ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 1 | // Copyright 2011 the V8 project authors. All rights reserved. |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 2 | // Redistribution and use in source and binary forms, with or without |
| 3 | // modification, are permitted provided that the following conditions are |
| 4 | // met: |
| 5 | // |
| 6 | // * Redistributions of source code must retain the above copyright |
| 7 | // notice, this list of conditions and the following disclaimer. |
| 8 | // * Redistributions in binary form must reproduce the above |
| 9 | // copyright notice, this list of conditions and the following |
| 10 | // disclaimer in the documentation and/or other materials provided |
| 11 | // with the distribution. |
| 12 | // * Neither the name of Google Inc. nor the names of its |
| 13 | // contributors may be used to endorse or promote products derived |
| 14 | // from this software without specific prior written permission. |
| 15 | // |
| 16 | // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 17 | // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 18 | // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 19 | // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 20 | // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 21 | // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 22 | // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 23 | // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 24 | // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 25 | // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 26 | // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 27 | |
| 28 | // Features shared by parsing and pre-parsing scanners. |
| 29 | |
erik.corry@gmail.com | 4a6c327 | 2010-11-18 12:04:40 +0000 | [diff] [blame] | 30 | #include "../include/v8stdint.h" |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 31 | #include "scanner-base.h" |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 32 | #include "char-predicates-inl.h" |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 33 | |
| 34 | namespace v8 { |
| 35 | namespace internal { |
| 36 | |
| 37 | // ---------------------------------------------------------------------------- |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 38 | // Scanner |
| 39 | |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 40 | Scanner::Scanner(UnicodeCache* unicode_cache) |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 41 | : unicode_cache_(unicode_cache) { } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 42 | |
| 43 | |
| 44 | uc32 Scanner::ScanHexEscape(uc32 c, int length) { |
| 45 | ASSERT(length <= 4); // prevent overflow |
| 46 | |
| 47 | uc32 digits[4]; |
| 48 | uc32 x = 0; |
| 49 | for (int i = 0; i < length; i++) { |
| 50 | digits[i] = c0_; |
| 51 | int d = HexValue(c0_); |
| 52 | if (d < 0) { |
| 53 | // According to ECMA-262, 3rd, 7.8.4, page 18, these hex escapes |
| 54 | // should be illegal, but other JS VMs just return the |
| 55 | // non-escaped version of the original character. |
| 56 | |
| 57 | // Push back digits read, except the last one (in c0_). |
| 58 | for (int j = i-1; j >= 0; j--) { |
| 59 | PushBack(digits[j]); |
| 60 | } |
| 61 | // Notice: No handling of error - treat it as "\u"->"u". |
| 62 | return c; |
| 63 | } |
| 64 | x = x * 16 + d; |
| 65 | Advance(); |
| 66 | } |
| 67 | |
| 68 | return x; |
| 69 | } |
| 70 | |
| 71 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 72 | |
| 73 | // ---------------------------------------------------------------------------- |
| 74 | // JavaScriptScanner |
| 75 | |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 76 | JavaScriptScanner::JavaScriptScanner(UnicodeCache* scanner_contants) |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 77 | : Scanner(scanner_contants), octal_pos_(Location::invalid()) { } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 78 | |
| 79 | |
| 80 | Token::Value JavaScriptScanner::Next() { |
| 81 | current_ = next_; |
| 82 | has_line_terminator_before_next_ = false; |
| 83 | Scan(); |
| 84 | return current_.token; |
| 85 | } |
| 86 | |
| 87 | |
| 88 | static inline bool IsByteOrderMark(uc32 c) { |
| 89 | // The Unicode value U+FFFE is guaranteed never to be assigned as a |
| 90 | // Unicode character; this implies that in a Unicode context the |
| 91 | // 0xFF, 0xFE byte pattern can only be interpreted as the U+FEFF |
| 92 | // character expressed in little-endian byte order (since it could |
| 93 | // not be a U+FFFE character expressed in big-endian byte |
| 94 | // order). Nevertheless, we check for it to be compatible with |
| 95 | // Spidermonkey. |
| 96 | return c == 0xFEFF || c == 0xFFFE; |
| 97 | } |
| 98 | |
| 99 | |
