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) |
danno@chromium.org | b645116 | 2011-08-17 14:33:23 +0000 | [diff] [blame^] | 77 | : Scanner(scanner_contants), |
| 78 | octal_pos_(Location::invalid()), |
| 79 | harmony_block_scoping_(false) { } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 80 | |
| 81 | |
lrn@chromium.org | ac2828d | 2011-06-23 06:29:21 +0000 | [diff] [blame] | 82 | void JavaScriptScanner::Initialize(UC16CharacterStream* source) { |
| 83 | source_ = source; |
| 84 | // Need to capture identifiers in order to recognize "get" and "set" |
| 85 | // in object literals. |
| 86 | Init(); |
| 87 | // Skip initial whitespace allowing HTML comment ends just like |
| 88 | // after a newline and scan first token. |
| 89 | has_line_terminator_before_next_ = true; |
| 90 | SkipWhiteSpace(); |
| 91 | Scan(); |
| 92 | } |
| 93 | |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 94 | |
| 95 | // Ensure that tokens can be stored in a byte. |
| 96 | STATIC_ASSERT(Token::NUM_TOKENS <= 0x100); |
| 97 | |
| 98 | // Table of one-character tokens, by character (0x00..0x7f only). |
| 99 | static const byte one_char_tokens[] = { |
| 100 | Token::ILLEGAL, |
| 101 | Token::ILLEGAL, |
| 102 | Token::ILLEGAL, |
| 103 | Token::ILLEGAL, |
| 104 | Token::ILLEGAL, |
| 105 | Token::ILLEGAL, |
| 106 | Token::ILLEGAL, |
| 107 | Token::ILLEGAL, |
| 108 | Token::ILLEGAL, |
| 109 | Token::ILLEGAL, |
| 110 | Token::ILLEGAL, |
| 111 | Token::ILLEGAL, |
| 112 | Token::ILLEGAL, |
| 113 | Token::ILLEGAL, |
| 114 | Token::ILLEGAL, |
| 115 | Token::ILLEGAL, |
| 116 | Token::ILLEGAL, |
| 117 | Token::ILLEGAL, |
| 118 | Token::ILLEGAL, |
| 119 | Token::ILLEGAL, |
| 120 | Token::ILLEGAL, |
| 121 | Token::ILLEGAL, |
| 122 | Token::ILLEGAL, |
| 123 | Token::ILLEGAL, |
| 124 | Token::ILLEGAL, |
| 125 | Token::ILLEGAL, |
| 126 | Token::ILLEGAL, |
| 127 | Token::ILLEGAL, |
| 128 | Token::ILLEGAL, |
| 129 | Token::ILLEGAL, |
| 130 | Token::ILLEGAL, |
| 131 | Token::ILLEGAL, |
| 132 | Token::ILLEGAL, |
| 133 | Token::ILLEGAL, |
| 134 | Token::ILLEGAL, |
| 135 | Token::ILLEGAL, |
| 136 | Token::ILLEGAL, |
| 137 | Token::ILLEGAL, |
| 138 | Token::ILLEGAL, |
| 139 | Token::ILLEGAL, |
| 140 | Token::LPAREN, // 0x28 |
| 141 | Token::RPAREN, // 0x29 |
| 142 | Token::ILLEGAL, |
| 143 | Token::ILLEGAL, |
| 144 | Token::COMMA, // 0x2c |
| 145 | Token::ILLEGAL, |
| 146 | Token::ILLEGAL, |
| 147 | Token::ILLEGAL, |
| 148 | Token::ILLEGAL, |
| 149 | Token::ILLEGAL, |
| 150 | Token::ILLEGAL, |
| 151 | Token::ILLEGAL, |
| 152 | Token::ILLEGAL, |
| 153 | Token::ILLEGAL, |
| 154 | Token::ILLEGAL, |
| 155 | Token::ILLEGAL, |
| 156 | Token::ILLEGAL, |
| 157 | Token::ILLEGAL, |
| 158 | Token::COLON, // 0x3a |
| 159 | Token::SEMICOLON, // 0x3b |
| 160 | Token::ILLEGAL, |
| 161 | Token::ILLEGAL, |
| 162 | Token::ILLEGAL, |
| 163 | Token::CONDITIONAL, // 0x3f |
| 164 | Token::ILLEGAL, |
| 165 | Token::ILLEGAL, |
| 166 | Token::ILLEGAL, |
| 167 | Token::ILLEGAL, |
| 168 | Token::ILLEGAL, |
| 169 | Token::ILLEGAL, |
| 170 | Token::ILLEGAL, |
| 171 | Token::ILLEGAL, |
| 172 | Token::ILLEGAL, |
| 173 | Token::ILLEGAL, |
| 174 | Token::ILLEGAL, |
| 175 | Token::ILLEGAL, |
| 176 | Token::ILLEGAL, |
| 177 | Token::ILLEGAL, |
| 178 | Token::ILLEGAL, |
| 179 | Token::ILLEGAL, |
| 180 | Token::ILLEGAL, |
| 181 | Token::ILLEGAL, |
| 182 | Token::ILLEGAL, |
| 183 | Token::ILLEGAL, |
| 184 | Token::ILLEGAL, |
| 185 | Token::ILLEGAL, |
| 186 | Token::ILLEGAL, |
| 187 | Token::ILLEGAL, |
| 188 | Token::ILLEGAL, |
| 189 | Token::ILLEGAL, |
| 190 | Token::ILLEGAL, |
| 191 | Token::LBRACK, // 0x5b |
| 192 | Token::ILLEGAL, |
| 193 | Token::RBRACK, // 0x5d |
| 194 | Token::ILLEGAL, |
| 195 | Token::ILLEGAL, |
| 196 | Token::ILLEGAL, |
| 197 | Token::ILLEGAL, |
| 198 | Token::ILLEGAL, |
| 199 | Token::ILLEGAL, |
| 200 | Token::ILLEGAL, |
| 201 | Token::ILLEGAL, |
| 202 | Token::ILLEGAL, |
| 203 | Token::ILLEGAL, |
| 204 | Token::ILLEGAL, |
| 205 | Token::ILLEGAL, |
| 206 | Token::ILLEGAL, |
| 207 | Token::ILLEGAL, |
| 208 | Token::ILLEGAL, |
| 209 | Token::ILLEGAL, |
| 210 | Token::ILLEGAL, |
| 211 | Token::ILLEGAL, |
| 212 | Token::ILLEGAL, |
| 213 | Token::ILLEGAL, |
| 214 | Token::ILLEGAL, |
| 215 | Token::ILLEGAL, |
| 216 | Token::ILLEGAL, |
| 217 | Token::ILLEGAL, |
| 218 | Token::ILLEGAL, |
| 219 | Token::ILLEGAL, |
| 220 | Token::ILLEGAL, |
| 221 | Token::ILLEGAL, |
| 222 | Token::ILLEGAL, |
| 223 | Token::LBRACE, // 0x7b |
| 224 | Token::ILLEGAL, |
| 225 | Token::RBRACE, // 0x7d |
| 226 | Token::BIT_NOT, // 0x7e |
| 227 | Token::ILLEGAL |
| 228 | }; |
| 229 | |
| 230 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 231 | Token::Value JavaScriptScanner::Next() { |
