[intl] Enhance Date parser to take Unicode SPACE

This is needed to prepare for the landing of ICU72.
Allow U+202F in the Date String, which the toLocaleString("en-US")
will generate w/ ICU72.

Bug: v8:13494
Change-Id: I41b83c4094ce3d0737a72dcd6310b52c68fdcdca
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4027341
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jungshik Shin <jshin@chromium.org>
Commit-Queue: Frank Tang <ftang@chromium.org>
Cr-Commit-Position: refs/heads/main@{#84308}

src/date/dateparser-inl.h

@@ -192,7 +192,7 @@ DateParser::DateToken DateParser::DateStringTokenizer<CharType>::Scan() {
   if (in_->Skip('+')) return DateToken::Symbol('+');
   if (in_->Skip('.')) return DateToken::Symbol('.');
   if (in_->Skip(')')) return DateToken::Symbol(')');
-  if (in_->IsAsciiAlphaOrAbove()) {
+  if (in_->IsAsciiAlphaOrAbove() && !in_->IsWhiteSpaceChar()) {
     DCHECK_EQ(KeywordTable::kPrefixLength, 3);
     uint32_t buffer[3] = {0, 0, 0};
     int length = in_->ReadWord(buffer, 3);

src/date/dateparser.h

@@ -91,7 +91,8 @@ class DateParser : public AllStatic {
     // Return word length.
     int ReadWord(uint32_t* prefix, int prefix_size) {
       int len;
-      for (len = 0; IsAsciiAlphaOrAbove(); Next(), len++) {
+      for (len = 0; IsAsciiAlphaOrAbove() && !IsWhiteSpaceChar();
+           Next(), len++) {
         if (len < prefix_size) prefix[len] = AsciiAlphaToLower(ch_);
       }
       for (int i = len; i < prefix_size; i++) prefix[i] = 0;
@@ -115,6 +116,7 @@ class DateParser : public AllStatic {
     bool IsEnd() const { return ch_ == 0; }
     bool IsAsciiDigit() const { return IsDecimalDigit(ch_); }
     bool IsAsciiAlphaOrAbove() const { return ch_ >= 'A'; }
+    bool IsWhiteSpaceChar() const { return IsWhiteSpace(ch_); }
     bool IsAsciiSign() const { return ch_ == '+' || ch_ == '-'; }
 
     // Return 1 for '+' and -1 for '-'.

test/intl/regress-13494.js

@@ -0,0 +1,47 @@
+// Copyright 2022 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.
+
+// Test the new Date( date.toLocaleString("en-US")) is not invalid.
+// This is not guaranteed by the standard but many code use that to set the
+// timezone as suggested in
+// https://stackoverflow.com/questions/15141762/how-to-initialize-a-javascript-date-to-a-particular-time-zone
+
+let d = new Date();
+
+// https://tc39.es/ecma262/#sec-todatestring
+// 21.4.4.41.4 ToDateString ( tv )
+// 1. If tv is NaN, return "Invalid Date".
+let invalid = "Invalid Date";
+let largestDiff = 25*60*60*1000;
+
+let garbage = new Date("garbage");
+assertTrue(invalid == garbage);
+assertEquals(NaN, garbage.getTime());
+
+let d1 = new Date(d.toLocaleString("en-US"));
+assertTrue(d1 != invalid);
+assertTrue(d1.getTime() != NaN);
+// The milliseconds are different between d1 and d.
+assertTrue(Math.abs(d1-d) < 1000);
+
+// Force a version of date string which have  U+202f before AM
+let nnbsp_am = new Date("11/16/2022, 9:04:55\u202fAM");
+assertTrue(nnbsp_am  != invalid);
+assertTrue(nnbsp_am.getTime() != NaN);
+// Force a version of date string which have  U+202f before PM
+let nnbsp_pm = new Date("11/16/2022, 9:04:55\u202fPM");
+assertTrue(nnbsp_pm  != invalid);
+assertTrue(nnbsp_pm.getTime() != NaN);
+
+let d2 = new Date(d.toLocaleString("en-US", {timeZone: "Asia/Taipei"}));
+assertTrue(d2 != invalid);
+assertTrue(d2.getTime() != NaN);
+// The differences should be within 25 hours.
+assertTrue(Math.abs(d2-d) < largestDiff);
+
+let d3 = new Date(d.toLocaleString("en-US", {timeZone: "Africa/Lusaka"}));
+assertTrue(d3 != invalid);
+assertTrue(d3.getTime() != NaN);
+// The differences should be within 25 hours.
+assertTrue(Math.abs(d3-d) < largestDiff);