[regexp, intl] Intl should not cause side effects to the RegExp object.

src/js/i18n.js

 // Copyright 2013 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 // ECMAScript 402 API implementation.
 
 /**
  * Intl object is a single object that has some named properties,
  * all of which are constructors.
  */
 (function(global, utils) {
 
 "use strict";
 
 %CheckIsBootstrapping();
 
 // -------------------------------------------------------------------
 // Imports
 
 var ArrayIndexOf;
 var ArrayJoin;
 var ArrayPush;
-var InstallFunctions = utils.InstallFunctions;
-var InstallGetter = utils.InstallGetter;
-var IsFinite;
-var IsNaN;
 var GlobalBoolean = global.Boolean;
 var GlobalDate = global.Date;
 var GlobalNumber = global.Number;
 var GlobalRegExp = global.RegExp;
 var GlobalString = global.String;
+var InstallFunctions = utils.InstallFunctions;
+var InstallGetter = utils.InstallGetter;
+var InternalPackedArray = utils.InternalPackedArray;
+var IsFinite;
+var IsNaN;
 var MakeError;
 var MakeRangeError;
 var MakeTypeError;
 var MathFloor;
 var ObjectDefineProperties = utils.ImportNow("ObjectDefineProperties");
 var ObjectDefineProperty = utils.ImportNow("ObjectDefineProperty");
 var ObjectHasOwnProperty = utils.ImportNow("ObjectHasOwnProperty");
 var patternSymbol = utils.ImportNow("intl_pattern_symbol");
-var RegExpTest;
 var resolvedSymbol = utils.ImportNow("intl_resolved_symbol");
 var SetFunctionName = utils.SetFunctionName;
+var RegExpMatch;
+var RegExpReplace
 var StringIndexOf;
 var StringLastIndexOf;
-var StringMatch;
-var StringReplace;
 var StringSplit;
 var StringSubstr;
 var StringSubstring;
 
 utils.Import(function(from) {
   ArrayIndexOf = from.ArrayIndexOf;
   ArrayJoin = from.ArrayJoin;
   ArrayPush = from.ArrayPush;
   IsFinite = from.IsFinite;
   IsNaN = from.IsNaN;
   MakeError = from.MakeError;
   MakeRangeError = from.MakeRangeError;
   MakeTypeError = from.MakeTypeError;
   MathFloor = from.MathFloor;
-  RegExpTest = from.RegExpTest;
+  RegExpMatch = from.InternalRegExpMatch;
+  RegExpReplace = from.InternalRegExpReplace;

[Dan Ehrenberg, 2016/03/23 15:55:41]

Could you use the original names, so it's clearer when reading the code below
that uses them?

[Yang, 2016/03/24 06:14:13]

Done.

   StringIndexOf = from.StringIndexOf;
   StringLastIndexOf = from.StringLastIndexOf;
-  StringMatch = from.StringMatch;
-  StringReplace = from.StringReplace;
   StringSplit = from.StringSplit;
   StringSubstr = from.StringSubstr;
   StringSubstring = from.StringSubstring;
 });
 
 // Utilities for definitions
 
 function OverrideFunction(object, name, f) {
   %CheckIsBootstrapping();
   ObjectDefineProperty(object, name, { value: f,
@@ skipping 192 matching lines @@
   }
   return TIMEZONE_NAME_LOCATION_PART_RE;
 }
 
 
 /**
  * Returns an intersection of locales and service supported locales.
  * Parameter locales is treated as a priority list.
  */
 function supportedLocalesOf(service, locales, options) {
-  if (IS_NULL(%_Call(StringMatch, service, GetServiceRE()))) {
+  if (IS_NULL(RegExpMatch(GetServiceRE(), service))) {
     throw MakeError(kWrongServiceType, service);
   }
 
   // Provide defaults if matcher was not specified.
   if (IS_UNDEFINED(options)) {
     options = {};
   } else {
     options = TO_OBJECT(options);
   }
 
