Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 1 | // Copyright 2006-2008 the V8 project authors. All rights reserved. |
| 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 | |
| 29 | // This file relies on the fact that the following declaration has been made |
| 30 | // in runtime.js: |
| 31 | // const $String = global.String; |
| 32 | // const $NaN = 0/0; |
| 33 | |
| 34 | |
| 35 | // Set the String function and constructor. |
| 36 | %SetCode($String, function(x) { |
| 37 | var value = %_ArgumentsLength() == 0 ? '' : ToString(x); |
| 38 | if (%_IsConstructCall()) { |
| 39 | %_SetValueOf(this, value); |
| 40 | } else { |
| 41 | return value; |
| 42 | } |
| 43 | }); |
| 44 | |
| 45 | %FunctionSetPrototype($String, new $String()); |
| 46 | |
| 47 | // ECMA-262 section 15.5.4.2 |
| 48 | function StringToString() { |
| 49 | if (!IS_STRING(this) && !IS_STRING_WRAPPER(this)) |
| 50 | throw new $TypeError('String.prototype.toString is not generic'); |
| 51 | return %_ValueOf(this); |
| 52 | } |
| 53 | |
| 54 | |
| 55 | // ECMA-262 section 15.5.4.3 |
| 56 | function StringValueOf() { |
| 57 | if (!IS_STRING(this) && !IS_STRING_WRAPPER(this)) |
| 58 | throw new $TypeError('String.prototype.valueOf is not generic'); |
| 59 | return %_ValueOf(this); |
| 60 | } |
| 61 | |
| 62 | |
| 63 | // ECMA-262, section 15.5.4.4 |
| 64 | function StringCharAt(pos) { |
| 65 | var char_code = %_FastCharCodeAt(this, pos); |
| 66 | if (!%_IsSmi(char_code)) { |
| 67 | var subject = ToString(this); |
| 68 | var index = TO_INTEGER(pos); |
| 69 | if (index >= subject.length || index < 0) return ""; |
| 70 | char_code = %StringCharCodeAt(subject, index); |
| 71 | } |
| 72 | return %CharFromCode(char_code); |
| 73 | } |
| 74 | |
| 75 | |
| 76 | // ECMA-262 section 15.5.4.5 |
| 77 | function StringCharCodeAt(pos) { |
| 78 | var fast_answer = %_FastCharCodeAt(this, pos); |
| 79 | if (%_IsSmi(fast_answer)) { |
| 80 | return fast_answer; |
| 81 | } |
| 82 | var subject = ToString(this); |
| 83 | var index = TO_INTEGER(pos); |
| 84 | return %StringCharCodeAt(subject, index); |
| 85 | } |
| 86 | |
| 87 | |
| 88 | // ECMA-262, section 15.5.4.6 |
| 89 | function StringConcat() { |
| 90 | var len = %_ArgumentsLength(); |
| 91 | var parts = new $Array(len + 1); |
| 92 | parts[0] = ToString(this); |
| 93 | for (var i = 0; i < len; i++) |
| 94 | parts[i + 1] = ToString(%_Arguments(i)); |
| 95 | return parts.join(''); |
| 96 | } |
| 97 | |
| 98 | // Match ES3 and Safari |
| 99 | %FunctionSetLength(StringConcat, 1); |
| 100 | |
| 101 | |
| 102 | // ECMA-262 section 15.5.4.7 |
| 103 | function StringIndexOf(searchString /* position */) { // length == 1 |
| 104 | var subject_str = ToString(this); |
| 105 | var pattern_str = ToString(searchString); |
| 106 | var subject_str_len = subject_str.length; |
| 107 | var pattern_str_len = pattern_str.length; |
| 108 | var index = 0; |
| 109 | if (%_ArgumentsLength() > 1) { |
| 110 | var arg1 = %_Arguments(1); // position |
| 111 | index = TO_INTEGER(arg1); |
| 112 | } |
| 113 | if (index < 0) index = 0; |
| 114 | if (index > subject_str_len) index = subject_str_len; |
| 115 | if (pattern_str_len + index > subject_str_len) return -1; |
| 116 | return %StringIndexOf(subject_str, pattern_str, index); |
| 117 | } |
| 118 | |
| 119 | |
| 120 | // ECMA-262 section 15.5.4.8 |
| 121 | function StringLastIndexOf(searchString /* position */) { // length == 1 |
| 122 | var sub = ToString(this); |
| 123 | var subLength = sub.length; |
| 124 | var pat = ToString(searchString); |
| 125 | var patLength = pat.length; |
| 126 | var index = subLength - patLength; |
| 127 | if (%_ArgumentsLength() > 1) { |
| 128 | var position = ToNumber(%_Arguments(1)); |
| 129 | if (!$isNaN(position)) { |
| 130 | position = TO_INTEGER(position); |
| 131 | if (position < 0) { |
| 132 | position = 0; |
| 133 | } |