| 232 | current_ = next_; |
| 233 | has_line_terminator_before_next_ = false; |
whesse@chromium.org | df8c03c | 2011-06-21 14:36:03 +0000 | [diff] [blame] | 234 | has_multiline_comment_before_next_ = false; |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 235 | if (static_cast<unsigned>(c0_) <= 0x7f) { |
| 236 | Token::Value token = static_cast<Token::Value>(one_char_tokens[c0_]); |
| 237 | if (token != Token::ILLEGAL) { |
| 238 | int pos = source_pos(); |
| 239 | next_.token = token; |
| 240 | next_.location.beg_pos = pos; |
| 241 | next_.location.end_pos = pos + 1; |
| 242 | Advance(); |
| 243 | return current_.token; |
| 244 | } |
| 245 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 246 | Scan(); |
| 247 | return current_.token; |
| 248 | } |
| 249 | |
| 250 | |
| 251 | static inline bool IsByteOrderMark(uc32 c) { |
| 252 | // The Unicode value U+FFFE is guaranteed never to be assigned as a |
| 253 | // Unicode character; this implies that in a Unicode context the |
| 254 | // 0xFF, 0xFE byte pattern can only be interpreted as the U+FEFF |
| 255 | // character expressed in little-endian byte order (since it could |
| 256 | // not be a U+FFFE character expressed in big-endian byte |
| 257 | // order). Nevertheless, we check for it to be compatible with |
| 258 | // Spidermonkey. |
| 259 | return c == 0xFEFF || c == 0xFFFE; |
| 260 | } |
| 261 | |
| 262 | |
| 263 | bool JavaScriptScanner::SkipWhiteSpace() { |
| 264 | int start_position = source_pos(); |
| 265 | |
| 266 | while (true) { |
| 267 | // We treat byte-order marks (BOMs) as whitespace for better |
| 268 | // compatibility with Spidermonkey and other JavaScript engines. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 269 | while (unicode_cache_->IsWhiteSpace(c0_) || IsByteOrderMark(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 270 | // IsWhiteSpace() includes line terminators! |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 271 | if (unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 272 | // Ignore line terminators, but remember them. This is necessary |
| 273 | // for automatic semicolon insertion. |
| 274 | has_line_terminator_before_next_ = true; |
| 275 | } |
| 276 | Advance(); |
| 277 | } |
| 278 | |
| 279 | // If there is an HTML comment end '-->' at the beginning of a |
| 280 | // line (with only whitespace in front of it), we treat the rest |
| 281 | // of the line as a comment. This is in line with the way |
| 282 | // SpiderMonkey handles it. |
| 283 | if (c0_ == '-' && has_line_terminator_before_next_) { |
| 284 | Advance(); |
| 285 | if (c0_ == '-') { |
| 286 | Advance(); |
| 287 | if (c0_ == '>') { |
| 288 | // Treat the rest of the line as a comment. |
| 289 | SkipSingleLineComment(); |
| 290 | // Continue skipping white space after the comment. |
| 291 | continue; |
| 292 | } |
| 293 | PushBack('-'); // undo Advance() |
| 294 | } |
| 295 | PushBack('-'); // undo Advance() |
| 296 | } |
| 297 | // Return whether or not we skipped any characters. |
| 298 | return source_pos() != start_position; |
| 299 | } |
| 300 | } |
| 301 | |
| 302 | |
| 303 | Token::Value JavaScriptScanner::SkipSingleLineComment() { |
| 304 | Advance(); |
| 305 | |
| 306 | // The line terminator at the end of the line is not considered |
| 307 | // to be part of the single-line comment; it is recognized |
| 308 | // separately by the lexical grammar and becomes part of the |
| 309 | // stream of input elements for the syntactic grammar (see |
whesse@chromium.org | 7b26015 | 2011-06-20 15:33:18 +0000 | [diff] [blame] | 310 | // ECMA-262, section 7.4). |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 311 | while (c0_ >= 0 && !unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 312 | Advance(); |
| 313 | } |
| 314 | |
| 315 | return Token::WHITESPACE; |
| 316 | } |
| 317 | |
| 318 | |
| 319 | Token::Value JavaScriptScanner::SkipMultiLineComment() { |
| 320 | ASSERT(c0_ == '*'); |
| 321 | Advance(); |
| 322 | |
| 323 | while (c0_ >= 0) { |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 324 | uc32 ch = c0_; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 325 | Advance(); |
whesse@chromium.org | 7b26015 | 2011-06-20 15:33:18 +0000 | [diff] [blame] | 326 | if (unicode_cache_->IsLineTerminator(ch)) { |
| 327 | // Following ECMA-262, section 7.4, a comment containing |
| 328 | // a newline will make the comment count as a line-terminator. |
whesse@chromium.org | df8c03c | 2011-06-21 14:36:03 +0000 | [diff] [blame] | 329 | has_multiline_comment_before_next_ = true; |
whesse@chromium.org | 7b26015 | 2011-06-20 15:33:18 +0000 | [diff] [blame] | 330 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 331 | // If we have reached the end of the multi-line comment, we |
