deps: cherry-pick 6989b3f6d7 from V8 upstream

Original commit message:
  Fix default Intl language tag handling

  With certain ICU data bundles (such as the Node.js "small-icu"),
  %GetDefaultICULocale() may return a more specific language tag (e.g.
  "en-US") than what's available (e.g. "en"). In those cases, consider the
  more specific language tag supported.

  This CL also resolves the following Node.js issue:
     nodejs#15223

  Bug: v8:7024
  Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
  Change-Id: Ifda0776b3418734d5caa8af4e50c17cda95add73
  Reviewed-on: https://chromium-review.googlesource.com/668350
  Commit-Queue: Daniel Ehrenberg <[email protected]>
  Reviewed-by: Daniel Ehrenberg <[email protected]>
  Cr-Commit-Position: refs/heads/master@{nodejs#52716}

PR-URL: nodejs#20826
Fixes: nodejs#15223
Refs: v8/v8@6989b3f
Reviewed-By: James M Snell <[email protected]>
Reviewed-By: Anatoli Papirovski <[email protected]>
Reviewed-By: Ben Noordhuis <[email protected]>

Author: TimothyGu

common.gypi

@@ -27,7 +27,7 @@
 
     # Reset this number to 0 on major V8 upgrades.
     # Increment by one for each non-official patch applied to deps/v8.
-    'v8_embedder_string': '-node.6',
+    'v8_embedder_string': '-node.7',
 
     # Enable disassembler for `--print-code` v8 options
     'v8_enable_disassembler': 1,

deps/v8/src/js/intl.js

@@ -152,18 +152,11 @@ var AVAILABLE_LOCALES = {
  */
 var DEFAULT_ICU_LOCALE = UNDEFINED;
 
-function GetDefaultICULocaleJS(service) {
+function GetDefaultICULocaleJS() {
   if (IS_UNDEFINED(DEFAULT_ICU_LOCALE)) {
     DEFAULT_ICU_LOCALE = %GetDefaultICULocale();
   }
-  // Check that this is a valid default for this service,
-  // otherwise fall back to "und"
-  // TODO(littledan,jshin): AvailableLocalesOf sometimes excludes locales
-  // which don't require tailoring, but work fine with root data. Look into
-  // exposing this fact in ICU or the way Chrome bundles data.
-  return (IS_UNDEFINED(service) ||
-          HAS_OWN_PROPERTY(getAvailableLocalesOf(service), DEFAULT_ICU_LOCALE))
-         ? DEFAULT_ICU_LOCALE : "und";
+  return DEFAULT_ICU_LOCALE;
 }
 
 /**
@@ -434,6 +427,48 @@ function resolveLocale(service, requestedLocales, options) {
 }
 
 
+/**
+ * Look up the longest non-empty prefix of |locale| that is an element of
+ * |availableLocales|. Returns undefined when the |locale| is completely
+ * unsupported by |availableLocales|.
+ */
+function bestAvailableLocale(availableLocales, locale) {
+  do {
+    if (!IS_UNDEFINED(availableLocales[locale])) {
+      return locale;
+    }
+    // Truncate locale if possible.
+    var pos = %StringLastIndexOf(locale, '-');
+    if (pos === -1) {
+      break;
+    }
+    locale = %_Call(StringSubstring, locale, 0, pos);
+  } while (true);
+
+  return UNDEFINED;
+}
+
+
+/**
+ * Try to match any mutation of |requestedLocale| against |availableLocales|.
+ */
+function attemptSingleLookup(availableLocales, requestedLocale) {
+  // Remove all extensions.
+  var noExtensionsLocale = %RegExpInternalReplace(
+      GetAnyExtensionRE(), requestedLocale, '');
+  var availableLocale = bestAvailableLocale(
+      availableLocales, requestedLocale);
+  if (!IS_UNDEFINED(availableLocale)) {
+    // Return the resolved locale and extension.
+    var extensionMatch = %regexp_internal_match(
+        GetUnicodeExtensionRE(), requestedLocale);
+    var extension = IS_NULL(extensionMatch) ? '' : extensionMatch[0];
+    return {locale: availableLocale, extension: extension};
+  }
+  return UNDEFINED;
+}
+
+
 /**
  * Returns best matched supported locale and extension info using basic
  * lookup algorithm.
@@ -446,31 +481,25 @@ function lookupMatcher(service, requestedLocales) {
   var availableLocales = getAvailableLocalesOf(service);
 
