/// <summary>
/// Performs image handling.
/// </summary>
/// <param name="content">The content.</param>
/// <param name="title">The title.</param>
/// <param name="rssKey">The RSS key.</param>
protected IEnumerable <Task> HandleImages(string content, string title, RssKey rssKey = null)
{
if (content != null && title != null)
{
var html = new HtmlDocument();
html.LoadHtml(content);
return(HandleImages(html, title, rssKey));
}
return(new Task[0]);
}
/// <summary>
/// Performs image handling.
/// </summary>
/// <param name="doc">The document.</param>
/// <param name="title">The title.</param>
/// <param name="rssKey">The RSS key.</param>
protected IEnumerable <Task> HandleImages(HtmlDocument doc, string title, RssKey rssKey = null)
{
var jobList = new List <Task>();
if (doc != null && title != null)
{
var images = doc.GetAllImages();
var link = rssKey != null ? rssKey.Link : Source.Uri;
var published = rssKey != null ? new DateTime(rssKey.Updated) : DateTime.UtcNow;
if (images != null)
{
foreach (var image in images)
{
jobList.Add(Observer.NotifyNewImageContentFoundThreadSafe(title, image.Item2, image.Item1, link,
published, Source));
}
}
}
return(jobList);
}
/// <summary>
/// Notifies the Observer that a (probably) new Rss feed item has been detected by a crawler.
/// Note: Must be thread safe
/// </summary>
/// <param name="item">The item to index</param>
/// <param name="rssKey">The belonging rss key.</param>
/// <param name="source">The source.</param>
/// <returns>If indexing is required: a task with the corresponding indexing job.
/// <para>Otherwise: a completed Task</para></returns>
public Task NotifyNewRssFeedFoundThreadSave(SyndicationItem item, RssKey rssKey, Source source)
{
var plainText = item.Summary?.Text?.GetTextFromHtml() ?? string.Empty;
var plainTitle = item.Title?.Text.GetTextFromHtml() ?? source.Name;
if (DoesNotViolateBlackList(source, plainText) && DoesNotViolateBlackList(source, plainTitle))
{
var article = new Article(Guid.NewGuid(),
plainTitle,
rssKey.Link,
null,
plainText,
new DateTime(rssKey.Updated),
DateTime.UtcNow,
source.Id);
var job = DatabaseAdapterFactory.GetControllerInstance <IIndexerService>()
.IndexRssFeedItemThreadSafeAsync(article, rssKey);
return(job ?? Task.CompletedTask);
}
return(Task.CompletedTask);
}
public void IndexRssItemTest()
{
var elasticController = DatabaseAdapterFactory.GetControllerInstance <IIndexerService>();
var milliseconds = DateTimeOffset.Now.ToUnixTimeMilliseconds();
var guid = Guid.NewGuid();
var indexedguid = Guid.NewGuid();
var articleOne = new Article(guid, "abgef:" + milliseconds,
"http://" + milliseconds + ".de", "http://" + milliseconds + ".de", "abgefssmt",
DateTime.Today, DateTime.Today, indexedguid);
var rsskey = new RssKey(milliseconds, articleOne.Link);
try
{
elasticController.IndexRssFeedItemThreadSafeAsync(articleOne, rsskey);
Assert.IsTrue(true);
}
catch (Exception)
{
Assert.Fail();
}
}
/// <summary>
/// Indexes an RSS article thread safe.
/// </summary>
/// <param name="article">The article.</param>
/// <param name="key">The rss key.</param>
public async Task IndexRssFeedItemThreadSafeAsync(Article article, RssKey key)
{
#if ST
lock (lockObj)
#endif
{
try
{
//using (new PerfTracer(nameof(IndexRssFeedItemThreadSafeAsync)))
{
var client = new ElasticClient(settings);
var speedKey = new ShortArticleKey(article.IndexingSourceId, key.Updated, key.Link);
#if ST
if (!(IsArticleDuplicate(speedKey, client).Result))
#else
if (!(await IsArticleDuplicate(speedKey, client)))
#endif
{
#if ST
var result = IndexArticle(article, speedKey, client).Result;
#else
var result = await IndexArticle(article, speedKey, client);
#endif
if (result.Status == SuccessState.UnknownError)
{
throw new Exception("IndexArticle failed: " + result.UserMessage);
}
}
}
}
catch (Exception e)
{
Console.WriteLine(e);
}
}
}
/// <summary>
/// Handles a given document (Xpath and regex)
/// </summary>
/// <param name="doc">The doc to handle.</param>
/// <param name="title">The title (optional).</param>
/// <param name="rssKey">The rss key (optional).</param>
internal IEnumerable <Task> HandleDocument(HtmlDocument doc, string title = null, RssKey rssKey = null)
{
if (string.Equals(doc.Text, this.lastFetchedContent, StringComparison.Ordinal))
{
return(new Task[0]);
}
this.lastFetchedContent = doc.Text;
//First possibility (makes nearly no sense): No filer -> direct index
var link = rssKey != null ? rssKey.Link : Source.Uri;
var published = rssKey != null ? new DateTime(rssKey.Updated) : DateTime.UtcNow;
var noXPath = string.IsNullOrEmpty(Source.XPath);
var noRegex = this.regex == null;
var jobList = new List <Task>();
if (noRegex && noXPath)
{
title = title ?? GetTitle(doc);
jobList.Add(Observer.NotifyNewWebPageContentFoundThreadSafe(title, doc.GetTextFromHtml(), link, published, Source));
jobList.AddRange(HandleImages(doc, title, rssKey));
return(jobList);
}
if (noXPath)
{
title = title ?? GetTitle(doc);
jobList.AddRange(HandleRegex(doc.Text, title, link, published));
}
else
{
var relevantNodes = doc.DocumentNode.SelectNodes(Source.XPath);
foreach (var node in relevantNodes)
{
if (noRegex)
{
title = title ?? GetTitle(doc);
jobList.Add(Observer.NotifyNewWebPageContentFoundThreadSafe(title,
node.InnerHtml.GetTextFromHtml(),
link, published, Source));
jobList.AddRange(HandleImages(node.InnerHtml, title, rssKey));
}
else
{
title = title ?? GetTitle(doc);
jobList.AddRange(HandleRegex(node.InnerHtml, title, link, published));
}
}
}
return(jobList);
}
