可以从线程调用返回类型为list的方法

时间:2013-05-01 13:26:41

标签: c# asp.net multithreading

我有一个方法,如下所示,它调用服务。

如何通过线程运行此方法?

public List<AccessDetails> GetAccessListOfMirror(string mirrorId,string server)
{
    List<AccessDetails> accessOfMirror = new List<AccessDetails>();
    string loginUserId = SessionManager.Session.Current.LoggedInUserName;
    string userPassword = SessionManager.Session.Current.Password;

    using (Service1Client client = new Service1Client())
    {
        client.Open();
        accessOfMirror = client.GetMirrorList1(mirrorId, server, null);
    }

    return accessOfMirror;
}

3 个答案:

答案 0 :(得分:3)

在C#3.5或4.0中,你可以这样做。

var task = Task.Factory.StartNew<List<AccessDetails>>(() => GetAccessListOfMirror(mirrorId,server))
.ContinueWith(tsk => ProcessResult(tsk));

private void ProcessResult(Task task)
{
    var result = task.Result;
}

在C#4.5中,有await / async关键字,这是上面的一些糖

public async Task<List<AccessDetails>> GetAccessListOfMirror(string mirrorId,string server)

var myResult = await GetAccessListOfMirror(mirrorId, server)

答案 1 :(得分:1)

尝试这样的事情:

public async Task<List<AccessDetails>> GetAccessListOfMirror(string mirrorId, string server)
    {
        List<AccessDetails> accessOfMirror = new List<AccessDetails>();
        string loginUserId = SessionManager.Session.Current.LoggedInUserName;
        string userPassword = SessionManager.Session.Current.Password;


        using (Service1Client client = new Service1Client())
        {
            client.Open();
            Task<List<AccessDetails>> Detail = client.GetMirrorList1(mirrorId, server, null);
            accessOfMirror = await Detail;

        }


        return accessOfMirror;
    }

答案 2 :(得分:0)

下面是我使用的辅助类,它引用了RX.NET

如果你在项目中包含它,那么你可以非常简单地编写内容 - 上面的代码你可以转移到一个单独的线程,如下所示:

int mirrorId = 0;
string server = "xxx";
ASync.Run<List<AccessDetails>>(GetAccessListOfMirror(mirrorId,server), resultList => {
   foreach(var accessDetail in resultList)
   {
         // do stuff with result
   }
}, error => { // if error occured on other thread, handle exception here  });

值得注意:将lambda表达式合并回原始调用线程 - 如果您从GUI线程启动异步操作,这非常方便。

它还有另一个非常方便的方法:Fork允许你分离多个工作线程并导致调用线程阻塞,直到所有子线程完成或出错。

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Concurrency;

namespace MyProject
{

    public static class ASync
    {
        public static void ThrowAway(Action todo)
        {
            ThrowAway(todo, null);
        }

        public static void ThrowAway(Action todo, Action<Exception> onException)
        {
            if (todo == null)
                return;

            Run<bool>(() =>
            {
                todo();
                return true;
            }, null, onException);
        }

        public static bool Fork(Action<Exception> onError, params Action[] toDo)
        {
            bool errors = false;
            var fork = Observable.ForkJoin(toDo.Select(t => Observable.Start(t).Materialize()));
            foreach (var x in fork.First())
                if (x.Kind == NotificationKind.OnError)
                {
                    if(onError != null)
                        onError(x.Exception);

                    errors = true;
                }

            return !errors;
        }

        public static bool Fork<T>(Action<Exception> onError, IEnumerable<T> args, Action<T> perArg)
        {
            bool errors = false;
            var fork = Observable.ForkJoin(args.Select(arg => Observable.Start(() => { perArg(arg); }).Materialize()));
            foreach (var x in fork.First())
                if (x.Kind == NotificationKind.OnError)
                {
                    if (onError != null)
                        onError(x.Exception);

                    errors = true;
                }

            return !errors;
        }


        public static void Run<TResult>(Func<TResult> todo, Action<TResult> continuation, Action<Exception> onException)
        {
            bool errored = false;
            IDisposable subscription = null;

            var toCall = Observable.ToAsync<TResult>(todo);
            var observable =
                Observable.CreateWithDisposable<TResult>(o => toCall().Subscribe(o)).ObserveOn(Scheduler.Dispatcher).Catch(
                (Exception err) =>
                {
                    errored = true;

                        if (onException != null)
                            onException(err);


                        return Observable.Never<TResult>();
                }).Finally(
                () =>
                {
                    if (subscription != null)
                        subscription.Dispose();
                });

            subscription = observable.Subscribe((TResult result) =>
            {
                if (!errored && continuation != null)
                {
                    try
                    {
                        continuation(result);
                    }
                    catch (Exception e)
                    {
                        if (onException != null)
                            onException(e);
                    }
                }
            });
        }
    }
}