链接
牛客OJ:从上往下打印二叉树
九度OJ:http://ac.jobdu.com/problem.php?pid=1523
GitHub代码: 023-从上往下打印二叉树
CSDN题解:剑指Offer–023-从上往下打印二叉树
牛客OJ | 九度OJ | CSDN题解 | GitHub代码 |
---|---|---|---|
从上往下打印二叉树 | 1523-从上往下打印二叉树 | 剑指Offer–023-从上往下打印二叉树 | 023-从上往下打印二叉树 |
题意
题目描述
从上往下打印出二叉树的每个节点,同层节点从左至右打印。
分析
其实就是层次遍历,这个我们在我之前的一篇博客里面将的很清楚了
其中调试的时候使用了
剑指Offer–006-重构二叉树来辅助我们通过前序和中序遍历的序列来生成二叉树
直接贴代码,只需要直接将我们层次遍历输出过程,变成将输入压入vector中即可
#include <iostream>
#include <vector>
#include <deque>
#include <queue>
using namespace std;
// 调试开关
#define __tmain main
#ifdef __tmain
#define debug cout
#else
#define debug 0 && cout
#endif // __tmain
#define undebug 0 && cout
#ifdef __tmain
struct TreeNode
{
int val;
struct TreeNode *left;
struct TreeNode *right;
TreeNode(int x)
:val(x), left(NULL), right(NULL)
{
}
};
#endif // __tmain
class Solution
{
vector<int> res;
public:
vector<int> PrintFromTopToBottom(TreeNode *root)
{
/// 方法1 -=> 循环调用递归打印每一层
LevelOrder(root);
/// 方法2 -=> 使用两个双端队列
LevelOrderDev(root);
/// 方法3 -=> 使用双指针标识当前指针和结束
LevelOrderUsePoint(root);
/// 方法4 -=> 使用aprent和childzsizesize标识前一层和当前层的节点个数
LevelOrderUseSize(root);
/// 方法4 -=> 在队列中插入结束标识来标识当前层结束
LevelOrderUseEnd(root);
return this->res;
}
/// 打印某一层的节点,递归实现
int PrintLevel(TreeNode *root, int level)
{
if(root == NULL || level < 0)
{
return 0;
}
else if(level == 0)
{
debug <<root->val;
res.push_back(root->val); /// add for result in vector
return 1;
}
else
{
return PrintLevel(root->left, level - 1) + PrintLevel(root->right, level - 1);
}
}
/// 循环每层输出其节点,调用递归函数
void LevelOrder(TreeNode *root)
{
res.clear( ); /// add for result in vector
if(root == NULL)
{
return;
}
for(int level = 0; ; level++)
{
if(PrintLevel(root, level) == 0)
{
break;
}
debug <<endl;
}
}
//
/// 使用两个双端队列
//
void LevelOrderDev(TreeNode *root)
{
res.clear( ); /// add for result in vector
/// deque双端队列,
/// 支持迭代器,有push_back()方法,
/// 跟vector差不多,比vector多了个pop_front,push_front方法
deque<TreeNode *> qFirst, qSecond;
qFirst.push_back(root);
while(qFirst.empty( ) != true)
{
while (qFirst.empty( ) != true)
{
TreeNode *temp = qFirst.front( );
qFirst.pop_front( );
debug << temp->val;
res.push_back(temp->val); /// add for result in vector
if (temp->left != NULL)
{
qSecond.push_back(temp->left);
}
if (temp->right != NULL)
{
qSecond.push_back(temp->right);
}
}
debug << endl;
qFirst.swap(qSecond);
}
}
//
/// 使用双指针标识当前指针和结束
//
void LevelOrderUsePoint(TreeNode *root)
{
res.clear( ); /// add for result in vector
vector<TreeNode*> vec;
vec.push_back(root);
int cur = 0;
int end = 1;
while (cur < vec.size())
{
end = vec.size(); /// 新的一行访问开始,重新定位end于当前行最后一个节点的下一个位置
while (cur < end)
{
debug << vec[cur]->val; /// 访问节点
res.push_back(vec[cur]->val); /// add for result in vector
if (vec[cur]->left != NULL) /// 压入左节点
{
vec.push_back(vec[cur]->left);
}
if (vec[cur]->right != NULL) /// 压入右节点
{
vec.push_back(vec[cur]->right);
}
cur++;
}
debug << endl;
}
}
//
/// 使用aprent和childzsizesize标识前一层和当前层的节点个数
//
void LevelOrderUseSize(TreeNode *root)
{
res.clear( ); /// add for result in vector
int parentSize = 1, childSize = 0;
TreeNode *temp = NULL;
queue<TreeNode *> q;
q.push(root);
while(q.empty( ) != true)
{
temp = q.front( );
debug <<temp->val;
res.push_back(temp->val); /// add for result in vector
