public void Start()
{
try
{
using (DBDocumentStorage ds = new DBDocumentStorage(_config.DBPath))
{
Threads = new Dictionary<Thread, ThreadContext>();
Dictionary<Guid, string> docs = ds.GetDocumentsList();
int c = 0;
Dictionary<Guid, string> ThrList = new Dictionary<Guid, string>();
foreach(var doc in docs)
{
c++;
if(c>= docs.Count/ ThreadCount)
{
Thread th = new Thread(Indexate);
IndexatorThreadContext tc = new IndexatorThreadContext(ThrList, th, Thread.CurrentThread, this);
th.IsBackground = true;
th.Start(tc);
Threads.Add(th, tc);
ThrList = new Dictionary<Guid, string>();
c = 0;
}
ThrList.Add(doc.Key, doc.Value);
}
Log("Threads started", "", LogType.Message);
}
}
catch (Exception ex)
{
Log(string.Format("{0} \r\n{1}", ex.Message, ex.StackTrace), "SpydeeCore", LogType.Error, "CoreConstructor");
}
}
private void Indexate(object context)
{
using (IndexatorThreadContext tc = (IndexatorThreadContext)context)
{
var kernel = tc.kernel;
DBDocumentStorage dbStorage = new DBDocumentStorage(kernel.DBPath);
foreach (var dcinfo in tc.DocList)
{
DBDocument d = dbStorage.GetDocumentByID(dcinfo.Key);
string t = d.Text;
//Log(string.Format("Doc [{0}, {1}] scanned.", d.ID, d.URL), "", LogType.Message);
}
}
}
private void LoadFromDB(string ConnectionString)
{
DBstorage = new DBDocumentStorage(ConnectionString);
docs = DBstorage.GetDocuments();
UpdateDocList();
}
public void SiteScan(object tc)
{
//ThreadContext mtc = (ThreadContext)tc;
using (CrawlerThreadContext mtc = (CrawlerThreadContext)tc)
{
CrawlerCore kernel = (CrawlerCore)mtc.kernel;
DBDocumentStorage dbStorage = new DBDocumentStorage(kernel.DBPath);
try
{
Log("Соединение с базой. Строка соединения: [" + kernel.DBPath + "] ...", "Core", LogType.Message);
string msg = string.Format("Поток ;{0} стартован, сканирование сайта '{1}'", Thread.CurrentThread.ManagedThreadId, mtc.SiteQueue.GetNextUnscanned().Adress);
kernel.Log(msg, "Core", LogType.Message);
if (kernel.UseRobots)
{
try
{
Uri robotsUrl = new Uri(mtc.Site.Adress, "robots.txt");
WebResponse wrr = kernel.MakeRequest(robotsUrl);
Stream ds = wrr.GetResponseStream();
Encoding fEncoding = Encoding.Default;
StreamReader r = new StreamReader(ds, fEncoding);
string content = r.ReadToEnd();
string hdr = HTMLUtils.GetHeaderAsString(wrr);
DBDocument d = dbStorage.CreateDocument(robotsUrl.OriginalString, content, hdr);
dbStorage.AddDocument(d);
//HtmlDocument doc = new HtmlDocument();
//doc.LoadHtml(robots);
}
catch (Exception e)
{
string pgmsg = string.Format("Robots.txt ОШИБКА: {0}", e.Message);
kernel.Log(pgmsg, "Core", LogType.Error);
// kernel.Log(e.StackTrace, "Core", LogType.Error);
}
}
while (mtc.SiteQueue.HasUnscanned && !mtc.Stop)
{
//получение адреса из очереди
Page pg = mtc.SiteQueue.GetNextUnscanned();
//запрос
try
{
WebResponse w = kernel.MakeRequest(pg.Adress);
System.Net.Mime.ContentType ct = new System.Net.Mime.ContentType(w.ContentType);
Encoding fEncoding = Encoding.Default;
if (!string.IsNullOrEmpty(ct.CharSet))
