在任务中运行Webbrowser线程

Question

I have a program that runs and starts 2 long running tasks. One of the tasks is a web scraper in which I have to use the WebBrowser ActiveX control so that I can get the rendered page. In order to use the control I have to start a thread so that I can set the apartment state for the message loop. When I do this, the proogram works fine, or at least for the first page that is fetched. Subsequent pages/calls, the webbrowser times out and it's state seems to remain at "uninitialized". In tracing my code, I never see the "HandleDestroyed" fire for the WebClient.

What do I need to do to Properly Destroy the WebBrowser control and or my own class in order for it to be reused again.

public static string AXFetch(string url, string ua)
{
    TestBrowser TB = new TestBrowser();
    Thread th = new Thread(() => TB.MakeLiRequest(url,ua));
    th.SetApartmentState(ApartmentState.STA);
    th.Start();
    th.Join(new TimeSpan(0, 0, 90)); //90 second timeout
    SiteData = TB.DocumentText;
    TB = null;
    return SiteData;
}

class TestBrowser
{
    public string DocumentText = "";
    private bool DocCompleted = false;

    public TestBrowser()
    {

    }

    private void reset_fetch_status()
    {
        this.DocCompleted = false;
        this.DocumentText = "";
    }


    public void MakeLiRequest(string url, string UA)
    {
        reset_fetch_status();
                using (WebBrowser wb = new WebBrowser())
        {
            wb.Visible = false;
            wb.AllowNavigation = true;
            wb.ScriptErrorsSuppressed = true;
            wb.DocumentCompleted += this.wb_DocumentCompleted;
            wb.Navigate(url, "_self", null, "User-Agent: " + UA + "\r\n");
            WaitForPage();
            wb.Url = null;
            wb.DocumentCompleted -= this.wb_DocumentCompleted;
        }
    }

    private void HandleDestroyed(Object sender, EventArgs e)
    {
                //This never seems to fire, I don't knwo why
        Logging.DoLog("You are in the Control.HandleDestroyed event.");

    }

    private bool WaitForPage()
    {
        int timer = 0;
        while (this.DocCompleted == false)
        {
            Application.DoEvents();
            System.Threading.Thread.Sleep(100);
            ++timer;
            if (timer == (PageTimeOut * 10))
            {
                Console.WriteLine("WebBrowser Timeout has been reached");
                Application.Exit();
                return false;
            }

        }
        return true;
    }

    private void wb_DocumentCompleted(object sender, WebBrowserDocumentCompletedEventArgs e)
    {
        WebBrowser wb = (WebBrowser)sender;
        if (wb.ReadyState == WebBrowserReadyState.Complete)
        {
            this.DocumentText = wb.DocumentText;
            this.DocCompleted = true;
        }
    }

}

Answer 1

On handle destroyed will only be called by the parent form.

If you were to try to access from the webbrowser control you would get this error:

Error   1   Cannot access protected member 
'System.Windows.Forms.Control.OnHandleDestroyed(System.EventArgs)' via a 
qualifier of type 'System.Windows.Forms.WebBrowser'; the qualifier must be of type 'stackoverflowpost47036339.Form1' (or derived from it)   

Also you are not hooking it up. But since you haven't given your web browser any parent form, It can't be called. This is how you would hook it up to the parent form.

     form1.HandleDestroyed += Form1_HandleDestroyed;
    }

    void Form1_HandleDestroyed(object sender, EventArgs e)
    {

    }