@@ skipping 27 matching lines @@
 
 /**
  * Returns the subset of the provided BCP 47 language priority list for which
  * this service has a matching locale when using the BCP 47 Lookup algorithm.
  * Locales appear in the same order in the returned list as in the input list.
  */
 function lookupSupportedLocalesOf(requestedLocales, availableLocales) {
   var matchedLocales = [];
   for (var i = 0; i < requestedLocales.length; ++i) {
     // Remove -u- extension.
-    var locale = %_Call(StringReplace,
-                        requestedLocales[i],
-                        GetUnicodeExtensionRE(),
-                        '');
+    var locale = RegExpReplace(
+        GetUnicodeExtensionRE(), requestedLocales[i], '');
     do {
       if (!IS_UNDEFINED(availableLocales[locale])) {
         // Push requested locale not the resolved one.
         %_Call(ArrayPush, matchedLocales, requestedLocales[i]);
         break;
       }
       // Truncate locale if possible, if not break.
       var pos = %_Call(StringLastIndexOf, locale, '-');
       if (pos === -1) {
         break;
@@ skipping 85 matching lines @@
 
   return resolved;
 }
 
 
 /**
  * Returns best matched supported locale and extension info using basic
  * lookup algorithm.
  */
 function lookupMatcher(service, requestedLocales) {
-  if (IS_NULL(%_Call(StringMatch, service, GetServiceRE()))) {
+  if (IS_NULL(RegExpMatch(GetServiceRE(), service))) {
     throw MakeError(kWrongServiceType, service);
   }
 
   // Cache these, they don't ever change per service.
   if (IS_UNDEFINED(AVAILABLE_LOCALES[service])) {
     AVAILABLE_LOCALES[service] = getAvailableLocalesOf(service);
   }
 
   for (var i = 0; i < requestedLocales.length; ++i) {
     // Remove all extensions.
-    var locale = %_Call(StringReplace, requestedLocales[i],
-                        GetAnyExtensionRE(), '');
+    var locale = RegExpReplace(GetAnyExtensionRE(), requestedLocales[i], '');
     do {
       if (!IS_UNDEFINED(AVAILABLE_LOCALES[service][locale])) {
         // Return the resolved locale and extension.
-        var extensionMatch =
-            %_Call(StringMatch, requestedLocales[i], GetUnicodeExtensionRE());
+        var extensionMatch = RegExpMatch(
+            GetUnicodeExtensionRE(), requestedLocales[i]);
         var extension = IS_NULL(extensionMatch) ? '' : extensionMatch[0];
         return {'locale': locale, 'extension': extension, 'position': i};
       }
       // Truncate locale if possible.
       var pos = %_Call(StringLastIndexOf, locale, '-');
       if (pos === -1) {
         break;
       }
       locale = %_Call(StringSubstring, locale, 0, pos);
     } while (true);
@@ skipping 153 matching lines @@
   if (original === resolved) {
     return original;
   }
 
   var locales = %GetLanguageTagVariants([original, resolved]);
   if (locales[0].maximized !== locales[1].maximized) {
     return resolved;
   }
 
   // Preserve extensions of resolved locale, but swap base tags with original.
-  var resolvedBase = new GlobalRegExp('^' + locales[1].base);
-  return %_Call(StringReplace, resolved, resolvedBase, locales[0].base);
+  var resolvedBase = new GlobalRegExp('^' + locales[1].base, 'g');
+  return RegExpReplace(resolvedBase, resolved, locales[0].base);
 }
 
 
 /**
  * Returns an Object that contains all of supported locales for a given
  * service.
  * In addition to the supported locales we add xx-ZZ locale for each xx-Yyyy-ZZ
  * that is supported. This is required by the spec.
  */
 function getAvailableLocalesOf(service) {
   var available = %AvailableLocalesOf(service);
 
   for (var i in available) {
     if (HAS_OWN_PROPERTY(available, i)) {
-      var parts =
-          %_Call(StringMatch, i, /^([a-z]{2,3})-([A-Z][a-z]{3})-([A-Z]{2})$/);
-      if (parts !== null) {
+      var parts = RegExpMatch(/^([a-z]{2,3})-([A-Z][a-z]{3})-([A-Z]{2})$/, i);
+      if (!IS_NULL(parts)) {
         // Build xx-ZZ. We don't care about the actual value,
         // as long it's not undefined.
         available[parts[1] + '-' + parts[3]] = null;
       }
     }
   }
 
   return available;
 }
 
@@ skipping 49 matching lines @@
          %StringToLowerCase(%_Call(StringSubstr, word, 1));
 }
 