| 332 | // consume the '/' and insert a whitespace. This way all |
whesse@chromium.org | 7b26015 | 2011-06-20 15:33:18 +0000 | [diff] [blame] | 333 | // multi-line comments are treated as whitespace. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 334 | if (ch == '*' && c0_ == '/') { |
| 335 | c0_ = ' '; |
| 336 | return Token::WHITESPACE; |
| 337 | } |
| 338 | } |
| 339 | |
| 340 | // Unterminated multi-line comment. |
| 341 | return Token::ILLEGAL; |
| 342 | } |
| 343 | |
| 344 | |
| 345 | Token::Value JavaScriptScanner::ScanHtmlComment() { |
| 346 | // Check for <!-- comments. |
| 347 | ASSERT(c0_ == '!'); |
| 348 | Advance(); |
| 349 | if (c0_ == '-') { |
| 350 | Advance(); |
| 351 | if (c0_ == '-') return SkipSingleLineComment(); |
| 352 | PushBack('-'); // undo Advance() |
| 353 | } |
| 354 | PushBack('!'); // undo Advance() |
| 355 | ASSERT(c0_ == '!'); |
| 356 | return Token::LT; |
| 357 | } |
| 358 | |
| 359 | |
| 360 | void JavaScriptScanner::Scan() { |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 361 | next_.literal_chars = NULL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 362 | Token::Value token; |
| 363 | do { |
| 364 | // Remember the position of the next token |
| 365 | next_.location.beg_pos = source_pos(); |
| 366 | |
| 367 | switch (c0_) { |
| 368 | case ' ': |
| 369 | case '\t': |
| 370 | Advance(); |
| 371 | token = Token::WHITESPACE; |
| 372 | break; |
| 373 | |
| 374 | case '\n': |
| 375 | Advance(); |
| 376 | has_line_terminator_before_next_ = true; |
| 377 | token = Token::WHITESPACE; |
| 378 | break; |
| 379 | |
| 380 | case '"': case '\'': |
| 381 | token = ScanString(); |
| 382 | break; |
| 383 | |
| 384 | case '<': |
| 385 | // < <= << <<= <!-- |
| 386 | Advance(); |
| 387 | if (c0_ == '=') { |
| 388 | token = Select(Token::LTE); |
| 389 | } else if (c0_ == '<') { |
| 390 | token = Select('=', Token::ASSIGN_SHL, Token::SHL); |
| 391 | } else if (c0_ == '!') { |
| 392 | token = ScanHtmlComment(); |
| 393 | } else { |
| 394 | token = Token::LT; |
| 395 | } |
| 396 | break; |
| 397 | |
| 398 | case '>': |
| 399 | // > >= >> >>= >>> >>>= |
| 400 | Advance(); |
| 401 | if (c0_ == '=') { |
| 402 | token = Select(Token::GTE); |
| 403 | } else if (c0_ == '>') { |
| 404 | // >> >>= >>> >>>= |
| 405 | Advance(); |
| 406 | if (c0_ == '=') { |
| 407 | token = Select(Token::ASSIGN_SAR); |
| 408 | } else if (c0_ == '>') { |
| 409 | token = Select('=', Token::ASSIGN_SHR, Token::SHR); |
| 410 | } else { |
| 411 | token = Token::SAR; |
| 412 | } |
| 413 | } else { |
| 414 | token = Token::GT; |
| 415 | } |
| 416 | break; |
| 417 | |
| 418 | case '=': |
| 419 | // = == === |
| 420 | Advance(); |
| 421 | if (c0_ == '=') { |
| 422 | token = Select('=', Token::EQ_STRICT, Token::EQ); |
| 423 | } else { |
| 424 | token = Token::ASSIGN; |
| 425 | } |
| 426 | break; |
| 427 | |
| 428 | case '!': |
| 429 | // ! != !== |
| 430 | Advance(); |
| 431 | if (c0_ == '=') { |
| 432 | token = Select('=', Token::NE_STRICT, Token::NE); |
| 433 | } else { |
| 434 | token = Token::NOT; |
| 435 | } |
| 436 | break; |
| 437 | |
| 438 | case '+': |
| 439 | // + ++ += |
| 440 | Advance(); |
| 441 | if (c0_ == '+') { |
| 442 | token = Select(Token::INC); |
| 443 | } else if (c0_ == '=') { |
| 444 | token = Select(Token::ASSIGN_ADD); |
| 445 | } else { |
| 446 | token = Token::ADD; |
| 447 | } |
| 448 | break; |
| 449 | |
| 450 | case '-': |
| 451 | // - -- --> -= |
| 452 | Advance(); |
| 453 | if (c0_ == '-') { |
| 454 | Advance(); |
| 455 | if (c0_ == '>' && has_line_terminator_before_next_) { |
| 456 | // For compatibility with SpiderMonkey, we skip lines that |
| 457 | // start with an HTML comment end '-->'. |
| 458 | token = SkipSingleLineComment(); |
| 459 | } else { |
| 460 | token = Token::DEC; |
| 461 | } |
| 462 | } else if (c0_ == '=') { |
| 463 | token = Select(Token::ASSIGN_SUB); |
| 464 | } else { |
| 465 | token = Token::SUB; |
| 466 | } |
| 467 | break; |
| 468 | |
| 469 | case '*': |
| 470 | // * *= |
| 471 | token = Select('=', Token::ASSIGN_MUL, Token::MUL); |
| 472 | break; |
| 473 | |
| 474 | case '%': |
| 475 | // % %= |
| 476 | token = Select('=', Token::ASSIGN_MOD, Token::MOD); |
| 477 | break; |
| 478 | |
| 479 | case '/': |
| 480 | // / // /* /= |
| 481 | Advance(); |
| 482 | if (c0_ == '/') { |
| 483 | token = SkipSingleLineComment(); |
| 484 | } else if (c0_ == '*') { |
| 485 | token = SkipMultiLineComment(); |
| 486 | } else if (c0_ == '=') { |
| 487 | token = Select(Token::ASSIGN_DIV); |
| 488 | } else { |
| 489 | token = Token::DIV; |
| 490 | } |
| 491 | break; |
| 492 | |
| 493 | case '&': |
| 494 | // & && &= |
| 495 | Advance(); |
| 496 | if (c0_ == '&') { |
