What is more expensive, storing multiple strings, or one large concatanated string?

Question

Hello.
Say I have a class, that stores information on a file. You have the path, the name of the file, and the extension. Would it be better to store a string for each of those, like this:

public class File {
    public string path;
    public string name;
    public string extension;

    public string FullName { get { return path + '\\' + name + extension; } }
}

Or, just storing the FullName and having a property for each of the things, like this:

public class File {
    public string fullName;

    public string Path { get { return fullName.Remove(fullName.LastIndexOf('\\')); } }
    //etc...
}

Thank you!

Answer 1

String are immutable in C#. Every concatenation or string function creates a new string.

Strings (C# Programming Guide)
$ - string interpolation (C# reference)
@ (C# Reference)

If the application use case requires path, name, and extension in then it makes more sense to leave these items as separate properties. Use a backing field for the FullName property so the concatenation is done once.

Properties (C# Programming Guide)

    class Program  
    {  
        static void Main(string[] args)  
        {  
            File file = new File()  
            {  
                Path = @"C:\temp",  
                Name = "myfile",  
                Extension = "txt"  
            };  
  
            Console.WriteLine(file.FullName);  
        }  
  
    }  
  
    public class File  
    {  
        public string Path { get; init; }  
        public string Name { get; init; }  
        public string Extension { get; init; }  
  
        private string _fullName;  
        public string FullName   
        {   
            get   
            {   
                if(String.IsNullOrEmpty(_fullName))  
                {  
                    _fullName = @$"{Path}\{Name}.{Extension}";  
                }  
                return _fullName;  
            }   
        }  
    }  

Answer 2

It depends on the use cases. Parsing is more expensive than concatenation. But which methods does the code call more, the parts or the full?

Also is this code called enough to even worry about optimization?