.value != TOKidentifier) | |
| 2142 { | |
| 2143 error("identifier expected for heredoc, not %s", t2.toChars()); | |
| 2144 delimright = c; | |
| 2145 } | |
| 2146 else | |
| 2147 { | |
| 2148 hereid = t2.ident; | |
| 2149 //printf("hereid = '%s'\n", hereid.toChars()); | |
| 2150 blankrol = 1; | |
| 2151 } | |
| 2152 nest = 0; | |
| 2153 } | |
| 2154 else | |
| 2155 { | |
| 2156 delimright = c; | |
| 2157 nest = 0; | |
| 2158 if (isspace(c)) | |
| 2159 error("delimiter cannot be whitespace"); | |
| 2160 } | |
| 2161 } | |
| 2162 else | |
| 2163 { | |
| 2164 if (blankrol) | |
| 2165 { | |
| 2166 error("heredoc rest of line should be blank"); | |
| 2167 blankrol = 0; | |
| 2168 continue; | |
| 2169 } | |
| 2170 if (nest == 1) | |
| 2171 { | |
| 2172 if (c == delimleft) | |
| 2173 nestcount++; | |
| 2174 else if (c == delimright) | |
| 2175 { nestcount--; | |
| 2176 if (nestcount == 0) | |
| 2177 goto Ldone; | |
| 2178 } | |
| 2179 } | |
| 2180 else if (c == delimright) | |
| 2181 goto Ldone; | |
| 2182 if (startline && isalpha(c) && hereid) | |
| 2183 { | |
| 2184 Token t2; | |
| 2185 ubyte* psave = p; | |
| 2186 p--; | |
| 2187 scan(&t2); // read in possible heredoc identifier | |
| 2188 //printf("endid = '%s'\n", t2.ident.toChars()); | |
| 2189 if (t2.value == TOKidentifier && t2.ident.equals(hereid)) | |
| 2190 { | |
| 2191 /* should check that rest of line is blank | |
| 2192 */ | |
| 2193 goto Ldone; | |
| 2194 } | |
| 2195 p = psave; | |
| 2196 } | |
| 2197 stringbuffer.writeUTF8(c); | |
| 2198 startline = 0; | |
| 2199 } | |
| 2200 } | |
| 2201 | |
| 2202 Ldone: | |
| 2203 if (*p == '"') | |
| 2204 p++; | |
| 2205 else | |
| 2206 error("delimited string must end in %c\"", delimright); | |
| 2207 t.len = stringbuffer.offset; | |
| 2208 stringbuffer.writeByte(0); | |
| 2209 void* mem = malloc(stringbuffer.offset); | |
| 2210 memcpy(mem, stringbuffer.data, stringbuffer.offset); | |
| 2211 t.ustring = cast(const(char)*)mem; | |
| 2212 stringPostfix(t); | |
| 2213 return TOKstring; | |
| 2214 | |
| 2215 Lerror: | |
| 2216 error("unterminated string constant starting at %s", start.toChars()); | |
| 2217 t.ustring = "".ptr; | |
| 2218 t.len = 0; | |
| 2219 t.postfix = 0; | |
| 2220 return TOKstring; | |
| 0 | 2221 } |
| 2222 | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2223 /************************************** |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2224 * Lex delimited strings: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2225 * q{ foo(xxx) } // " foo(xxx) " |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2226 * q{foo(} // "foo(" |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2227 * q{{foo}"}"} // "{foo}"}"" |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2228 * Input: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2229 * p is on the q |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2230 */ |
| 0 | 2231 TOK tokenStringConstant(Token* t) |
| 2232 { | |
|
8
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2233 uint nest = 1; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2234 Loc start = loc; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2235 ubyte* pstart = ++p; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2236 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2237 while (true) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2238 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2239 Token tok; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2240 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2241 scan(&tok); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2242 switch (tok.value) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2243 { |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2244 case TOKlcurly: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2245 nest++; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2246 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2247 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2248 case TOKrcurly: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2249 if (--nest == 0) |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2250 goto Ldone; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2251 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2252 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2253 case TOKeof: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2254 goto Lerror; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2255 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2256 default: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2257 continue; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2258 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2259 } |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2260 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2261 Ldone: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2262 t.len = p - 1 - pstart; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2263 char* tmp = cast(char*)GC.malloc(t.len + 1); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2264 memcpy(tmp, pstart, t.len); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2265 tmp[t.len] = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2266 t.ustring = tmp; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2267 stringPostfix(t); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2268 return TOKstring; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2269 |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2270 Lerror: |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2271 error("unterminated token string constant starting at %s", start.toChars()); |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2272 t.ustring = "".ptr; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2273 t.len = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2274 t.postfix = 0; |