| 497 | token = Select(Token::AND); |
| 498 | } else if (c0_ == '=') { |
| 499 | token = Select(Token::ASSIGN_BIT_AND); |
| 500 | } else { |
| 501 | token = Token::BIT_AND; |
| 502 | } |
| 503 | break; |
| 504 | |
| 505 | case '|': |
| 506 | // | || |= |
| 507 | Advance(); |
| 508 | if (c0_ == '|') { |
| 509 | token = Select(Token::OR); |
| 510 | } else if (c0_ == '=') { |
| 511 | token = Select(Token::ASSIGN_BIT_OR); |
| 512 | } else { |
| 513 | token = Token::BIT_OR; |
| 514 | } |
| 515 | break; |
| 516 | |
| 517 | case '^': |
| 518 | // ^ ^= |
| 519 | token = Select('=', Token::ASSIGN_BIT_XOR, Token::BIT_XOR); |
| 520 | break; |
| 521 | |
| 522 | case '.': |
| 523 | // . Number |
| 524 | Advance(); |
| 525 | if (IsDecimalDigit(c0_)) { |
| 526 | token = ScanNumber(true); |
| 527 | } else { |
| 528 | token = Token::PERIOD; |
| 529 | } |
| 530 | break; |
| 531 | |
| 532 | case ':': |
| 533 | token = Select(Token::COLON); |
| 534 | break; |
| 535 | |
| 536 | case ';': |
| 537 | token = Select(Token::SEMICOLON); |
| 538 | break; |
| 539 | |
| 540 | case ',': |
| 541 | token = Select(Token::COMMA); |
| 542 | break; |
| 543 | |
| 544 | case '(': |
| 545 | token = Select(Token::LPAREN); |
| 546 | break; |
| 547 | |
| 548 | case ')': |
| 549 | token = Select(Token::RPAREN); |
| 550 | break; |
| 551 | |
| 552 | case '[': |
| 553 | token = Select(Token::LBRACK); |
| 554 | break; |
| 555 | |
| 556 | case ']': |
| 557 | token = Select(Token::RBRACK); |
| 558 | break; |
| 559 | |
| 560 | case '{': |
| 561 | token = Select(Token::LBRACE); |
| 562 | break; |
| 563 | |
| 564 | case '}': |
| 565 | token = Select(Token::RBRACE); |
| 566 | break; |
| 567 | |
| 568 | case '?': |
| 569 | token = Select(Token::CONDITIONAL); |
| 570 | break; |
| 571 | |
| 572 | case '~': |
| 573 | token = Select(Token::BIT_NOT); |
| 574 | break; |
| 575 | |
| 576 | default: |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 577 | if (unicode_cache_->IsIdentifierStart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 578 | token = ScanIdentifierOrKeyword(); |
| 579 | } else if (IsDecimalDigit(c0_)) { |
| 580 | token = ScanNumber(false); |
| 581 | } else if (SkipWhiteSpace()) { |
| 582 | token = Token::WHITESPACE; |
| 583 | } else if (c0_ < 0) { |
| 584 | token = Token::EOS; |
| 585 | } else { |
| 586 | token = Select(Token::ILLEGAL); |
| 587 | } |
| 588 | break; |
| 589 | } |
| 590 | |
| 591 | // Continue scanning for tokens as long as we're just skipping |
| 592 | // whitespace. |
| 593 | } while (token == Token::WHITESPACE); |
| 594 | |
| 595 | next_.location.end_pos = source_pos(); |
| 596 | next_.token = token; |
| 597 | } |
| 598 | |
| 599 | |
| 600 | void JavaScriptScanner::SeekForward(int pos) { |
ager@chromium.org | 5f0c45f | 2010-12-17 08:51:21 +0000 | [diff] [blame] | 601 | // After this call, we will have the token at the given position as |
| 602 | // the "next" token. The "current" token will be invalid. |
| 603 | if (pos == next_.location.beg_pos) return; |
| 604 | int current_pos = source_pos(); |
| 605 | ASSERT_EQ(next_.location.end_pos, current_pos); |
| 606 | // Positions inside the lookahead token aren't supported. |
| 607 | ASSERT(pos >= current_pos); |
| 608 | if (pos != current_pos) { |
| 609 | source_->SeekForward(pos - source_->pos()); |
| 610 | Advance(); |
| 611 | // This function is only called to seek to the location |
| 612 | // of the end of a function (at the "}" token). It doesn't matter |
| 613 | // whether there was a line terminator in the part we skip. |
| 614 | has_line_terminator_before_next_ = false; |
whesse@chromium.org | df8c03c | 2011-06-21 14:36:03 +0000 | [diff] [blame] | 615 | has_multiline_comment_before_next_ = false; |
ager@chromium.org | 5f0c45f | 2010-12-17 08:51:21 +0000 | [diff] [blame] | 616 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 617 | Scan(); |
| 618 | } |
| 619 | |
| 620 | |
| 621 | void JavaScriptScanner::ScanEscape() { |
| 622 | uc32 c = c0_; |
| 623 | Advance(); |
| 624 | |
| 625 | // Skip escaped newlines. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 626 | if (unicode_cache_->IsLineTerminator(c)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 627 | // Allow CR+LF newlines in multiline string literals. |
| 628 | if (IsCarriageReturn(c) && IsLineFeed(c0_)) Advance(); |
| 629 | // Allow LF+CR newlines in multiline string literals. |
| 630 | if (IsLineFeed(c) && IsCarriageReturn(c0_)) Advance(); |
| 631 | return; |
| 632 | } |
| 633 | |
| 634 | switch (c) { |
| 635 | case '\'': // fall through |
| 636 | case '"' : // fall through |
| 637 | case '\\': break; |
| 638 | case 'b' : c = '\b'; break; |
| 639 | case 'f' : c = '\f'; break; |
| 640 | case 'n' : c = '\n'; break; |
