[regexp] Port split

This CL ports RegExp.prototype[@@split] to C++. Performance
regressions are expected due to:

* Slow RegExpImpl::Exec implementation instead of RegExpExec
  stub. We should be able to improve this by straight-lining
  RegExpImpl::Exec.
* Slow Factory::NewSubString instead of SubStringStub.
* Slow elements access to lastMatchInfo.

These points will be addressed in a follow-up CL.

BUG=v8:5339

Review-Url: https://codereview.chromium.org/2401643002
Cr-Commit-Position: refs/heads/master@{#40161}

src/bootstrapper.cc

@@ -1748,6 +1748,13 @@ void Genesis::InitializeGlobal(Handle<JSGlobalObject> global_object,
             Builtins::kRegExpPrototypeSearch, 1, false);
         InstallFunction(prototype, fun, factory->search_symbol(), DONT_ENUM);
       }
+
+      {
+        Handle<JSFunction> fun = SimpleCreateFunction(
+            isolate, factory->InternalizeUtf8String("[Symbol.split]"),
+            Builtins::kRegExpPrototypeSplit, 2, false);
+        InstallFunction(prototype, fun, factory->split_symbol(), DONT_ENUM);
+      }
     }
 
     {

src/builtins/builtins.h

@@ -590,6 +590,7 @@ namespace internal {
   CPP(RegExpPrototypeSearch)                                                  \
   CPP(RegExpPrototypeSourceGetter)                                            \
   CPP(RegExpPrototypeSpeciesGetter)                                           \
+  CPP(RegExpPrototypeSplit)                                                   \
   TFJ(RegExpPrototypeStickyGetter, 1)                                         \
   CPP(RegExpPrototypeTest)                                                    \
   CPP(RegExpPrototypeToString)                                                \

src/js/regexp.js

@@ -121,159 +121,6 @@ function RegExpSubclassExec(regexp, string, exec) {
 %SetForceInlineFlag(RegExpSubclassExec);
 
