Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(923)

Issues Search
    My Issues | Starred     Open | Closed | All

Side by Side Diff: src/regexp-delay.js

Issue 1094014: Merge the partial_snapshots branch back into bleeding_edge. For... (Closed) Base URL: http://v8.googlecode.com/svn/branches/bleeding_edge/
Patch Set: Created 10 years, 9 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch | Annotate | Revision Log
« src/mksnapshot.cc ('K') | « src/regexp.js ('k') | src/runtime.cc » ('j') | src/x64/stub-cache-x64.cc » ('J')
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
(Empty)
1 // Copyright 2006-2009 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 // Expect $Object = global.Object;
29 // Expect $Array = global.Array;
30
31 const $RegExp = global.RegExp;
32
33 // A recursive descent parser for Patterns according to the grammar of
34 // ECMA-262 15.10.1, with deviations noted below.
35 function DoConstructRegExp(object, pattern, flags, isConstructorCall) {
36 // RegExp : Called as constructor; see ECMA-262, section 15.10.4.
37 if (IS_REGEXP(pattern)) {
38 if (!IS_UNDEFINED(flags)) {
39 throw MakeTypeError('regexp_flags', []);
40 }
41 flags = (pattern.global ? 'g' : '')
42 + (pattern.ignoreCase ? 'i' : '')
43 + (pattern.multiline ? 'm' : '');
44 pattern = pattern.source;
45 }
46
47 pattern = IS_UNDEFINED(pattern) ? '' : ToString(pattern);
48 flags = IS_UNDEFINED(flags) ? '' : ToString(flags);
49
50 var global = false;
51 var ignoreCase = false;
52 var multiline = false;
53
54 for (var i = 0; i < flags.length; i++) {
55 var c = StringCharAt.call(flags, i);
56 switch (c) {
57 case 'g':
58 // Allow duplicate flags to be consistent with JSC and others.
59 global = true;
60 break;
61 case 'i':
62 ignoreCase = true;
63 break;
64 case 'm':
65 multiline = true;
66 break;
67 default:
68 // Ignore flags that have no meaning to be consistent with
69 // JSC.
70 break;
71 }
72 }
73
74 if (isConstructorCall) {
75 // ECMA-262, section 15.10.7.1.
76 %SetProperty(object, 'source', pattern,
77 DONT_DELETE | READ_ONLY | DONT_ENUM);
78
79 // ECMA-262, section 15.10.7.2.
80 %SetProperty(object, 'global', global, DONT_DELETE | READ_ONLY | DONT_ENUM);
81
82 // ECMA-262, section 15.10.7.3.
83 %SetProperty(object, 'ignoreCase', ignoreCase,
84 DONT_DELETE | READ_ONLY | DONT_ENUM);
85
86 // ECMA-262, section 15.10.7.4.
87 %SetProperty(object, 'multiline', multiline,
88 DONT_DELETE | READ_ONLY | DONT_ENUM);
89
90 // ECMA-262, section 15.10.7.5.
91 %SetProperty(object, 'lastIndex', 0, DONT_DELETE | DONT_ENUM);
92 } else { // RegExp is being recompiled via RegExp.prototype.compile.
93 %IgnoreAttributesAndSetProperty(object, 'source', pattern);
94 %IgnoreAttributesAndSetProperty(object, 'global', global);
95 %IgnoreAttributesAndSetProperty(object, 'ignoreCase', ignoreCase);
96 %IgnoreAttributesAndSetProperty(object, 'multiline', multiline);
97 %IgnoreAttributesAndSetProperty(object, 'lastIndex', 0);
98 regExpCache.type = 'none';
99 }
100
101 // Call internal function to compile the pattern.
102 %RegExpCompile(object, pattern, flags);
103 }
104
105
106 function RegExpConstructor(pattern, flags) {
107 if (%_IsConstructCall()) {
108 DoConstructRegExp(this, pattern, flags, true);
109 } else {
110 // RegExp : Called as function; see ECMA-262, section 15.10.3.1.