| 100 | bool JavaScriptScanner::SkipWhiteSpace() { |
| 101 | int start_position = source_pos(); |
| 102 | |
| 103 | while (true) { |
| 104 | // We treat byte-order marks (BOMs) as whitespace for better |
| 105 | // compatibility with Spidermonkey and other JavaScript engines. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 106 | while (unicode_cache_->IsWhiteSpace(c0_) || IsByteOrderMark(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 107 | // IsWhiteSpace() includes line terminators! |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 108 | if (unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 109 | // Ignore line terminators, but remember them. This is necessary |
| 110 | // for automatic semicolon insertion. |
| 111 | has_line_terminator_before_next_ = true; |
| 112 | } |
| 113 | Advance(); |
| 114 | } |
| 115 | |
| 116 | // If there is an HTML comment end '-->' at the beginning of a |
| 117 | // line (with only whitespace in front of it), we treat the rest |
| 118 | // of the line as a comment. This is in line with the way |
| 119 | // SpiderMonkey handles it. |
| 120 | if (c0_ == '-' && has_line_terminator_before_next_) { |
| 121 | Advance(); |
| 122 | if (c0_ == '-') { |
| 123 | Advance(); |
| 124 | if (c0_ == '>') { |
| 125 | // Treat the rest of the line as a comment. |
| 126 | SkipSingleLineComment(); |
| 127 | // Continue skipping white space after the comment. |
| 128 | continue; |
| 129 | } |
| 130 | PushBack('-'); // undo Advance() |
| 131 | } |
| 132 | PushBack('-'); // undo Advance() |
| 133 | } |
| 134 | // Return whether or not we skipped any characters. |
| 135 | return source_pos() != start_position; |
| 136 | } |
| 137 | } |
| 138 | |
| 139 | |
| 140 | Token::Value JavaScriptScanner::SkipSingleLineComment() { |
| 141 | Advance(); |
| 142 | |
| 143 | // The line terminator at the end of the line is not considered |
| 144 | // to be part of the single-line comment; it is recognized |
| 145 | // separately by the lexical grammar and becomes part of the |
| 146 | // stream of input elements for the syntactic grammar (see |
| 147 | // ECMA-262, section 7.4, page 12). |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 148 | while (c0_ >= 0 && !unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 149 | Advance(); |
| 150 | } |
| 151 | |
| 152 | return Token::WHITESPACE; |
| 153 | } |
| 154 | |
| 155 | |
| 156 | Token::Value JavaScriptScanner::SkipMultiLineComment() { |
| 157 | ASSERT(c0_ == '*'); |
| 158 | Advance(); |
| 159 | |
| 160 | while (c0_ >= 0) { |
| 161 | char ch = c0_; |
| 162 | Advance(); |
| 163 | // If we have reached the end of the multi-line comment, we |
| 164 | // consume the '/' and insert a whitespace. This way all |
| 165 | // multi-line comments are treated as whitespace - even the ones |
| 166 | // containing line terminators. This contradicts ECMA-262, section |
| 167 | // 7.4, page 12, that says that multi-line comments containing |
| 168 | // line terminators should be treated as a line terminator, but it |
| 169 | // matches the behaviour of SpiderMonkey and KJS. |
| 170 | if (ch == '*' && c0_ == '/') { |
| 171 | c0_ = ' '; |
| 172 | return Token::WHITESPACE; |
| 173 | } |
| 174 | } |
| 175 | |
| 176 | // Unterminated multi-line comment. |
| 177 | return Token::ILLEGAL; |
| 178 | } |
| 179 | |
| 180 | |
| 181 | Token::Value JavaScriptScanner::ScanHtmlComment() { |
| 182 | // Check for <!-- comments. |
| 183 | ASSERT(c0_ == '!'); |
| 184 | Advance(); |
| 185 | if (c0_ == '-') { |
| 186 | Advance(); |
| 187 | if (c0_ == '-') return SkipSingleLineComment(); |
| 188 | PushBack('-'); // undo Advance() |
| 189 | } |
| 190 | PushBack('!'); // undo Advance() |
| 191 | ASSERT(c0_ == '!'); |
| 192 | return Token::LT; |
| 193 | } |
| 194 | |
| 195 | |
| 196 | void JavaScriptScanner::Scan() { |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 197 | next_.literal_chars = NULL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 198 | Token::Value token; |
| 199 | do { |
| 200 | // Remember the position of the next token |
| 201 | next_.location.beg_pos = source_pos(); |
| 202 | |
| 203 | switch (c0_) { |
| 204 | case ' ': |
| 205 | case '\t': |
| 206 | Advance(); |
| 207 | token = Token::WHITESPACE; |
| 208 | break; |
| 209 | |
| 210 | case '\n': |
| 211 | Advance(); |
| 212 | has_line_terminator_before_next_ = true; |
| 213 | token = Token::WHITESPACE; |
| 214 | break; |
| 215 | |
| 216 | case '"': case '\'': |
| 217 | token = ScanString(); |
| 218 | break; |
| 219 | |
| 220 | case '<': |
| 221 | // < <= << <<= <!-- |
| 222 | Advance(); |
| 223 | if (c0_ == '=') { |
| 224 | token = Select(Token::LTE); |
| 225 | } else if (c0_ == '<') { |
| 226 | token = Select('=', Token::ASSIGN_SHL, Token::SHL); |
| 227 | } else if (c0_ == '!') { |
