程序Ⅰ：通过Task类创建新线程
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
namespace Test000
{
class Program
{
/// <summary>
/// Task类封装
/// </summary>
class MyTask
{
/// <summary>
/// 第一个Task类，每秒计数，计数5次
/// </summary>
public static void T1()
{
Console.WriteLine("Task #{0}: Begin!", Task.CurrentId);
for (int i = 0; i < 5; i++)
{
Thread.Sleep(1000);
Console.WriteLine("Task #{0}: {1}", Task.CurrentId, i);
}
Console.WriteLine("Task #{0}: Terminated!", Task.CurrentId);
}
/// <summary>
/// 第二个Task类，每秒计数，计数5次
/// </summary>
public static void T2()
{
Console.WriteLine("Task #{0}: Begin!", Task.CurrentId);
Thread.Sleep(500);
for (int i = 0; i < 5; i++)
{
Thread.Sleep(1000);
Console.WriteLine("Task #{0}: {1}", Task.CurrentId, i);
}
Console.WriteLine("Task #{0}: Terminated!", Task.CurrentId);
}
}
static void Main(string[] args)
{
//建立两个Task
Task tsk1 = new Task(MyTask.T1);
Console.WriteLine("Task #{0}: Constructed!", tsk1.Id);
Task tsk2 = new Task(MyTask.T2);
Console.WriteLine("Task #{0}: Constructed!", tsk1.Id);
//运行Task
tsk1.Start();
tsk2.Start();
//等待Task运行结束
WaitAll(tsk1, tsk2);
Console.ReadLine();
}
/// <summary>
/// 等待所有的Task运行结束
/// </summary>
/// <param name="tsks">等待的Task类</param>
public static void WaitAll(params Task[] tsks)
{
foreach (var t in tsks)
{
t.Wait();
}
}
}
}
运行结果
程序Ⅱ：通过TaskFactory启动任务并接收任务的返回值
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace Test001
{
class Program
{
/// <summary>
/// Task类封装
/// </summary>
class MyTask
{
/// <summary>
/// 求1+2+...+n的和
/// </summary>
/// <param name="n">数n</param>
/// <returns></returns>
public static int Sum(object n)
{
int x = (int)n;
int sum = 0;
for (int i = 1; i <= x; i++)
{
sum += i;
}
return sum;
}
}
static void Main(string[] args)
{
Task<int> tsk = Task<int>.Factory.StartNew(MyTask.Sum, 100);
Console.WriteLine("Result is: " + tsk.Result);
Console.ReadLine();
}
}
}
运行结果
程序Ⅲ：通过Parallel类的Invoke函数，并行调用多个Task
本程序中通过Lambda表达式来建立新的Task
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
namespace Test002
{
class Program
{
static void Main(string[] args)
{
Parallel.Invoke
(
() =>
{
Console.WriteLine("Task #{0}: Begin!", Task.CurrentId);
for (int i = 0; i < 5; i++)
{
Thread.Sleep(1000);
Console.WriteLine("Task #{0}: {1}", Task.CurrentId, i);
}
Console.WriteLine("Task #{0}: Terminated!", Task.CurrentId);
},
() =>
{
Console.WriteLine("Task #{0}: Begin!", Task.CurrentId);
Thread.Sleep(500);
for (int i = 0; i < 5; i++)
{
Thread.Sleep(1000);
Console.WriteLine("Task #{0}: {1}", Task.CurrentId, i);
}
Console.WriteLine("Task #{0}: Terminated!", Task.CurrentId);
}
);
Console.ReadLine();
}
}
}
运行结果
程序Ⅳ：通过Parallel类的For和FoeEach函数，并行调用多个Task
本程序中，通过Stopwatch类统计程序段的运行时间
从例中可以看出：不是所有的循环在并行化时都是有效的。通常，对于小型循环或执行非常简单的操作的循环来说，使用顺序循环比并行循环更加快速
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace Test003
{
class Program
{
/// <summary>
/// 示例函数
/// </summary>
/// <param name="n">参数</param>
public static void DoSomeThing(int n)
{
int sum = 0;
for (int i = 0; i < n * 100; i++)
{
sum += i;
}
}
static void Main(string[] args)
{
//计时工具，需要System.Diagnostics
Stopwatch sw = new Stopwatch();
//统计依次顺序调用函数的时间
sw.Start();
for (int i = 100; i < 105; i++)
{
DoSomeThing(100);
}
sw.Stop();
Console.WriteLine("TotalTime: {0}", sw.Elapsed.TotalMilliseconds);
sw.Reset();
Console.WriteLine("===========");
//统计并行调用函数的时间
sw.Start();
Parallel.For(100, 105, DoSomeThing);
sw.Stop();
Console.WriteLine("TotalTime: {0}", sw.Elapsed.TotalMilliseconds);
sw.Reset();
Console.WriteLine("===========");
//统计并行调用函数的时间
sw.Start();
Parallel.ForEach(new int[5] { 100, 101, 102, 103, 104 }, DoSomeThing);
sw.Stop();
Console.WriteLine("TotalTime: {0}", sw.Elapsed.TotalMilliseconds);
sw.Reset();
Console.ReadLine();
}
}
}
运行结果
END