計算哈夫曼編碼長度
阿新 • • 發佈：2018-12-26
#include <iostream>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
using namespace std;
#define BUFF_SIZE 4096
#define HASH_SIZE 256
char buff[BUFF_SIZE]; //緩衝區
int
int heap[(HASH_SIZE<<1)+2];
int pos[HASH_SIZE+1][3];//內部節點 , 0和1記錄子結點位置,3節錄當前的深度
int tlen[HASH_SIZE+1]; //記錄每個葉子結點的深度
int fd; //檔案描述符
int sym_num ; //檔案中出現的符號數量
int SUM(0); //檔案中字元總數
//初始化程式
void init(const char * pathname)
{
memset(buff , 0
memset(hash , 0 , sizeof(hash));
memset(heap , 0 , sizeof(heap));
memset(pos , 0 , sizeof(pos));
memset(tlen , 0 , sizeof(tlen));
//開啟檔案
fd = open(pathname , O_RDONLY);
if(fd < 0){
printf("init: %s dont exit!\n" , pathname);
exit(1
}
}
//統計檔案中每個符號出現的次數
void count_symbol()
{
lseek(fd , 0 , SEEK_SET);
while(read(fd , buff , BUFF_SIZE)){
SUM += strlen(buff);
for(int i=strlen(buff) - 1;i>=0;i--)
hash[(unsigned int)(buff[i] & 0xFF)]++;
}
//記錄出現的符號數量;
for(int i = HASH_SIZE - 1; i >= 0; i--)
if(hash[i])sym_num++;
}
//建立一個最小堆
void build_min_heap()
{
for(int i=sym_num;i>0;i--){
int p = i >> 1 , j = i;
while(p >= 1){
if(heap[heap[p]] > heap[heap[j]])
std::swap(heap[j] , heap[p]);
j = p; p >>= 1;
}
}
}
//每次取出最小數之後重新調整堆,
//h 指推中元素的個數
void heap_adjust(int h)
{
int t = 1 , p , q , l;
while(t<h){
p = t<<1; q = p + 1; l = t;
if(p <= h && heap[heap[p]] < heap[heap[t]])l = p;
if(q <= h && heap[heap[q]] < heap[heap[l]])l = q;
if(l == t)break;
std::swap(heap[l] , heap[t]);
t = l;
}
}
//計算每個字元編碼的長度
void huff_length()
{
int i , j , p , h , m1 , m2;
for(i=1 , p=0;i<=sym_num;i++){
while(!hash[p]) p++;
heap[sym_num + i] = hash[p];
heap[i] = sym_num + i;
p++;
}
h = sym_num;
//對1到n建立最小堆
build_min_heap();
while(h>1){
//取出最小數
m1 = heap[heap[1]];
pos[h][0] = heap[1];
heap[1] = heap[h];
h--;
heap_adjust(h);
//取出次小數
m2 = heap[heap[1]];
pos[h+1][1] = heap[1];
//最後數和次小數之和放在堆的最後一個位置
heap[h+1] = m1 + m2;
//重新指向最新合併的結點
heap[1] = h+1;
heap_adjust(h);
}
//統計編碼長度 , 線性時間統計
int ts = sym_num << 1;
for(int i=2;i<=sym_num;i++){
if(pos[i][0] <= sym_num) pos[pos[i][0]][2] = pos[i][2] + 1;
else tlen[pos[i][0] - sym_num] = pos[i][2] + 1;
if(pos[i][1] <= sym_num) pos[pos[i][1]][2] = pos[i][2] + 1;
else tlen[pos[i][1] - sym_num] = pos[i][2] + 1;
}
}
int main()
{
init("data.dat");
count_symbol();
huff_length();
unsigned int sum = 0;
for(int i=1;i<=sym_num;i++)
sum += tlen[i] * heap[sym_num + i];
cout<<SUM <<"\t\t"<<sum<<"\t\t"<<sum*1.0/SUM<<endl;