 
-function AtSurrogatePair(subject, index) {
-  if (index + 1 >= subject.length) return false;
-  var first = %_StringCharCodeAt(subject, index);
-  if (first < 0xD800 || first > 0xDBFF) return false;
-  var second = %_StringCharCodeAt(subject, index + 1);
-  return second >= 0xDC00 && second <= 0xDFFF;
-}
-
-
-// Fast path implementation of RegExp.prototype[Symbol.split] which
-// doesn't properly call the underlying exec, @@species methods
-function RegExpSplit(string, limit) {
-  if (!IS_REGEXP(this)) {
-    throw %make_type_error(kIncompatibleMethodReceiver,
-                        "RegExp.prototype.@@split", this);
-  }
-  var separator = this;
-  var subject = TO_STRING(string);
-
-  limit = (IS_UNDEFINED(limit)) ? kMaxUint32 : TO_UINT32(limit);
-  var length = subject.length;
-
-  if (limit === 0) return [];
-
-  if (length === 0) {
-    if (DoRegExpExec(separator, subject, 0, 0) !== null) return [];
-    return [subject];
-  }
-
-  var currentIndex = 0;
-  var startIndex = 0;
-  var startMatch = 0;
-  var result = new InternalArray();
-
-  outer_loop:
-  while (true) {
-    if (startIndex === length) {
-      result[result.length] = %_SubString(subject, currentIndex, length);
-      break;
-    }
-
-    var matchInfo = DoRegExpExec(separator, subject, startIndex);
-    if (matchInfo === null || length === (startMatch = matchInfo[CAPTURE0])) {
-      result[result.length] = %_SubString(subject, currentIndex, length);
-      break;
-    }
-    var endIndex = matchInfo[CAPTURE1];
-
-    // We ignore a zero-length match at the currentIndex.
-    if (startIndex === endIndex && endIndex === currentIndex) {
-      if (REGEXP_UNICODE(this) && AtSurrogatePair(subject, startIndex)) {
-        startIndex += 2;
-      } else {
-        startIndex++;
-      }
-      continue;
-    }
-
-    result[result.length] = %_SubString(subject, currentIndex, startMatch);
-
-    if (result.length === limit) break;
-
-    var matchinfo_len = NUMBER_OF_CAPTURES(matchInfo) + REGEXP_FIRST_CAPTURE;
-    for (var i = REGEXP_FIRST_CAPTURE + 2; i < matchinfo_len; ) {
-      var start = matchInfo[i++];
-      var end = matchInfo[i++];
-      if (end != -1) {
-        result[result.length] = %_SubString(subject, start, end);
-      } else {
-        result[result.length] = UNDEFINED;
-      }
-      if (result.length === limit) break outer_loop;
-    }
-
-    startIndex = currentIndex = endIndex;
-  }
-
-  var array_result = [];
-  %MoveArrayContents(result, array_result);
-  return array_result;
-}
-
-
-// ES#sec-regexp.prototype-@@split
-// RegExp.prototype [ @@split ] ( string, limit )
-function RegExpSubclassSplit(string, limit) {
-  if (!IS_RECEIVER(this)) {
-    throw %make_type_error(kIncompatibleMethodReceiver,
-                        "RegExp.prototype.@@split", this);
-  }
-  string = TO_STRING(string);
-  var constructor = SpeciesConstructor(this, GlobalRegExp);
-  var flags = TO_STRING(this.flags);
-
-  // TODO(adamk): this fast path is wrong as we doesn't ensure that 'exec'
-  // is actually a data property on RegExp.prototype.
-  if (IS_REGEXP(this) && constructor === GlobalRegExp) {
-    var exec = this.exec;
-    if (exec === RegExpExecJS) {
-      return %_Call(RegExpSplit, this, string, limit);
-    }
-  }
-
-  var unicode = %StringIndexOf(flags, 'u', 0) >= 0;
-  var sticky = %StringIndexOf(flags, 'y', 0) >= 0;
-  var newFlags = sticky ? flags : flags + "y";
-  var splitter = new constructor(this, newFlags);
-  var array = new GlobalArray();
-  var arrayIndex = 0;
-  var lim = (IS_UNDEFINED(limit)) ? kMaxUint32 : TO_UINT32(limit);
-  var size = string.length;
-  var prevStringIndex = 0;
-  if (lim === 0) return array;
-  var result;
-  if (size === 0) {
-    result = RegExpSubclassExec(splitter, string);
-    if (IS_NULL(result)) %AddElement(array, 0, string);
-    return array;
-  }
-  var stringIndex = prevStringIndex;
-  while (stringIndex < size) {
-    splitter.lastIndex = stringIndex;
-    result = RegExpSubclassExec(splitter, string);
-    if (IS_NULL(result)) {
-      stringIndex += AdvanceStringIndex(string, stringIndex, unicode);
-    } else {
-      var end = MinSimple(TO_LENGTH(splitter.lastIndex), size);
-      if (end === prevStringIndex) {
-        stringIndex += AdvanceStringIndex(string, stringIndex, unicode);
-      } else {
-        %AddElement(
-            array, arrayIndex,
-            %_SubString(string, prevStringIndex, stringIndex));
-        arrayIndex++;
-        if (arrayIndex === lim) return array;
-        prevStringIndex = end;
-        var numberOfCaptures = MaxSimple(TO_LENGTH(result.length), 0);
-        for (var i = 1; i < numberOfCaptures; i++) {
-          %AddElement(array, arrayIndex, result[i]);
-          arrayIndex++;
-          if (arrayIndex === lim) return array;
-        }
-        stringIndex = prevStringIndex;
-      }
-    }
-  }
-  %AddElement(array, arrayIndex,
-                     %_SubString(string, prevStringIndex, size));
-  return array;
-}
-%FunctionRemovePrototype(RegExpSubclassSplit);
-
-
 // Legacy implementation of RegExp.prototype[Symbol.replace] which
 // doesn't properly call the underlying exec method.
 
@@ -673,7 +520,6 @@ function RegExpSubclassReplace(string, replace) {
 
 utils.InstallFunctions(GlobalRegExp.prototype, DONT_ENUM, [
   replaceSymbol, RegExpSubclassReplace,
-  splitSymbol, RegExpSubclassSplit,
 ]);
 
 %InstallToContext(["regexp_last_match_info", RegExpLastMatchInfo]);