| 641 | case 'r' : c = '\r'; break; |
| 642 | case 't' : c = '\t'; break; |
| 643 | case 'u' : c = ScanHexEscape(c, 4); break; |
| 644 | case 'v' : c = '\v'; break; |
| 645 | case 'x' : c = ScanHexEscape(c, 2); break; |
| 646 | case '0' : // fall through |
| 647 | case '1' : // fall through |
| 648 | case '2' : // fall through |
| 649 | case '3' : // fall through |
| 650 | case '4' : // fall through |
| 651 | case '5' : // fall through |
| 652 | case '6' : // fall through |
| 653 | case '7' : c = ScanOctalEscape(c, 2); break; |
| 654 | } |
| 655 | |
| 656 | // According to ECMA-262, 3rd, 7.8.4 (p 18ff) these |
| 657 | // should be illegal, but they are commonly handled |
| 658 | // as non-escaped characters by JS VMs. |
| 659 | AddLiteralChar(c); |
| 660 | } |
| 661 | |
| 662 | |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 663 | // Octal escapes of the forms '\0xx' and '\xxx' are not a part of |
| 664 | // ECMA-262. Other JS VMs support them. |
| 665 | uc32 JavaScriptScanner::ScanOctalEscape(uc32 c, int length) { |
| 666 | uc32 x = c - '0'; |
| 667 | int i = 0; |
| 668 | for (; i < length; i++) { |
| 669 | int d = c0_ - '0'; |
| 670 | if (d < 0 || d > 7) break; |
| 671 | int nx = x * 8 + d; |
| 672 | if (nx >= 256) break; |
| 673 | x = nx; |
| 674 | Advance(); |
| 675 | } |
| 676 | // Anything except '\0' is an octal escape sequence, illegal in strict mode. |
| 677 | // Remember the position of octal escape sequences so that an error |
| 678 | // can be reported later (in strict mode). |
| 679 | // We don't report the error immediately, because the octal escape can |
| 680 | // occur before the "use strict" directive. |
| 681 | if (c != '0' || i > 0) { |
| 682 | octal_pos_ = Location(source_pos() - i - 1, source_pos() - 1); |
| 683 | } |
| 684 | return x; |
| 685 | } |
| 686 | |
| 687 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 688 | Token::Value JavaScriptScanner::ScanString() { |
| 689 | uc32 quote = c0_; |
| 690 | Advance(); // consume quote |
| 691 | |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 692 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 693 | while (c0_ != quote && c0_ >= 0 |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 694 | && !unicode_cache_->IsLineTerminator(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 695 | uc32 c = c0_; |
| 696 | Advance(); |
| 697 | if (c == '\\') { |
| 698 | if (c0_ < 0) return Token::ILLEGAL; |
| 699 | ScanEscape(); |
| 700 | } else { |
| 701 | AddLiteralChar(c); |
| 702 | } |
| 703 | } |
| 704 | if (c0_ != quote) return Token::ILLEGAL; |
| 705 | literal.Complete(); |
| 706 | |
| 707 | Advance(); // consume quote |
| 708 | return Token::STRING; |
| 709 | } |
| 710 | |
| 711 | |
| 712 | void JavaScriptScanner::ScanDecimalDigits() { |
| 713 | while (IsDecimalDigit(c0_)) |
| 714 | AddLiteralCharAdvance(); |
| 715 | } |
| 716 | |
| 717 | |
| 718 | Token::Value JavaScriptScanner::ScanNumber(bool seen_period) { |
| 719 | ASSERT(IsDecimalDigit(c0_)); // the first digit of the number or the fraction |
| 720 | |
| 721 | enum { DECIMAL, HEX, OCTAL } kind = DECIMAL; |
| 722 | |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 723 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 724 | if (seen_period) { |
| 725 | // we have already seen a decimal point of the float |
| 726 | AddLiteralChar('.'); |
| 727 | ScanDecimalDigits(); // we know we have at least one digit |
| 728 | |
| 729 | } else { |
| 730 | // if the first character is '0' we must check for octals and hex |
| 731 | if (c0_ == '0') { |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 732 | int start_pos = source_pos(); // For reporting octal positions. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 733 | AddLiteralCharAdvance(); |
| 734 | |
| 735 | // either 0, 0exxx, 0Exxx, 0.xxx, an octal number, or a hex number |
| 736 | if (c0_ == 'x' || c0_ == 'X') { |
| 737 | // hex number |
| 738 | kind = HEX; |
| 739 | AddLiteralCharAdvance(); |
| 740 | if (!IsHexDigit(c0_)) { |
| 741 | // we must have at least one hex digit after 'x'/'X' |
| 742 | return Token::ILLEGAL; |
| 743 | } |
| 744 | while (IsHexDigit(c0_)) { |
| 745 | AddLiteralCharAdvance(); |
| 746 | } |
| 747 | } else if ('0' <= c0_ && c0_ <= '7') { |
| 748 | // (possible) octal number |
| 749 | kind = OCTAL; |
| 750 | while (true) { |
| 751 | if (c0_ == '8' || c0_ == '9') { |
| 752 | kind = DECIMAL; |
| 753 | break; |
| 754 | } |
ager@chromium.org | 0ee099b | 2011-01-25 14:06:47 +0000 | [diff] [blame] | 755 | if (c0_ < '0' || '7' < c0_) { |
| 756 | // Octal literal finished. |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 757 | octal_pos_ = Location(start_pos, source_pos()); |