|
d42cd5917df4
wysiwyg strings, alias this, templates, TypeSlice implementation
dkoroskin <>
parents:
4
diff
changeset
|
2275 return TOKstring; |
| 0 | 2276 } |
| 2277 } | |
| 2278 TOK escapeStringConstant(Token* t, int wide) | |
| 2279 { | |
| 2280 uint c; | |
| 2281 Loc start = loc; | |
| 2282 | |
| 2283 p++; | |
| 2284 stringbuffer.reset(); | |
| 2285 while (true) | |
| 2286 { | |
| 2287 c = *p++; | |
| 2288 switch (c) | |
| 2289 { | |
| 2290 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 2291 } else { | |
| 2292 case '\\': | |
| 2293 switch (*p) | |
| 2294 { | |
| 2295 case 'u': | |
| 2296 case 'U': | |
| 2297 case '&': | |
| 2298 c = escapeSequence(); | |
| 2299 stringbuffer.writeUTF8(c); | |
| 2300 continue; | |
| 2301 | |
| 2302 default: | |
| 2303 c = escapeSequence(); | |
| 2304 break; | |
| 2305 } | |
| 2306 break; | |
| 2307 } | |
| 2308 case '\n': | |
| 2309 loc.linnum++; | |
| 2310 break; | |
| 2311 | |
| 2312 case '\r': | |
| 2313 if (*p == '\n') | |
| 2314 continue; // ignore | |
| 2315 c = '\n'; // treat EndOfLine as \n character | |
| 2316 loc.linnum++; | |
| 2317 break; | |
| 2318 | |
| 2319 case '"': | |
| 2320 t.len = stringbuffer.offset; | |
| 2321 stringbuffer.writeByte(0); | |
| 2 | 2322 char* tmp = cast(char*)GC.malloc(stringbuffer.offset); |
| 0 | 2323 memcpy(tmp, stringbuffer.data, stringbuffer.offset); |
| 2324 t.ustring = tmp; | |
| 2325 stringPostfix(t); | |
| 2326 return TOK.TOKstring; | |
| 2327 | |
| 2328 case 0: | |
| 2329 case 0x1A: | |
| 2330 p--; | |
| 2331 error("unterminated string constant starting at %s", start.toChars()); | |
| 2332 t.ustring = "".ptr; | |
| 2333 t.len = 0; | |
| 2334 t.postfix = 0; | |
| 2335 return TOK.TOKstring; | |
| 2336 | |
| 2337 default: | |
| 2338 if (c & 0x80) | |
| 2339 { | |
| 2340 p--; | |
| 2341 c = decodeUTF(); | |
| 2342 if (c == LS || c == PS) | |
| 2343 { c = '\n'; | |
| 2344 loc.linnum++; | |
| 2345 } | |
| 2346 p++; | |
| 2347 stringbuffer.writeUTF8(c); | |
| 2348 continue; | |
| 2349 } | |
| 2350 break; | |
| 2351 } | |
| 2352 stringbuffer.writeByte(c); | |
| 2353 } | |
| 2354 | |
| 2355 assert(false); | |
| 2356 } | |
| 2357 | |
| 2358 TOK charConstant(Token* t, int wide) | |
| 2359 { | |
| 2360 uint c; | |
| 2361 TOK tk = TOKcharv; | |
| 2362 | |
| 2363 //printf("Lexer.charConstant\n"); | |
| 2364 p++; | |
| 2365 c = *p++; | |
| 2366 switch (c) | |
| 2367 { | |
| 2368 version (TEXTUAL_ASSEMBLY_OUT) { | |
| 2369 } else { | |
| 2370 case '\\': | |
| 2371 switch (*p) | |
| 2372 { | |
| 2373 case 'u': | |
| 2374 t.uns64value = escapeSequence(); | |
| 2375 tk = TOKwcharv; | |
| 2376 break; | |
| 2377 | |
| 2378 case 'U': | |
| 2379 case '&': | |
| 2380 t.uns64value = escapeSequence(); | |
| 2381 tk = TOKdcharv; | |
| 2382 break; | |
| 2383 | |
| 2384 default: | |
| 2385 t.uns64value = escapeSequence(); | |
| 2386 break; | |
| 2387 } | |
| 2388 break; | |
| 2389 } | |
| 2390 case '\n': | |
| 2391 L1: | |
| 2392 loc.linnum++; | |
| 2393 case '\r': | |
| 2394 case 0: | |
| 2395 case 0x1A: | |
| 2396 case '\'': | |
| 2397 error("unterminated character constant"); | |
| 2398 return tk; | |
| 2399 | |
| 2400 default: | |
| 2401 if (c & 0x80) | |
| 2402 { | |
| 2403 p--; | |
| 2404 c = decodeUTF(); | |
| 2405 p++; | |
| 2406 if (c == LS || c == PS) | |
| 2407 goto L1; | |
| 2408 if (c < 0xD800 || (c >= 0xE000 && c < 0xFFFE)) | |
| 2409 tk = TOKwcharv; | |
| 2410 else | |
| 2411 tk = TOKdcharv; | |
| 2412 } | |
| 2413 t.uns64value = c; | |
| 2414 break; | |
| 2415 } | |
| 2416 | |
| 2417 if (*p != '\'') | |
| 2418 { | |
| 2419 error("unterminated character constant"); | |
| 2420 return tk; | |
| 2421 } | |
| 2422 p++; | |
| 2423 return tk; | |
| 2424 } | |
| 2425 | |
| 2426 /*************************************** | |
| 2427 * Get postfix of string literal. | |
| 2428 */ | |
| 2429 void stringPostfix(Token* t) | |
| 2430 { | |
| 2431 switch (*p) | |
| 2432 { | |
| 2433 case 'c': | |
| 2434 case 'w': | |
| 2435 case 'd': | |
| 2436 t.postfix = *p; | |
| 2437 p++; | |
| 2438 break; | |
| 2439 | |
| 2440 default: | |
| 2441 t.postfix = 0; | |
| 2442 break; | |
| 2443 } | |
| 2444 } | |
| 2445 | |
| 2446 uint wchar_(uint u) | |
| 2447 { | |
| 2448 assert(false); | |
| 2449 } | |
| 2450 | |
| 2451 /************************************** | |
| 2452 * Read in a number. | |
| 2453 * If it's an integer, store it in tok.TKutok.Vlong. | |
| 2454 * integers can be decimal, octal or hex | |
| 2455 * Handle the suffixes U, UL, LU, L, etc. | |
| 2456 * If it's double, store it in tok.TKutok.Vdouble. | |
| 2457 * Returns: | |
| 2458 * TKnum | |
| 2459 * TKdouble,... | |
| 2460 */ | |
| 2461 | |
| 2462 TOK number(Token* t) | |
| 2463 { | |
| 2464 // We use a state machine to collect numbers | |
| 2465 enum STATE { STATE_initial, STATE_0, STATE_decimal, STATE_octal, STATE_octale, | |
| 2466 STATE_hex, STATE_binary, STATE_hex0, STATE_binary0, | |
