Leetcode-Construct Binary Tree from inorder and postorder travesal

LiBlog發表於2014-11-09

Given inorder and postorder traversal of a tree, construct the binary tree.

Solution:

 1 /**
 2  * Definition for binary tree
 3  * public class TreeNode {
 4  *     int val;
 5  *     TreeNode left;
 6  *     TreeNode right;
 7  *     TreeNode(int x) { val = x; }
 8  * }
 9  */
10 public class Solution {
11     public TreeNode buildTree(int[] inorder, int[] postorder) {
12          if (inorder.length==0)
13                 return null;
14                 
15             int len = inorder.length;
16             TreeNode root = buildTreeRecur(inorder,postorder,0,len-1,0,len-1);
17             return root;
18     }
19     
20     //Build tree for current list, i.e., inorder[inHead] to inorder[inEnd].
21     public TreeNode buildTreeRecur(int[] inorder, int[] postorder, int inHead, int inEnd, int postHead, int postEnd){
22         if (inHead==inEnd){
23                 TreeNode root = new TreeNode(inorder[inHead]);
24                 return root;
25             }
26             
27             int curRoot = postorder[postEnd];
28             int index = -1;
29             for (int i=inHead;i<=inEnd;i++)
30                 if (inorder[i]==curRoot){
31                     index = i;
32                     break;
33                 }
34             int leftNodeNum = index-inHead;
35             
36             
37             int leftInHead = inHead;
38             int leftInEnd = inHead+leftNodeNum-1;
39             int rightInHead = index+1;
40             int rightInEnd = inEnd;      
41             
42             
43             int leftPostHead = postHead;
44             int leftPostEnd = postHead+leftNodeNum-1;
45             int rightPostHead = leftPostEnd+1;
46             int rightPostEnd = postEnd-1;
47             
48             TreeNode root = new TreeNode(curRoot);
49             TreeNode leftChild = null;
50             if (leftInEnd>=inHead){
51                 leftChild = buildTreeRecur(inorder,postorder,leftInHead,leftInEnd,leftPostHead,leftPostEnd);
52                 root.left = leftChild;
53             }
54                 
55             TreeNode rightChild = null;
56             if (rightInHead<=inEnd){
57                 rightChild = buildTreeRecur(inorder,postorder,rightInHead,rightInEnd,rightPostHead,rightPostEnd);
58                 root.right = rightChild;
59             }
60             
61             return root;
62     }
63 }

We need to be very carefull about how to count the start and end of the left sub-tree and the right-sub tree. Especially detecting the case that some sub-tree is void.

A better way is to calculate the number of nodes in left and right tree first, then find out the range, like this:

 1 /**
 2  * Definition for binary tree
 3  * public class TreeNode {
 4  *     int val;
 5  *     TreeNode left;
 6  *     TreeNode right;
 7  *     TreeNode(int x) { val = x; }
 8  * }
 9  */
10 public class Solution {
11     public TreeNode buildTree(int[] inorder, int[] postorder) {
12          if (inorder.length==0)
13                 return null;
14                 
15             int len = inorder.length;
16             TreeNode root = buildTreeRecur(inorder,postorder,0,len-1,0,len-1);
17             return root;
18     }
19     
20     //Build tree for current list, i.e., inorder[inHead] to inorder[inEnd].
21     public TreeNode buildTreeRecur(int[] inorder, int[] postorder, int inHead, int inEnd, int postHead, int postEnd){
22         if (inHead==inEnd){
23                 TreeNode root = new TreeNode(inorder[inHead]);
24                 return root;
25             }
26             
27             int curRoot = postorder[postEnd];
28             TreeNode root = new TreeNode(curRoot);
29             TreeNode leftChild = null;
30             TreeNode rightChild = null;
31             
32             int index = -1;
33             for (int i=inHead;i<=inEnd;i++)
34                 if (inorder[i]==curRoot){
35                     index = i;
36                     break;
37                 }
38             int leftNodeNum = index-inHead;
39             int rightNodeNum = inEnd-index;
40             
41             if (leftNodeNum>0){
42                 int leftInHead = inHead;
43                 int leftInEnd = inHead+leftNodeNum-1;
44                 int leftPostHead = postHead;
45                 int leftPostEnd = postHead+leftNodeNum-1;
46                 leftChild = buildTreeRecur(inorder,postorder,leftInHead,leftInEnd,leftPostHead,leftPostEnd);
47                 root.left = leftChild;
48             }
49                 
50             
51             if (rightNodeNum>0){
52                 int rightInHead = index+1;
53                 int rightInEnd = inEnd;
54                 int rightPostHead = postEnd-rightNodeNum;
55                 int rightPostEnd = postEnd-1;
56                 rightChild = buildTreeRecur(inorder,postorder,rightInHead,rightInEnd,rightPostHead,rightPostEnd);
57                 root.right = rightChild;
58             }
59             
60             return root;
61     }
62 }

 

相關文章