Fixes scanning of Utf16 surrogate pairs in string literals.

runtime/vm/scanner.cc

 // Copyright (c) 2012, the Dart project authors.  Please see the AUTHORS file
 // for details. All rights reserved. Use of this source code is governed by a
 // BSD-style license that can be found in the LICENSE file.
 
 #include "vm/scanner.h"
 
 #include "platform/assert.h"
 #include "vm/flags.h"
 #include "vm/object.h"
 #include "vm/object_store.h"
@@ skipping 563 matching lines @@
           // Strings are canonicalized: Allocate a symbol.
           current_token_.literal = &String::ZoneHandle(
               Symbols::FromUTF32(string_chars.data(), string_chars.length()));
         }
         EndStringLiteral();
         return;
       } else {
         string_chars.Add(string_delimiter_);
       }
     } else {
-      string_chars.Add(c0_);
+      // Test for a two part utf16 sequence, and decode to a code point
+      // if we find one.
+      int32_t ch1 = c0_;
+      if (Utf16::IsLeadSurrogate(ch1)) {
+        ReadChar();
+        const int32_t ch2 = c0_;
+        if (Utf16::IsTrailSurrogate(ch2)) {
+          ch1 = Utf16::Decode(ch1, ch2);
+        } else {
+          string_chars.Add(ch1);
+          continue;

[Ivan Posva, 2013/08/14 05:21:52]

Can you restructure this to avoid the continue?

[zra, 2013/08/14 14:59:11]

Yes. Missed the LookaheadChar function before.

+        }
+      }
+      string_chars.Add(ch1);
     }
     ReadChar();
   }
 }
 
 
 void Scanner::Scan() {
   newline_seen_ = false;
 
   do {
@@ skipping 345 matching lines @@
               "%c%#"Px"", kPrivateKeySeparator, key_value);
   const String& result = String::Handle(String::New(private_key, Heap::kOld));
   return result.raw();
 }
 
 
 void Scanner::InitOnce() {
 }
 
 }  // namespace dart