如何在C＃中实现此并行模式？

Question

What I'm ultimately trying to accomplish is to get the HTML from an unknown but limited number of webpages where GetPage(i) returns the HTML for page i and I want to stop as soon as I found a non-page.

The exact pattern I'm going for is like this:

• Start N parallel tasks that are GetPage(0) , ..., GetPage(N-1) .
• As soon as a task GetPage(i) completes, if the task was able to get the page, add it to a collection of pages and try to get the next largest page that hasn't tried to be processed yet; or if the task was not able to get the page, cancel all tasks GetPage(j) where j>i .

So my attempted implementation is like

        var docs = new LinkedList<HtmlDocument>();
        int tlimit = 20;
        var tasks = new Task<HtmlDocument>[tlimit];
        for(int i = 0; i < tlimit; ++i)
        {
            tasks[i] = Task<HtmlDocument>.Factory.StartNew(() => BoardScanner.GetBoardPage(i));
        }
       /// ???

Answer 1

This might be what you're looking for. It will treat your documents in parallel and async possibly starting with the largest one (not guaranteed though). If you really want to go by the largest one, then you'll have to tweak this quite a bit. And I am not sure if you will end up with parallel processing for parallel sort of means that you give more control to the machine in terms of deciding what gets treated first:

var docs = new List<XmlDocument>();

            var tasks = docs.OrderBy(p => p.InnerXml.Length).Select(file => Task.Run(async () =>
           {
               await BoardScanner.GetBoardPage(file);
                 // your document treatment logic here
             }
                 ));

            await Task.WhenAll(tasks);
            // your logic upon all of your documents were treated;