| 228 | token = ScanHtmlComment(); |
| 229 | } else { |
| 230 | token = Token::LT; |
| 231 | } |
| 232 | break; |
| 233 | |
| 234 | case '>': |
| 235 | // > >= >> >>= >>> >>>= |
| 236 | Advance(); |
| 237 | if (c0_ == '=') { |
| 238 | token = Select(Token::GTE); |
| 239 | } else if (c0_ == '>') { |
| 240 | // >> >>= >>> >>>= |
| 241 | Advance(); |
| 242 | if (c0_ == '=') { |
| 243 | token = Select(Token::ASSIGN_SAR); |
| 244 | } else if (c0_ == '>') { |
| 245 | token = Select('=', Token::ASSIGN_SHR, Token::SHR); |
| 246 | } else { |
| 247 | token = Token::SAR; |
| 248 | } |
| 249 | } else { |
| 250 | token = Token::GT; |
| 251 | } |
| 252 | break; |
| 253 | |
| 254 | case '=': |
| 255 | // = == === |
| 256 | Advance(); |
| 257 | if (c0_ == '=') { |
| 258 | token = Select('=', Token::EQ_STRICT, Token::EQ); |
| 259 | } else { |
| 260 | token = Token::ASSIGN; |
| 261 | } |
| 262 | break; |
| 263 | |
| 264 | case '!': |
| 265 | // ! != !== |
| 266 | Advance(); |
| 267 | if (c0_ == '=') { |
| 268 | token = Select('=', Token::NE_STRICT, Token::NE); |
| 269 | } else { |
| 270 | token = Token::NOT; |
| 271 | } |
| 272 | break; |
| 273 | |
| 274 | case '+': |
| 275 | // + ++ += |
| 276 | Advance(); |
| 277 | if (c0_ == '+') { |
| 278 | token = Select(Token::INC); |
| 279 | } else if (c0_ == '=') { |
| 280 | token = Select(Token::ASSIGN_ADD); |
| 281 | } else { |
| 282 | token = Token::ADD; |
| 283 | } |
| 284 | break; |
| 285 | |
| 286 | case '-': |
| 287 | // - -- --> -= |
| 288 | Advance(); |
| 289 | if (c0_ == '-') { |
| 290 | Advance(); |
| 291 | if (c0_ == '>' && has_line_terminator_before_next_) { |
| 292 | // For compatibility with SpiderMonkey, we skip lines that |
| 293 | // start with an HTML comment end '-->'. |
| 294 | token = SkipSingleLineComment(); |
| 295 | } else { |
| 296 | token = Token::DEC; |
| 297 | } |
| 298 | } else if (c0_ == '=') { |
| 299 | token = Select(Token::ASSIGN_SUB); |
| 300 | } else { |
| 301 | token = Token::SUB; |
| 302 | } |
| 303 | break; |
| 304 | |
| 305 | case '*': |
| 306 | // * *= |
| 307 | token = Select('=', Token::ASSIGN_MUL, Token::MUL); |
| 308 | break; |
| 309 | |
| 310 | case '%': |
| 311 | // % %= |
| 312 | token = Select('=', Token::ASSIGN_MOD, Token::MOD); |
| 313 | break; |
| 314 | |
| 315 | case '/': |
| 316 | // / // /* /= |
| 317 | Advance(); |
| 318 | if (c0_ == '/') { |
| 319 | token = SkipSingleLineComment(); |
| 320 | } else if (c0_ == '*') { |
| 321 | token = SkipMultiLineComment(); |
| 322 | } else if (c0_ == '=') { |
| 323 | token = Select(Token::ASSIGN_DIV); |
| 324 | } else { |
| 325 | token = Token::DIV; |
| 326 | } |
| 327 | break; |
| 328 | |
| 329 | case '&': |
| 330 | // & && &= |
| 331 | Advance(); |
| 332 | if (c0_ == '&') { |
| 333 | token = Select(Token::AND); |
| 334 | } else if (c0_ == '=') { |
| 335 | token = Select(Token::ASSIGN_BIT_AND); |
| 336 | } else { |
| 337 | token = Token::BIT_AND; |
| 338 | } |
| 339 | break; |
| 340 | |
| 341 | case '|': |
| 342 | // | || |= |
| 343 | Advance(); |
| 344 | if (c0_ == '|') { |
| 345 | token = Select(Token::OR); |
| 346 | } else if (c0_ == '=') { |
| 347 | token = Select(Token::ASSIGN_BIT_OR); |
| 348 | } else { |
| 349 | token = Token::BIT_OR; |
| 350 | } |
| 351 | break; |
| 352 | |
| 353 | case '^': |
| 354 | // ^ ^= |
| 355 | token = Select('=', Token::ASSIGN_BIT_XOR, Token::BIT_XOR); |
| 356 | break; |
| 357 | |
| 358 | case '.': |
| 359 | // . Number |
| 360 | Advance(); |
| 361 | if (IsDecimalDigit(c0_)) { |
| 362 | token = ScanNumber(true); |
| 363 | } else { |
| 364 | token = Token::PERIOD; |
| 365 | } |
| 366 | break; |
| 367 | |
| 368 | case ':': |
| 369 | token = Select(Token::COLON); |
| 370 | break; |
| 371 | |
| 372 | case ';': |
| 373 | token = Select(Token::SEMICOLON); |
| 374 | break; |
| 375 | |
| 376 | case ',': |
| 377 | token = Select(Token::COMMA); |
| 378 | break; |
| 379 | |
| 380 | case '(': |
| 381 | token = Select(Token::LPAREN); |
| 382 | break; |
| 383 | |
| 384 | case ')': |
| 385 | token = Select(Token::RPAREN); |
| 386 | break; |
| 387 | |
| 388 | case '[': |
| 389 | token = Select(Token::LBRACK); |
| 390 | break; |
| 391 | |
| 392 | case ']': |
| 393 | token = Select(Token::RBRACK); |
| 394 | break; |
| 395 | |
| 396 | case '{': |
| 397 | token = Select(Token::LBRACE); |
| 398 | break; |
| 399 | |
| 400 | case '}': |
| 401 | token = Select(Token::RBRACE); |
| 402 | break; |
| 403 | |
| 404 | case '?': |
| 405 | token = Select(Token::CONDITIONAL); |
| 406 | break; |
| 407 | |
| 408 | case '~': |
| 409 | token = Select(Token::BIT_NOT); |
| 410 | break; |
| 411 | |
| 412 | default: |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 413 | if (unicode_cache_->IsIdentifierStart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 414 | token = ScanIdentifierOrKeyword(); |
