I guess you could do it like this, though (i'm sure more optimization could be done)
using System;
using System.Collections.Generic;
public class Program
{
public class IndexedTree {
private readonly IDictionary<string, IndexedTree> _me;
private object _value;
private readonly string _splitKey = ".";
public IndexedTree this[string key] {
get {
return _me[key];
}
}
public object Value { get; set; }
public void Add(string dottedItem) {
if ( string.IsNullOrWhiteSpace( dottedItem ) ) {
throw new ArgumentException("dottedItem cannot be empty");
}
int index;
if ( (index = dottedItem.IndexOf( _splitKey ) ) < 0 ) {
throw new ArgumentException("dottedItem didn't contain " + _splitKey);
}
string key = dottedItem.Substring(0, index), rest = dottedItem.Substring(index + 1);
IndexedTree child;
if (_me.ContainsKey(key)) {
child = _me[key];
} else {
child = new IndexedTree( _splitKey );
_me.Add(key, child);
}
if (rest.IndexOf(_splitKey) >= 0) {
child.Add(rest);
} else {
// maybe it can be checked if there is already a value set here or not
// in case there is a warning or error might be more appropriate
child.Value = rest;
}
}
public IndexedTree(string splitKey) {
_splitKey = splitKey;
_me = new Dictionary<string, IndexedTree>();
}
}
public static void Main()
{
IndexedTree tree = new IndexedTree(".");
tree.Add("Level1.Level2.Level3.Item");
tree.Add("Level1.Level2.Value");
Console.WriteLine(tree["Level1"]["Level2"].Value);
Console.WriteLine(tree["Level1"]["Level2"]["Level3"].Value);
}
}
You could see the result here:
https://dotnetfiddle.net/EGagoz