| 134 | if (position + patLength < subLength) { |
| 135 | index = position |
| 136 | } |
| 137 | } |
| 138 | } |
| 139 | if (index < 0) { |
| 140 | return -1; |
| 141 | } |
| 142 | return %StringLastIndexOf(sub, pat, index); |
| 143 | } |
| 144 | |
| 145 | |
| 146 | // ECMA-262 section 15.5.4.9 |
| 147 | // |
| 148 | // This function is implementation specific. For now, we do not |
| 149 | // do anything locale specific. |
| 150 | function StringLocaleCompare(other) { |
| 151 | if (%_ArgumentsLength() === 0) return 0; |
| 152 | |
| 153 | var this_str = ToString(this); |
| 154 | var other_str = ToString(other); |
| 155 | return %StringLocaleCompare(this_str, other_str); |
| 156 | } |
| 157 | |
| 158 | |
| 159 | // ECMA-262 section 15.5.4.10 |
| 160 | function StringMatch(regexp) { |
| 161 | if (!IS_REGEXP(regexp)) regexp = new ORIGINAL_REGEXP(regexp); |
| 162 | var subject = ToString(this); |
| 163 | |
| 164 | if (!regexp.global) return regexp.exec(subject); |
| 165 | %_Log('regexp', 'regexp-match,%0S,%1r', [subject, regexp]); |
| 166 | // lastMatchInfo is defined in regexp-delay.js. |
| 167 | return %StringMatch(subject, regexp, lastMatchInfo); |
| 168 | } |
| 169 | |
| 170 | |
| 171 | // SubString is an internal function that returns the sub string of 'string'. |
| 172 | // If resulting string is of length 1, we use the one character cache |
| 173 | // otherwise we call the runtime system. |
| 174 | function SubString(string, start, end) { |
| 175 | // Use the one character string cache. |
| 176 | if (start + 1 == end) { |
| 177 | var char_code = %_FastCharCodeAt(string, start); |
| 178 | if (!%_IsSmi(char_code)) { |
| 179 | char_code = %StringCharCodeAt(string, start); |
| 180 | } |
| 181 | return %CharFromCode(char_code); |
| 182 | } |
| 183 | return %StringSlice(string, start, end); |
| 184 | } |
| 185 | |
| 186 | |
| 187 | // This has the same size as the lastMatchInfo array, and can be used for |
| 188 | // functions that expect that structure to be returned. It is used when the |
| 189 | // needle is a string rather than a regexp. In this case we can't update |
| 190 | // lastMatchArray without erroneously affecting the properties on the global |
| 191 | // RegExp object. |
| 192 | var reusableMatchInfo = [2, "", "", -1, -1]; |
| 193 | |
| 194 | |
| 195 | // ECMA-262, section 15.5.4.11 |
| 196 | function StringReplace(search, replace) { |
| 197 | var subject = ToString(this); |
| 198 | |
| 199 | // Delegate to one of the regular expression variants if necessary. |
| 200 | if (IS_REGEXP(search)) { |
| 201 | %_Log('regexp', 'regexp-replace,%0r,%1S', [search, subject]); |
| 202 | if (IS_FUNCTION(replace)) { |
| 203 | return StringReplaceRegExpWithFunction(subject, search, replace); |
| 204 | } else { |
| 205 | return StringReplaceRegExp(subject, search, replace); |
| 206 | } |
| 207 | } |
| 208 | |
| 209 | // Convert the search argument to a string and search for it. |
| 210 | search = ToString(search); |
| 211 | var start = %StringIndexOf(subject, search, 0); |
| 212 | if (start < 0) return subject; |
| 213 | var end = start + search.length; |
| 214 | |
| 215 | var builder = new ReplaceResultBuilder(subject); |
| 216 | // prefix |
| 217 | builder.addSpecialSlice(0, start); |
| 218 | |
| 219 | // Compute the string to replace with. |
| 220 | if (IS_FUNCTION(replace)) { |
| 221 | builder.add(replace.call(null, search, start, subject)); |
| 222 | } else { |
| 223 | reusableMatchInfo[CAPTURE0] = start; |
| 224 | reusableMatchInfo[CAPTURE1] = end; |
| 225 | ExpandReplacement(ToString(replace), subject, reusableMatchInfo, builder); |
| 226 | } |
| 227 | |
| 228 | // suffix |
| 229 | builder.addSpecialSlice(end, subject.length); |
| 230 | |
| 231 | return builder.generate(); |
| 232 | } |
| 233 | |
| 234 | |
| 235 | // Helper function for regular expressions in String.prototype.replace. |
| 236 | function StringReplaceRegExp(subject, regexp, replace) { |