| 415 | } else if (IsDecimalDigit(c0_)) { |
| 416 | token = ScanNumber(false); |
| 417 | } else if (SkipWhiteSpace()) { |
| 418 | token = Token::WHITESPACE; |
| 419 | } else if (c0_ < 0) { |
| 420 | token = Token::EOS; |
| 421 | } else { |
| 422 | token = Select(Token::ILLEGAL); |
| 423 | } |
| 424 | break; |
| 425 | } |
| 426 | |
| 427 | // Continue scanning for tokens as long as we're just skipping |
| 428 | // whitespace. |
| 429 | } while (token == Token::WHITESPACE); |
| 430 | |
| 431 | next_.location.end_pos = source_pos(); |
| 432 | next_.token = token; |
| 433 | } |
| 434 | |
| 435 | |
| 436 | void JavaScriptScanner::SeekForward(int pos) { |
ager@chromium.org | 5f0c45f | 2010-12-17 08:51:21 +0000 | [diff] [blame] | 437 | // After this call, we will have the token at the given position as |
| 438 | // the "next" token. The "current" token will be invalid. |
| 439 | if (pos == next_.location.beg_pos) return; |
| 440 | int current_pos = source_pos(); |
| 441 | ASSERT_EQ(next_.location.end_pos, current_pos); |
| 442 | // Positions inside the lookahead token aren't supported. |
| 443 | ASSERT(pos >= current_pos); |
| 444 | if (pos != current_pos) { |
| 445 | source_->SeekForward(pos - source_->pos()); |
| 446 | Advance(); |
| 447 | // This function is only called to seek to the location |
| 448 | // of the end of a function (at the "}" token). It doesn't matter |
| 449 | // whether there was a line terminator in the part we skip. |
| 450 | has_line_terminator_before_next_ = false; |
| 451 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 452 | Scan(); |
| 453 | } |
| 454 | |
| 455 | |
| 456 | void JavaScriptScanner::ScanEscape() { |
| 457 | uc32 c = c0_; |
| 458 | Advance(); |
| 459 | |
| 460 | // Skip escaped newlines. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 461 | if (unicode_cache_->IsLineTerminator(c)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 462 | // Allow CR+LF newlines in multiline string literals. |
| 463 | if (IsCarriageReturn(c) && IsLineFeed(c0_)) Advance(); |
| 464 | // Allow LF+CR newlines in multiline string literals. |
| 465 | if (IsLineFeed(c) && IsCarriageReturn(c0_)) Advance(); |
| 466 | return; |
| 467 | } |
| 468 | |
| 469 | switch (c) { |
| 470 | case '\'': // fall through |
| 471 | case '"' : // fall through |
| 472 | case '\\': break; |
| 473 | case 'b' : c = '\b'; break; |
| 474 | case 'f' : c = '\f'; break; |
| 475 | case 'n' : c = '\n'; break; |
| 476 | case 'r' : c = '\r'; break; |
| 477 | case 't' : c = '\t'; break; |
| 478 | case 'u' : c = ScanHexEscape(c, 4); break; |
| 479 | case 'v' : c = '\v'; break; |
| 480 | case 'x' : c = ScanHexEscape(c, 2); break; |
| 481 | case '0' : // fall through |
| 482 | case '1' : // fall through |
| 483 | case '2' : // fall through |
| 484 | case '3' : // fall through |
| 485 | case '4' : // fall through |
| 486 | case '5' : // fall through |
| 487 | case '6' : // fall through |
| 488 | case '7' : c = ScanOctalEscape(c, 2); break; |
| 489 | } |
| 490 | |
| 491 | // According to ECMA-262, 3rd, 7.8.4 (p 18ff) these |
| 492 | // should be illegal, but they are commonly handled |
| 493 | // as non-escaped characters by JS VMs. |
| 494 | AddLiteralChar(c); |
| 495 | } |
| 496 | |
| 497 | |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 498 | // Octal escapes of the forms '\0xx' and '\xxx' are not a part of |
| 499 | // ECMA-262. Other JS VMs support them. |
| 500 | uc32 JavaScriptScanner::ScanOctalEscape(uc32 c, int length) { |
| 501 | uc32 x = c - '0'; |
| 502 | int i = 0; |
| 503 | for (; i < length; i++) { |
| 504 | int d = c0_ - '0'; |
| 505 | if (d < 0 || d > 7) break; |
| 506 | int nx = x * 8 + d; |
| 507 | if (nx >= 256) break; |
| 508 | x = nx; |
| 509 | Advance(); |
| 510 | } |
| 511 | // Anything except '\0' is an octal escape sequence, illegal in strict mode. |
| 512 | // Remember the position of octal escape sequences so that an error |
| 513 | // can be reported later (in strict mode). |
| 514 | // We don't report the error immediately, because the octal escape can |
| 515 | // occur before the "use strict" directive. |
| 516 | if (c != '0' || i > 0) { |
| 517 | octal_pos_ = Location(source_pos() - i - 1, source_pos() - 1); |
| 518 | } |
| 519 | return x; |
| 520 | } |
| 521 | |
| 522 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 523 | Token::Value JavaScriptScanner::ScanString() { |
| 524 | uc32 quote = c0_; |
| 525 | Advance(); // consume quote |
| 526 | |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 527 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 528 | while (c0_ != quote && c0_ >= 0 |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 529 | && !unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 530 | uc32 c = c0_; |
