Deleting a node in C# tree

Question

I was wondering if someone could help me.

I want to delete a node in a C# tree and the code below does it perfectly. but what I don't fully understand is all of the nodes that are declared in the process of traversing and deleting are just copies of the actual nodes and I can't see how they can affect the actual nodes and swap them with each other. I had no problem using this kind of method in C++ because I used pointers and two pointers could simultaneously point to one chunk of memory however we don't do it in C#.

class BinarySearchTree<T>
{
    class Node
    {
        public Node Left;
        public T Info;
        public Node Right;
    }
    public bool DeleteItem(T item)
    {
        Node t = Root, s = null;
        while (t != null && Compare(item, t.Info) != 0)
        {
            s = t;
            t = (Compare(item, t.Info) <= 0) ? t.Left : t.Right;
        }

        if (t == null)
            return false;

        if (t.Left != null && t.Right != null) //node has 2 children
        {
            Node suc = t.Right, parent_suc = null;
            while (suc.Left != null)
            {
                parent_suc = suc;
                suc = suc.Left;
            }
            if (parent_suc == null)
                t.Right = suc.Right;
            else
                parent_suc.Left = suc.Right;
            t.Info = suc.Info;
        }
        else //node has either one child or no child at all
        {
            Node child = t.Left;
            if (t.Right != null)
                child = t.Right;
            if (s == null)
                Root = child;
            else
            {
                if (s.Left == t)
                    s.Left = child;
                else
                    s.Right = child;
            }
        }
        return true;
    }
}

Answer 1

Your Node type is a class, which is a reference type. That means that when you assign or copy it to a new variable, it will create a new reference to the original data, instead of copying the data itself (the opposite would be value type). It is indeed very similar to C++ pointers, with some differences (without pointer arithmetics, but with automatic garbage collection).

See this MSDN article about C# types, and this one about C++ pointers and C# references.