| 237 | replace = ToString(replace); |
| 238 | return %StringReplaceRegExpWithString(subject, |
| 239 | regexp, |
| 240 | replace, |
| 241 | lastMatchInfo); |
| 242 | }; |
| 243 | |
| 244 | |
| 245 | // Expand the $-expressions in the string and return a new string with |
| 246 | // the result. |
| 247 | function ExpandReplacement(string, subject, matchInfo, builder) { |
| 248 | var next = %StringIndexOf(string, '$', 0); |
| 249 | if (next < 0) { |
| 250 | builder.add(string); |
| 251 | return; |
| 252 | } |
| 253 | |
| 254 | // Compute the number of captures; see ECMA-262, 15.5.4.11, p. 102. |
| 255 | var m = NUMBER_OF_CAPTURES(matchInfo) >> 1; // Includes the match. |
| 256 | |
| 257 | if (next > 0) builder.add(SubString(string, 0, next)); |
| 258 | var length = string.length; |
| 259 | |
| 260 | while (true) { |
| 261 | var expansion = '$'; |
| 262 | var position = next + 1; |
| 263 | if (position < length) { |
| 264 | var peek = %_FastCharCodeAt(string, position); |
| 265 | if (!%_IsSmi(peek)) { |
| 266 | peek = %StringCharCodeAt(string, position); |
| 267 | } |
| 268 | if (peek == 36) { // $$ |
| 269 | ++position; |
| 270 | builder.add('$'); |
| 271 | } else if (peek == 38) { // $& - match |
| 272 | ++position; |
| 273 | builder.addSpecialSlice(matchInfo[CAPTURE0], |
| 274 | matchInfo[CAPTURE1]); |
| 275 | } else if (peek == 96) { // $` - prefix |
| 276 | ++position; |
| 277 | builder.addSpecialSlice(0, matchInfo[CAPTURE0]); |
| 278 | } else if (peek == 39) { // $' - suffix |
| 279 | ++position; |
| 280 | builder.addSpecialSlice(matchInfo[CAPTURE1], subject.length); |
| 281 | } else if (peek >= 48 && peek <= 57) { // $n, 0 <= n <= 9 |
| 282 | ++position; |
| 283 | var n = peek - 48; |
| 284 | if (position < length) { |
| 285 | peek = %_FastCharCodeAt(string, position); |
| 286 | if (!%_IsSmi(peek)) { |
| 287 | peek = %StringCharCodeAt(string, position); |
| 288 | } |
| 289 | // $nn, 01 <= nn <= 99 |
| 290 | if (n != 0 && peek == 48 || peek >= 49 && peek <= 57) { |
| 291 | var nn = n * 10 + (peek - 48); |
| 292 | if (nn < m) { |
| 293 | // If the two digit capture reference is within range of |
| 294 | // the captures, we use it instead of the single digit |
| 295 | // one. Otherwise, we fall back to using the single |
| 296 | // digit reference. This matches the behavior of |
| 297 | // SpiderMonkey. |
| 298 | ++position; |
| 299 | n = nn; |
| 300 | } |
| 301 | } |
| 302 | } |
| 303 | if (0 < n && n < m) { |
| 304 | addCaptureString(builder, matchInfo, n); |
| 305 | } else { |
| 306 | // Because of the captures range check in the parsing of two |
| 307 | // digit capture references, we can only enter here when a |
| 308 | // single digit capture reference is outside the range of |
| 309 | // captures. |
| 310 | builder.add('$'); |
| 311 | --position; |
| 312 | } |
| 313 | } else { |
| 314 | builder.add('$'); |
| 315 | } |
| 316 | } else { |
| 317 | builder.add('$'); |
| 318 | } |
| 319 | |
| 320 | // Go the the next $ in the string. |
| 321 | next = %StringIndexOf(string, '$', position); |
| 322 | |
| 323 | // Return if there are no more $ characters in the string. If we |
| 324 | // haven't reached the end, we need to append the suffix. |
| 325 | if (next < 0) { |
| 326 | if (position < length) { |
| 327 | builder.add(SubString(string, position, length)); |
| 328 | } |
| 329 | return; |
| 330 | } |
| 331 | |
| 332 | // Append substring between the previous and the next $ character. |
| 333 | builder.add(SubString(string, position, next)); |
| 334 | } |
| 335 | }; |
| 336 | |
| 337 | |
| 338 | // Compute the string of a given regular expression capture. |
| 339 | function CaptureString(string, lastCaptureInfo, index) { |
| 340 | // Scale the index. |
| 341 | var scaled = index << 1; |
| 342 | // Compute start and end. |
| 343 | var start = lastCaptureInfo[CAPTURE(scaled)]; |