| 531 | Advance(); |
| 532 | if (c == '\\') { |
| 533 | if (c0_ < 0) return Token::ILLEGAL; |
| 534 | ScanEscape(); |
| 535 | } else { |
| 536 | AddLiteralChar(c); |
| 537 | } |
| 538 | } |
| 539 | if (c0_ != quote) return Token::ILLEGAL; |
| 540 | literal.Complete(); |
| 541 | |
| 542 | Advance(); // consume quote |
| 543 | return Token::STRING; |
| 544 | } |
| 545 | |
| 546 | |
| 547 | void JavaScriptScanner::ScanDecimalDigits() { |
| 548 | while (IsDecimalDigit(c0_)) |
| 549 | AddLiteralCharAdvance(); |
| 550 | } |
| 551 | |
| 552 | |
| 553 | Token::Value JavaScriptScanner::ScanNumber(bool seen_period) { |
| 554 | ASSERT(IsDecimalDigit(c0_)); // the first digit of the number or the fraction |
| 555 | |
| 556 | enum { DECIMAL, HEX, OCTAL } kind = DECIMAL; |
| 557 | |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 558 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 559 | if (seen_period) { |
| 560 | // we have already seen a decimal point of the float |
| 561 | AddLiteralChar('.'); |
| 562 | ScanDecimalDigits(); // we know we have at least one digit |
| 563 | |
| 564 | } else { |
| 565 | // if the first character is '0' we must check for octals and hex |
| 566 | if (c0_ == '0') { |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 567 | int start_pos = source_pos(); // For reporting octal positions. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 568 | AddLiteralCharAdvance(); |
| 569 | |
| 570 | // either 0, 0exxx, 0Exxx, 0.xxx, an octal number, or a hex number |
| 571 | if (c0_ == 'x' || c0_ == 'X') { |
| 572 | // hex number |
| 573 | kind = HEX; |
| 574 | AddLiteralCharAdvance(); |
| 575 | if (!IsHexDigit(c0_)) { |
| 576 | // we must have at least one hex digit after 'x'/'X' |
| 577 | return Token::ILLEGAL; |
| 578 | } |
| 579 | while (IsHexDigit(c0_)) { |
| 580 | AddLiteralCharAdvance(); |
| 581 | } |
| 582 | } else if ('0' <= c0_ && c0_ <= '7') { |
| 583 | // (possible) octal number |
| 584 | kind = OCTAL; |
| 585 | while (true) { |
| 586 | if (c0_ == '8' || c0_ == '9') { |
| 587 | kind = DECIMAL; |
| 588 | break; |
| 589 | } |
ager@chromium.org | 0ee099b | 2011-01-25 14:06:47 +0000 | [diff] [blame] | 590 | if (c0_ < '0' || '7' < c0_) { |
| 591 | // Octal literal finished. |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 592 | octal_pos_ = Location(start_pos, source_pos()); |
ager@chromium.org | 0ee099b | 2011-01-25 14:06:47 +0000 | [diff] [blame] | 593 | break; |
| 594 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 595 | AddLiteralCharAdvance(); |
| 596 | } |
| 597 | } |
| 598 | } |
| 599 | |
| 600 | // Parse decimal digits and allow trailing fractional part. |
| 601 | if (kind == DECIMAL) { |
| 602 | ScanDecimalDigits(); // optional |
| 603 | if (c0_ == '.') { |
| 604 | AddLiteralCharAdvance(); |
| 605 | ScanDecimalDigits(); // optional |
| 606 | } |
| 607 | } |
| 608 | } |
| 609 | |
| 610 | // scan exponent, if any |
| 611 | if (c0_ == 'e' || c0_ == 'E') { |
| 612 | ASSERT(kind != HEX); // 'e'/'E' must be scanned as part of the hex number |
| 613 | if (kind == OCTAL) return Token::ILLEGAL; // no exponent for octals allowed |
| 614 | // scan exponent |
| 615 | AddLiteralCharAdvance(); |
| 616 | if (c0_ == '+' || c0_ == '-') |
| 617 | AddLiteralCharAdvance(); |
| 618 | if (!IsDecimalDigit(c0_)) { |
| 619 | // we must have at least one decimal digit after 'e'/'E' |
| 620 | return Token::ILLEGAL; |
| 621 | } |
| 622 | ScanDecimalDigits(); |
| 623 | } |
| 624 | |
| 625 | // The source character immediately following a numeric literal must |
| 626 | // not be an identifier start or a decimal digit; see ECMA-262 |
| 627 | // section 7.8.3, page 17 (note that we read only one decimal digit |
| 628 | // if the value is 0). |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 629 | if (IsDecimalDigit(c0_) || unicode_cache_->IsIdentifierStart(c0_)) |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 630 | return Token::ILLEGAL; |
| 631 | |
| 632 | literal.Complete(); |
| 633 | |
| 634 | return Token::NUMBER; |
| 635 | } |
| 636 | |
| 637 | |
| 638 | uc32 JavaScriptScanner::ScanIdentifierUnicodeEscape() { |
| 639 | Advance(); |
| 640 | if (c0_ != 'u') return unibrow::Utf8::kBadChar; |
| 641 | Advance(); |
| 642 | uc32 c = ScanHexEscape('u', 4); |
| 643 | // We do not allow a unicode escape sequence to start another |
| 644 | // unicode escape sequence. |
| 645 | if (c == '\\') return unibrow::Utf8::kBadChar; |
| 646 | return c; |
| 647 | } |
| 648 | |
| 649 | |
| 650 | Token::Value JavaScriptScanner::ScanIdentifierOrKeyword() { |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 651 | ASSERT(unicode_cache_->IsIdentifierStart(c0_)); |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 652 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 653 | KeywordMatcher keyword_match; |
