c# - 任务没有第二次完成 [间歇性问题] – 异步/等待

Question

我有一个 WebBrowser 控件是 Windows 窗体项目。它浏览“MyTableTest.html”中所有可用的 url。此页面中有四个 url，网络浏览器逐一浏览。一旦到达最后一个，它应该再次转到第一个。它在第一次迭代中运行良好——但在第二次迭代中无法访问 url。这是一个间歇性问题 - 某些时候它会起作用。

似乎(从日志中)他等待的任务没有完成。可以做些什么来使其在第二次迭代中也能正常工作？

注:下面给出MyTableTest.html

注意:这是基于帖子 Get ReadyState from WebBrowser control without DoEvents

问题

       startNavigation();

       WriteLogFunction("Location 1");

       // wait for DOM onload event, throw if cancelled
       await onloadTcs.Task;

       //ISSUE: Not reaching this location at second time navigation
       WriteLogFunction("Location 2");

代码

public partial class Form1 : Form
{
    public Form1()
    {
        InitializeComponent();
        this.Load += MainForm_Load;
    }

    List<string> visitedProducts = new List<string>();
    string nextNavigationUrl = String.Empty;

    // Form Load event handler
    async void MainForm_Load(object sender, EventArgs e)
    {
        // cancel the whole operation in 20 sec
        var cts = new CancellationTokenSource(20000);

        //urlStore.Add(@"C:\Samples_L\MyTableTest.html");
        nextNavigationUrl = GetHomoePageUrl();
        await NavigateInLoopAsync(cts.Token);
    }

    // navigate to each URL in a loop
    async Task NavigateInLoopAsync(CancellationToken ct)
    {

        bool isIterationComplete = false;

        while (!isIterationComplete)
        {

            string url = String.Empty;

            if (String.IsNullOrEmpty(nextNavigationUrl))
            {
                WriteLogFunction("Close");
                isIterationComplete = true;
            }
            else
            {
                url = nextNavigationUrl;
                ct.ThrowIfCancellationRequested();


                WriteLogFunction("Calling NavigateAsync");

                Action startNavigation = () => this.webBrowser1.Navigate(url);
                var html = await NavigateAsync(ct, startNavigation);


            }
        }
    }

    // asynchronous navigation
    async Task<string> NavigateAsync(CancellationToken ct, Action startNavigation)
    {
        var onloadTcs = new TaskCompletionSource<bool>();
        EventHandler onloadEventHandler = null;

        WriteLogFunction("Inside Function NavigateAsync");

        WebBrowserDocumentCompletedEventHandler documentCompletedHandler = delegate
        {
            // DocumentCompleted may be called several time for the same page,
            // if the page has frames
            if (onloadEventHandler != null)
                return;

            // so, observe DOM onload event to make sure the document is fully loaded
            onloadEventHandler = (s, e) =>
                onloadTcs.TrySetResult(true);
            this.webBrowser1.Document.Window.AttachEventHandler("onload", onloadEventHandler);
        };

        this.webBrowser1.DocumentCompleted += documentCompletedHandler;

        try
        {
            using (ct.Register(() => onloadTcs.TrySetCanceled(), useSynchronizationContext: true))
            {
                startNavigation();

                WriteLogFunction("Location 1");

                // wait for DOM onload event, throw if cancelled
                await onloadTcs.Task;

                //ISSUE: Not reaching this location at second time navigation
                WriteLogFunction("Location 2");
            }
        }
        finally
        {
            this.webBrowser1.DocumentCompleted -= documentCompletedHandler;
            if (onloadEventHandler != null)
                this.webBrowser1.Document.Window.DetachEventHandler("onload", onloadEventHandler);
        }

        WriteLogFunction("Place 3");

        // the page has fully loaded by now

        // optional: let the page run its dynamic AJAX code,
        // we might add another timeout for this loop
        do { await Task.Delay(500, ct); }
        while (this.webBrowser1.IsBusy);

        //Call Processing -- Added By Lijo
        ExerciseApp(this.webBrowser1, null);