| 344 | var end = lastCaptureInfo[CAPTURE(scaled + 1)]; |
| 345 | // If either start or end is missing return undefined. |
| 346 | if (start < 0 || end < 0) return; |
| 347 | return SubString(string, start, end); |
| 348 | }; |
| 349 | |
| 350 | |
| 351 | // Add the string of a given regular expression capture to the |
| 352 | // ReplaceResultBuilder |
| 353 | function addCaptureString(builder, matchInfo, index) { |
| 354 | // Scale the index. |
| 355 | var scaled = index << 1; |
| 356 | // Compute start and end. |
| 357 | var start = matchInfo[CAPTURE(scaled)]; |
| 358 | var end = matchInfo[CAPTURE(scaled + 1)]; |
| 359 | // If either start or end is missing return. |
| 360 | if (start < 0 || end <= start) return; |
| 361 | builder.addSpecialSlice(start, end); |
| 362 | }; |
| 363 | |
| 364 | |
| 365 | // Helper function for replacing regular expressions with the result of a |
| 366 | // function application in String.prototype.replace. The function application |
| 367 | // must be interleaved with the regexp matching (contrary to ECMA-262 |
| 368 | // 15.5.4.11) to mimic SpiderMonkey and KJS behavior when the function uses |
| 369 | // the static properties of the RegExp constructor. Example: |
| 370 | // 'abcd'.replace(/(.)/g, function() { return RegExp.$1; } |
| 371 | // should be 'abcd' and not 'dddd' (or anything else). |
| 372 | function StringReplaceRegExpWithFunction(subject, regexp, replace) { |
| 373 | var matchInfo = DoRegExpExec(regexp, subject, 0); |
| 374 | if (IS_NULL(matchInfo)) return subject; |
| 375 | |
| 376 | var result = new ReplaceResultBuilder(subject); |
| 377 | // There's at least one match. If the regexp is global, we have to loop |
| 378 | // over all matches. The loop is not in C++ code here like the one in |
| 379 | // RegExp.prototype.exec, because of the interleaved function application. |
| 380 | // Unfortunately, that means this code is nearly duplicated, here and in |
| 381 | // jsregexp.cc. |
| 382 | if (regexp.global) { |
| 383 | var previous = 0; |
| 384 | do { |
| 385 | result.addSpecialSlice(previous, matchInfo[CAPTURE0]); |
| 386 | var startOfMatch = matchInfo[CAPTURE0]; |
| 387 | previous = matchInfo[CAPTURE1]; |
| 388 | result.add(ApplyReplacementFunction(replace, matchInfo, subject)); |
| 389 | // Can't use matchInfo any more from here, since the function could |
| 390 | // overwrite it. |
| 391 | // Continue with the next match. |
| 392 | // Increment previous if we matched an empty string, as per ECMA-262 |
| 393 | // 15.5.4.10. |
| 394 | if (previous == startOfMatch) { |
| 395 | // Add the skipped character to the output, if any. |
| 396 | if (previous < subject.length) { |
| 397 | result.addSpecialSlice(previous, previous + 1); |
| 398 | } |
| 399 | previous++; |
| 400 | } |
| 401 | |
| 402 | // Per ECMA-262 15.10.6.2, if the previous index is greater than the |
| 403 | // string length, there is no match |
| 404 | matchInfo = (previous > subject.length) |
| 405 | ? null |
| 406 | : DoRegExpExec(regexp, subject, previous); |
| 407 | } while (!IS_NULL(matchInfo)); |
| 408 | |
| 409 | // Tack on the final right substring after the last match, if necessary. |
| 410 | if (previous < subject.length) { |
| 411 | result.addSpecialSlice(previous, subject.length); |
| 412 | } |
| 413 | } else { // Not a global regexp, no need to loop. |
| 414 | result.addSpecialSlice(0, matchInfo[CAPTURE0]); |
| 415 | var endOfMatch = matchInfo[CAPTURE1]; |
| 416 | result.add(ApplyReplacementFunction(replace, matchInfo, subject)); |
| 417 | // Can't use matchInfo any more from here, since the function could |
| 418 | // overwrite it. |
| 419 | result.addSpecialSlice(endOfMatch, subject.length); |
| 420 | } |
| 421 | |
| 422 | return result.generate(); |
| 423 | } |
| 424 | |
| 425 | |
| 426 | // Helper function to apply a string replacement function once. |