q.pop( );
if (temp->left != NULL)
{
q.push(temp->left);
childSize++;
}
if (temp->right != NULL)
{
q.push(temp->right);
childSize++;
}
parentSize--;
if (parentSize == 0)
{
parentSize = childSize;
childSize = 0;
debug << endl;
}
}
}
//
/// 在队列中插入结束标识来标识当前层结束
//
void LevelOrderUseEnd(TreeNode *root)
{
res.clear( ); /// add for result in vector
queue<TreeNode *> q;
q.push(root);
q.push(NULL);
while(q.empty( ) != true)
{
TreeNode* node = q.front();
q.pop();
if (node)
{
debug << node->val;
res.push_back(node->val); /// add for result in vector
if (node->left != NULL)
{
q.push(node->left);
}
if (node->right != NULL)
{
q.push(node->right);
}
}
else if (q.empty( ) != true)
{
q.push(NULL);
debug << endl;
}
}
}
struct TreeNode* reConstructBinaryTree(vector<int> pre,vector<int> in)
{
// 前序遍历的长度跟中序遍历的长度应该相同
if(pre.size( ) != in.size( ))
{
undebug <<"the length of PRE and IN should be smae" <<endl;
return NULL;
}
// 长度不能为0
int size = pre.size( );
if(size == 0)
{
undebug <<"it's a NULL tree(length = 0)" <<endl;
return NULL;
}
int length = pre.size( );
undebug <<"the length of your tree = " <<length <<endl;
int value = pre[0]; // 前序遍历的第一个结点是根节点
TreeNode *root = new TreeNode(value);
undebug <<"the root is" <<root->val <<endl;
// 在中序遍历中查找到根的位置
int rootIndex = 0;
for(rootIndex = 0; rootIndex < length; rootIndex++)
{
if(in[rootIndex] == value)
{
undebug <<"find the root at " <<rootIndex <<" in IN" <<endl;
break;
}
}
if(rootIndex >= length)
{
undebug <<"can't find root (value = " <<value <<") in IN" <<endl;
return NULL;
}
/// 区分左子树和右子树
/// 中序遍历中, 根左边的就是左子数, 右边的就是右子树
/// 前序遍历中, 根后面是先遍历左子树, 然后是右子树
/// 首先确定左右子树的长度, 从中序遍历in中确定
int leftLength = rootIndex;
int rightLength = length - 1 - rootIndex;
undebug <<"left length = " <<leftLength <<", rightLength = " <<rightLength <<endl;
vector<int> preLeft(leftLength), inLeft(leftLength);
vector<int> preRight(rightLength), inRight(rightLength);
for(int i = 0; i < length; i++)
{
if(i < rootIndex)
{
// 前序遍历的第一个是根节点, 根后面的(leftLegnth = rootIndex) - 1个节点是左子树, 因此是i+1
preLeft[i] = pre[i + 1];
// 中序遍历前(leftLength = rootIndex) - 1个节点是左子树, 第rootIndex个节点是根
inLeft[i] = in[i];
undebug <<preLeft[i] <<inLeft[i] <<" ";
}
else if(i > rootIndex)
{
// 前序遍历的第一个是根节点, 根后面的(leftLegnth = rootIndex) - 1个节点是左子树, 后面是右子树
preRight[i - rootIndex - 1] = pre[i];
// 中序遍历前(leftLength = rootIndex) - 1个节点是左子树, 第rootIndex个节点是根, 然后是右子树
inRight[i - rootIndex - 1] = in[i];
undebug <<preRight[i - rootIndex - 1] <<inRight[i - rootIndex - 1] <<" ";
}
}
undebug <<endl <<"the left tree" <<endl;
for(int i = 0; i < leftLength; i++)
{
undebug <<preLeft[i] <<inLeft[i] <<" ";
}
undebug <<endl;
undebug <<"the right tree" <<endl;
for(int i = 0; i < rightLength; i++)
{
undebug <<preRight[i] <<inRight[i] <<" ";
}
undebug <<endl;
root->left = reConstructBinaryTree(preLeft, inLeft);
root->right = reConstructBinaryTree(preRight, inRight);
return root;
}
};
int __tmain( )
{
int pre[] = { 1, 2, 4, 7, 3, 5, 6, 8 };
int in[] = { 4, 7, 2, 1, 5, 3, 8, 6 };
vector<int> preOrder(pre, pre + 8);
vector<int> inOrder( in, in + 8);
Solution solu;
TreeNode *root = solu.reConstructBinaryTree(preOrder, inOrder);
solu.PrintFromTopToBottom(root);
return 0;
}