   for (var i = 0; i < requestedLocales.length; ++i) {
-    // Remove all extensions.
-    var locale = %RegExpInternalReplace(
-        GetAnyExtensionRE(), requestedLocales[i], '');
-    do {
-      if (!IS_UNDEFINED(availableLocales[locale])) {
-        // Return the resolved locale and extension.
-        var extensionMatch = %regexp_internal_match(
-            GetUnicodeExtensionRE(), requestedLocales[i]);
-        var extension = IS_NULL(extensionMatch) ? '' : extensionMatch[0];
-        return {locale: locale, extension: extension, position: i};
-      }
-      // Truncate locale if possible.
-      var pos = %StringLastIndexOf(locale, '-');
-      if (pos === -1) {
-        break;
-      }
-      locale = %_Call(StringSubstring, locale, 0, pos);
-    } while (true);
+    var result = attemptSingleLookup(availableLocales, requestedLocales[i]);
+    if (!IS_UNDEFINED(result)) {
+      return result;
+    }
+  }
+
+  var defLocale = GetDefaultICULocaleJS();
+
+  // While ECMA-402 returns defLocale directly, we have to check if it is
+  // supported, as such support is not guaranteed.
+  var result = attemptSingleLookup(availableLocales, defLocale);
+  if (!IS_UNDEFINED(result)) {
+    return result;
   }
 
   // Didn't find a match, return default.
   return {
-    locale: GetDefaultICULocaleJS(service),
-    extension: '',
-    position: -1
+    locale: 'und',
+    extension: ''
   };
 }
 

deps/v8/test/intl/assert.js

@@ -132,6 +132,16 @@ function assertFalse(value, user_message = '') {
 }
 
 
+/**
+ * Throws if value is null.
+ */
+function assertNotNull(value, user_message = '') {
+  if (value === null) {
+    fail("not null", value, user_message);
+  }
+}
+
+
 /**
  * Runs code() and asserts that it throws the specified exception.
  */
@@ -189,3 +199,34 @@ function assertInstanceof(obj, type) {
                     (actualTypeName ? ' but of < ' + actualTypeName + '>' : ''));
   }
 }
+
+
+/**
+ * Split a BCP 47 language tag into locale and extension.
+ */
+function splitLanguageTag(tag) {
+  var extRe = /(-[0-9A-Za-z](-[0-9A-Za-z]{2,8})+)+$/;
+  var match = %regexp_internal_match(extRe, tag);
+  if (match) {
+    return { locale: tag.slice(0, match.index), extension: match[0] };
+  }
+
+  return { locale: tag, extension: '' };
+}
+
+
+/**
+ * Throw if |parent| is not a more general language tag of |child|, nor |child|
+ * itself, per BCP 47 rules.
+ */
+function assertLanguageTag(child, parent) {
+  var childSplit = splitLanguageTag(child);
+  var parentSplit = splitLanguageTag(parent);
+
+  // Do not compare extensions at this moment, as %GetDefaultICULocale()
+  // doesn't always output something we support.
+  if (childSplit.locale !== parentSplit.locale &&
+      !childSplit.locale.startsWith(parentSplit.locale + '-')) {
+    fail(child, parent, 'language tag comparison');
+  }
+}

deps/v8/test/intl/break-iterator/default-locale.js

@@ -37,8 +37,8 @@ assertFalse(options.locale === 'und');
 assertFalse(options.locale === '');
 assertFalse(options.locale === undefined);
 
-// Then check for equality.
-assertEquals(options.locale, %GetDefaultICULocale());
+// Then check for legitimacy.
+assertLanguageTag(%GetDefaultICULocale(), options.locale);
 
 var iteratorNone = new Intl.v8BreakIterator();
 assertEquals(options.locale, iteratorNone.resolvedOptions().locale);

deps/v8/test/intl/break-iterator/wellformed-unsupported-locale.js

@@ -29,4 +29,4 @@
 
 var iterator = Intl.v8BreakIterator(['xx']);
 