 /**
  * Returns titlecased location, bueNos_airES -> Buenos_Aires
  * or ho_cHi_minH -> Ho_Chi_Minh. It is locale-agnostic and only
  * deals with ASCII only characters.
  * 'of', 'au' and 'es' are special-cased and lowercased.
  */
 function toTitleCaseTimezoneLocation(location) {
-  var match = %_Call(StringMatch, location, GetTimezoneNameLocationPartRE());
+  var match = RegExpMatch(GetTimezoneNameLocationPartRE(), location)
   if (IS_NULL(match)) throw MakeRangeError(kExpectedLocation, location);
 
   var result = toTitleCaseWord(match[1]);
   if (!IS_UNDEFINED(match[2]) && 2 < match.length) {
     // The first character is a separator, '_' or '-'.
     // None of IANA zone names has both '_' and '-'.
     var separator = %_Call(StringSubstring, match[2], 0, 1);
     var parts = %_Call(StringSplit, match[2], separator);
     for (var i = 1; i < parts.length; i++) {
       var part = parts[i]
@@ skipping 76 matching lines @@
  * Validates the language tag. Section 2.2.9 of the bcp47 spec
  * defines a valid tag.
  *
  * ICU is too permissible and lets invalid tags, like
  * hant-cmn-cn, through.
  *
  * Returns false if the language tag is invalid.
  */
 function isValidLanguageTag(locale) {
   // Check if it's well-formed, including grandfadered tags.
-  if (!%_Call(RegExpTest, GetLanguageTagRE(), locale)) {
+  if (IS_NULL(RegExpMatch(GetLanguageTagRE(), locale))) {
     return false;
   }
 
   // Just return if it's a x- form. It's all private.
   if (%_Call(StringIndexOf, locale, 'x-') === 0) {
     return true;
   }
 
   // Check if there are any duplicate variants or singletons (extensions).
 
   // Remove private use section.
-  locale = %_Call(StringSplit, locale, /-x-/)[0];
+  locale = %_Call(StringSplit, locale, '-x-')[0];
 
   // Skip language since it can match variant regex, so we start from 1.
   // We are matching i-klingon here, but that's ok, since i-klingon-klingon
   // is not valid and would fail LANGUAGE_TAG_RE test.
   var variants = [];
   var extensions = [];
-  var parts = %_Call(StringSplit, locale, /-/);
+  var parts = %_Call(StringSplit, locale, '-');
   for (var i = 1; i < parts.length; i++) {
     var value = parts[i];
-    if (%_Call(RegExpTest, GetLanguageVariantRE(), value) &&
+    if (!IS_NULL(RegExpMatch(GetLanguageVariantRE(), value)) &&
         extensions.length === 0) {
       if (%_Call(ArrayIndexOf, variants, value) === -1) {
         %_Call(ArrayPush, variants, value);
       } else {
         return false;
       }
     }
 
-    if (%_Call(RegExpTest, GetLanguageSingletonRE(), value)) {
+    if (!IS_NULL(RegExpMatch(GetLanguageSingletonRE(), value))) {
       if (%_Call(ArrayIndexOf, extensions, value) === -1) {
         %_Call(ArrayPush, extensions, value);
       } else {
         return false;
       }
     }
   }
 
   return true;
  }
@@ skipping 235 matching lines @@
 
 
 AddBoundMethod(Intl.Collator, 'compare', compare, 2);
 
 /**
  * Verifies that the input is a well-formed ISO 4217 currency code.
  * Don't uppercase to test. It could convert invalid code into a valid one.
  * For example \u00DFP (Eszett+P) becomes SSP.
  */
 function isWellFormedCurrencyCode(currency) {
-  return typeof currency == "string" &&
-      currency.length == 3 &&
-      %_Call(StringMatch, currency, /[^A-Za-z]/) == null;
+  return typeof currency == "string" && currency.length == 3 &&
+      IS_NULL(RegExpMatch(/[^A-Za-z]/, currency));
 }
 
 
 /**
  * Returns the valid digit count for a property, or throws RangeError on
  * a value out of the range.
  */
 function getNumberOption(options, property, min, max, fallback) {
   var value = options[property];
   if (!IS_UNDEFINED(value)) {
@@ skipping 309 matching lines @@
     return '';
   }
 }
 
 
 /**
  * Returns object that matches LDML representation of the date.
  */
 function fromLDMLString(ldmlString) {
   // First remove '' quoted text, so we lose 'Uhr' strings.
-  ldmlString = %_Call(StringReplace, ldmlString, GetQuotedStringRE(), '');
+  ldmlString = RegExpReplace(GetQuotedStringRE(), ldmlString, '');
 
   var options = {};
-  var match = %_Call(StringMatch, ldmlString, /E{3,5}/g);
+  var match = RegExpMatch(/E{3,5}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'weekday', match, {EEEEE: 'narrow', EEE: 'short', EEEE: 'long'});
 