| 2467 STATE_hexh, STATE_error }; | |
| 2468 STATE state; | |
| 2469 | |
| 2470 enum FLAGS | |
| 2471 { | |
| 2472 FLAGS_undefined = 0, | |
| 2473 FLAGS_decimal = 1, // decimal | |
| 2474 FLAGS_unsigned = 2, // u or U suffix | |
| 2475 FLAGS_long = 4, // l or L suffix | |
| 2476 }; | |
| 2477 | |
| 2478 FLAGS flags = FLAGS.FLAGS_decimal; | |
| 2479 | |
| 2480 int i; | |
| 2481 int base; | |
| 2482 uint c; | |
| 2483 ubyte *start; | |
| 2484 TOK result; | |
| 2485 | |
| 2486 //printf("Lexer.number()\n"); | |
| 2487 state = STATE.STATE_initial; | |
| 2488 base = 0; | |
| 2489 stringbuffer.reset(); | |
| 2490 start = p; | |
| 2491 while (1) | |
| 2492 { | |
| 2493 c = *p; | |
| 2494 switch (state) | |
| 2495 { | |
| 2496 case STATE.STATE_initial: // opening state | |
| 2497 if (c == '0') | |
| 2498 state = STATE.STATE_0; | |
| 2499 else | |
| 2500 state = STATE.STATE_decimal; | |
| 2501 break; | |
| 2502 | |
| 2503 case STATE.STATE_0: | |
| 2504 flags = (flags & ~FLAGS.FLAGS_decimal); | |
| 2505 switch (c) | |
| 2506 { | |
| 2507 version (ZEROH) { | |
| 2508 case 'H': // 0h | |
| 2509 case 'h': | |
| 2510 goto hexh; | |
| 2511 } | |
| 2512 case 'X': | |
| 2513 case 'x': | |
| 2514 state = STATE.STATE_hex0; | |
| 2515 break; | |
| 2516 | |
| 2517 case '.': | |
| 2518 if (p[1] == '.') // .. is a separate token | |
| 2519 goto done; | |
| 2520 case 'i': | |
| 2521 case 'f': | |
| 2522 case 'F': | |
| 2523 goto real_; | |
| 2524 version (ZEROH) { | |
| 2525 case 'E': | |
| 2526 case 'e': | |
| 2527 goto case_hex; | |
| 2528 } | |
| 2529 case 'B': | |
| 2530 case 'b': | |
| 2531 state = STATE.STATE_binary0; | |
| 2532 break; | |
| 2533 | |
| 2534 case '0': case '1': case '2': case '3': | |
| 2535 case '4': case '5': case '6': case '7': | |
| 2536 state = STATE.STATE_octal; | |
| 2537 break; | |
| 2538 | |
| 2539 version (ZEROH) { | |
| 2540 case '8': case '9': case 'A': | |
| 2541 case 'C': case 'D': case 'F': | |
| 2542 case 'a': case 'c': case 'd': case 'f': | |
| 2543 case_hex: | |
| 2544 state = STATE.STATE_hexh; | |
| 2545 break; | |
| 2546 } | |
| 2547 case '_': | |
| 2548 state = STATE.STATE_octal; | |
| 2549 p++; | |
| 2550 continue; | |
| 2551 | |
| 2552 case 'L': | |
| 2553 if (p[1] == 'i') | |
| 2554 goto real_; | |
| 2555 goto done; | |
| 2556 | |
| 2557 default: | |
| 2558 goto done; | |
| 2559 } | |
| 2560 break; | |
| 2561 | |
| 2562 case STATE.STATE_decimal: // reading decimal number | |
| 2563 if (!isdigit(c)) | |
| 2564 { | |
| 2565 version (ZEROH) { | |
| 2566 if (ishex(c) | |
| 2567 || c == 'H' || c == 'h' | |
| 2568 ) | |
| 2569 goto hexh; | |
| 2570 } | |
| 2571 if (c == '_') // ignore embedded _ | |
| 2572 { p++; | |
| 2573 continue; | |
| 2574 } | |
| 2575 if (c == '.' && p[1] != '.') | |
| 2576 goto real_; | |
| 2577 else if (c == 'i' || c == 'f' || c == 'F' || | |
| 2578 c == 'e' || c == 'E') | |
| 2579 { | |
| 2580 real_: // It's a real number. Back up and rescan as a real | |
| 2581 p = start; | |
| 2582 return inreal(t); | |
| 2583 } | |
| 2584 else if (c == 'L' && p[1] == 'i') | |
| 2585 goto real_; | |
| 2586 goto done; | |
| 2587 } | |
| 2588 break; | |
| 2589 | |
| 2590 case STATE.STATE_hex0: // reading hex number | |
| 2591 case STATE.STATE_hex: | |
| 2592 if (! ishex(cast(ubyte)c)) | |
| 2593 { | |
| 2594 if (c == '_') // ignore embedded _ | |
| 2595 { p++; | |
| 2596 continue; | |
| 2597 } | |
| 2598 if (c == '.' && p[1] != '.') | |
| 2599 goto real_; | |
| 2600 if (c == 'P' || c == 'p' || c == 'i') | |
| 2601 goto real_; | |
| 2602 if (state == STATE.STATE_hex0) | |
| 2603 error("Hex digit expected, not '%c'", c); | |
| 2604 goto done; | |
| 2605 } | |
| 2606 state = STATE.STATE_hex; | |
| 2607 break; | |
| 2608 | |
| 2609 version (ZEROH) { | |
| 2610 hexh: | |
| 2611 state = STATE.STATE_hexh; | |
| 2612 case STATE.STATE_hexh: // parse numbers like 0FFh | |
| 2613 if (!ishex(c)) | |
| 2614 { | |
| 2615 if (c == 'H' || c == 'h') | |
| 2616 { | |
| 2617 p++; | |
| 2618 base = 16; | |
| 2619 goto done; | |
| 2620 } | |
| 2621 else | |
| 2622 { | |
| 2623 // Check for something like 1E3 or 0E24 | |
| 2624 if (memchr(cast(char*)stringbuffer.data, 'E', stringbuffer.offset) || | |
| 2625 memchr(cast(char*)stringbuffer.data, 'e', stringbuffer.offset)) | |
| 2626 goto real_; | |
| 2627 error("Hex digit expected, not '%c'", c); | |
| 2628 goto done; | |
| 2629 } | |
| 2630 } | |
| 2631 break; | |
| 2632 } | |
| 2633 | |
| 2634 case STATE.STATE_octal: // reading octal number | |
| 2635 case STATE.STATE_octale: // reading octal number with non-octal digits | |
| 2636 if (!isoctal(cast(ubyte)c)) | |
| 2637 { | |
| 2638 version (ZEROH) { | |
| 2639 if (ishex(c) | |
| 2640 || c == 'H' || c == 'h' | |
| 2641 ) | |
| 2642 goto hexh; | |
| 2643 } | |
| 2644 if (c == '_') // ignore embedded _ | |
| 2645 { p++; | |
| 2646 continue; | |
| 2647 } | |
| 2648 if (c == '.' && p[1] != '.') | |
| 2649 goto real_; | |
| 2650 if (c == 'i') | |
| 2651 goto real_; | |
| 2652 if (isdigit(c)) | |
| 2653 { | |
| 2654 state = STATE.STATE_octale; | |
| 2655 } | |
| 2656 else | |
| 2657 goto done; | |
| 2658 } | |
| 2659 break; | |
| 2660 | |
| 2661 case STATE.STATE_binary0: // starting binary number | |
| 2662 case STATE.STATE_binary: // reading binary number | |
| 2663 if (c != '0' && c != '1') | |
| 2664 { | |
| 2665 version (ZEROH) { | |
| 2666 if (ishex(c) | |
| 2667 || c == 'H' || c == 'h' | |