| 654 | // Scan identifier start character. |
| 655 | if (c0_ == '\\') { |
| 656 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 657 | // Only allow legal identifier start characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 658 | if (!unicode_cache_->IsIdentifierStart(c)) return Token::ILLEGAL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 659 | AddLiteralChar(c); |
| 660 | return ScanIdentifierSuffix(&literal); |
| 661 | } |
| 662 | |
| 663 | uc32 first_char = c0_; |
| 664 | Advance(); |
| 665 | AddLiteralChar(first_char); |
| 666 | if (!keyword_match.AddChar(first_char)) { |
| 667 | return ScanIdentifierSuffix(&literal); |
| 668 | } |
| 669 | |
| 670 | // Scan the rest of the identifier characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 671 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 672 | if (c0_ != '\\') { |
| 673 | uc32 next_char = c0_; |
| 674 | Advance(); |
| 675 | AddLiteralChar(next_char); |
| 676 | if (keyword_match.AddChar(next_char)) continue; |
| 677 | } |
| 678 | // Fallthrough if no loner able to complete keyword. |
| 679 | return ScanIdentifierSuffix(&literal); |
| 680 | } |
| 681 | literal.Complete(); |
| 682 | |
| 683 | return keyword_match.token(); |
| 684 | } |
| 685 | |
| 686 | |
| 687 | Token::Value JavaScriptScanner::ScanIdentifierSuffix(LiteralScope* literal) { |
| 688 | // Scan the rest of the identifier characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 689 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 690 | if (c0_ == '\\') { |
| 691 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 692 | // Only allow legal identifier part characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 693 | if (!unicode_cache_->IsIdentifierPart(c)) return Token::ILLEGAL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 694 | AddLiteralChar(c); |
| 695 | } else { |
| 696 | AddLiteralChar(c0_); |
| 697 | Advance(); |
| 698 | } |
| 699 | } |
| 700 | literal->Complete(); |
| 701 | |
| 702 | return Token::IDENTIFIER; |
| 703 | } |
| 704 | |
| 705 | |
| 706 | bool JavaScriptScanner::ScanRegExpPattern(bool seen_equal) { |
| 707 | // Scan: ('/' | '/=') RegularExpressionBody '/' RegularExpressionFlags |
| 708 | bool in_character_class = false; |
| 709 | |
| 710 | // Previous token is either '/' or '/=', in the second case, the |
| 711 | // pattern starts at =. |
| 712 | next_.location.beg_pos = source_pos() - (seen_equal ? 2 : 1); |
| 713 | next_.location.end_pos = source_pos() - (seen_equal ? 1 : 0); |
| 714 | |
| 715 | // Scan regular expression body: According to ECMA-262, 3rd, 7.8.5, |
| 716 | // the scanner should pass uninterpreted bodies to the RegExp |
| 717 | // constructor. |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 718 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 719 | if (seen_equal) |
| 720 | AddLiteralChar('='); |
| 721 | |
| 722 | while (c0_ != '/' || in_character_class) { |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 723 | if (unicode_cache_->IsLineTerminator(c0_) || c0_ < 0) return false; |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 724 | if (c0_ == '\\') { // Escape sequence. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 725 | AddLiteralCharAdvance(); |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 726 | if (unicode_cache_->IsLineTerminator(c0_) || c0_ < 0) return false; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 727 | AddLiteralCharAdvance(); |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 728 | // If the escape allows more characters, i.e., \x??, \u????, or \c?, |
| 729 | // only "safe" characters are allowed (letters, digits, underscore), |
| 730 | // otherwise the escape isn't valid and the invalid character has |
| 731 | // its normal meaning. I.e., we can just continue scanning without |
| 732 | // worrying whether the following characters are part of the escape |
| 733 | // or not, since any '/', '\\' or '[' is guaranteed to not be part |
| 734 | // of the escape sequence. |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 735 | |
| 736 | // TODO(896): At some point, parse RegExps more throughly to capture |
| 737 | // octal esacpes in strict mode. |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 738 | } else { // Unescaped character. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 739 | if (c0_ == '[') in_character_class = true; |
| 740 | if (c0_ == ']') in_character_class = false; |
| 741 | AddLiteralCharAdvance(); |
| 742 | } |
| 743 | } |
| 744 | Advance(); // consume '/' |
| 745 | |
| 746 | literal.Complete(); |
| 747 | |
| 748 | return true; |
| 749 | } |
| 750 | |
| 751 | |
| 752 | bool JavaScriptScanner::ScanRegExpFlags() { |