111 if (IS_REGEXP(pattern) && IS_UNDEFINED(flags)) {
112 return pattern;
113 }
114 return new $RegExp(pattern, flags);
115 }
116 }
117
118
119 // Deprecated RegExp.prototype.compile method. We behave like the constructor
120 // were called again. In SpiderMonkey, this method returns the regexp object.
121 // In JSC, it returns undefined. For compatibility with JSC, we match their
122 // behavior.
123 function CompileRegExp(pattern, flags) {
124 // Both JSC and SpiderMonkey treat a missing pattern argument as the
125 // empty subject string, and an actual undefined value passed as the
126 // pattern as the string 'undefined'. Note that JSC is inconsistent
127 // here, treating undefined values differently in
128 // RegExp.prototype.compile and in the constructor, where they are
129 // the empty string. For compatibility with JSC, we match their
130 // behavior.
131 if (IS_UNDEFINED(pattern) && %_ArgumentsLength() != 0) {
132 DoConstructRegExp(this, 'undefined', flags, false);
133 } else {
134 DoConstructRegExp(this, pattern, flags, false);
135 }
136 }
137
138
139 function DoRegExpExec(regexp, string, index) {
140 return %_RegExpExec(regexp, string, index, lastMatchInfo);
141 }
142
143
144 function RegExpCache() {
145 this.type = 'none';
146 this.regExp = 0;
147 this.subject = 0;
148 this.replaceString = 0;
149 this.lastIndex = 0;
150 this.answer = 0;
151 }
152
153
154 var regExpCache = new RegExpCache();
155
156
157 function CloneRegexpAnswer(array) {
158 var len = array.length;
159 var answer = new $Array(len);
160 for (var i = 0; i < len; i++) {
161 answer[i] = array[i];
162 }
163 answer.index = array.index;
164 answer.input = array.input;
165 return answer;
166 }
167
168
169 function RegExpExec(string) {
170 if (!IS_REGEXP(this)) {
171 throw MakeTypeError('incompatible_method_receiver',
172 ['RegExp.prototype.exec', this]);
173 }
174
175 var cache = regExpCache;
176
177 if (%_ObjectEquals(cache.type, 'exec') &&
178 %_ObjectEquals(cache.lastIndex, this.lastIndex) &&
179 %_ObjectEquals(cache.regExp, this) &&
180 %_ObjectEquals(cache.subject, string)) {
181 var last = cache.answer;
182 if (last == null) {
183 return last;
184 } else {
185 return CloneRegexpAnswer(last);
186 }
187 }
188
189 if (%_ArgumentsLength() == 0) {
190 var regExpInput = LAST_INPUT(lastMatchInfo);
191 if (IS_UNDEFINED(regExpInput)) {
192 throw MakeError('no_input_to_regexp', [this]);
193 }
194 string = regExpInput;
195 }
196 var s;
197 if (IS_STRING(string)) {
198 s = string;
199 } else {
200 s = ToString(string);
201 }
202 var lastIndex = this.lastIndex;
203
204 var i = this.global ? TO_INTEGER(lastIndex) : 0;
205
206 if (i < 0 || i > s.length) {
207 this.lastIndex = 0;
208 return null;
209 }
210
211 %_Log('regexp', 'regexp-exec,%0r,%1S,%2i', [this, s, lastIndex]);
212 // matchIndices is either null or the lastMatchInfo array.
213 var matchIndices = %_RegExpExec(this, s, i, lastMatchInfo);
214
215 if (matchIndices == null) {
216 if (this.global) this.lastIndex = 0;
217 cache.lastIndex = lastIndex;
218 cache.regExp = this;
219 cache.subject = s;
220 cache.answer = matchIndices; // Null.
221 cache.type = 'exec';
222 return matchIndices; // No match.
223 }
224
225 var numResults = NUMBER_OF_CAPTURES(lastMatchInfo) >> 1;
226 var result;
227 if (numResults === 1) {
228 var matchStart = lastMatchInfo[CAPTURE(0)];
229 var matchEnd = lastMatchInfo[CAPTURE(1)];
230 result = [SubString(s, matchStart, matchEnd)];
231 } else {
232 result = new $Array(numResults);
233 for (var i = 0; i < numResults; i++) {
234 var matchStart = lastMatchInfo[CAPTURE(i << 1)];
235 var matchEnd = lastMatchInfo[CAPTURE((i << 1) + 1)];
236 if (matchStart != -1 && matchEnd != -1) {
237 result[i] = SubString(s, matchStart, matchEnd);
238 } else {
239 // Make sure the element is present. Avoid reading the undefined
240 // property from the global object since this may change.
