ValueSerializer: add kOneByteString to expected key fast path.

This was missed when Latin-1 encoding replaced UTF-8 encoding when one-byte
strings (like most keys) are serialized.

BUG=chromium:686159

Review-Url: https://codereview.chromium.org/2784423002
Cr-Commit-Position: refs/heads/master@{#44320}

src/value-serializer.cc

@@ -1261,10 +1261,9 @@ bool ValueDeserializer::ReadExpectedString(Handle<String> expected) {
 
   // If the bytes are verbatim what is in the flattened string, then the string
   // is successfully consumed.
-  if (tag == SerializationTag::kUtf8String && flat.IsOneByte()) {
+  if (tag == SerializationTag::kOneByteString && flat.IsOneByte()) {
     Vector<const uint8_t> chars = flat.ToOneByteVector();
     if (byte_length == static_cast<size_t>(chars.length()) &&
-        String::IsAscii(chars.begin(), chars.length()) &&
         memcmp(bytes.begin(), chars.begin(), byte_length) == 0) {
       return true;
     }
@@ -1274,6 +1273,13 @@ bool ValueDeserializer::ReadExpectedString(Handle<String> expected) {
         memcmp(bytes.begin(), chars.begin(), byte_length) == 0) {
       return true;
     }
+  } else if (tag == SerializationTag::kUtf8String && flat.IsOneByte()) {
+    Vector<const uint8_t> chars = flat.ToOneByteVector();
+    if (byte_length == static_cast<size_t>(chars.length()) &&
+        String::IsAscii(chars.begin(), chars.length()) &&
+        memcmp(bytes.begin(), chars.begin(), byte_length) == 0) {
+      return true;
+    }
   }
 
   position_ = original_position;