[regexp] step back if starting unicode regexp within surrogate pair.

test/mjsunit/harmony/unicode-regexp-last-index.js

Index: test/mjsunit/harmony/unicode-regexp-last-index.js
diff --git a/test/mjsunit/harmony/unicode-regexp-last-index.js b/test/mjsunit/harmony/unicode-regexp-last-index.js
new file mode 100644
index 0000000000000000000000000000000000000000..a986d3020c23407120ada3f96c2b312f0fef87f3
--- /dev/null
+++ b/test/mjsunit/harmony/unicode-regexp-last-index.js
@@ -0,0 +1,59 @@
+// Copyright 2016 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.
+
+// Flags: --harmony-unicode-regexps --harmony-regexp-lookbehind
+
+var r = /./ug;

[erikcorry, 2016/01/22 10:10:10]

All these tests should be duplicated without the /u flag, to ensure that non-u
behaviour did not change.

[Yang, 2016/01/25 07:38:41]

Done.

+assertEquals(["\ud800\udc00"], r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+r.lastIndex = 1;
+assertEquals(["\ud800\udc00"], r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+assertEquals(["\ud801\udc01"], r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 3;
+assertEquals(["\ud801\udc01"], r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 4;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 5;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+
+r.lastIndex = 3;
+assertEquals(["\ud802"], r.exec("\ud800\udc00\ud801\ud802"));
+r.lastIndex = 4;
+assertNull(r.exec("\ud800\udc00\ud801\ud802"));
+
+r = /^./ug;
+assertEquals(["\ud800\udc00"], r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+r.lastIndex = 1;
+assertEquals(["\ud800\udc00"], r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 3;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 4;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 5;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+
+r = /(?:(^.)|.)/ug;
+assertEquals(["\ud800\udc00", "\ud800\udc00"],
+             r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+r.lastIndex = 1;
+assertEquals(["\ud800\udc00", "\ud800\udc00"],
+             r.exec("\ud800\udc00\ud801\udc01"));
+assertEquals(2, r.lastIndex);
+assertEquals(["\ud801\udc01", undefined], r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 3;
+assertEquals(["\ud801\udc01", undefined], r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 4;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+r.lastIndex = 5;
+assertNull(r.exec("\ud800\udc00\ud801\udc01"));
+
+r.lastIndex = 3;
+assertEquals(["\ud802", undefined], r.exec("\ud800\udc00\ud801\ud802"));
+r.lastIndex = 4;
+assertNull(r.exec("\ud800\udc00\ud801\ud802"));