        // return the page's HTML content
        return this.webBrowser1.Document.GetElementsByTagName("html")[0].OuterHtml;
    }

    private void ExerciseApp(object sender, WebBrowserDocumentCompletedEventArgs e)
    {
        WriteLogFunction("ExerciseApp");
        var wb = sender as WebBrowser;
        int catalogElementIterationCounter = 0;
        var elementsToConsider = wb.Document.All;
        string productUrl = String.Empty;
        bool isClicked = false;

        foreach (HtmlElement e1 in elementsToConsider)
        {

            catalogElementIterationCounter++;

            string x = e1.TagName;
            String idStr = e1.GetAttribute("id");

            if (!String.IsNullOrWhiteSpace(idStr))
            {
                //Each Product Navigation
                if (idStr.Contains("catalogEntry_img"))
                {
                    productUrl = e1.GetAttribute("href");
                    if (!visitedProducts.Contains(productUrl))
                    {
                        WriteLogFunction("productUrl -- " + productUrl);
                        visitedProducts.Add(productUrl);
                        isClicked = true;

                        //e1.InvokeMember("Click");
                        nextNavigationUrl = productUrl;

                        break;
                    }

                }
            }
        }

        WriteLogFunction(visitedProducts.Count.ToString());
        WriteLogFunction(nextNavigationUrl);

        if (visitedProducts.Count == 4)
        {
            WriteLogFunction("Condition B");
            visitedProducts = new List<string>();
        }

        if (!isClicked)
        {
            WriteLogFunction("Condition C");
            nextNavigationUrl = GetHomoePageUrl();
        }
    }

    private void HomoePageNavigate()
    {
        webBrowser1.Navigate(GetHomoePageUrl());
    }

    private string GetHomoePageUrl()
    {
       return @"C:\Samples_L\MyTableTest.html";
    }

    private void WriteLogFunction(string strMessage)
    {
        using (StreamWriter w = File.AppendText("log.txt"))
        {
            w.WriteLine("\r\n{0} ..... {1} ", DateTime.Now.ToLongTimeString(), strMessage);
        }
    }



}

MyTableTest.html

<html>
<head>

    <style type="text/css">
        table {
            border: 2px solid blue;
        }

        td {
            border: 1px solid teal;
        }
    </style>

</head>
<body>

    <table id="four-grid">
         <tr>
            <td>
                <a href="https://www.wikipedia.org/" id="catalogEntry_img63666">

                    <img src="ssss"
                        alt="B" width="70" />
                </a>
            </td>
            <td>
                <a href="http://www.keralatourism.org/" id="catalogEntry_img63667">

                    <img src="ssss"
                        alt="A" width="70" />
                </a>
            </td>
        </tr>
        <tr>
            <td>
                <a href="https://stackoverflow.com/users/696627/lijo" id="catalogEntry_img63664">

                    <img src="ssss"
                        alt="G" width="70" />
                </a>
            </td>
            <td>
                <a href="http://msdn.microsoft.com/en-US/#fbid=zgGLygxrE84" id="catalogEntry_img63665">

                    <img src="ssss"
                        alt="Y" width="70" />
                </a>
            </td>
        </tr>

    </table>
</body>

</html>

Answer 1

"Location 1" is printed but not "Location 2". What is your suggestion to overcome this? Were you able to reproduce this issue?

我没有尝试重现，但就像我在评论中所说的那样，很明显 DocumentCompleted 或 window.onload 不会针对该特定 URL 被触发.如果页面无法完全加载，或者例如如果 URL 无效或服务器变得不可访问。这就是超时逻辑的用途。

在documentCompletedHandler开头多放一根，在onloadEventHandler里多放一根:

onloadEventHandler = (s, e) => { Log("inside onloadEventHandler");
    onloadTcs.TrySetResult(true); }

看看什么被解雇了，什么没有。

此外，你还应该实现WebBrowser Feature Control启用现代 IE 功能。这通常也会影响网页的加载方式。从 here 复制 SetFeatureBrowserEmulation .