ager@chromium.org | 0ee099b | 2011-01-25 14:06:47 +0000 | [diff] [blame] | 758 | break; |
| 759 | } |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 760 | AddLiteralCharAdvance(); |
| 761 | } |
| 762 | } |
| 763 | } |
| 764 | |
| 765 | // Parse decimal digits and allow trailing fractional part. |
| 766 | if (kind == DECIMAL) { |
| 767 | ScanDecimalDigits(); // optional |
| 768 | if (c0_ == '.') { |
| 769 | AddLiteralCharAdvance(); |
| 770 | ScanDecimalDigits(); // optional |
| 771 | } |
| 772 | } |
| 773 | } |
| 774 | |
| 775 | // scan exponent, if any |
| 776 | if (c0_ == 'e' || c0_ == 'E') { |
| 777 | ASSERT(kind != HEX); // 'e'/'E' must be scanned as part of the hex number |
| 778 | if (kind == OCTAL) return Token::ILLEGAL; // no exponent for octals allowed |
| 779 | // scan exponent |
| 780 | AddLiteralCharAdvance(); |
| 781 | if (c0_ == '+' || c0_ == '-') |
| 782 | AddLiteralCharAdvance(); |
| 783 | if (!IsDecimalDigit(c0_)) { |
| 784 | // we must have at least one decimal digit after 'e'/'E' |
| 785 | return Token::ILLEGAL; |
| 786 | } |
| 787 | ScanDecimalDigits(); |
| 788 | } |
| 789 | |
| 790 | // The source character immediately following a numeric literal must |
| 791 | // not be an identifier start or a decimal digit; see ECMA-262 |
| 792 | // section 7.8.3, page 17 (note that we read only one decimal digit |
| 793 | // if the value is 0). |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 794 | if (IsDecimalDigit(c0_) || unicode_cache_->IsIdentifierStart(c0_)) |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 795 | return Token::ILLEGAL; |
| 796 | |
| 797 | literal.Complete(); |
| 798 | |
| 799 | return Token::NUMBER; |
| 800 | } |
| 801 | |
| 802 | |
| 803 | uc32 JavaScriptScanner::ScanIdentifierUnicodeEscape() { |
| 804 | Advance(); |
| 805 | if (c0_ != 'u') return unibrow::Utf8::kBadChar; |
| 806 | Advance(); |
| 807 | uc32 c = ScanHexEscape('u', 4); |
| 808 | // We do not allow a unicode escape sequence to start another |
| 809 | // unicode escape sequence. |
| 810 | if (c == '\\') return unibrow::Utf8::kBadChar; |
| 811 | return c; |
| 812 | } |
| 813 | |
| 814 | |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 815 | // ---------------------------------------------------------------------------- |
| 816 | // Keyword Matcher |
| 817 | |
danno@chromium.org | b645116 | 2011-08-17 14:33:23 +0000 | [diff] [blame^] | 818 | #define KEYWORDS(KEYWORD_GROUP, KEYWORD) \ |
| 819 | KEYWORD_GROUP('b') \ |
| 820 | KEYWORD("break", Token::BREAK) \ |
| 821 | KEYWORD_GROUP('c') \ |
| 822 | KEYWORD("case", Token::CASE) \ |
| 823 | KEYWORD("catch", Token::CATCH) \ |
| 824 | KEYWORD("class", Token::FUTURE_RESERVED_WORD) \ |
| 825 | KEYWORD("const", Token::CONST) \ |
| 826 | KEYWORD("continue", Token::CONTINUE) \ |
| 827 | KEYWORD_GROUP('d') \ |
| 828 | KEYWORD("debugger", Token::DEBUGGER) \ |
| 829 | KEYWORD("default", Token::DEFAULT) \ |
| 830 | KEYWORD("delete", Token::DELETE) \ |
| 831 | KEYWORD("do", Token::DO) \ |
| 832 | KEYWORD_GROUP('e') \ |
| 833 | KEYWORD("else", Token::ELSE) \ |
| 834 | KEYWORD("enum", Token::FUTURE_RESERVED_WORD) \ |
| 835 | KEYWORD("export", Token::FUTURE_RESERVED_WORD) \ |
| 836 | KEYWORD("extends", Token::FUTURE_RESERVED_WORD) \ |
| 837 | KEYWORD_GROUP('f') \ |
| 838 | KEYWORD("false", Token::FALSE_LITERAL) \ |
| 839 | KEYWORD("finally", Token::FINALLY) \ |
| 840 | KEYWORD("for", Token::FOR) \ |
| 841 | KEYWORD("function", Token::FUNCTION) \ |
| 842 | KEYWORD_GROUP('i') \ |
| 843 | KEYWORD("if", Token::IF) \ |
| 844 | KEYWORD("implements", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 845 | KEYWORD("import", Token::FUTURE_RESERVED_WORD) \ |
| 846 | KEYWORD("in", Token::IN) \ |
| 847 | KEYWORD("instanceof", Token::INSTANCEOF) \ |
| 848 | KEYWORD("interface", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 849 | KEYWORD_GROUP('l') \ |
| 850 | KEYWORD("let", harmony_block_scoping \ |
| 851 | ? Token::LET : Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 852 | KEYWORD_GROUP('n') \ |
| 853 | KEYWORD("new", Token::NEW) \ |
| 854 | KEYWORD("null", Token::NULL_LITERAL) \ |
| 855 | KEYWORD_GROUP('p') \ |
| 856 | KEYWORD("package", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 857 | KEYWORD("private", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 858 | KEYWORD("protected", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 859 | KEYWORD("public", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 860 | KEYWORD_GROUP('r') \ |
| 861 | KEYWORD("return", Token::RETURN) \ |
| 862 | KEYWORD_GROUP('s') \ |
| 863 | KEYWORD("static", Token::FUTURE_STRICT_RESERVED_WORD) \ |
| 864 | KEYWORD("super", Token::FUTURE_RESERVED_WORD) \ |
| 865 | KEYWORD("switch", Token::SWITCH) \ |
| 866 | KEYWORD_GROUP('t') \ |
