Improving scalability of the modified preorder tree traversal algorithm

Question

I\'ve been thinking about the modified preorder tree traversal algorithm for storing trees within a flat table (such as SQL).

One property I dislike about the standard a

Answer 1

You can split your table into two: the first is (node ID, node value), the second (node ID, child ID), which stores all the edges of the tree. Insertion and deletion then become O(tree depth) (you have to navigate to the element and fix what is below it).

The solution you propose looks like a B-tree. If you can estimate the total number of nodes in your tree, then you can choose the depth of the tree beforehand.