| 2668 ) | |
| 2669 goto hexh; | |
| 2670 } | |
| 2671 if (c == '_') // ignore embedded _ | |
| 2672 { p++; | |
| 2673 continue; | |
| 2674 } | |
| 2675 if (state == STATE.STATE_binary0) | |
| 2676 { error("binary digit expected"); | |
| 2677 state = STATE.STATE_error; | |
| 2678 break; | |
| 2679 } | |
| 2680 else | |
| 2681 goto done; | |
| 2682 } | |
| 2683 state = STATE.STATE_binary; | |
| 2684 break; | |
| 2685 | |
| 2686 case STATE.STATE_error: // for error recovery | |
| 2687 if (!isdigit(c)) // scan until non-digit | |
| 2688 goto done; | |
| 2689 break; | |
| 2690 | |
| 2691 default: | |
| 2692 assert(0); | |
| 2693 } | |
| 2694 stringbuffer.writeByte(c); | |
| 2695 p++; | |
| 2696 } | |
| 2697 done: | |
| 2698 stringbuffer.writeByte(0); // terminate string | |
| 2699 if (state == STATE.STATE_octale) | |
| 2700 error("Octal digit expected"); | |
| 2701 | |
| 2702 ulong n; // unsigned >=64 bit integer type | |
| 2703 | |
| 2704 if (stringbuffer.offset == 2 && (state == STATE.STATE_decimal || state == STATE.STATE_0)) | |
| 2705 n = stringbuffer.data[0] - '0'; | |
| 2706 else | |
| 2707 { | |
| 2708 // Convert string to integer | |
| 2709 version (__DMC__) { | |
| 2710 errno = 0; | |
| 2711 n = strtoull(cast(char*)stringbuffer.data,null,base); | |
| 2712 if (errno == ERANGE) | |
| 2713 error("integer overflow"); | |
| 2714 } else { | |
| 2715 // Not everybody implements strtoull() | |
| 2716 char* p = cast(char*)stringbuffer.data; | |
| 2717 int r = 10, d; | |
| 2718 | |
| 2719 if (*p == '0') | |
| 2720 { | |
| 2721 if (p[1] == 'x' || p[1] == 'X') | |
| 2722 p += 2, r = 16; | |
| 2723 else if (p[1] == 'b' || p[1] == 'B') | |
| 2724 p += 2, r = 2; | |
| 2725 else if (isdigit(p[1])) | |
| 2726 p += 1, r = 8; | |
| 2727 } | |
| 2728 | |
| 2729 n = 0; | |
| 2730 while (1) | |
| 2731 { | |
| 2732 if (*p >= '0' && *p <= '9') | |
| 2733 d = *p - '0'; | |
| 2734 else if (*p >= 'a' && *p <= 'z') | |
| 2735 d = *p - 'a' + 10; | |
| 2736 else if (*p >= 'A' && *p <= 'Z') | |
| 2737 d = *p - 'A' + 10; | |
| 2738 else | |
| 2739 break; | |
| 2740 if (d >= r) | |
| 2741 break; | |
| 2742 ulong n2 = n * r; | |
| 2743 //printf("n2 / r = %llx, n = %llx\n", n2/r, n); | |
| 2744 if (n2 / r != n || n2 + d < n) | |
| 2745 { | |
| 2746 error ("integer overflow"); | |
| 2747 break; | |
| 2748 } | |
| 2749 | |
| 2750 n = n2 + d; | |
| 2751 p++; | |
| 2752 } | |
| 2753 } | |
| 2754 if (n.sizeof > 8 && | |
| 2755 n > 0xFFFFFFFFFFFFFFFF) // if n needs more than 64 bits | |
| 2756 error("integer overflow"); | |
| 2757 } | |
| 2758 | |
| 2759 // Parse trailing 'u', 'U', 'l' or 'L' in any combination | |
| 2760 while (1) | |
| 2761 { FLAGS f; | |
| 2762 | |
| 2763 switch (*p) | |
| 2764 { case 'U': | |
| 2765 case 'u': | |
| 2766 f = FLAGS.FLAGS_unsigned; | |
| 2767 goto L1; | |
| 2768 | |
| 2769 case 'l': | |
| 2770 if (1 || !global.params.useDeprecated) | |
| 2771 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2772 case 'L': | |
| 2773 f = FLAGS.FLAGS_long; | |
| 2774 L1: | |
| 2775 p++; | |
| 2776 if (flags & f) | |
| 2777 error("unrecognized token"); | |
| 2778 flags = (flags | f); | |
| 2779 continue; | |
| 2780 default: | |
| 2781 break; | |
| 2782 } | |
| 2783 break; | |
| 2784 } | |
| 2785 | |
| 2786 switch (flags) | |
| 2787 { | |
| 2788 case FLAGS.FLAGS_undefined: | |
| 2789 /* Octal or Hexadecimal constant. | |
| 2790 * First that fits: int, uint, long, ulong | |
| 2791 */ | |
| 2792 if (n & 0x8000000000000000) | |
| 2793 result = TOK.TOKuns64v; | |
| 2794 else if (n & 0xFFFFFFFF00000000) | |
| 2795 result = TOK.TOKint64v; | |
| 2796 else if (n & 0x80000000) | |
| 2797 result = TOK.TOKuns32v; | |
| 2798 else | |
| 2799 result = TOK.TOKint32v; | |
| 2800 break; | |
| 2801 | |
| 2802 case FLAGS.FLAGS_decimal: | |
| 2803 /* First that fits: int, long, long long | |
| 2804 */ | |
| 2805 if (n & 0x8000000000000000) | |
| 2806 { error("signed integer overflow"); | |
| 2807 result = TOK.TOKuns64v; | |
| 2808 } | |
| 2809 else if (n & 0xFFFFFFFF80000000) | |
| 2810 result = TOK.TOKint64v; | |
| 2811 else | |
| 2812 result = TOK.TOKint32v; | |
| 2813 break; | |
| 2814 | |
| 2815 case FLAGS.FLAGS_unsigned: | |
| 2816 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_unsigned: | |
| 2817 /* First that fits: uint, ulong | |
| 2818 */ | |
| 2819 if (n & 0xFFFFFFFF00000000) | |
| 2820 result = TOK.TOKuns64v; | |
| 2821 else | |
| 2822 result = TOK.TOKuns32v; | |
| 2823 break; | |
| 2824 | |
| 2825 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_long: | |
| 2826 if (n & 0x8000000000000000) | |
| 2827 { error("signed integer overflow"); | |
| 2828 result = TOK.TOKuns64v; | |
| 2829 } | |
| 2830 else | |
| 2831 result = TOK.TOKint64v; | |
| 2832 break; | |
| 2833 | |
| 2834 case FLAGS.FLAGS_long: | |
| 2835 if (n & 0x8000000000000000) | |
| 2836 result = TOK.TOKuns64v; | |
| 2837 else | |
| 2838 result = TOK.TOKint64v; | |
| 2839 break; | |
| 2840 | |
| 2841 case FLAGS.FLAGS_unsigned | FLAGS.FLAGS_long: | |
| 2842 case FLAGS.FLAGS_decimal | FLAGS.FLAGS_unsigned | FLAGS.FLAGS_long: | |
| 2843 result = TOK.TOKuns64v; | |
| 2844 break; | |
| 2845 | |
| 2846 default: | |
| 2847 debug { | |
| 2848 printf("%x\n",flags); | |
| 2849 } | |
| 2850 assert(0); | |
| 2851 } | |
| 2852 t.uns64value = n; | |
| 2853 return result; | |
| 2854 } | |
| 2855 | |
| 2856 /************************************** | |