| 867 | KEYWORD("this", Token::THIS) \ |
| 868 | KEYWORD("throw", Token::THROW) \ |
| 869 | KEYWORD("true", Token::TRUE_LITERAL) \ |
| 870 | KEYWORD("try", Token::TRY) \ |
| 871 | KEYWORD("typeof", Token::TYPEOF) \ |
| 872 | KEYWORD_GROUP('v') \ |
| 873 | KEYWORD("var", Token::VAR) \ |
| 874 | KEYWORD("void", Token::VOID) \ |
| 875 | KEYWORD_GROUP('w') \ |
| 876 | KEYWORD("while", Token::WHILE) \ |
| 877 | KEYWORD("with", Token::WITH) \ |
| 878 | KEYWORD_GROUP('y') \ |
| 879 | KEYWORD("yield", Token::FUTURE_STRICT_RESERVED_WORD) |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 880 | |
| 881 | |
| 882 | static Token::Value KeywordOrIdentifierToken(const char* input, |
danno@chromium.org | b645116 | 2011-08-17 14:33:23 +0000 | [diff] [blame^] | 883 | int input_length, |
| 884 | bool harmony_block_scoping) { |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 885 | ASSERT(input_length >= 1); |
| 886 | const int kMinLength = 2; |
| 887 | const int kMaxLength = 10; |
| 888 | if (input_length < kMinLength || input_length > kMaxLength) { |
| 889 | return Token::IDENTIFIER; |
| 890 | } |
| 891 | switch (input[0]) { |
| 892 | default: |
| 893 | #define KEYWORD_GROUP_CASE(ch) \ |
| 894 | break; \ |
| 895 | case ch: |
| 896 | #define KEYWORD(keyword, token) \ |
| 897 | { \ |
| 898 | /* 'keyword' is a char array, so sizeof(keyword) is */ \ |
| 899 | /* strlen(keyword) plus 1 for the NUL char. */ \ |
| 900 | const int keyword_length = sizeof(keyword) - 1; \ |
| 901 | STATIC_ASSERT(keyword_length >= kMinLength); \ |
| 902 | STATIC_ASSERT(keyword_length <= kMaxLength); \ |
| 903 | if (input_length == keyword_length && \ |
| 904 | input[1] == keyword[1] && \ |
| 905 | (keyword_length <= 2 || input[2] == keyword[2]) && \ |
| 906 | (keyword_length <= 3 || input[3] == keyword[3]) && \ |
| 907 | (keyword_length <= 4 || input[4] == keyword[4]) && \ |
| 908 | (keyword_length <= 5 || input[5] == keyword[5]) && \ |
| 909 | (keyword_length <= 6 || input[6] == keyword[6]) && \ |
| 910 | (keyword_length <= 7 || input[7] == keyword[7]) && \ |
| 911 | (keyword_length <= 8 || input[8] == keyword[8]) && \ |
| 912 | (keyword_length <= 9 || input[9] == keyword[9])) { \ |
danno@chromium.org | b645116 | 2011-08-17 14:33:23 +0000 | [diff] [blame^] | 913 | return token; \ |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 914 | } \ |
| 915 | } |
| 916 | KEYWORDS(KEYWORD_GROUP_CASE, KEYWORD) |
| 917 | } |
| 918 | return Token::IDENTIFIER; |
| 919 | } |
| 920 | |
| 921 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 922 | Token::Value JavaScriptScanner::ScanIdentifierOrKeyword() { |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 923 | ASSERT(unicode_cache_->IsIdentifierStart(c0_)); |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 924 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 925 | // Scan identifier start character. |
| 926 | if (c0_ == '\\') { |
| 927 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 928 | // Only allow legal identifier start characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 929 | if (!unicode_cache_->IsIdentifierStart(c)) return Token::ILLEGAL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 930 | AddLiteralChar(c); |
| 931 | return ScanIdentifierSuffix(&literal); |
| 932 | } |
| 933 | |
| 934 | uc32 first_char = c0_; |
| 935 | Advance(); |
| 936 | AddLiteralChar(first_char); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 937 | |
| 938 | // Scan the rest of the identifier characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 939 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 940 | if (c0_ != '\\') { |
| 941 | uc32 next_char = c0_; |
| 942 | Advance(); |
| 943 | AddLiteralChar(next_char); |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 944 | continue; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 945 | } |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 946 | // Fallthrough if no longer able to complete keyword. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 947 | return ScanIdentifierSuffix(&literal); |
| 948 | } |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 949 | |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 950 | literal.Complete(); |
| 951 | |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 952 | if (next_.literal_chars->is_ascii()) { |
| 953 | Vector<const char> chars = next_.literal_chars->ascii_literal(); |
danno@chromium.org | b645116 | 2011-08-17 14:33:23 +0000 | [diff] [blame^] | 954 | return KeywordOrIdentifierToken(chars.start(), |
| 955 | chars.length(), |
| 956 | harmony_block_scoping_); |
kmillikin@chromium.org | 7c2628c | 2011-08-10 11:27:35 +0000 | [diff] [blame] | 957 | } |
| 958 | |