-  match = %_Call(StringMatch, ldmlString, /G{3,5}/g);
+  match = RegExpMatch(/G{3,5}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'era', match, {GGGGG: 'narrow', GGG: 'short', GGGG: 'long'});
 
-  match = %_Call(StringMatch, ldmlString, /y{1,2}/g);
+  match = RegExpMatch(/y{1,2}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'year', match, {y: 'numeric', yy: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /M{1,5}/g);
+  match = RegExpMatch(/M{1,5}/, ldmlString);
   options = appendToDateTimeObject(options, 'month', match, {MM: '2-digit',
       M: 'numeric', MMMMM: 'narrow', MMM: 'short', MMMM: 'long'});
 
   // Sometimes we get L instead of M for month - standalone name.
-  match = %_Call(StringMatch, ldmlString, /L{1,5}/g);
+  match = RegExpMatch(/L{1,5}/, ldmlString);
   options = appendToDateTimeObject(options, 'month', match, {LL: '2-digit',
       L: 'numeric', LLLLL: 'narrow', LLL: 'short', LLLL: 'long'});
 
-  match = %_Call(StringMatch, ldmlString, /d{1,2}/g);
+  match = RegExpMatch(/d{1,2}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'day', match, {d: 'numeric', dd: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /h{1,2}/g);
+  match = RegExpMatch(/h{1,2}/, ldmlString);
   if (match !== null) {
     options['hour12'] = true;
   }
   options = appendToDateTimeObject(
       options, 'hour', match, {h: 'numeric', hh: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /H{1,2}/g);
+  match = RegExpMatch(/H{1,2}/, ldmlString);
   if (match !== null) {
     options['hour12'] = false;
   }
   options = appendToDateTimeObject(
       options, 'hour', match, {H: 'numeric', HH: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /m{1,2}/g);
+  match = RegExpMatch(/m{1,2}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'minute', match, {m: 'numeric', mm: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /s{1,2}/g);
+  match = RegExpMatch(/s{1,2}/, ldmlString);
   options = appendToDateTimeObject(
       options, 'second', match, {s: 'numeric', ss: '2-digit'});
 
-  match = %_Call(StringMatch, ldmlString, /z|zzzz/g);
+  match = RegExpMatch(/z|zzzz/, ldmlString);
   options = appendToDateTimeObject(
       options, 'timeZoneName', match, {z: 'short', zzzz: 'long'});
 
   return options;
 }
 
 
 function appendToDateTimeObject(options, option, match, pairs) {
   if (IS_NULL(match)) {
     if (!HAS_OWN_PROPERTY(options, option)) {
@@ skipping 306 matching lines @@
   var upperID = %StringToUpperCase(tzID);
   if (upperID === 'UTC' || upperID === 'GMT' ||
       upperID === 'ETC/UTC' || upperID === 'ETC/GMT') {
     return 'UTC';
   }
 
   // TODO(jshin): Add support for Etc/GMT[+-]([1-9]|1[0-2])
 
   // We expect only _, '-' and / beside ASCII letters.
   // All inputs should conform to Area/Location(/Location)* from now on.
-  var match = %_Call(StringMatch, tzID, GetTimezoneNameCheckRE());
+  var match = RegExpMatch(GetTimezoneNameCheckRE(), tzID);
   if (IS_NULL(match)) throw MakeRangeError(kExpectedTimezoneID, tzID);
 
   var result = toTitleCaseTimezoneLocation(match[1]) + '/' +
                toTitleCaseTimezoneLocation(match[2]);
 
   if (!IS_UNDEFINED(match[3]) && 3 < match.length) {
     var locations = %_Call(StringSplit, match[3], '/');
     // The 1st element is empty. Starts with i=1.
     for (var i = 1; i < locations.length; i++) {
       result = result + '/' + toTitleCaseTimezoneLocation(locations[i]);
@@ skipping 330 matching lines @@
     }
 
     var locales = arguments[0];
     var options = arguments[1];
     return toLocaleDateTime(
         this, locales, options, 'time', 'time', 'dateformattime');
   }
 );
 
 })