给一千万个数据量的文件排序

标签：

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <cstdlib>
#include <set>
#include <bitset>
#include <ctime>
#include <cassert>
#include <climits>
#include <algorithm>
#define MAXN 5000000
#define RST(N)memset(N, 0, sizeof(N))
using namespace std;

typedef unsigned long long ULL;

int main()
{
    bitset <MAXN> bit_map;
    bit_map.reset();

    //time record
    clock_t begin = clock();

    //read data from data.in file
    FILE *wait_sort_file = fopen("data.in", "r");
    assert(wait_sort_file);
    ULL Num;
    while(~fscanf(wait_sort_file, "%d ", &Num))
    {
        if(Num < MAXN) bit_map.set(Num, 1);
    }
    //write data to data.out file
    FILE *sorted_file = fopen("data.out", "w");
    assert(sorted_file);
    for(int i=0; i<MAXN; i++)
    {
        if(bit_map[i] == 1) fprintf(sorted_file, "%d ", i);
    }
    //seek the offset of file
    int offset = fseek(wait_sort_file, 0, SEEK_SET);
    if(offset) perror("fseek fail");
    else
    {
        //read data again
        bit_map.reset();
        while(~fscanf(wait_sort_file, "%d ", &Num))
        {
            if(Num >= MAXN) bit_map.set(Num-MAXN, 1);
        }
        //write again
        for(int i=0; i<MAXN; i++)
        {
            if(bit_map[i] == 1)
                fprintf(sorted_file, "%d ", i+MAXN);
        }
    }

    clock_t end = clock();
    cout << "cost time: " << (double)(end-begin)/CLOCKS_PER_SEC << endl;

    fclose(wait_sort_file);
    fclose(sorted_file);
    return 0;
}

 

 

 

给一千万个数据量的文件排序

标签：

原文地址：http://blog.csdn.net/u011225629/article/details/48109639