| 753 | // Scan regular expression flags. |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 754 | LiteralScope literal(this); |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 755 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 756 | if (c0_ == '\\') { |
| 757 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 758 | if (c != static_cast<uc32>(unibrow::Utf8::kBadChar)) { |
| 759 | // We allow any escaped character, unlike the restriction on |
| 760 | // IdentifierPart when it is used to build an IdentifierName. |
| 761 | AddLiteralChar(c); |
| 762 | continue; |
| 763 | } |
| 764 | } |
| 765 | AddLiteralCharAdvance(); |
| 766 | } |
| 767 | literal.Complete(); |
| 768 | |
| 769 | next_.location.end_pos = source_pos() - 1; |
| 770 | return true; |
| 771 | } |
| 772 | |
| 773 | // ---------------------------------------------------------------------------- |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 774 | // Keyword Matcher |
| 775 | |
| 776 | KeywordMatcher::FirstState KeywordMatcher::first_states_[] = { |
| 777 | { "break", KEYWORD_PREFIX, Token::BREAK }, |
| 778 | { NULL, C, Token::ILLEGAL }, |
| 779 | { NULL, D, Token::ILLEGAL }, |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 780 | { NULL, E, Token::ILLEGAL }, |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 781 | { NULL, F, Token::ILLEGAL }, |
| 782 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 783 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 784 | { NULL, I, Token::ILLEGAL }, |
| 785 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 786 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 787 | { "let", KEYWORD_PREFIX, Token::FUTURE_RESERVED_WORD }, |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 788 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 789 | { NULL, N, Token::ILLEGAL }, |
| 790 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 791 | { NULL, P, Token::ILLEGAL }, |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 792 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 793 | { "return", KEYWORD_PREFIX, Token::RETURN }, |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 794 | { NULL, S, Token::ILLEGAL }, |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 795 | { NULL, T, Token::ILLEGAL }, |
| 796 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 797 | { NULL, V, Token::ILLEGAL }, |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 798 | { NULL, W, Token::ILLEGAL }, |
| 799 | { NULL, UNMATCHABLE, Token::ILLEGAL }, |
| 800 | { "yield", KEYWORD_PREFIX, Token::FUTURE_RESERVED_WORD } |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 801 | }; |
| 802 | |
| 803 | |
| 804 | void KeywordMatcher::Step(unibrow::uchar input) { |
| 805 | switch (state_) { |
| 806 | case INITIAL: { |
| 807 | // matching the first character is the only state with significant fanout. |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 808 | // Match only lower-case letters in range 'b'..'y'. |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 809 | unsigned int offset = input - kFirstCharRangeMin; |
| 810 | if (offset < kFirstCharRangeLength) { |
| 811 | state_ = first_states_[offset].state; |
| 812 | if (state_ == KEYWORD_PREFIX) { |
| 813 | keyword_ = first_states_[offset].keyword; |
| 814 | counter_ = 1; |
| 815 | keyword_token_ = first_states_[offset].token; |
| 816 | } |
| 817 | return; |
| 818 | } |
| 819 | break; |
| 820 | } |
| 821 | case KEYWORD_PREFIX: |
| 822 | if (static_cast<unibrow::uchar>(keyword_[counter_]) == input) { |
| 823 | counter_++; |
| 824 | if (keyword_[counter_] == '\0') { |
| 825 | state_ = KEYWORD_MATCHED; |
| 826 | token_ = keyword_token_; |
| 827 | } |
| 828 | return; |
| 829 | } |
| 830 | break; |
| 831 | case KEYWORD_MATCHED: |
| 832 | token_ = Token::IDENTIFIER; |
| 833 | break; |
| 834 | case C: |
| 835 | if (MatchState(input, 'a', CA)) return; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 836 | if (MatchKeywordStart(input, "class", 1, |
| 837 | Token::FUTURE_RESERVED_WORD)) return; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 838 | if (MatchState(input, 'o', CO)) return; |
| 839 | break; |
| 840 | case CA: |
| 841 | if (MatchKeywordStart(input, "case", 2, Token::CASE)) return; |
| 842 | if (MatchKeywordStart(input, "catch", 2, Token::CATCH)) return; |
| 843 | break; |
| 844 | case CO: |
| 845 | if (MatchState(input, 'n', CON)) return; |
| 846 | break; |
| 847 | case CON: |
| 848 | if (MatchKeywordStart(input, "const", 3, Token::CONST)) return; |
| 849 | if (MatchKeywordStart(input, "continue", 3, Token::CONTINUE)) return; |
| 850 | break; |
| 851 | case D: |
| 852 | if (MatchState(input, 'e', DE)) return; |
| 853 | if (MatchKeyword(input, 'o', KEYWORD_MATCHED, Token::DO)) return; |
