Question

刚刚开始使用AsyncEnumerator库并且不确定我缺少什么。我试图让图书馆作者自己提供an example，但不知道该行中的“执行”是什么：

ae.Execute(ProcessAllAndEachOps(ae, urls));

方法参考。有人能给我一些线索吗？

更新：我设法通过进行一些已经包含在下面的代码中的更改来使其运行。值得注意的是，Peter注意到AsyncEnumerator对象的Execute（）方法已经过时，应该用一些辅助函数替换。

using System;
using System.Collections.Generic;
using System.Net;
using Wintellect.Threading.AsyncProgModel;


public static class AsyncEnumeratorPatterns {

    private static void Execute(AsyncEnumerator ae, IEnumerator<Int32> enumerator){
                    ae.EndExecute(ae.BeginExecute(enumerator, null));
    }

      public static void Main() {
        String[] urls = new String[] { 
          "http://Wintellect.com/", 
          "http://1.1.1.1/",   // Demonstrates error recovery
          "http://www.Devscovery.com/" 
        };

        // Demonstrate process
        AsyncEnumerator ae = new AsyncEnumerator();
        Execute(ae, ProcessAllAndEachOps(ae, urls));
      }

      private static IEnumerator<Int32> ProcessAllAndEachOps(
           AsyncEnumerator ae, String[] urls) {
        Int32 numOps = urls.Length;

        // Issue all the asynchronous operation(s) so they run concurrently
        for (Int32 n = 0; n < numOps; n++) {
          WebRequest wr = WebRequest.Create(urls[n]);
          wr.BeginGetResponse(ae.End(), wr);
        }

        // Have AsyncEnumerator wait until ALL operations complete
        yield return numOps;

        Console.WriteLine("All the operations completed:");
        for (Int32 n = 0; n < numOps; n++) {
          ProcessCompletedWebRequest(ae.DequeueAsyncResult());
        }

        Console.WriteLine(); // *** Blank line between demos ***

        // Issue all the asynchronous operation(s) so they run concurrently
        for (Int32 n = 0; n < numOps; n++) {
          WebRequest wr = WebRequest.Create(urls[n]);
          wr.BeginGetResponse(ae.End(), wr);
        }

        for (Int32 n = 0; n < numOps; n++) {
          // Have AsyncEnumerator wait until EACH operation completes
          yield return 1;

          Console.WriteLine("An operation completed:");
          ProcessCompletedWebRequest(ae.DequeueAsyncResult());
        }
      }

      private static void ProcessCompletedWebRequest(IAsyncResult ar) {
        WebRequest wr = (WebRequest)ar.AsyncState;
        try {
          Console.Write("   Uri=" + wr.RequestUri + "    ");
          using (WebResponse response = wr.EndGetResponse(ar)) {
            Console.WriteLine("ContentLength=" + response.ContentLength);
          }
        }
        catch (WebException e) {
          Console.WriteLine("WebException=" + e.Message);
        }
    }
}

Answer 1

正如其他人所说，Execute方法不久前已被弃用，现在已完全消失，MSDN示例已经过时了。

该方法稍微违背了AsyncEnumerator类的要点，因为进行调用的线程将运行迭代器直到第一个yield，然后阻塞直到迭代器完成处理。

因此可以更明确地将其替换为：

var asyncResult = ae.BeginExecute(ProcessAllAndEachOps(ae, urls), null);
asyncResult.AsyncWaitHandle.WaitOne();
ae.EndExecute(asyncResult);

这看起来更像是在AsyncEnumerator迭代器中进行常规APM调用，并使阻塞更加明确。

AsyncEnumerator用法示例

1 个答案: