private static perDocument extractLinks(string srUrl, HtmlDocument doc)
{
var baseUri = new Uri(srUrl);
perDocument myDoc = new perDocument();
// extracting all links
var vrNodes = doc.DocumentNode.SelectNodes("//a[@href]");
if (vrNodes != null)
{
foreach (HtmlNode link in vrNodes)//xpath notation
{
HtmlAttribute att = link.Attributes["href"];
//this is used to convert from relative path to absolute path
var absoluteUri = new Uri(baseUri, att.Value.ToString());
if (!absoluteUri.ToString().StartsWith("http://") && !absoluteUri.ToString().StartsWith("https://"))
{
continue;
}
myDoc.lstExtractedUrls.Add(absoluteUri.ToString().Split('#').FirstOrDefault());
}
}
myDoc.lstExtractedUrls = myDoc.lstExtractedUrls.Distinct().ToList();
var vrDocTitle = doc.DocumentNode.SelectSingleNode("//title")?.InnerText.ToString().Trim();
vrDocTitle = System.Net.WebUtility.HtmlDecode(vrDocTitle);
myDoc.srDocTitle = vrDocTitle;
return(myDoc);
}
public static void crawlPage(string srUrlToCrawl, int irUrlDepthLevel, string _srParentUrl, DateTime _dtDiscoverDate)
{
var vrLocalUrl = srUrlToCrawl;
crawlingResult crawlResult = new crawlingResult();
crawlResult.srCrawledUrl = vrLocalUrl;
crawlResult.srParentUrlHash = _srParentUrl;
crawlResult.dtDiscoverDate = _dtDiscoverDate;
Stopwatch swTimerCrawling = new Stopwatch();
swTimerCrawling.Start();
HtmlWeb wbClient = new HtmlWeb();//you should use httpwebrequest for more control and better performance
wbClient.AutoDetectEncoding = true;
HtmlAgilityPack.HtmlDocument doc = new HtmlAgilityPack.HtmlDocument();
try
{
doc = wbClient.Load(crawlResult.srCrawledUrl);
crawlResult.srCrawledSourceCode = doc.Text;
}
catch (Exception E)
{
crawlResult.blcrawlSuccess = false;
logError(E, "crawlPage");
}
swTimerCrawling.Stop();
crawlResult.irCrawlingTimeMS = Convert.ToInt32(swTimerCrawling.ElapsedMilliseconds);
if (crawlResult.blcrawlSuccess)
{
perDocument docResults = extractLinks(crawlResult.srCrawledUrl, doc);
crawlResult.lstDiscoveredLinks.AddRange(docResults.lstExtractedUrls);
crawlResult.srTitleofPage = docResults.srDocTitle;
}
doc = null;
saveCrawlInDatabase(crawlResult);
}
private static perDocument extractLinks(string srUrl)
{
lock (swCrawling)
swCrawling.WriteLine(DateTime.Now + "\t" + srUrl);
var baseUri = new Uri(srUrl);
HtmlWeb web = new HtmlWeb();
web.AutoDetectEncoding = true;
HtmlAgilityPack.HtmlDocument doc = new HtmlAgilityPack.HtmlDocument();
doc = web.Load(srUrl);
perDocument myDoc = new perDocument();
// extracting all links
foreach (HtmlNode link in doc.DocumentNode.SelectNodes("//a[@href]"))//xpath notation
{
HtmlAttribute att = link.Attributes["href"];
//this is used to convert from relative path to absolute path
var absoluteUri = new Uri(baseUri, att.Value.ToString());
if (!absoluteUri.ToString().StartsWith("http://") && !absoluteUri.ToString().StartsWith("https://"))
{
continue;
}
myDoc.lstExtractedUrls.Add(absoluteUri.ToString());
}
myDoc.lstExtractedUrls = myDoc.lstExtractedUrls.Distinct().ToList();
var vrDocTitle = doc.DocumentNode.SelectSingleNode("//title")?.InnerText.ToString().Trim();
vrDocTitle = System.Net.WebUtility.HtmlDecode(vrDocTitle);
myDoc.srDocTitle = vrDocTitle;
return(myDoc);
}