-assertEquals(iterator.resolvedOptions().locale, %GetDefaultICULocale());
+assertLanguageTag(%GetDefaultICULocale(), iterator.resolvedOptions().locale);

deps/v8/test/intl/collator/default-locale.js

@@ -37,8 +37,8 @@ assertFalse(options.locale === 'und');
 assertFalse(options.locale === '');
 assertFalse(options.locale === undefined);
 
-// Then check for equality.
-assertEquals(options.locale, %GetDefaultICULocale());
+// Then check for legitimacy.
+assertLanguageTag(%GetDefaultICULocale(), options.locale);
 
 var collatorNone = new Intl.Collator();
 assertEquals(options.locale, collatorNone.resolvedOptions().locale);
@@ -48,5 +48,8 @@ var collatorBraket = new Intl.Collator({});
 assertEquals(options.locale, collatorBraket.resolvedOptions().locale);
 
 var collatorWithOptions = new Intl.Collator(undefined, {usage: 'search'});
-assertEquals(%GetDefaultICULocale() + '-u-co-search',
-             collatorWithOptions.resolvedOptions().locale);
+assertLanguageTag(%GetDefaultICULocale(),
+                  collatorWithOptions.resolvedOptions().locale);
+assertNotNull(
+    %regexp_internal_match(/-u(-[a-zA-Z]+-[a-zA-Z]+)*-co-search/,
+        collatorWithOptions.resolvedOptions().locale));

deps/v8/test/intl/collator/wellformed-unsupported-locale.js

@@ -29,4 +29,4 @@
 
 var collator = Intl.Collator(['xx']);
 
-assertEquals(collator.resolvedOptions().locale, %GetDefaultICULocale());
+assertLanguageTag(%GetDefaultICULocale(), collator.resolvedOptions().locale);

deps/v8/test/intl/date-format/default-locale.js

@@ -37,8 +37,8 @@ assertFalse(options.locale === 'und');
 assertFalse(options.locale === '');
 assertFalse(options.locale === undefined);
 
-// Then check for equality.
-assertEquals(options.locale, %GetDefaultICULocale());
+// Then check for legitimacy.
+assertLanguageTag(%GetDefaultICULocale(), options.locale);
 
 var dtfNone = new Intl.DateTimeFormat();
 assertEquals(options.locale, dtfNone.resolvedOptions().locale);

deps/v8/test/intl/date-format/wellformed-unsupported-locale.js

@@ -29,4 +29,4 @@
 
 var dtf = Intl.DateTimeFormat(['xx']);
 
-assertEquals(dtf.resolvedOptions().locale, %GetDefaultICULocale());
+assertLanguageTag(%GetDefaultICULocale(), dtf.resolvedOptions().locale);

deps/v8/test/intl/number-format/default-locale.js

@@ -37,8 +37,8 @@ assertFalse(options.locale === 'und');
 assertFalse(options.locale === '');
 assertFalse(options.locale === undefined);
 
-// Then check for equality.
-assertEquals(options.locale, %GetDefaultICULocale());
+// Then check for legitimacy.
+assertLanguageTag(%GetDefaultICULocale(), options.locale);
 
 var nfNone = new Intl.NumberFormat();
 assertEquals(options.locale, nfNone.resolvedOptions().locale);

deps/v8/test/intl/number-format/wellformed-unsupported-locale.js

@@ -29,4 +29,4 @@
 
 var nf = Intl.NumberFormat(['xx']);
 
-assertEquals(nf.resolvedOptions().locale, %GetDefaultICULocale());
+assertLanguageTag(%GetDefaultICULocale(), nf.resolvedOptions().locale);

deps/v8/test/mjsunit/regress/regress-6288.js

@@ -8,6 +8,6 @@
 // DateTimeFormat but not Collation
 
 if (this.Intl) {
-  assertEquals('und', Intl.Collator().resolvedOptions().locale);
+  assertEquals('pt', Intl.Collator().resolvedOptions().locale);
   assertEquals('pt-BR', Intl.DateTimeFormat().resolvedOptions().locale);
 }