Question

How do you ignore a null statement, and only apply a method to remove special characters to only strings that are populated.

Answer1 = RemoveSpecialChars(doc.SelectSingleNode("/Main/Answer[@answerid='1']").Attributes["keypress"].Value);                 
Answer2 = RemoveSpecialChars(doc.SelectSingleNode("/Main/Answer[@answerid='2']").Attributes["keypress"].Value);

 public string RemoveSpecialChars(string input)
       {

           return Regex.Replace(input, @"[^0-9a-zA-Z\._]", string.Empty);
       }

What’s happening, is when the user presses and sends an answer one, and nothing for answer two I get an exception, because the method is trying to run on an empty string. What is the best way to pass answer1, if answer 2 is empty?

Answer 1

It sounds like your problem is not in the RemoveSpecialChars method, but rather in the return value of SelectSingleNode (which may be null) or the Attributes["keypress"] attribute (which may also be null).

Any of the above will result in a NullReferenceException. Here’s rewritten code to guard against the first, which is probably causing the issue:

var node1 = doc.SelectSingleNode("/Main/Answer[@answerid='1']");
var node2 = doc.SelectSingleNode("/Main/Answer[@answerid='2']");

Answer1 = node1 == null ? null : RemoveSpecialChars(node1.Attributes["keypress"].Value);
Answer2 = node2 == null ? null : RemoveSpecialChars(node2.Attributes["keypress"].Value);

Update:

To guard against a null keypress attribute, you would do

Answer1 = node1 == null || node1.Attributes["keypress"] == null
            ? null 
            : RemoveSpecialChars(node1.Attributes["keypress"].Value);

and the same for Answer2.