For IsAtDotDotToken, ensure the current token is a DotToken before parsing the next token to see if it's also a DotToken. (#78338)

I see LanguageParser.IsAtDotToken accounting for about 1 second of CPU time (0.5%) in the C# editing speedometer test. Local testing indicates about 85% of calls into this method will be filtered by this check.

Test insertion: https://dev.azure.com/devdiv/DevDiv/_git/VS/pullrequest/631973

See PR for details on the perf characteristics with and without the change.

Author: ToddGrun

src/Compilers/CSharp/Portable/Parser/LanguageParser.cs

@@ -11477,7 +11477,13 @@ private AssignmentExpressionSyntax ParseAssignmentExpression(SyntaxKind operator
 
         /// <summary>Check if we're currently at a .. sequence that can then be parsed out as a <see cref="SyntaxKind.DotDotToken"/>.</summary>
         public bool IsAtDotDotToken()
-            => IsAtDotDotToken(this.CurrentToken, this.PeekToken(1));
+        {
+            if (this.CurrentToken.Kind != SyntaxKind.DotToken)
+                return false;
+
+            var nextToken = this.PeekToken(1);
+            return nextToken.Kind == SyntaxKind.DotToken && NoTriviaBetween(this.CurrentToken, nextToken);
+        }
 
         public static bool IsAtDotDotToken(SyntaxToken token1, SyntaxToken token2)
             => token1.Kind == SyntaxKind.DotToken &&