| 959 | return Token::IDENTIFIER; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 960 | } |
| 961 | |
| 962 | |
| 963 | Token::Value JavaScriptScanner::ScanIdentifierSuffix(LiteralScope* literal) { |
| 964 | // Scan the rest of the identifier characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 965 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 966 | if (c0_ == '\\') { |
| 967 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 968 | // Only allow legal identifier part characters. |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 969 | if (!unicode_cache_->IsIdentifierPart(c)) return Token::ILLEGAL; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 970 | AddLiteralChar(c); |
| 971 | } else { |
| 972 | AddLiteralChar(c0_); |
| 973 | Advance(); |
| 974 | } |
| 975 | } |
| 976 | literal->Complete(); |
| 977 | |
| 978 | return Token::IDENTIFIER; |
| 979 | } |
| 980 | |
| 981 | |
| 982 | bool JavaScriptScanner::ScanRegExpPattern(bool seen_equal) { |
| 983 | // Scan: ('/' | '/=') RegularExpressionBody '/' RegularExpressionFlags |
| 984 | bool in_character_class = false; |
| 985 | |
| 986 | // Previous token is either '/' or '/=', in the second case, the |
| 987 | // pattern starts at =. |
| 988 | next_.location.beg_pos = source_pos() - (seen_equal ? 2 : 1); |
| 989 | next_.location.end_pos = source_pos() - (seen_equal ? 1 : 0); |
| 990 | |
| 991 | // Scan regular expression body: According to ECMA-262, 3rd, 7.8.5, |
| 992 | // the scanner should pass uninterpreted bodies to the RegExp |
| 993 | // constructor. |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 994 | LiteralScope literal(this); |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 995 | if (seen_equal) |
| 996 | AddLiteralChar('='); |
| 997 | |
| 998 | while (c0_ != '/' || in_character_class) { |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 999 | if (unicode_cache_->IsLineTerminator(c0_) || c0_ < 0) return false; |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 1000 | if (c0_ == '\\') { // Escape sequence. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 1001 | AddLiteralCharAdvance(); |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 1002 | if (unicode_cache_->IsLineTerminator(c0_) || c0_ < 0) return false; |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 1003 | AddLiteralCharAdvance(); |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 1004 | // If the escape allows more characters, i.e., \x??, \u????, or \c?, |
| 1005 | // only "safe" characters are allowed (letters, digits, underscore), |
| 1006 | // otherwise the escape isn't valid and the invalid character has |
| 1007 | // its normal meaning. I.e., we can just continue scanning without |
| 1008 | // worrying whether the following characters are part of the escape |
| 1009 | // or not, since any '/', '\\' or '[' is guaranteed to not be part |
| 1010 | // of the escape sequence. |
lrn@chromium.org | 1c09276 | 2011-05-09 09:42:16 +0000 | [diff] [blame] | 1011 | |
| 1012 | // TODO(896): At some point, parse RegExps more throughly to capture |
| 1013 | // octal esacpes in strict mode. |
kmillikin@chromium.org | d2c22f0 | 2011-01-10 08:15:37 +0000 | [diff] [blame] | 1014 | } else { // Unescaped character. |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 1015 | if (c0_ == '[') in_character_class = true; |
| 1016 | if (c0_ == ']') in_character_class = false; |
| 1017 | AddLiteralCharAdvance(); |
| 1018 | } |
| 1019 | } |
| 1020 | Advance(); // consume '/' |
| 1021 | |
| 1022 | literal.Complete(); |
| 1023 | |
| 1024 | return true; |
| 1025 | } |
| 1026 | |
| 1027 | |
| 1028 | bool JavaScriptScanner::ScanRegExpFlags() { |
| 1029 | // Scan regular expression flags. |
fschneider@chromium.org | 9e3e0b6 | 2011-01-03 10:16:46 +0000 | [diff] [blame] | 1030 | LiteralScope literal(this); |
ager@chromium.org | a9aa5fa | 2011-04-13 08:46:07 +0000 | [diff] [blame] | 1031 | while (unicode_cache_->IsIdentifierPart(c0_)) { |
vegorov@chromium.org | 21b5e95 | 2010-11-23 10:24:40 +0000 | [diff] [blame] | 1032 | if (c0_ == '\\') { |
| 1033 | uc32 c = ScanIdentifierUnicodeEscape(); |
| 1034 | if (c != static_cast<uc32>(unibrow::Utf8::kBadChar)) { |
| 1035 | // We allow any escaped character, unlike the restriction on |
| 1036 | // IdentifierPart when it is used to build an IdentifierName. |
| 1037 | AddLiteralChar(c); |
| 1038 | continue; |
| 1039 | } |
| 1040 | } |
| 1041 | AddLiteralCharAdvance(); |
| 1042 | } |
| 1043 | literal.Complete(); |
| 1044 | |
| 1045 | next_.location.end_pos = source_pos() - 1; |
| 1046 | return true; |
| 1047 | } |
| 1048 | |
whesse@chromium.org | f0ac72d | 2010-11-08 12:47:26 +0000 | [diff] [blame] | 1049 | } } // namespace v8::internal |