| 854 | break; |
| 855 | case DE: |
| 856 | if (MatchKeywordStart(input, "debugger", 2, Token::DEBUGGER)) return; |
| 857 | if (MatchKeywordStart(input, "default", 2, Token::DEFAULT)) return; |
| 858 | if (MatchKeywordStart(input, "delete", 2, Token::DELETE)) return; |
| 859 | break; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 860 | case E: |
| 861 | if (MatchKeywordStart(input, "else", 1, Token::ELSE)) return; |
| 862 | if (MatchKeywordStart(input, "enum", 1, |
| 863 | Token::FUTURE_RESERVED_WORD)) return; |
| 864 | if (MatchState(input, 'x', EX)) return; |
| 865 | break; |
| 866 | case EX: |
| 867 | if (MatchKeywordStart(input, "export", 2, |
| 868 | Token::FUTURE_RESERVED_WORD)) return; |
| 869 | if (MatchKeywordStart(input, "extends", 2, |
| 870 | Token::FUTURE_RESERVED_WORD)) return; |
| 871 | break; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 872 | case F: |
| 873 | if (MatchKeywordStart(input, "false", 1, Token::FALSE_LITERAL)) return; |
| 874 | if (MatchKeywordStart(input, "finally", 1, Token::FINALLY)) return; |
| 875 | if (MatchKeywordStart(input, "for", 1, Token::FOR)) return; |
| 876 | if (MatchKeywordStart(input, "function", 1, Token::FUNCTION)) return; |
| 877 | break; |
| 878 | case I: |
| 879 | if (MatchKeyword(input, 'f', KEYWORD_MATCHED, Token::IF)) return; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 880 | if (MatchState(input, 'm', IM)) return; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 881 | if (MatchKeyword(input, 'n', IN, Token::IN)) return; |
| 882 | break; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 883 | case IM: |
| 884 | if (MatchState(input, 'p', IMP)) return; |
| 885 | break; |
| 886 | case IMP: |
| 887 | if (MatchKeywordStart(input, "implements", 3, |
| 888 | Token::FUTURE_RESERVED_WORD )) return; |
| 889 | if (MatchKeywordStart(input, "import", 3, |
| 890 | Token::FUTURE_RESERVED_WORD)) return; |
| 891 | break; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 892 | case IN: |
| 893 | token_ = Token::IDENTIFIER; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 894 | if (MatchKeywordStart(input, "interface", 2, |
| 895 | Token::FUTURE_RESERVED_WORD)) return; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 896 | if (MatchKeywordStart(input, "instanceof", 2, Token::INSTANCEOF)) return; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 897 | break; |
| 898 | case N: |
| 899 | if (MatchKeywordStart(input, "native", 1, Token::NATIVE)) return; |
| 900 | if (MatchKeywordStart(input, "new", 1, Token::NEW)) return; |
| 901 | if (MatchKeywordStart(input, "null", 1, Token::NULL_LITERAL)) return; |
| 902 | break; |
ricow@chromium.org | 83aa549 | 2011-02-07 12:42:56 +0000 | [diff] [blame] | 903 | case P: |
| 904 | if (MatchKeywordStart(input, "package", 1, |
| 905 | Token::FUTURE_RESERVED_WORD)) return; |
| 906 | if (MatchState(input, 'r', PR)) return; |
| 907 | if (MatchKeywordStart(input, "public", 1, |
| 908 | Token::FUTURE_RESERVED_WORD)) return; |
| 909 | break; |
| 910 | case PR: |
| 911 | if (MatchKeywordStart(input, "private", 2, |
| 912 | Token::FUTURE_RESERVED_WORD)) return; |
| 913 | if (MatchKeywordStart(input, "protected", 2, |
| 914 | Token::FUTURE_RESERVED_WORD)) return; |
| 915 | break; |
| 916 | case S: |
| 917 | if (MatchKeywordStart(input, "static", 1, |
| 918 | Token::FUTURE_RESERVED_WORD)) return; |
| 919 | if (MatchKeywordStart(input, "super", 1, |
| 920 | Token::FUTURE_RESERVED_WORD)) return; |
| 921 | if (MatchKeywordStart(input, "switch", 1, |
| 922 | Token::SWITCH)) return; |
| 923 | break; |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 924 | case T: |
| 925 | if (MatchState(input, 'h', TH)) return; |
| 926 | if (MatchState(input, 'r', TR)) return; |
| 927 | if (MatchKeywordStart(input, "typeof", 1, Token::TYPEOF)) return; |
| 928 | break; |
| 929 | case TH: |
| 930 | if (MatchKeywordStart(input, "this", 2, Token::THIS)) return; |
| 931 | if (MatchKeywordStart(input, "throw", 2, Token::THROW)) return; |
| 932 | break; |
| 933 | case TR: |
| 934 | if (MatchKeywordStart(input, "true", 2, Token::TRUE_LITERAL)) return; |
| 935 | if (MatchKeyword(input, 'y', KEYWORD_MATCHED, Token::TRY)) return; |
| 936 | break; |
| 937 | case V: |
| 938 | if (MatchKeywordStart(input, "var", 1, Token::VAR)) return; |
| 939 | if (MatchKeywordStart(input, "void", 1, Token::VOID)) return; |
| 940 | break; |
| 941 | case W: |
| 942 | if (MatchKeywordStart(input, "while", 1, Token::WHILE)) return; |
| 943 | if (MatchKeywordStart(input, "with", 1, Token::WITH)) return; |
| 944 | break; |
| 945 | case UNMATCHABLE: |
| 946 | break; |
| 947 | } |
| 948 | // On fallthrough, it's a failure. |
| 949 | state_ = UNMATCHABLE; |
| 950 | } |
| 951 | |
| 952 | } } // namespace v8::internal |