| 2857 * Read in characters, converting them to real. | |
| 2858 * Bugs: | |
| 2859 * Exponent overflow not detected. | |
| 2860 * Too much requested precision is not detected. | |
| 2861 */ | |
| 2862 TOK inreal(Token* t) | |
| 2863 in | |
| 2864 { | |
| 2865 assert(*p == '.' || isdigit(*p)); | |
| 2866 } | |
| 2867 out (result) | |
| 2868 { | |
| 2869 switch (result) | |
| 2870 { | |
| 2871 case TOKfloat32v: | |
| 2872 case TOKfloat64v: | |
| 2873 case TOKfloat80v: | |
| 2874 case TOKimaginary32v: | |
| 2875 case TOKimaginary64v: | |
| 2876 case TOKimaginary80v: | |
| 2877 break; | |
| 2878 | |
| 2879 default: | |
| 2880 assert(0); | |
| 2881 } | |
| 2882 } | |
| 2883 body | |
| 2884 { | |
| 2885 int dblstate; | |
| 2886 uint c; | |
| 2887 char hex; // is this a hexadecimal-floating-constant? | |
| 2888 TOK result; | |
| 2889 | |
| 2890 //printf("Lexer.inreal()\n"); | |
| 2891 stringbuffer.reset(); | |
| 2892 dblstate = 0; | |
| 2893 hex = 0; | |
| 2894 Lnext: | |
| 2895 while (true) | |
| 2896 { | |
| 2897 // Get next char from input | |
| 2898 c = *p++; | |
| 2899 //printf("dblstate = %d, c = '%c'\n", dblstate, c); | |
| 2900 while (true) | |
| 2901 { | |
| 2902 switch (dblstate) | |
| 2903 { | |
| 2904 case 0: // opening state | |
| 2905 if (c == '0') | |
| 2906 dblstate = 9; | |
| 2907 else if (c == '.') | |
| 2908 dblstate = 3; | |
| 2909 else | |
| 2910 dblstate = 1; | |
| 2911 break; | |
| 2912 | |
| 2913 case 9: | |
| 2914 dblstate = 1; | |
| 2915 if (c == 'X' || c == 'x') | |
| 2916 { | |
| 2917 hex++; | |
| 2918 break; | |
| 2919 } | |
| 2920 case 1: // digits to left of . | |
| 2921 case 3: // digits to right of . | |
| 2922 case 7: // continuing exponent digits | |
| 2923 if (!isdigit(c) && !(hex && isxdigit(c))) | |
| 2924 { | |
| 2925 if (c == '_') | |
| 2926 goto Lnext; // ignore embedded '_' | |
| 2927 dblstate++; | |
| 2928 continue; | |
| 2929 } | |
| 2930 break; | |
| 2931 | |
| 2932 case 2: // no more digits to left of . | |
| 2933 if (c == '.') | |
| 2934 { | |
| 2935 dblstate++; | |
| 2936 break; | |
| 2937 } | |
| 2938 case 4: // no more digits to right of . | |
| 2939 if ((c == 'E' || c == 'e') || | |
| 2940 hex && (c == 'P' || c == 'p')) | |
| 2941 { | |
| 2942 dblstate = 5; | |
| 2943 hex = 0; // exponent is always decimal | |
| 2944 break; | |
| 2945 } | |
| 2946 if (hex) | |
| 2947 error("binary-exponent-part required"); | |
| 2948 goto done; | |
| 2949 | |
| 2950 case 5: // looking immediately to right of E | |
| 2951 dblstate++; | |
| 2952 if (c == '-' || c == '+') | |
| 2953 break; | |
| 2954 case 6: // 1st exponent digit expected | |
| 2955 if (!isdigit(c)) | |
| 2956 error("exponent expected"); | |
| 2957 dblstate++; | |
| 2958 break; | |
| 2959 | |
| 2960 case 8: // past end of exponent digits | |
| 2961 goto done; | |
| 79 | 2962 |
| 2963 default: | |
| 2964 assert(0, "inreal.dblstate has unexpected value"); | |
| 0 | 2965 } |
| 2966 break; | |
| 2967 } | |
| 2968 stringbuffer.writeByte(c); | |
| 2969 } | |
| 2970 done: | |
| 2971 p--; | |
| 2972 | |
| 2973 stringbuffer.writeByte(0); | |
| 2974 | |
| 114 | 2975 version (Windows) { /// && __DMC__ |
| 0 | 2976 char* save = __locale_decpoint; |
| 2977 __locale_decpoint = cast(char*)".".ptr; | |
| 2978 } | |
| 2979 t.float80value = strtold(cast(char*)stringbuffer.data, null); | |
| 2980 | |
| 2981 errno = 0; | |
| 2982 switch (*p) | |
| 2983 { | |
| 2984 case 'F': | |
| 2985 case 'f': | |
| 2986 strtof(cast(char*)stringbuffer.data, null); | |
| 2987 result = TOKfloat32v; | |
| 2988 p++; | |
| 2989 break; | |
| 2990 | |
| 2991 default: | |
| 2992 strtod(cast(char*)stringbuffer.data, null); | |
| 2993 result = TOKfloat64v; | |
| 2994 break; | |
| 2995 | |
| 2996 case 'l': | |
| 2997 if (!global.params.useDeprecated) | |
| 2998 error("'l' suffix is deprecated, use 'L' instead"); | |
| 2999 case 'L': | |
| 3000 result = TOKfloat80v; | |
| 3001 p++; | |
| 3002 break; | |
| 3003 } | |
| 3004 if (*p == 'i' || *p == 'I') | |
| 3005 { | |
| 3006 if (!global.params.useDeprecated && *p == 'I') | |
| 3007 error("'I' suffix is deprecated, use 'i' instead"); | |
| 3008 p++; | |
| 3009 switch (result) | |
| 3010 { | |
| 3011 case TOKfloat32v: | |
| 3012 result = TOKimaginary32v; | |
| 3013 break; | |
| 3014 case TOKfloat64v: | |
| 3015 result = TOKimaginary64v; | |
| 3016 break; | |
| 3017 case TOKfloat80v: | |
| 3018 result = TOKimaginary80v; | |
| 3019 break; | |
|
157
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3020 default: |
| 0 | 3021 } |
| 3022 } | |
| 3023 | |
| 114 | 3024 version (Windows) { ///&& __DMC__ |
| 0 | 3025 __locale_decpoint = save; |
| 3026 } | |
| 3027 if (errno == ERANGE) | |
| 3028 error("number is not representable"); | |
| 3029 | |
| 3030 return result; | |
| 3031 } | |
| 3032 | |
| 3033 void error(T...)(string format, T t) | |
| 3034 { | |
| 3035 error(this.loc, format, t); | |
| 3036 } | |
| 3037 | |
| 3038 void error(T...)(Loc loc, string format, T t) | |
| 3039 { | |
| 3040 if (mod && !global.gag) | |
| 3041 { | |
| 3042 string p = loc.toChars(); | |
| 3043 if (p.length != 0) | |
| 3044 writef("%s: ", p); | |
| 3045 | |
| 3046 writefln(format, t); | |
| 3047 | |
| 3048 if (global.errors >= 20) // moderate blizzard of cascading messages | |
| 3049 fatal(); | |
| 3050 } | |
| 3051 | |
| 3052 global.errors++; | |
| 3053 } | |
| 3054 | |
| 162 | 3055 /********************************************* |
