Implement a binary tree (#5431)

* Add code for the article 'Java Primitives versus Objects'

* Use JMH for benchmarking the primitive and wrapper classes

* Uncomment the benchmarks and remove unused class

* Add a binary search tree implementation

* Add an example of how to use the binary tree class

* Adjust the print statements

core-kotlin/src/main/kotlin/com/baeldung/datastructures/Main.kt

@@ -0,0 +1,19 @@
+package com.baeldung.datastructures
+
+/**
+ * Example of how to use the {@link Node} class.
+ *
+ */
+fun main(args: Array<String>) {
+    val tree = Node(4)
+    val keys = arrayOf(8, 15, 21, 3, 7, 2, 5, 10, 2, 3, 4, 6, 11)
+    for (key in keys) {
+        tree.insert(key)
+    }
+    val node = tree.find(4)!!
+    println("Node with value ${node.key} [left = ${node.left?.key}, right = ${node.right?.key}]")
+    println("Delete node with key = 3")
+    node.delete(3)
+    print("Tree content after the node elimination: ")
+    println(tree.visit().joinToString { it.toString() })
+}

core-kotlin/src/main/kotlin/com/baeldung/datastructures/Node.kt

@@ -0,0 +1,167 @@
+package com.baeldung.datastructures
+
+/**
+ * An ADT for a binary search tree.
+ * Note that this data type is neither immutable nor thread safe.
+ */
+class Node(
+        var key: Int,
+        var left: Node? = null,
+        var right: Node? = null) {
+
+    /**
+     * Return a node with given value. If no such node exists, return null.
+     * @param value
+     */
+    fun find(value: Int): Node? = when {
+        this.key > value -> left?.find(value)
+        this.key < value -> right?.find(value)
+        else -> this
+
+    }
+
+    /**
+     * Insert a given value into the tree.
+     * After insertion, the tree should contain a node with the given value.
+     * If the tree already contains the given value, nothing is performed.
+     * @param value
+     */
+    fun insert(value: Int) {
+        if (value > this.key) {
+            if (this.right == null) {
+                this.right = Node(value)
+            } else {
+                this.right?.insert(value)
+            }
+        } else if (value < this.key) {
+            if (this.left == null) {
+                this.left = Node(value)
+            } else {
+                this.left?.insert(value)
+            }
+        }
+    }
+
+    /**
+     * Delete the value from the given tree. If the tree does not contain the value, the tree remains unchanged.
+     * @param value
+     */
+    fun delete(value: Int) {
+        when {
+            value > key -> scan(value, this.right, this)
+            value < key -> scan(value, this.left, this)
+            else -> removeNode(this, null)
+        }
+    }
+
+    /**
+     * Scan the tree in the search of the given value.
+     * @param value
+     * @param node sub-tree that potentially might contain the sought value
+     * @param parent node's parent
+     */
+    private fun scan(value: Int, node: Node?, parent: Node?) {
+        if (node == null) {
+            System.out.println("value " + value
+                    + " seems not present in the tree.")
+            return
+        }
+        when {
+            value > node.key -> scan(value, node.right, node)
+            value < node.key -> scan(value, node.left, node)
+            value == node.key -> removeNode(node, parent)
+        }
+
+    }
+
+    /**
+     * Remove the node.
+     *
+     * Removal process depends on how many children the node has.
+     *
+     * @param node node that is to be removed
+     * @param parent parent of the node to be removed
+     */
+    private fun removeNode(node: Node, parent: Node?) {
+        node.left?.let { leftChild ->
+            run {
+                node.right?.let {
+                    removeTwoChildNode(node)
+                } ?: removeSingleChildNode(node, leftChild)
+            }
+        } ?: run {
+            node.right?.let { rightChild -> removeSingleChildNode(node, rightChild) } ?: removeNoChildNode(node, parent)
+        }
+
+
+    }
+
+    /**
+     * Remove the node without children.
+     * @param node
+     * @param parent
+     */
+    private fun removeNoChildNode(node: Node, parent: Node?) {
+        parent?.let { p ->
+            if (node == p.left) {
+                p.left = null
+            } else if (node == p.right) {
+                p.right = null
+            }
+        } ?: throw IllegalStateException(
+                "Can not remove the root node without child nodes")
+
+    }
+
+    /**
+     * Remove a node that has two children.
+     *
+     * The process of elimination is to find the biggest key in the left sub-tree and replace the key of the
+     * node that is to be deleted with that key.
+     */
+    private fun removeTwoChildNode(node: Node) {
+        val leftChild = node.left!!
+        leftChild.right?.let {
+            val maxParent = findParentOfMaxChild(leftChild)
+            maxParent.right?.let {
+                node.key = it.key
+                maxParent.right = null
+            } ?: throw IllegalStateException("Node with max child must have the right child!")
+
+        } ?: run {
+            node.key = leftChild.key
+            node.left = leftChild.left
+        }
+
+    }
+
+    /**
+     * Return a node whose right child contains the biggest value in the given sub-tree.
+     * Assume that the node n has a non-null right child.
+     *
+     * @param n
+     */
+    private fun findParentOfMaxChild(n: Node): Node {
+        return n.right?.let { r -> r.right?.let { findParentOfMaxChild(r) } ?: n }
+                ?: throw IllegalArgumentException("Right child must be non-null")
+
+    }
+
+    /**
+     * Remove a parent that has only one child.
+     * Removal is effectively is just coping the data from the child parent to the parent parent.
+     * @param parent Node to be deleted. Assume that it has just one child
+     * @param child Assume it is a child of the parent
+     */
+    private fun removeSingleChildNode(parent: Node, child: Node) {
+        parent.key = child.key
+        parent.left = child.left
+        parent.right = child.right
+    }
+
+    fun visit(): Array<Int> {
+        val a = left?.visit() ?: emptyArray()
+        val b = right?.visit() ?: emptyArray()
+        return a + arrayOf(key) + b
+    }
+}