241 result[i] = void 0;
242 }
243 }
244 }
245
246 result.index = lastMatchInfo[CAPTURE0];
247 result.input = s;
248 if (this.global) {
249 this.lastIndex = lastMatchInfo[CAPTURE1];
250 return result;
251 } else {
252 cache.regExp = this;
253 cache.subject = s;
254 cache.lastIndex = lastIndex;
255 cache.answer = result;
256 cache.type = 'exec';
257 return CloneRegexpAnswer(result);
258 }
259 }
260
261
262 // Section 15.10.6.3 doesn't actually make sense, but the intention seems to be
263 // that test is defined in terms of String.prototype.exec. However, it probably
264 // means the original value of String.prototype.exec, which is what everybody
265 // else implements.
266 function RegExpTest(string) {
267 if (!IS_REGEXP(this)) {
268 throw MakeTypeError('incompatible_method_receiver',
269 ['RegExp.prototype.test', this]);
270 }
271 if (%_ArgumentsLength() == 0) {
272 var regExpInput = LAST_INPUT(lastMatchInfo);
273 if (IS_UNDEFINED(regExpInput)) {
274 throw MakeError('no_input_to_regexp', [this]);
275 }
276 string = regExpInput;
277 }
278 var s;
279 if (IS_STRING(string)) {
280 s = string;
281 } else {
282 s = ToString(string);
283 }
284
285 var lastIndex = this.lastIndex;
286
287 var cache = regExpCache;
288
289 if (%_ObjectEquals(cache.type, 'test') &&
290 %_ObjectEquals(cache.regExp, this) &&
291 %_ObjectEquals(cache.subject, string) &&
292 %_ObjectEquals(cache.lastIndex, lastIndex)) {
293 return cache.answer;
294 }
295
296 var length = s.length;
297 var i = this.global ? TO_INTEGER(lastIndex) : 0;
298
299 cache.type = 'test';
300 cache.regExp = this;
301 cache.subject = s;
302 cache.lastIndex = i;
303
304 if (i < 0 || i > s.length) {
305 this.lastIndex = 0;
306 cache.answer = false;
307 return false;
308 }
309
310 %_Log('regexp', 'regexp-exec,%0r,%1S,%2i', [this, s, lastIndex]);
311 // matchIndices is either null or the lastMatchInfo array.
312 var matchIndices = %_RegExpExec(this, s, i, lastMatchInfo);
313
314 if (matchIndices == null) {
315 if (this.global) this.lastIndex = 0;
316 cache.answer = false;
317 return false;
318 }
319
320 if (this.global) this.lastIndex = lastMatchInfo[CAPTURE1];
321 cache.answer = true;
322 return true;
323 }
324
325
326 function RegExpToString() {
327 // If this.source is an empty string, output /(?:)/.
328 // http://bugzilla.mozilla.org/show_bug.cgi?id=225550
329 // ecma_2/RegExp/properties-001.js.
330 var src = this.source ? this.source : '(?:)';
331 var result = '/' + src + '/';
332 if (this.global)
333 result += 'g';
334 if (this.ignoreCase)
335 result += 'i';
336 if (this.multiline)
337 result += 'm';
338 return result;
339 }
340
341
342 // Getters for the static properties lastMatch, lastParen, leftContext, and
343 // rightContext of the RegExp constructor. The properties are computed based
344 // on the captures array of the last successful match and the subject string
345 // of the last successful match.
346 function RegExpGetLastMatch() {
347 var regExpSubject = LAST_SUBJECT(lastMatchInfo);
348 return SubString(regExpSubject,
349 lastMatchInfo[CAPTURE0],
350 lastMatchInfo[CAPTURE1]);
351 }
352
353
354 function RegExpGetLastParen() {
355 var length = NUMBER_OF_CAPTURES(lastMatchInfo);
356 if (length <= 2) return ''; // There were no captures.
