Fix inconsistencies wrt whitespaces.

src/scanner.cc

Index: src/scanner.cc
diff --git a/src/scanner.cc b/src/scanner.cc
index 26f840b23a5914e9e5d52c207d5f9ed904e15a9d..27768547fb7b56ab70d51b2df0aed7a093510d2e 100644
--- a/src/scanner.cc
+++ b/src/scanner.cc
@@ -246,7 +246,8 @@ Token::Value Scanner::Next() {
 }
 
 
-static inline bool IsByteOrderMark(uc32 c) {
+// TODO(yangguo): check whether this is actually necessary.
+static inline bool IsLittleEndianByteOrderMark(uc32 c) {
   // The Unicode value U+FFFE is guaranteed never to be assigned as a
   // Unicode character; this implies that in a Unicode context the
   // 0xFF, 0xFE byte pattern can only be interpreted as the U+FEFF
@@ -254,7 +255,7 @@ static inline bool IsByteOrderMark(uc32 c) {
   // not be a U+FFFE character expressed in big-endian byte
   // order). Nevertheless, we check for it to be compatible with
   // Spidermonkey.
-  return c == 0xFEFF || c == 0xFFFE;
+  return c == 0xFFFE;
 }
 
 
@@ -262,14 +263,14 @@ bool Scanner::SkipWhiteSpace() {
   int start_position = source_pos();
 
   while (true) {
-    // We treat byte-order marks (BOMs) as whitespace for better
-    // compatibility with Spidermonkey and other JavaScript engines.
-    while (unicode_cache_->IsWhiteSpace(c0_) || IsByteOrderMark(c0_)) {
-      // IsWhiteSpace() includes line terminators!
+    while (true) {
+      // Advance as long as character is a WhiteSpace or LineTerminator.
+      // Remember if the latter is the case.
       if (unicode_cache_->IsLineTerminator(c0_)) {
-        // Ignore line terminators, but remember them. This is necessary
-        // for automatic semicolon insertion.
         has_line_terminator_before_next_ = true;
+      } else if (!unicode_cache_->IsWhiteSpace(c0_) &&
+                 !IsLittleEndianByteOrderMark(c0_)) {
+        break;
       }
       Advance();
     }