| 427 | function ApplyReplacementFunction(replace, matchInfo, subject) { |
| 428 | // Compute the parameter list consisting of the match, captures, index, |
| 429 | // and subject for the replace function invocation. |
| 430 | var index = matchInfo[CAPTURE0]; |
| 431 | // The number of captures plus one for the match. |
| 432 | var m = NUMBER_OF_CAPTURES(matchInfo) >> 1; |
| 433 | if (m == 1) { |
| 434 | var s = CaptureString(subject, matchInfo, 0); |
| 435 | // Don't call directly to avoid exposing the built-in global object. |
| 436 | return replace.call(null, s, index, subject); |
| 437 | } |
| 438 | var parameters = $Array(m + 2); |
| 439 | for (var j = 0; j < m; j++) { |
| 440 | parameters[j] = CaptureString(subject, matchInfo, j); |
| 441 | } |
| 442 | parameters[j] = index; |
| 443 | parameters[j + 1] = subject; |
| 444 | return replace.apply(null, parameters); |
| 445 | } |
| 446 | |
| 447 | |
| 448 | // ECMA-262 section 15.5.4.12 |
| 449 | function StringSearch(re) { |
| 450 | var regexp = new ORIGINAL_REGEXP(re); |
| 451 | var s = ToString(this); |
| 452 | var last_idx = regexp.lastIndex; // keep old lastIndex |
| 453 | regexp.lastIndex = 0; // ignore re.global property |
| 454 | var result = regexp.exec(s); |
| 455 | regexp.lastIndex = last_idx; // restore lastIndex |
| 456 | if (result == null) |
| 457 | return -1; |
| 458 | else |
| 459 | return result.index; |
| 460 | } |
| 461 | |
| 462 | |
| 463 | // ECMA-262 section 15.5.4.13 |
| 464 | function StringSlice(start, end) { |
| 465 | var s = ToString(this); |
| 466 | var s_len = s.length; |
| 467 | var start_i = TO_INTEGER(start); |
| 468 | var end_i = s_len; |
| 469 | if (end !== void 0) |
| 470 | end_i = TO_INTEGER(end); |
| 471 | |
| 472 | if (start_i < 0) { |
| 473 | start_i += s_len; |
| 474 | if (start_i < 0) |
| 475 | start_i = 0; |
| 476 | } else { |
| 477 | if (start_i > s_len) |
| 478 | start_i = s_len; |
| 479 | } |
| 480 | |
| 481 | if (end_i < 0) { |
| 482 | end_i += s_len; |
| 483 | if (end_i < 0) |
| 484 | end_i = 0; |
| 485 | } else { |
| 486 | if (end_i > s_len) |
| 487 | end_i = s_len; |
| 488 | } |
| 489 | |
| 490 | var num_c = end_i - start_i; |
| 491 | if (num_c < 0) |
| 492 | num_c = 0; |
| 493 | |
| 494 | return SubString(s, start_i, start_i + num_c); |
| 495 | } |
| 496 | |
| 497 | |
| 498 | // ECMA-262 section 15.5.4.14 |
| 499 | function StringSplit(separator, limit) { |
| 500 | var subject = ToString(this); |
| 501 | limit = (limit === void 0) ? 0xffffffff : ToUint32(limit); |
| 502 | if (limit === 0) return []; |
| 503 | |
| 504 | // ECMA-262 says that if separator is undefined, the result should |
| 505 | // be an array of size 1 containing the entire string. SpiderMonkey |
| 506 | // and KJS have this behaviour only when no separator is given. If |
| 507 | // undefined is explicitly given, they convert it to a string and |
| 508 | // use that. We do as SpiderMonkey and KJS. |
| 509 | if (%_ArgumentsLength() === 0) { |
| 510 | return [subject]; |
| 511 | } |
| 512 | |
| 513 | var length = subject.length; |
| 514 | if (IS_REGEXP(separator)) { |
| 515 | %_Log('regexp', 'regexp-split,%0S,%1r', [subject, separator]); |
| 516 | } else { |
| 517 | separator = ToString(separator); |
| 518 | // If the separator string is empty then return the elements in the subject. |
| 519 | if (separator.length == 0) { |
| 520 | var result = $Array(length); |
| 521 | for (var i = 0; i < length; i++) result[i] = subject[i]; |
| 522 | return result; |
| 523 | } |
| 524 | } |
| 525 | |
| 526 | if (length === 0) { |
| 527 | if (splitMatch(separator, subject, 0, 0) != null) return []; |
| 528 | return [subject]; |
| 529 | } |
| 530 | |
| 531 | var currentIndex = 0; |
| 532 | var startIndex = 0; |
| 533 | var result = []; |
| 534 | |
| 535 | while (true) { |
| 536 | |
| 537 | if (startIndex === length) { |
| 538 | result[result.length] = subject.slice(currentIndex, length); |
