单线程与多线程读取文件

来源：互联网发布：淘宝卖办公用品编辑：程序博客网时间：2024/06/06 09:28

//实验多个线程读取多个文件和单个线程读取多个文件

//当文件数量是2-3个时，单线程比多线程更快，可能是没有创建线程之类的开销

//文件数量增加到10多个以后，多个线程稍微快一点，但是并没有快多少

//按照网上的说法，读取文件的性能瓶颈是磁盘io的速度，多个线程并不能提高性能

#include <iostream>#include <fstream>#include <thread>#include <string>#include <windows.h>using namespace std;void ReadFileS(const char **path,int len){for (int i = 0; i < len;i++){fstream fs(*(path + i), ios_base::in | ios_base::out);string line;int lineCount = 1;while (fs && getline(fs, line));//cout << "File: " << *(path + i) << "line:" << lineCount++ << ", " << line.c_str() << endl;}}void _tReadFile(const char* path){fstream fs(path, ios_base::in | ios_base::out);string line;int lineCount = 1;while (fs && getline(fs, line));//cout << "File: " << path << "line:" << lineCount++ << ", " << line.c_str() << endl;}void ReadFileM(const char **path, int len){for (int i = 0; i < len; i++){std::thread t(_tReadFile,*(path+i));t.join();}}void main(){const char *path[13] = { "D:\\a.txt", "D:\\b.txt", "D:\\a.txt", "D:\\a.txt", "D:\\a.txt", "D:\\a.txt", "D:\\a.txt", "D:\\a.txt","D:\\a.txt","D:\\a.txt","D:\\a.txt","D:\\a.txt","D:\\a.txt" };double cost;double start = GetTickCount();ReadFileS(path, 13);double end = GetTickCount();cost = difftime(end, start);cout << "time diff 1------------------->" << cost << endl;start = GetTickCount();ReadFileM(path, 13);end = GetTickCount();cost = difftime(end, start);cout << "time diff 2------------------->" << cost << endl;system("pause");}

阅读全文

0 0