We want to delete a node i.
- If i has zero kids, delete it.
- If i has one kid, delete i and move i's kid into i's place.
- If i has two kids, let j be the successor of i. Delete j
in place (it only has one kid, so that's easy). Now, move j into
i's place.
Why is this well defined? In particular, how do we know that j
only has one child?
Next: Successor Tree Walk
Up: BINARY SEARCH TREES
Previous: Where is the Successor?