我正在编写一个名为RootedBinaryTree
的模板类,它有一个类似于链表的结构,其元素的类型为Node
,这是我在下面的头文件中定义的结构。二叉树中的每个节点都有parent Node *
,可以有leftChild Node *
和rightChild Node *
或没有子节点。 (例如:如果你要绘制一个带根的二叉树,它应该看起来像这个http://mathworld.wolfram.com/images/eps-gif/CompleteBinaryTree_1000.gif)。
有根二进制树有两个成员:root
和currentPosition
,它们是node类型。
当我重载operator=()
时,我必须做currentPosition = RHS.currentPosition;
之类的事情(这是我目前所写的)。我知道这是不正确的,我需要做的是在*this
树中找到与currentPosition Node *
树中的RHS
对应的节点。
我的问题是:遍历RHS树以查找其currentPosition Node *
并在调用树中找到相应的Node
的算法是什么?
我的想法是创建一个空字符串并使用某种深度优先搜索算法遍历RHS
,从根目录开始直到我找到currentPosition,并通过附加一个跟踪我到达那里的路径0到字符串的结尾如果我向下走到树的左边,或者如果我向下走到树的右边,然后使用该字符串来遍历*this
树,这应该会带我到相应的Node
。但是,我知道必须有更好的方法去做(部分来自直觉,部分是因为我的导师告诉我有更好的方法哈哈)。
以下是相关文件:
RootedBinaryTree.h
#ifndef ROOTEDBINARYTREE_H
#define ROOTEDBINARYTREE_H
template <class T>
class RootedBinaryTree
{
private:
template <class T>
struct Node
{
T nodeData;
Node<T> * parent;
Node<T> * leftChild;
Node<T> * rightChild;
Node<T>::Node()
{
parent = leftChild = rightChild = 0L;
}
};
Node<T> * root;
Node<T> * currentPosition;
void copySubtree(Node<T> * & target, Node<T> * const & original);
void deleteSubtree(Node<T> * n);
public:
RootedBinaryTree(const T & rootData);
RootedBinaryTree(const RootedBinaryTree<T> & original);
~RootedBinaryTree();
void toRoot();
bool moveLeft();
bool moveRight();
bool moveUp();
T getData() const {return currentPosition->nodeData;};
RootedBinaryTree<T> & operator=(const RootedBinaryTree<T> & RHS);
void combineTrees(const RootedBinaryTree<T> & leftTree, const RootedBinaryTree<T> & rightTree);
void setNodeData(const T & nodeData);
};
#endif
RootedBinaryTree.cpp
#ifndef ROOTEDBINARYTREE_CPP
#define ROOTEDBINARYTREE_CPP
#include "RootedBinaryTree.h"
template<class T>
void RootedBinaryTree<T>::copySubtree(Node<T> * & target, Node<T> * const & original)
{// Assumes that target's leftChild = rightChild = 0L. I.e. target is a leaf.
target = new Node<T>;
if(original->leftChild != 0L)
{
target->leftChild->parent = target;
copySubtree(target->leftChild, original->leftChild);
}
else
{
target->leftChild = 0L;
}
// ^^^ copy targets left (and right) children to originals
if(original->rightChild != 0L)
{
target->rightChild->parent = target;
copySubtree(target->rightChild, original->rightChild);
}
else
{
target->rightChild = 0L;
}
target->nodeData = original->nodeData;
}
template <class T>
void RootedBinaryTree<T>::deleteSubtree(Node<T> * n) // Done
{// Assumes that n is a valid node.
if(n->leftChild != 0L) deleteSubtree(n->leftChild); // Delete all nodes in left subtree
if(n->rightChild != 0L) deleteSubtree(n->rightChild); // Delete all nodes in right subtree
delete n;
}
template <class T>
RootedBinaryTree<T>::RootedBinaryTree(const T & rootData) // Done
{
root = new Node <T>; // Roots parent = leftChild = rightChild = 0L
root->nodeData = rootData;
currentPosition = root;
}
template <class T>
RootedBinaryTree<T>::RootedBinaryTree(const RootedBinaryTree<T> & original) // done
{
root = currentPosition = new Node<T>;
*this = original;
}
template <class T>
RootedBinaryTree<T>::~RootedBinaryTree() // done
{
deleteSubtree(root); // root will be valid because of our constructor and other methods
root = currentPosition = 0L;
}
template <class T>
void RootedBinaryTree<T>::toRoot() // Done
{
currentPosition = root;
}
template <class T>
bool RootedBinaryTree<T>::moveUp() // Done
{
if(currentPosition->parent == 0L) return false; // If we are at the root of the tree, we cannot move up it.
currentPosition = currentPosition->parent;
return true;
}
template <class T>
bool RootedBinaryTree<T>::moveLeft() // Done
{
if(currentPosition->leftChild == 0L) return false;
currentPosition = currentPosition->leftChild;
return true;
}
template <class T>
bool RootedBinaryTree<T>::moveRight() // Done
{
if(currentPosition->rightChild == 0L) return false;
currentPosition = currentPosition->rightChild;
return true;
}
template <class T>
RootedBinaryTree<T> & RootedBinaryTree<T>::operator=(const RootedBinaryTree<T> & RHS)
{
if(&RHS == this)
{
return *this;
}
this->~RootedBinaryTree();
copySubtree(root, RHS.root);
currentPosition = RHS.currentPosition; // This is wrong.
return *this;
}
template <class T>
void RootedBinaryTree<T>::combineTrees(const RootedBinaryTree<T> & leftTree, const RootedBinaryTree<T> & rightTree)
{ // Copies leftTree into root's left tree and rightTree into root's right tree.
if(this == &leftTree || this == &rightTree)
{
throw "A rooted binary tree cannot be combined with itself.";
}
if(root->leftChild != 0L) deleteSubtree(root->leftChild);
if(root->rightChild != 0L) deleteSubtree(root->rightChild);
copySubtree(root->leftChild, leftTree.root);
copySubtree(root->rightChild, rightTree.root);
}
template <class T>
void RootedBinaryTree<T>::setNodeData(const T & nodeData)
{
currentPosition->nodeData = nodeData;
}
#endif
提前感谢任何回答此问题的人!非常感谢您的帮助。
答案 0 :(得分:0)
我是通过实现私有方法getPath()
来执行此操作,该方法返回从root
到currentPosition
的移动列表(左侧或右侧):
template <class T>
list<bool> RootedBinaryTree<T>::getPath() const
{
list<bool> path;
Node<T> *p1=currentPosition;
Node<T> *p2 = p1->parent;
while(p2 != 0L)
{
if(p2->leftChild==p1)
path.push_front(true);
else
path.push_front(false); // yes, I know this can be done more concisely; I'm going for clarity
p1=p2;
p2=p1->parent;
}
return(path);
}
在另一棵树上调用此方法,获取路径,然后在此树中遵循该路径:
template <class T>
void RootedBinaryTree<T>::followPath(list<bool> path)
{
currentPosition = root;
for(list<bool>::const_iterator itr=path.begin(); itr !=path.end() ; ++itr)
if(*itr)
moveLeft();
else
moveRight();
}