Use ICU for ID_START and ID_CONTINUE for Unicode 9 data

test/webkit/ToNumber.js

Index: test/webkit/ToNumber.js
diff --git a/test/webkit/ToNumber.js b/test/webkit/ToNumber.js
index 8d1a40b14cb04024284d2877a09b490fa1d80151..4a287600e33c5f97c8d54c384c32c229c94a8113 100644
--- a/test/webkit/ToNumber.js
+++ b/test/webkit/ToNumber.js
@@ -105,7 +105,7 @@ shouldBe("+lf", "0");
 shouldBe("+ls", "0");
 shouldBe("+ps", "0");
 shouldBe("+oghamSpaceMark", "0");
-shouldBe("+mongolianVowelSeparator", "0");
+shouldBe("+mongolianVowelSeparator", "NaN");

[Dan Ehrenberg, 2017/06/10 06:00:12]

This test expectations change seems appropriate to me, as a continuation of the
work Yang started in https://codereview.chromium.org/2720953003 . Maybe you
could reference this earlier partial work in the commit message when you're
discussing this change.

[jungshik at Google, 2017/06/13 09:37:29]

will do

 shouldBe("+enQuad", "0");
 shouldBe("+emQuad", "0");
 shouldBe("+enSpace", "0");
@@ -129,7 +129,7 @@ shouldBe("+(lf + '1')", "1");
 shouldBe("+(ls + '1')", "1");
 shouldBe("+(ps + '1')", "1");
 shouldBe("+(oghamSpaceMark + '1')", "1");
-shouldBe("+(mongolianVowelSeparator + '1')", "1");
+shouldBe("+(mongolianVowelSeparator + '1')", "NaN");
 shouldBe("+(enQuad + '1')", "1");
 shouldBe("+(emQuad + '1')", "1");
 shouldBe("+(enSpace + '1')", "1");
@@ -153,7 +153,7 @@ shouldBe("+('1' + lf)", "1");
 shouldBe("+('1' + ls)", "1");
 shouldBe("+('1' + ps)", "1");
 shouldBe("+('1' + oghamSpaceMark)", "1");
-shouldBe("+('1' + mongolianVowelSeparator)", "1");
+shouldBe("+('1' + mongolianVowelSeparator)", "NaN");
 shouldBe("+('1' + enQuad)", "1");
 shouldBe("+('1' + emQuad)", "1");
 shouldBe("+('1' + enSpace)", "1");