（一）递归实现：

完整代码： 

#include
#include
#include
using namespace std;#define MaxSize 100
typedef int ElemType;
typedef struct BiNode {ElemType data;BiNode* lchild;BiNode* rchild;}BiNode, * BiTree;//构建二叉树
BiNode* Create(BiNode* bt) {static int i = 0;char ch;//string str = "AB#D##C##";//string str = "124##56##7##3##";//string str = "ABD#G##E##CF###";string str = "ABD#GH##I##E##CF###";ch = str[i++];if (ch == '#')bt = NULL;//建立一棵空树 else {bt = (BiTree)malloc(sizeof(BiNode)); bt->data = ch;//生成一个结点，数据域为chbt->lchild = Create(bt->lchild);//递归建立左子树bt->rchild = Create(bt->rchild);//递归建立右子树}return bt;
}typedef struct {ElemType data[MaxSize];//存放栈中元素int top;//栈顶指针
}SqStack;//（1）初始化
void InitStack(SqStack& S) {S.top = -1;//初始化栈顶指针
}
//（2）判栈空
bool IsEmpty(SqStack& S) {if (S.top == -1) {//栈空return true;}else {//不空return false;}
}
//（3）进栈
bool Push(SqStack& S, ElemType data) {if (S.top == MaxSize - 1) {//栈满，报错return false;}S.data[++S.top] = data;//指针先加1，再加入栈return true;
}
//（4）出栈
bool Pop(SqStack& S, ElemType &data) {if (S.top == -1) {//栈空，报错return false;}data = S.data[S.top--];//先出栈，指针再减1return true;
}
//（5）读栈顶元素
bool GetTop(SqStack& S, ElemType data) {if (S.top == -1) {//栈空，报错return false;}data = S.data[S.top];//先出栈，指针再减1return true;
}bool Ancestors(BiNode* root,ElemType x, SqStack &S) {if (!root) return false;if (root->data == x) return true;if (Ancestors(root->lchild, x,S) || Ancestors(root->rchild, x, S)) {//printf("%c ", root->data);//G D B APush(S, root->data);return true;}return false;
}int main() {//创建一棵二叉树BiTree T = (BiTree)malloc(sizeof(BiNode));//创建一颗二叉树T = Create(T);SqStack S;InitStack(S);Ancestors(T, 'I',S);ElemType ch;while (!IsEmpty(S)) {Pop(S, ch);printf("%c ", ch); // A B D G}
}

 （二）非递归实现：

因为查找的过程就是后序遍历的过程，因此使用的栈的深度不超过树的深度

完整的代码：

#include
#include
#include
using namespace std;#define MaxSize 100
typedef char ElemType;
typedef struct BiNode {ElemType data;BiNode* lchild;BiNode* rchild;}BiNode, * BiTree;//构建二叉树
BiNode* Create(BiNode* bt) {static int i = 0;char ch;//string str = "AB#D##C##";//string str = "124##56##7##3##";//string str = "ABD#G##E##CF###";string str = "ABD#GH##I##E##CF###";ch = str[i++];if (ch == '#')bt = NULL;//建立一棵空树 else {bt = (BiTree)malloc(sizeof(BiNode)); bt->data = ch;//生成一个结点，数据域为chbt->lchild = Create(bt->lchild);//递归建立左子树bt->rchild = Create(bt->rchild);//递归建立右子树}return bt;
}//算法思想：采用非递归后序遍历，最后访问根结点，访问到值为x的结点时，栈
//中所有元素均为该结点的祖先，依次出栈打印即可。算法实现如下：
typedef struct {BiTree t;int tag;
}stack;//tag = 0 表示左子女被访问，tag = 1表示右子女被访问
bool Search(BiTree bt,ElemType x) {//在二叉树bt中，查找值为x的结点，并打印其所有祖先stack s[MaxSize];int top = 0;while (bt != NULL || top > 0) {while (bt != NULL && bt->data != x) {//结点入栈s[++top].t = bt;s[top].tag = 0;bt = bt->lchild;//沿左分支向下}if (bt != NULL && bt->data == x) {printf("所查结点的所有祖先结点的值为：\n");//找到xfor (int i = 1; i <= top; i++) {printf("%c ",s[i].t->data);//输出祖先值后结束}exit(1);}while (top != 0 && s[top].tag == 1)top--;//退栈（空遍历）if (top != 0) {s[top].tag = 1;bt = s[top].t->rchild;//沿右分支向下遍历}}
}int main() {//创建一棵二叉树BiTree bt = (BiTree)malloc(sizeof(BiNode));//创建一颗二叉树bt = Create(bt);Search(bt, 'I');
}