| 3056 * Do pragma. | |
| 3057 * Currently, the only pragma supported is: | |
| 3058 * #line linnum [filespec] | |
| 3059 */ | |
| 0 | 3060 void pragma_() |
| 3061 { | |
| 162 | 3062 Token tok; |
| 3063 int linnum; | |
| 3064 string filespec = null; | |
| 3065 Loc loc = this.loc; | |
| 3066 | |
| 3067 scan(&tok); | |
| 3068 if (tok.value != TOKidentifier || tok.ident != Id.line) | |
| 3069 goto Lerr; | |
| 3070 | |
| 3071 scan(&tok); | |
| 3072 if (tok.value == TOKint32v || tok.value == TOKint64v) | |
| 3073 linnum = cast(int)(tok.uns64value - 1); /// | |
| 3074 else | |
| 3075 goto Lerr; | |
| 3076 | |
| 3077 while (1) | |
| 3078 { | |
| 3079 switch (*p) | |
| 3080 { | |
| 3081 case 0: | |
| 3082 case 0x1A: | |
| 3083 case '\n': | |
| 3084 Lnewline: | |
| 3085 this.loc.linnum = linnum; | |
| 3086 if (filespec != null) | |
| 3087 this.loc.filename = filespec; | |
| 3088 return; | |
| 3089 | |
| 3090 case '\r': | |
| 3091 p++; | |
| 3092 if (*p != '\n') | |
| 3093 { p--; | |
| 3094 goto Lnewline; | |
| 3095 } | |
| 3096 continue; | |
| 3097 | |
| 3098 case ' ': | |
| 3099 case '\t': | |
| 3100 case '\v': | |
| 3101 case '\f': | |
| 3102 p++; | |
| 3103 continue; // skip white space | |
| 3104 | |
| 3105 case '_': | |
| 3106 if (mod && memcmp(p, "__FILE__".ptr, 8) == 0) | |
| 3107 { | |
| 3108 p += 8; | |
| 3109 filespec = (loc.filename ? loc.filename : mod.ident.toChars()); | |
| 3110 } | |
| 3111 continue; | |
| 3112 | |
| 3113 case '"': | |
| 3114 if (filespec) | |
| 3115 goto Lerr; | |
| 3116 stringbuffer.reset(); | |
| 3117 p++; | |
| 3118 while (1) | |
| 3119 { | |
| 3120 uint c; | |
| 3121 | |
| 3122 c = *p; | |
| 3123 switch (c) | |
| 3124 { | |
| 3125 case '\n': | |
| 3126 case '\r': | |
| 3127 case 0: | |
| 3128 case 0x1A: | |
| 3129 goto Lerr; | |
| 3130 | |
| 3131 case '"': | |
| 3132 stringbuffer.writeByte(0); | |
| 3133 filespec = stringbuffer.extractString(); /// | |
| 3134 p++; | |
| 3135 break; | |
| 3136 | |
| 3137 default: | |
| 3138 if (c & 0x80) | |
| 3139 { | |
| 3140 uint u = decodeUTF(); | |
| 3141 if (u == PS || u == LS) | |
| 3142 goto Lerr; | |
| 3143 } | |
| 3144 stringbuffer.writeByte(c); | |
| 3145 p++; | |
| 3146 continue; | |
| 3147 } | |
| 3148 break; | |
| 3149 } | |
| 3150 continue; | |
| 3151 | |
| 3152 default: | |
| 3153 if (*p & 0x80) | |
| 3154 { | |
| 3155 uint u = decodeUTF(); | |
| 3156 if (u == PS || u == LS) | |
| 3157 goto Lnewline; | |
| 3158 } | |
| 3159 goto Lerr; | |
| 3160 } | |
| 3161 } | |
| 3162 | |
| 3163 Lerr: | |
| 3164 error(loc, "#line integer [\"filespec\"]\\n expected"); | |
| 0 | 3165 } |
| 3166 | |
| 49 | 3167 /******************************************** |
| 3168 * Decode UTF character. | |
| 3169 * Issue error messages for invalid sequences. | |
| 3170 * Return decoded character, advance p to last character in UTF sequence. | |
| 3171 */ | |
| 0 | 3172 uint decodeUTF() |
| 3173 { | |
| 49 | 3174 dchar u; |
| 3175 ubyte c; | |
| 3176 ubyte* s = p; | |
| 3177 size_t len; | |
| 3178 size_t idx; | |
| 3179 string msg; | |
| 3180 | |
| 3181 c = *s; | |
| 3182 assert(c & 0x80); | |
| 3183 | |
| 3184 // Check length of remaining string up to 6 UTF-8 characters | |
| 3185 for (len = 1; len < 6 && s[len]; len++) { | |
| 3186 ; | |
| 3187 } | |
| 3188 | |
| 3189 idx = 0; | |
| 3190 msg = utf_decodeChar(cast(string)s[0..len], &idx, &u); | |
| 3191 p += idx - 1; | |
| 3192 if (msg) | |
| 3193 { | |
| 3194 error("%s", msg); | |
| 3195 } | |
| 3196 return u; | |
| 0 | 3197 } |
| 3198 | |
|
157
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3199 /*************************************************** |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3200 * Parse doc comment embedded between t.ptr and p. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3201 * Remove trailing blanks and tabs from lines. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3202 * Replace all newlines with \n. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3203 * Remove leading comment character from each line. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3204 * Decide if it's a lineComment or a blockComment. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3205 * Append to previous one for this token. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3206 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3207 void getDocComment(Token* t, uint lineComment) |
| 0 | 3208 { |
|
157
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3209 /* ct tells us which kind of comment it is: '!', '/', '*', or '+' |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3210 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3211 ubyte ct = t.ptr[2]; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3212 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3213 /* Start of comment text skips over / * *, / + +, or / / / |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3214 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3215 ubyte* q = t.ptr + 3; // start of comment text |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3216 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3217 ubyte* qend = p; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3218 if (ct == '*' || ct == '+') |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3219 qend -= 2; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3220 