357 // We match the SpiderMonkey behavior: return the substring defined by the
358 // last pair (after the first pair) of elements of the capture array even if
359 // it is empty.
360 var regExpSubject = LAST_SUBJECT(lastMatchInfo);
361 var start = lastMatchInfo[CAPTURE(length - 2)];
362 var end = lastMatchInfo[CAPTURE(length - 1)];
363 if (start != -1 && end != -1) {
364 return SubString(regExpSubject, start, end);
365 }
366 return "";
367 }
368
369
370 function RegExpGetLeftContext() {
371 return SubString(LAST_SUBJECT(lastMatchInfo),
372 0,
373 lastMatchInfo[CAPTURE0]);
374 }
375
376
377 function RegExpGetRightContext() {
378 var subject = LAST_SUBJECT(lastMatchInfo);
379 return SubString(subject,
380 lastMatchInfo[CAPTURE1],
381 subject.length);
382 }
383
384
385 // The properties $1..$9 are the first nine capturing substrings of the last
386 // successful match, or ''. The function RegExpMakeCaptureGetter will be
387 // called with indices from 1 to 9.
388 function RegExpMakeCaptureGetter(n) {
389 return function() {
390 var index = n * 2;
391 if (index >= NUMBER_OF_CAPTURES(lastMatchInfo)) return '';
392 var matchStart = lastMatchInfo[CAPTURE(index)];
393 var matchEnd = lastMatchInfo[CAPTURE(index + 1)];
394 if (matchStart == -1 || matchEnd == -1) return '';
395 return SubString(LAST_SUBJECT(lastMatchInfo), matchStart, matchEnd);
396 };
397 }
398
399
400 // Property of the builtins object for recording the result of the last
401 // regexp match. The property lastMatchInfo includes the matchIndices
402 // array of the last successful regexp match (an array of start/end index
403 // pairs for the match and all the captured substrings), the invariant is
404 // that there are at least two capture indeces. The array also contains
405 // the subject string for the last successful match.
406 var lastMatchInfo = [
407 2, // REGEXP_NUMBER_OF_CAPTURES
408 "", // Last subject.
409 void 0, // Last input - settable with RegExpSetInput.
410 0, // REGEXP_FIRST_CAPTURE + 0
411 0, // REGEXP_FIRST_CAPTURE + 1
412 ];
413
414 // -------------------------------------------------------------------
415
416 function SetupRegExp() {
417 %FunctionSetInstanceClassName($RegExp, 'RegExp');
418 %FunctionSetPrototype($RegExp, new $Object());
419 %SetProperty($RegExp.prototype, 'constructor', $RegExp, DONT_ENUM);
420 %SetCode($RegExp, RegExpConstructor);
421
422 InstallFunctions($RegExp.prototype, DONT_ENUM, $Array(
423 "exec", RegExpExec,
424 "test", RegExpTest,
425 "toString", RegExpToString,
426 "compile", CompileRegExp
427 ));
428
429 // The length of compile is 1 in SpiderMonkey.
430 %FunctionSetLength($RegExp.prototype.compile, 1);
431
432 // The properties input, $input, and $_ are aliases for each other. When this
433 // value is set the value it is set to is coerced to a string.
434 // Getter and setter for the input.
435 function RegExpGetInput() {
436 var regExpInput = LAST_INPUT(lastMatchInfo);
437 return IS_UNDEFINED(regExpInput) ? "" : regExpInput;
438 }
439 function RegExpSetInput(string) {
440 regExpCache.type = 'none';
441 LAST_INPUT(lastMatchInfo) = ToString(string);
442 };
443
444 %DefineAccessor($RegExp, 'input', GETTER, RegExpGetInput, DONT_DELETE);
445 %DefineAccessor($RegExp, 'input', SETTER, RegExpSetInput, DONT_DELETE);
446 %DefineAccessor($RegExp, '$_', GETTER, RegExpGetInput, DONT_ENUM | DONT_DELETE );
447 %DefineAccessor($RegExp, '$_', SETTER, RegExpSetInput, DONT_ENUM | DONT_DELETE );
448 %DefineAccessor($RegExp, '$input', GETTER, RegExpGetInput, DONT_ENUM | DONT_DE LETE);
449 %DefineAccessor($RegExp, '$input', SETTER, RegExpSetInput, DONT_ENUM | DONT_DE LETE);
450
451 // The properties multiline and $* are aliases for each other. When this
452 // value is set in SpiderMonkey, the value it is set to is coerced to a
453 // boolean. We mimic that behavior with a slight difference: in SpiderMonkey
454 // the value of the expression 'RegExp.multiline = null' (for instance) is the
455 // boolean false (ie, the value after coercion), while in V8 it is the value
456 // null (ie, the value before coercion).
