pku 1521 赫夫曼编码 Entropy 解题报告

pku 1521 赫夫曼编码 Entropy 解题报告

题意:要快速明白此题,那么严蔚敏编写的<<数据结构>>有关赫夫曼编码的内容看明白才行;明白了赫夫曼编码此题就很容易AC了,题目本来就是要我们写一个赫夫曼树.

算法:赫夫曼编码算法.

AC代码:

#include <stdio.h>

#include <string.h>

#include <stdlib.h>

typedef struct

{

       int weight;    

       int parent, lchild, rchild; 

}HTNode, *HuffmanTree;      //动态分配数组存储赫夫曼树   

typedef char** HuffmanCode; //动态分配数组存储赫夫曼编码

int answer;

//赫夫曼编码的算法实现

void SetHuffmanTree(HuffmanTree &HT, int weight, int parent, int lchild, int rchild)

{

       HT->weight = weight;

    HT->parent = parent;

       HT->lchild = lchild;

       HT->rchild = rchild;

}

void Select(HuffmanTree HT, int n, int &s1, int &s2)

{

       //在HT[1...n]选择parent为0且weight最小的两个节点,分别赋给s1,s2

    //s1最小,s2次小

       int i, min = 0, mun = 0;

       HT[0].weight = 30000;

    for (i = 1; i <= n; i++)

       {

              if (HT[i].parent == 0)

              {

            if (HT[i].weight < HT[min].weight) 

                     {

                            mun = min;

                            min = i;

                     }

            else if (HT[i].weight < HT[mun].weight)

                     {

                            mun = i;

                     }

              }       

       }

       s1 = min;

       s2 = mun;

}

int ans, result[100];

void HuffmanCoding(HuffmanTree &HT, HuffmanCode &HC, int *w, int n)

{  

       int i, j, m, start, c, f, s1, s2;

       HuffmanTree p;

       //w存放n个字符的权值(均>0),构造赫夫曼树HT,并求出n个字符的赫夫曼编码

    if (n <= 1)

       {

              return;

       }

       //m:总共节点数

    m = 2 * n - 1;

       HT = (HuffmanTree) malloc((m + 1) * sizeof(HTNode)); //0号单元未用       

    for (p = HT + 1, i = 1; i <= n; i++, p++, w++)

       {

              //HT叶子的初值

              SetHuffmanTree(p, *w, 0, 0, 0);

       }

    for ( ; i <= m; i++, p++)

       {

              //HT非叶子节点的初值

              SetHuffmanTree(p, 0, 0, 0, 0);                    

    }

       for (i = n + 1; i <= m; i++)

       {

              //建赫夫曼树 

              //找出weight值最小的两个节点

              Select(HT, i - 1, s1, s2);

              HT[s1].parent = i; HT[s2].parent = i;

              HT[i].lchild = s1; HT[i].rchild = s2;

              HT[i].weight = HT[s1].weight + HT[s2].weight;

       }

    //从叶子到根逆向求每个字符的赫夫曼编码

       //分配n个字符编码的头指针向量

       HC = (HuffmanCode) malloc((n + 1) * sizeof(char *));      

    char *cd = (char *) malloc(n * sizeof(char));

       //编码结束符

       cd[n - 1] = '/0';                 

    for (i = 1; i <= n; i++)

       {

              //编码结束的位置

              start = n - 1;

              //从叶子到根逆向求编码

        for (c = i, f = HT[i].parent; f != 0; c = f, f = HT[f].parent)

              {   

                     //从叶子到根逆向的逆向过程

                     if (HT[f].lchild == c)

                     {

                            cd[--start] = '0';

                     }

            else

                     {

                            cd[--start] = '1';

                     }

              }

              //为第i个字符编码分配空间

              HC[i] = (char * ) malloc ((n - start) * sizeof(char));

              result[i] = 0;

        for (j = start; j <= n; j++)

              {

                     HC[i][j - start] = cd[j];

                     if (cd[j] == '1' || cd[j] == '0')

                     {

                            result[i]++;

                     }

              } 

              //从cd复制编码(串)到HC

              //输出编码

              ans += result[i] * HT[i].weight;

       }

       free(cd);

}

int total;

char s[100];

int get_id(char c)

{

       int i;

       for (i = 0; i < total; i++)

       {

              if (c == s[i])

              {

                     return i;

              }

       }

       s[i] = c;

       total++;

       return i;

}

int cmp(const void *a, const void *b)

{

       return *(int *)a - *(int *)b;

}

int main()

{

       HuffmanTree HT;

       HuffmanCode HC;

       int i, j, w[100], n, sum;

       char str[100];

       while (scanf("%s", str))

       {

              if (strcmp(str, "END") == 0)

              {

                     break;

              }

              sum = 0; total = 0;

              n = strlen(str);

              for (i = 0; i < n; i++)

              {

                     w[i] = 0;

              }

              for (i = 0; i < n; i++, sum += 8)

              {

                     j = get_id(str[i]);

                     w[j]++;

              }

              ans = 0;

              if (total == 1)

              {

                     ans = n;

                     printf("%d %d %.1lf/n", sum, ans, (float) sum / (float) ans);

              }

              else

              {

                     HuffmanCoding(HT, HC, w, total);

                     printf("%d %d %.1lf/n", sum, ans, (float) sum / (float) ans);

              }

              getchar();

       }

       return 0;

}