fEncoding = Encoding.GetEncoding(ct.CharSet);
Stream dataStream = w.GetResponseStream();
StreamReader reader = new StreamReader(dataStream, fEncoding);
string content = reader.ReadToEnd();
string scanmsg = string.Format("Сканирую страницу: {1}", Thread.CurrentThread.ManagedThreadId, pg.Adress);
kernel.Log(scanmsg, "Core", LogType.Message);
List<FoundText> ft = null;
if (ct.MediaType.StartsWith("text", StringComparison.InvariantCultureIgnoreCase))
{
HtmlDocument doc = new HtmlDocument();
doc.LoadHtml(content);
//получение ссылок на другие страницы
SiteList links = GetLinksFromHTMLDocument(doc);
foreach (string link in links)
{
string[] parts = link.Split('?');
parts = parts[0].Split('#');
mtc.SiteQueue.Add(parts[0]);
}
//получение "чистого" текста
string plainText = HTMLUtils.GetPlainText(doc);
ft = HTMLUtils.GetMatches(plainText, kernel.SearchWords.ToArray());
}
if ((ft != null) && (ft.Count > 0))
{
string hdr = HTMLUtils.GetHeaderAsString(w);
//запись в базу
DBDocument d = dbStorage.CreateDocument(pg.Adress.OriginalString, content, hdr);
dbStorage.AddDocument(d);
}
//kernel.Log(string.Format("Document added: {0}", pg.Adress), "Core", LogType.Message);
mtc.SiteQueue[pg] = true;
Thread.Sleep(mtc.kernel.SiteScanDelay);
}
catch (Exception e)
{
string pgmsg = string.Format("Страница: '{0}' Ошибка: {1}", pg.Adress, e.Message);
kernel.Log(pgmsg, "Core", LogType.Error);
// kernel.Log(e.StackTrace, "Core", LogType.Error);
}
}
}
catch (Exception e)
{
kernel.Log(e.Message, "Core", LogType.Error);
kernel.Log(e.StackTrace, "Core", LogType.Error);
}
finally
{
dbStorage.Dispose();
}
}
}
private void FilterDocList(object _df)
{
AddDocumentGridDelegate deleg = new AddDocumentGridDelegate(AddDocumentGrid);
DocumentFilter df = (DocumentFilter)_df;
if (df.Count > 0)
{
DBDocumentStorage ds = new DBDocumentStorage(DBstorage.DB.ConnectionString);
List<Guid> docids = ds.GetDocumentsIDs();
docs.Clear();
int i = 0;
int scancount = 0;
foreach (Guid docid in docids)
{
if (StopSearch)
break;
bool flag = false;
DBDocument d = ds.GetDocumentByID(docid);
try
{
string ct = d.LastVersion.HeadersCollection[HttpResponseHeader.ContentType];
if (ct.StartsWith("text"))
{
string src = HTMLUtils.GetPlainText(d.LastVersion.DataAsText);
/*foreach (string word in df)
{
if (src.IndexOf(word) > -1)
{
flag = true;
if (foundword != "")
foundword = foundword + ", ";
foundword += word;
//break;
}
}*/
List<FoundText> matches = HTMLUtils.GetMatches(src, df.ToArray());
if (matches.Count >0 )
{
docs.Add(d.ID, d.URL);
string FoundWords = "";
foreach(FoundText m in matches)
{
if (FoundWords != "")
FoundWords = FoundWords+ ", " ;
FoundWords = FoundWords + string.Format("({0},{1}) {2}", m.Index, m.Length, src.Substring(m.Index, m.Length));
}
dgv.Invoke(deleg, i++, d.ID, d.URL, FoundWords);
}
}
}
catch (Exception ex) { Log(ex); }
finally
{
scancount++;
string message = string.Format("Scanned {0}/{1} documents. Found {2} documents.", scancount, docids.Count, i);
Log(message);
d.Dispose();
}
}
}
}
