/// <summary>
/// 线程执行方法
/// </summary>
void StartCrawlePage()
{
// 循环抓取URL
while (MainForm.UrlManager.Urls.Count > 0)
{
UrlItem url;
lock (MainForm.UrlManager.Urls)
{
if (MainForm.UrlManager.Urls != null && MainForm.UrlManager.Urls.Count > 0)
{
url = MainForm.UrlManager.Urls[0];
MainForm.UrlManager.Urls.Remove(url);
}
else
{
return;
}
}
// 抓取单个网页并分词
Dictionary <string, ulong> words = null;
try
{
words = HtmlHelper.SetupSingleUrl(url.Url, false, Service.JWordSegmentor, Service.DbHelper.WebRuleCollections);
}
catch (Exception)
{
// 抓取一个网页
Interlocked.Increment(ref _failedPage);
FailedUrls.Add(url.Url);
MainForm.BeginInvoke(new Action(FailCrawle));
// 如果已经抓取完毕
if ((GetedPage + FailedPage) == TotalPages)
{
MainForm.BeginInvoke(new Action(PostCrawlerProcess));
}
continue;
}
// 合并所有线程抓回的单词
if (words != null && words.Count > 0)
{
lock (NewWords)
{
foreach (KeyValuePair <string, ulong> word in words)
{
if (NewWords.ContainsKey(word.Key))
{
NewWords[word.Key] = NewWords[word.Key] + word.Value;
}
else
{
NewWords.Add(word.Key, word.Value);
}
}
//foreach (KeyValuePair<string, ulong> word in words)
//{
// if (JWordSegmentorService.IsExist(word.Key, ""))
// {
// if (ExistWords.ContainsKey(word.Key))
// {
// ExistWords[word.Key] = ExistWords[word.Key] + word.Value;
// }
// else
// {
// ExistWords.Add(word.Key, word.Value);
// }
// continue;
// }
// if (NewWords.ContainsKey(word.Key))
// {
// NewWords[word.Key] = NewWords[word.Key] + word.Value;
// }
// else
// {
// NewWords.Add(word.Key, word.Value);
// }
//}
}
}
// 抓取一个网页
Interlocked.Increment(ref _getedPage);
GetedUrls.Add(url.Url);
MainForm.BeginInvoke(new Action(SucceedCrawle));
// 如果已经抓取完毕
if ((GetedPage + FailedPage) == TotalPages)
{
MainForm.BeginInvoke(new Action(PostCrawlerProcess));
}
}//while
}
public async Task Crawl()
{
List <Task> downloadTask = new List <Task>();
pendingUrls.Enqueue(InitialUrl);
while (pendingUrls.Count > 0)
{
if (!pendingUrls.TryDequeue(out var current))
{
continue;
}
;
if (CompletedUrls.Count + DownloadingUrls.Count > MaxCount)
{
break;
}
//限制并发
if (MaxParallel > 0 && DownloadingUrls.Count >= MaxParallel)
{
await Task.Delay(100);
continue;
}
//下载开始事件
var downEvent = new BeforeDownloadEventArgs()
{
Url = current,
Cancelled = false
};
BeforeDownload?.Invoke(this, downEvent);
if (!downEvent.Cancelled)
{
lock (urlLock)
{
DownloadingUrls.Add(current);
}
var down = DownLoad(current, nameIndex++.ToString())
.ContinueWith(taks =>
{
if (taks.IsFaulted)
{
FailedUrls.Add(current);
lock (urlLock)
{
DownloadingUrls.Remove(current);
}
return;
}
CompletedUrls.Add(current);
lock (urlLock)
{
DownloadingUrls.Remove(current);
}
var ev = new DownloadedEventArgs()
{
Html = taks.Result,
OverrideUrlParse = false,
Url = current
};
Downloaded?.Invoke(this, ev);
//覆盖默认的url解析
if (ev.OverrideUrlParse && ev.NextUrls != null)
{
foreach (var u in ev.NextUrls)
{
AddPending(u);
}
}
else
{
Parse(taks.Result, current);//解析,并加入新的链接
}
});
downloadTask.Add(down);
Downloading?.Invoke(this, new DownloadingEventArgs()
{
Url = current
});
}
if (pendingUrls.Count <= 0)
{
await Task.WhenAny(downloadTask);
downloadTask.RemoveAll(t => t.IsCompleted);
}
}
await Task.WhenAll(downloadTask);
this.CrawlerCompleted?.Invoke(this, new CrawlerCompletedEventArgs()
{
FailedCount = FailedUrls.Count,
SuccessCount = CompletedUrls.Count
});
}