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3221 /* Scan over initial row of ****'s or ++++'s or ////'s |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3222 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3223 for (; q < qend; q++) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3224 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3225 if (*q != ct) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3226 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3227 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3228 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3229 /* Remove trailing row of ****'s or ++++'s |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3230 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3231 if (ct != '/' && ct != '!') |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3232 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3233 for (; q < qend; qend--) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3234 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3235 if (qend[-1] != ct) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3236 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3237 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3238 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3239 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3240 /* Comment is now [q .. qend]. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3241 * Canonicalize it into buf[]. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3242 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3243 OutBuffer buf = new OutBuffer; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3244 int linestart = 0; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3245 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3246 for (; q < qend; q++) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3247 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3248 ubyte c = *q; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3249 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3250 switch (c) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3251 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3252 case '*': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3253 case '+': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3254 if (linestart && c == ct) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3255 { linestart = 0; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3256 /* Trim preceding whitespace up to preceding \n |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3257 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3258 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3259 buf.offset--; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3260 continue; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3261 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3262 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3263 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3264 case ' ': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3265 case '\t': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3266 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3267 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3268 case '\r': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3269 if (q[1] == '\n') |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3270 continue; // skip the \r |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3271 goto Lnewline; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3272 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3273 default: |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3274 if (c == 226) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3275 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3276 // If LS or PS |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3277 if (q[1] == 128 && |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3278 (q[2] == 168 || q[2] == 169)) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3279 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3280 q += 2; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3281 goto Lnewline; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3282 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3283 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3284 linestart = 0; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3285 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3286 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3287 Lnewline: |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3288 c = '\n'; // replace all newlines with \n |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3289 case '\n': |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3290 linestart = 1; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3291 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3292 /* Trim trailing whitespace |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3293 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3294 while (buf.offset && (buf.data[buf.offset - 1] == ' ' || buf.data[buf.offset - 1] == '\t')) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3295 buf.offset--; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3296 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3297 break; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3298 