Implement ES6 Template Literals

src/scanner.cc

Index: src/scanner.cc
diff --git a/src/scanner.cc b/src/scanner.cc
index ddcd937584d50fc45fa921a25e2a5d21fcf5697e..8e05006688c0084c98719c553cfc7634a2f3cec8 100644
--- a/src/scanner.cc
+++ b/src/scanner.cc
@@ -404,6 +404,7 @@ Token::Value Scanner::ScanHtmlComment() {
 void Scanner::Scan() {
   next_.literal_chars = NULL;
   Token::Value token;
+
   do {
     // Remember the position of the next token
     next_.location.beg_pos = source_pos();
@@ -626,6 +627,12 @@ void Scanner::Scan() {
         token = Select(Token::BIT_NOT);
         break;
 
+      case '`':
+        if (HarmonyTemplates()) {
+          token = ScanTemplateSpan();
+          break;
+        }
+
       default:
         if (c0_ < 0) {
           token = Token::EOS;
@@ -673,6 +680,7 @@ void Scanner::SeekForward(int pos) {
 
 bool Scanner::ScanEscape() {
   uc32 c = c0_;
+  bool singleCharEscape = true;

[arv (Not doing code reviews), 2014/11/11 17:15:15]

This is no longer used

   Advance();
 
   // Skip escaped newlines.
@@ -694,12 +702,14 @@ bool Scanner::ScanEscape() {
     case 'r' : c = '\r'; break;
     case 't' : c = '\t'; break;
     case 'u' : {
+      singleCharEscape = false;
       c = ScanHexNumber(4);
       if (c < 0) return false;
       break;
     }
     case 'v' : c = '\v'; break;
     case 'x' : {
+      singleCharEscape = false;
       c = ScanHexNumber(2);
       if (c < 0) return false;
       break;
@@ -711,7 +721,10 @@ bool Scanner::ScanEscape() {
     case '4' :  // fall through
     case '5' :  // fall through
     case '6' :  // fall through
-    case '7' : c = ScanOctalEscape(c, 2); break;
+    case '7':
+      singleCharEscape = false;
+      c = ScanOctalEscape(c, 2);
+      break;
   }
 
   // According to ECMA-262, section 7.8.4, characters not covered by the
@@ -770,6 +783,80 @@ Token::Value Scanner::ScanString() {
 }
 
 
+Token::Value Scanner::ScanTemplateSpan() {
+  // When scanning a TemplateSpan, we are looking for the following construct:
+  // TEMPLATE_SPAN ::
+  //     ` LiteralChars* ${
+  //   | } LiteralChars* ${
+  //
+  // TEMPLATE_TAIL ::
+  //     ` LiteralChars* `
+  //   | } LiteralChar* `
+  //
+  // A TEMPLATE_SPAN should always be followed by an Expression, while a
+  // TEMPLATE_TAIL terminates a TemplateLiteral and does not need to be
+  // followed by an Expression.
+  //
+
+  if (next_.token == Token::RBRACE) {
+    // After parsing an Expression, the source position is incorrect due to
+    // having scanned the brace. Push the RBRACE back into the stream.
+    PushBack('}');
+  }
+
+  next_.location.beg_pos = source_pos();
+  Token::Value result = Token::ILLEGAL;
+  DCHECK(c0_ == '`' || c0_ == '}');
+  Advance();  // Consume ` or }
+
+  LiteralScope literal(this);
+  while (true) {
+    uc32 c = c0_;
+    Advance();
+    if (c == '`') {
+      result = Token::TEMPLATE_TAIL;
+      break;
+    } else if (c == '$' && c0_ == '{') {
+      Advance();  // Consume '{'
+      result = Token::TEMPLATE_SPAN;
+      break;
+    } else if (c == '\\') {
+      if (unicode_cache_->IsLineTerminator(c0_)) {
+        // The TV of LineContinuation :: \ LineTerminatorSequence is the empty
+        // code unit sequence.
+        do {
+          uc32 lastChar = c0_;
+          Advance();
+          if (lastChar == '\r' && c0_ == '\n') Advance();
+        } while (unicode_cache_->IsLineTerminator(c0_));
+      } else if (c0_ == '0') {
+        Advance();
+        AddLiteralChar('0');
+      } else {
+        ScanEscape();
+      }
+    } else if (c < 0) {
+      // Unterminated template literal
+      PushBack(c);
+      break;
+    } else {
+      // The TRV of LineTerminatorSequence :: <CR> is the CV 0x000A.
+      // The TRV of LineTerminatorSequence :: <CR><LF> is the sequence
+      // consisting of the CV 0x000A.
+      if (c == '\r') {
+        if (c0_ == '\n') Advance();
+        c = '\n';
+      }
+      AddLiteralChar(c);
+    }
+  }
+  literal.Complete();
+  next_.location.end_pos = source_pos();
+  next_.token = result;
+  return result;
+}
+
+
 void Scanner::ScanDecimalDigits() {
   while (IsDecimalDigit(c0_))
     AddLiteralCharAdvance();