457
458 // Getter and setter for multiline.
459 var multiline = false;
460 function RegExpGetMultiline() { return multiline; };
461 function RegExpSetMultiline(flag) { multiline = flag ? true : false; };
462
463 %DefineAccessor($RegExp, 'multiline', GETTER, RegExpGetMultiline, DONT_DELETE) ;
464 %DefineAccessor($RegExp, 'multiline', SETTER, RegExpSetMultiline, DONT_DELETE) ;
465 %DefineAccessor($RegExp, '$*', GETTER, RegExpGetMultiline, DONT_ENUM | DONT_DE LETE);
466 %DefineAccessor($RegExp, '$*', SETTER, RegExpSetMultiline, DONT_ENUM | DONT_DE LETE);
467
468
469 function NoOpSetter(ignored) {}
470
471
472 // Static properties set by a successful match.
473 %DefineAccessor($RegExp, 'lastMatch', GETTER, RegExpGetLastMatch, DONT_DELETE) ;
474 %DefineAccessor($RegExp, 'lastMatch', SETTER, NoOpSetter, DONT_DELETE);
475 %DefineAccessor($RegExp, '$&', GETTER, RegExpGetLastMatch, DONT_ENUM | DONT_DE LETE);
476 %DefineAccessor($RegExp, '$&', SETTER, NoOpSetter, DONT_ENUM | DONT_DELETE);
477 %DefineAccessor($RegExp, 'lastParen', GETTER, RegExpGetLastParen, DONT_DELETE) ;
478 %DefineAccessor($RegExp, 'lastParen', SETTER, NoOpSetter, DONT_DELETE);
479 %DefineAccessor($RegExp, '$+', GETTER, RegExpGetLastParen, DONT_ENUM | DONT_DE LETE);
480 %DefineAccessor($RegExp, '$+', SETTER, NoOpSetter, DONT_ENUM | DONT_DELETE);
481 %DefineAccessor($RegExp, 'leftContext', GETTER, RegExpGetLeftContext, DONT_DEL ETE);
482 %DefineAccessor($RegExp, 'leftContext', SETTER, NoOpSetter, DONT_DELETE);
483 %DefineAccessor($RegExp, '$`', GETTER, RegExpGetLeftContext, DONT_ENUM | DONT_ DELETE);
484 %DefineAccessor($RegExp, '$`', SETTER, NoOpSetter, DONT_ENUM | DONT_DELETE);
485 %DefineAccessor($RegExp, 'rightContext', GETTER, RegExpGetRightContext, DONT_D ELETE);
486 %DefineAccessor($RegExp, 'rightContext', SETTER, NoOpSetter, DONT_DELETE);
487 %DefineAccessor($RegExp, "$'", GETTER, RegExpGetRightContext, DONT_ENUM | DONT _DELETE);
488 %DefineAccessor($RegExp, "$'", SETTER, NoOpSetter, DONT_ENUM | DONT_DELETE);
489
490 for (var i = 1; i < 10; ++i) {
491 %DefineAccessor($RegExp, '$' + i, GETTER, RegExpMakeCaptureGetter(i), DONT_D ELETE);
492 %DefineAccessor($RegExp, '$' + i, SETTER, NoOpSetter, DONT_DELETE);
493 }
494 }
495
496
497 SetupRegExp();
OLDNEW
« src/mksnapshot.cc ('K') | « src/regexp.js ('k') | src/runtime.cc » ('j') | src/x64/stub-cache-x64.cc » ('J')

Powered by Google App Engine
This is Rietveld 408576698