| 539 | return result; |
| 540 | } |
| 541 | |
| 542 | var matchInfo = splitMatch(separator, subject, currentIndex, startIndex); |
| 543 | |
| 544 | if (IS_NULL(matchInfo)) { |
| 545 | result[result.length] = subject.slice(currentIndex, length); |
| 546 | return result; |
| 547 | } |
| 548 | |
| 549 | var endIndex = matchInfo[CAPTURE1]; |
| 550 | |
| 551 | // We ignore a zero-length match at the currentIndex. |
| 552 | if (startIndex === endIndex && endIndex === currentIndex) { |
| 553 | startIndex++; |
| 554 | continue; |
| 555 | } |
| 556 | |
| 557 | result[result.length] = SubString(subject, currentIndex, matchInfo[CAPTURE0]); |
| 558 | if (result.length === limit) return result; |
| 559 | |
| 560 | for (var i = 2; i < NUMBER_OF_CAPTURES(matchInfo); i += 2) { |
| 561 | var start = matchInfo[CAPTURE(i)]; |
| 562 | var end = matchInfo[CAPTURE(i + 1)]; |
| 563 | if (start != -1 && end != -1) { |
| 564 | result[result.length] = SubString(subject, start, end); |
| 565 | } else { |
| 566 | result[result.length] = void 0; |
| 567 | } |
| 568 | if (result.length === limit) return result; |
| 569 | } |
| 570 | |
| 571 | startIndex = currentIndex = endIndex; |
| 572 | } |
| 573 | } |
| 574 | |
| 575 | |
| 576 | // ECMA-262 section 15.5.4.14 |
| 577 | // Helper function used by split. This version returns the matchInfo |
| 578 | // instead of allocating a new array with basically the same information. |
| 579 | function splitMatch(separator, subject, current_index, start_index) { |
| 580 | if (IS_REGEXP(separator)) { |
| 581 | var matchInfo = DoRegExpExec(separator, subject, start_index); |
| 582 | if (matchInfo == null) return null; |
| 583 | // Section 15.5.4.14 paragraph two says that we do not allow zero length |
| 584 | // matches at the end of the string. |
| 585 | if (matchInfo[CAPTURE0] === subject.length) return null; |
| 586 | return matchInfo; |
| 587 | } |
| 588 | |
| 589 | var separatorIndex = subject.indexOf(separator, start_index); |
| 590 | if (separatorIndex === -1) return null; |
| 591 | |
| 592 | reusableMatchInfo[CAPTURE0] = separatorIndex; |
| 593 | reusableMatchInfo[CAPTURE1] = separatorIndex + separator.length; |
| 594 | return reusableMatchInfo; |
| 595 | }; |
| 596 | |
| 597 | |
| 598 | // ECMA-262 section 15.5.4.15 |
| 599 | function StringSubstring(start, end) { |
| 600 | var s = ToString(this); |
| 601 | var s_len = s.length; |
| 602 | var start_i = TO_INTEGER(start); |
| 603 | var end_i = s_len; |
| 604 | if (!IS_UNDEFINED(end)) |
| 605 | end_i = TO_INTEGER(end); |
| 606 | |
| 607 | if (start_i < 0) start_i = 0; |
| 608 | if (start_i > s_len) start_i = s_len; |
| 609 | if (end_i < 0) end_i = 0; |
| 610 | if (end_i > s_len) end_i = s_len; |
| 611 | |
| 612 | if (start_i > end_i) { |
| 613 | var tmp = end_i; |
| 614 | end_i = start_i; |
| 615 | start_i = tmp; |
| 616 | } |
| 617 | |
| 618 | return SubString(s, start_i, end_i); |
| 619 | } |
| 620 | |
| 621 | |
| 622 | // This is not a part of ECMA-262. |
| 623 | function StringSubstr(start, n) { |
| 624 | var s = ToString(this); |
| 625 | var len; |
| 626 | |
| 627 | // Correct n: If not given, set to string length; if explicitly |
| 628 | // set to undefined, zero, or negative, returns empty string. |
| 629 | if (n === void 0) { |
| 630 | len = s.length; |
| 631 | } else { |
| 632 | len = TO_INTEGER(n); |
| 633 | if (len <= 0) return ''; |
| 634 | } |
| 635 | |
| 636 | // Correct start: If not given (or undefined), set to zero; otherwise |
| 637 | // convert to integer and handle negative case. |
| 638 | if (start === void 0) { |
| 639 | start = 0; |
| 640 | } else { |
| 641 | start = TO_INTEGER(start); |
| 642 | // If positive, and greater than or equal to the string length, |
| 643 | // return empty string. |
| 644 | if (start >= s.length) return ''; |
| 645 | // If negative and absolute value is larger than the string length, |