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3299 buf.writeByte(c); |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3300 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3301 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3302 // Always end with a newline |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3303 if (!buf.offset || buf.data[buf.offset - 1] != '\n') |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3304 buf.writeByte('\n'); |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3305 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3306 buf.writeByte(0); |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3307 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3308 // It's a line comment if the start of the doc comment comes |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3309 // after other non-whitespace on the same line. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3310 string* dc = (lineComment && anyToken) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3311 ? &t.lineComment |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3312 : &t.blockComment; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3313 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3314 // Combine with previous doc comment, if any |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3315 if (*dc) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3316 *dc = combineComments(*dc, cast(string) buf.data[0 .. buf.size]); // TODO: utf decode etc? |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3317 else |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3318 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3319 auto bufsize = buf.size; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3320 *dc = cast(string) buf.extractData()[0..bufsize]; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3321 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3322 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3323 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3324 /******************************************** |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3325 * Combine two document comments into one, |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3326 * separated by a newline. |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3327 */ |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3328 static string combineComments(string c1, string c2) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3329 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3330 //printf("Lexer::combineComments('%s', '%s')\n", c1, c2); |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3331 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3332 string c = c2; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3333 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3334 if (c1) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3335 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3336 c = c1; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3337 if (c2) |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3338 { |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3339 size_t len1 = c1.length; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3340 size_t len2 = c2.length; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3341 |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3342 c = c1.idup; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3343 if (len1 && c1[$-1] != '\n') |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3344 c ~= '\n'; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3345 c ~= c2; |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3346 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3347 } |
|
b7b61140701d
* added all missing default cases in switch statements
trass3r
parents:
135
diff
changeset
|
3348 return c; |
| 0 | 3349 } |
| 3350 | |
| 3351 static bool isValidIdentifier(string p) | |
| 3352 { | |
| 3353 if (p.length == 0) { | |
| 3354 return false; | |
| 3355 } | |
| 3356 | |
| 3357 if (p[0] >= '0' && p[0] <= '9') { // beware of isdigit() on signed chars | |
| 3358 return false; | |
| 3359 } | |
| 3360 | |
| 3361 size_t idx = 0; | |
| 3362 while (idx < p.length) | |
| 3363 { | |
| 3364 dchar dc; | |
| 3365 | |
| 3366 if (utf_decodeChar(p, &idx, &dc) !is null) { | |
| 3367 return false; | |
| 3368 } | |
| 3369 | |
| 3370 if (!((dc >= 0x80 && isUniAlpha(dc)) || isalnum(dc) || dc == '_')) { | |
| 3371 return false; | |
| 3372 } | |
| 3373 } | |
| 3374 | |
| 3375 return true; | |
| 3376 } | |
| 3377 | |
| 79 | 3378 /// TODO: use normal string append when GC works |
| 3379 static string combineComments(const(char)[] c1, const(char)[] c2) | |
| 0 | 3380 { |
| 79 | 3381 //writef("Lexer.combineComments('%s', '%s')\n", c1, c2); |
| 0 | 3382 |
| 79 | 3383 char[] c = cast(char[]) c2; |
| 3384 | |
| 3385 if (c1 !is null) | |
| 0 | 3386 { |
| 79 | 3387 c = cast(char[]) c1; |
| 3388 if (c2 !is null) | |
| 0 | 3389 { |
| 79 | 3390 c = cast(char[]) (GC.malloc(c1.length + 1 + c2.length)[0 .. c1.length + 1 + c2.length]); |
| 3391 size_t len1 = c1.length; | |
| 3392 c[0..len1] = c1[]; | |
| 3393 c[len1++] = '\n'; | |
| 3394 c[len1 .. len1 + c2.length] = c2[]; | |
| 0 | 3395 } |
| 3396 } | |
| 79 | 3397 |
| 3398 return cast(string)c; | |
| 0 | 3399 } |
| 3400 } |