return 0;
}
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
using namespace std;
#define BUFF_SIZE 4096
#define HASH_SIZE 256
char buff[BUFF_SIZE]; //緩衝區
int
int heap[(HASH_SIZE<<1)+2];
int pos[HASH_SIZE+1][3];//內部節點 , 0和1記錄子結點位置,3節錄當前的深度
int tlen[HASH_SIZE+1]; //記錄每個葉子結點的深度
int fd; //檔案描述符
int sym_num ; //檔案中出現的符號數量
int SUM(0); //檔案中字元總數
//初始化程式
void init(const char * pathname)
{
memset(buff , 0
memset(hash , 0 , sizeof(hash));
memset(heap , 0 , sizeof(heap));
memset(pos , 0 , sizeof(pos));
memset(tlen , 0 , sizeof(tlen));
//開啟檔案
fd = open(pathname , O_RDONLY);
if(fd < 0){
printf("init: %s dont exit!\n" , pathname);
exit(1
}
}
//統計檔案中每個符號出現的次數
void count_symbol()
{
lseek(fd , 0 , SEEK_SET);
while(read(fd , buff , BUFF_SIZE)){
SUM += strlen(buff);
for(int i=strlen(buff) - 1;i>=0;i--)
hash[(unsigned int)(buff[i] & 0xFF)]++;
}
//記錄出現的符號數量;
for(int i = HASH_SIZE - 1; i >= 0; i--)
if(hash[i])sym_num++;
}
//建立一個最小堆
void build_min_heap()
{
for(int i=sym_num;i>0;i--){
int p = i >> 1 , j = i;
while(p >= 1){
if(heap[heap[p]] > heap[heap[j]])
std::swap(heap[j] , heap[p]);
j = p; p >>= 1;
}
}
}
//每次取出最小數之後重新調整堆,
//h 指推中元素的個數
void heap_adjust(int h)
{
int t = 1 , p , q , l;
while(t<h){
p = t<<1; q = p + 1; l = t;
if(p <= h && heap[heap[p]] < heap[heap[t]])l = p;
if(q <= h && heap[heap[q]] < heap[heap[l]])l = q;
if(l == t)break;
std::swap(heap[l] , heap[t]);
t = l;
}
}
//計算每個字元編碼的長度
void huff_length()
{
int i , j , p , h , m1 , m2;
for(i=1 , p=0;i<=sym_num;i++){
while(!hash[p]) p++;
heap[sym_num + i] = hash[p];
heap[i] = sym_num + i;
p++;
}
h = sym_num;
//對1到n建立最小堆
build_min_heap();
while(h>1){
//取出最小數
m1 = heap[heap[1]];
pos[h][0] = heap[1];
heap[1] = heap[h];
h--;
heap_adjust(h);
//取出次小數
m2 = heap[heap[1]];
pos[h+1][1] = heap[1];
//最後數和次小數之和放在堆的最後一個位置
heap[h+1] = m1 + m2;
//重新指向最新合併的結點
heap[1] = h+1;
heap_adjust(h);
}
//統計編碼長度 , 線性時間統計
int ts = sym_num << 1;
for(int i=2;i<=sym_num;i++){
if(pos[i][0] <= sym_num) pos[pos[i][0]][2] = pos[i][2] + 1;
else tlen[pos[i][0] - sym_num] = pos[i][2] + 1;
if(pos[i][1] <= sym_num) pos[pos[i][1]][2] = pos[i][2] + 1;
else tlen[pos[i][1] - sym_num] = pos[i][2] + 1;
}
}
int main()
{
init("data.dat");
count_symbol();
huff_length();
unsigned int sum = 0;
for(int i=1;i<=sym_num;i++)
sum += tlen[i] * heap[sym_num + i];
cout<<SUM <<"\t\t"<<sum<<"\t\t"<<sum*1.0/SUM<<endl;
return 0;
}