| 646 | // use zero. |
| 647 | if (start < 0) { |
| 648 | start += s.length; |
| 649 | if (start < 0) start = 0; |
| 650 | } |
| 651 | } |
| 652 | |
| 653 | var end = start + len; |
| 654 | if (end > s.length) end = s.length; |
| 655 | |
| 656 | return SubString(s, start, end); |
| 657 | } |
| 658 | |
| 659 | |
| 660 | // ECMA-262, 15.5.4.16 |
| 661 | function StringToLowerCase() { |
| 662 | return %StringToLowerCase(ToString(this)); |
| 663 | } |
| 664 | |
| 665 | |
| 666 | // ECMA-262, 15.5.4.17 |
| 667 | function StringToLocaleLowerCase() { |
| 668 | return %StringToLowerCase(ToString(this)); |
| 669 | } |
| 670 | |
| 671 | |
| 672 | // ECMA-262, 15.5.4.18 |
| 673 | function StringToUpperCase() { |
| 674 | return %StringToUpperCase(ToString(this)); |
| 675 | } |
| 676 | |
| 677 | |
| 678 | // ECMA-262, 15.5.4.19 |
| 679 | function StringToLocaleUpperCase() { |
| 680 | return %StringToUpperCase(ToString(this)); |
| 681 | } |
| 682 | |
Steve Block | 3ce2e20 | 2009-11-05 08:53:23 +0000 | [diff] [blame^] | 683 | // ES5, 15.5.4.20 |
| 684 | function StringTrim() { |
| 685 | return %StringTrim(ToString(this), true, true); |
| 686 | } |
| 687 | |
| 688 | function StringTrimLeft() { |
| 689 | return %StringTrim(ToString(this), true, false); |
| 690 | } |
| 691 | |
| 692 | function StringTrimRight() { |
| 693 | return %StringTrim(ToString(this), false, true); |
| 694 | } |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 695 | |
| 696 | // ECMA-262, section 15.5.3.2 |
| 697 | function StringFromCharCode(code) { |
| 698 | var n = %_ArgumentsLength(); |
| 699 | if (n == 1) return %CharFromCode(ToNumber(code) & 0xffff) |
| 700 | |
| 701 | // NOTE: This is not super-efficient, but it is necessary because we |
| 702 | // want to avoid converting to numbers from within the virtual |
| 703 | // machine. Maybe we can find another way of doing this? |
| 704 | var codes = new $Array(n); |
| 705 | for (var i = 0; i < n; i++) codes[i] = ToNumber(%_Arguments(i)); |
| 706 | return %StringFromCharCodeArray(codes); |
| 707 | } |
| 708 | |
| 709 | |
| 710 | // Helper function for very basic XSS protection. |
| 711 | function HtmlEscape(str) { |
| 712 | return ToString(str).replace(/</g, "<") |
| 713 | .replace(/>/g, ">") |
| 714 | .replace(/"/g, """) |
| 715 | .replace(/'/g, "'"); |
| 716 | }; |
| 717 | |
| 718 | |
| 719 | // Compatibility support for KJS. |
| 720 | // Tested by mozilla/js/tests/js1_5/Regress/regress-276103.js. |
| 721 | function StringLink(s) { |
| 722 | return "<a href=\"" + HtmlEscape(s) + "\">" + this + "</a>"; |
| 723 | } |
| 724 | |
| 725 | |
| 726 | function StringAnchor(name) { |
| 727 | return "<a name=\"" + HtmlEscape(name) + "\">" + this + "</a>"; |
| 728 | } |
| 729 | |
| 730 | |
| 731 | function StringFontcolor(color) { |
| 732 | return "<font color=\"" + HtmlEscape(color) + "\">" + this + "</font>"; |
| 733 | } |
| 734 | |
| 735 | |
| 736 | function StringFontsize(size) { |
| 737 | return "<font size=\"" + HtmlEscape(size) + "\">" + this + "</font>"; |
| 738 | } |
| 739 | |
| 740 | |
| 741 | function StringBig() { |
| 742 | return "<big>" + this + "</big>"; |
| 743 | } |
| 744 | |
| 745 | |
| 746 | function StringBlink() { |
| 747 | return "<blink>" + this + "</blink>"; |
| 748 | } |
| 749 | |
| 750 | |
| 751 | function StringBold() { |
| 752 | return "<b>" + this + "</b>"; |
| 753 | } |
| 754 | |
| 755 | |
| 756 | function StringFixed() { |
| 757 | return "<tt>" + this + "</tt>"; |
| 758 | } |
| 759 | |
| 760 | |
| 761 | function StringItalics() { |
| 762 | return "<i>" + this + "</i>"; |
| 763 | } |
| 764 | |
| 765 | |
| 766 | function StringSmall() { |
| 767 | return "<small>" + this + "</small>"; |
| 768 | } |
| 769 | |
| 770 | |
| 771 | function StringStrike() { |
| 772 | return "<strike>" + this + "</strike>"; |
| 773 | } |
| 774 | |
| 775 | |
| 776 | function StringSub() { |
| 777 | return "<sub>" + this + "</sub>"; |
| 778 | } |
| 779 | |
| 780 | |
| 781 | function StringSup() { |
| 782 | return "<sup>" + this + "</sup>"; |
| 783 | } |
| 784 | |
| 785 | |
| 786 | // StringBuilder support. |
| 787 | |
| 788 | function StringBuilder() { |
| 789 | this.elements = new $Array(); |
| 790 | } |
| 791 | |
| 792 | |
| 793 | function ReplaceResultBuilder(str) { |
| 794 | this.elements = new $Array(); |
| 795 | this.special_string = str; |
| 796 | } |
| 797 | |
| 798 | |
| 799 | ReplaceResultBuilder.prototype.add = |
| 800 | StringBuilder.prototype.add = function(str) { |
| 801 | if (!IS_STRING(str)) str = ToString(str); |
| 802 | if (str.length > 0) { |
| 803 | var elements = this.elements; |
| 804 | elements[elements.length] = str; |
| 805 | } |
| 806 | } |
| 807 | |
| 808 | |
| 809 | ReplaceResultBuilder.prototype.addSpecialSlice = function(start, end) { |
| 810 | var len = end - start; |
| 811 | if (len == 0) return; |
| 812 | var elements = this.elements; |
| 813 | if (start >= 0 && len >= 0 && start < 0x80000 && len < 0x800) { |
| 814 | elements[elements.length] = (start << 11) + len; |
| 815 | } else { |
| 816 | elements[elements.length] = SubString(this.special_string, start, end); |
| 817 | } |
| 818 | } |
| 819 | |
| 820 | |
| 821 | StringBuilder.prototype.generate = function() { |
| 822 | return %StringBuilderConcat(this.elements, ""); |
| 823 | } |
| 824 | |
| 825 | |
| 826 | ReplaceResultBuilder.prototype.generate = function() { |
| 827 | return %StringBuilderConcat(this.elements, this.special_string); |
| 828 | } |
| 829 | |
| 830 | |
| 831 | function StringToJSON(key) { |
| 832 | return CheckJSONPrimitive(this.valueOf()); |
| 833 | } |
| 834 | |
| 835 | |
| 836 | // ------------------------------------------------------------------- |
| 837 | |
| 838 | function SetupString() { |
| 839 | // Setup the constructor property on the String prototype object. |
| 840 | %SetProperty($String.prototype, "constructor", $String, DONT_ENUM); |
| 841 | |
| 842 | |
| 843 | // Setup the non-enumerable functions on the String object. |
| 844 | InstallFunctions($String, DONT_ENUM, $Array( |
| 845 | "fromCharCode", StringFromCharCode |
| 846 | )); |
| 847 | |
| 848 | |
| 849 | // Setup the non-enumerable functions on the String prototype object. |
| 850 | InstallFunctionsOnHiddenPrototype($String.prototype, DONT_ENUM, $Array( |
| 851 | "valueOf", StringValueOf, |
| 852 | "toString", StringToString, |
| 853 | "charAt", StringCharAt, |
| 854 | "charCodeAt", StringCharCodeAt, |
| 855 | "concat", StringConcat, |
| 856 | "indexOf", StringIndexOf, |
| 857 | "lastIndexOf", StringLastIndexOf, |
| 858 | "localeCompare", StringLocaleCompare, |
| 859 | "match", StringMatch, |
| 860 | "replace", StringReplace, |
| 861 | "search", StringSearch, |
| 862 | "slice", StringSlice, |
| 863 | "split", StringSplit, |
| 864 | "substring", StringSubstring, |
| 865 | "substr", StringSubstr, |
| 866 | "toLowerCase", StringToLowerCase, |
| 867 | "toLocaleLowerCase", StringToLocaleLowerCase, |
| 868 | "toUpperCase", StringToUpperCase, |
| 869 | "toLocaleUpperCase", StringToLocaleUpperCase, |
Steve Block | 3ce2e20 | 2009-11-05 08:53:23 +0000 | [diff] [blame^] | 870 | "trim", StringTrim, |
| 871 | "trimLeft", StringTrimLeft, |
| 872 | "trimRight", StringTrimRight, |
Steve Block | a7e24c1 | 2009-10-30 11:49:00 +0000 | [diff] [blame] | 873 | "link", StringLink, |
| 874 | "anchor", StringAnchor, |
| 875 | "fontcolor", StringFontcolor, |
| 876 | "fontsize", StringFontsize, |
| 877 | "big", StringBig, |
| 878 | "blink", StringBlink, |
| 879 | "bold", StringBold, |
| 880 | "fixed", StringFixed, |
| 881 | "italics", StringItalics, |
| 882 | "small", StringSmall, |
| 883 | "strike", StringStrike, |
| 884 | "sub", StringSub, |
| 885 | "sup", StringSup, |
| 886 | "toJSON", StringToJSON |
| 887 | )); |
| 888 | } |
| 889 | |
| 890 | |
| 891 | SetupString(); |