/// <summary>
/// Crawls a Url and creates a <see cref="UrlCrawlData"/> object that is stored in
/// the internal crawledUrls <see cref="ArrayList"/>. Since it runs in one of the
/// crawling threads that may be interrupted or aborted at any time it must be able
/// to handle ThreadAbortException and ThreadInterruptedException.
/// </summary>
/// <param name="urlToCrawl">A reference to the <see cref="InternetUrlToCrawl"/>
/// object that encapsulates the url that must be crawled.</param>
private void CrawlUrl(ref InternetUrlToCrawl urlToCrawl)
{
try
{
UrlCrawlData urlData = new UrlCrawlData();
HiResTimer timer = new HiResTimer();
//create the web request and download the data
HttpWebRequest pageRequest = null;
try
{
pageRequest = (HttpWebRequest)HttpWebRequest.Create(urlToCrawl.Url);
}
catch
{
urlData.HttpStatusCode=HttpStatusCode.BadRequest;//TODO comment
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.OutLinks = new InternetUrlToIndex[0];
urlData.Data = String.Empty;
UpdateStatistics(HttpStatusCode.BadRequest, 0);
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
UpdateStatistics(HttpStatusCode.BadRequest, 0);
return;
}
pageRequest.UserAgent = globals.UserAgent;
pageRequest.Timeout=ExponentialBackoff.DefaultBackoff; //page timeout = 30 seconds
HttpWebResponse pageResponse=null;
try
{
timer.Start();
pageResponse = (HttpWebResponse)pageRequest.GetResponse();
//the above line might throw either WebException or UriFormatException
}
catch(WebException we)
{
HttpWebResponse response=(HttpWebResponse)we.Response;
if (response!=null)
{
//although an exception occured we're able to get the Status Code
urlData.HttpStatusCode=response.StatusCode;
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
UpdateStatistics(response.StatusCode, response.ContentLength);
}
else
{
urlData.HttpStatusCode=HttpStatusCode.BadRequest;//TODO comment
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
UpdateStatistics(HttpStatusCode.BadRequest, 0);
}
}
catch(UriFormatException)
{
//this will occur if the url is not valid
urlData.HttpStatusCode=HttpStatusCode.BadRequest;
urlData.Updated=false;
urlData.Data = String.Empty;
urlData.UrlToCrawl = urlToCrawl;
urlData.OutLinks = new InternetUrlToIndex[0];
}
finally
{
timer.Stop();
urlData.TimeStamp = DateTime.UtcNow;
}
if(pageResponse !=null)
{
//update the fields
urlData.HttpStatusCode = pageResponse.StatusCode;
//download and parse the contents of the url
Stream receiveStream=pageResponse.GetResponseStream();
StreamReader receivedBytes=new StreamReader(receiveStream,defaultEncoding);
string contents = String.Empty;
try
{
contents=receivedBytes.ReadToEnd();
}
catch
{
//it should be response timeout not request timeout
urlData.HttpStatusCode = HttpStatusCode.RequestTimeout;
urlData.Updated = true;
urlData.RetrievalTime = (int)timer.Duration;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
urlData.UrlToCrawl = urlToCrawl;
try
{
receivedBytes.Close();
receiveStream.Close();
pageResponse.Close();
}
catch
{}
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
UpdateStatistics(HttpStatusCode.RequestTimeout, 0);
return;
}
byte []buffer=Encoding.ASCII.GetBytes(contents);
receiveStream.Close();
receivedBytes.Close();
UpdateStatistics(pageResponse.StatusCode, contents.Length);
string redirectUrl = string.Empty;
if (pageResponse.ResponseUri.AbsoluteUri!=urlToCrawl.Url)
{//now that was a bloody BUGBUG
redirectUrl = pageResponse.ResponseUri.AbsoluteUri;
urlData.RedirectedPriority = CleanupRedirectUrl(ref redirectUrl);
if(urlToCrawl.Url != redirectUrl)
{
urlData.Redirected=true;
urlToCrawl.Url=redirectUrl;
}
}
Parser parser = SelectParser(pageResponse.ContentType);
pageResponse.Close();
long CRC = CompressionUtils.BufferCRC(buffer);
if(CRC != urlToCrawl.CRC)
{
urlData.Updated = true;
urlToCrawl.CRC = CRC;
}
if(urlData.Updated)
{
urlData.RetrievalTime = (int)timer.Duration;
//if redirected, calculate robots, domain & priority for redirect url
if(urlData.Redirected)
{
urlData.RedirectedFlagDomain = domainFilter.FilterUrl(ref redirectUrl);
urlData.RedirectedFlagRobots = robotsFilter.FilterUrl(ref redirectUrl, ref urlToCrawl, RobotsMetaTagValue.NoMeta);
}
//perform link extraction and content extraction
ArrayList outlinks = null;
try
{
if((parser == htmlParser)||(parser == textParser))
{
urlData.Data = parser.ExtractContent(ref contents, false);
if(urlData.Data == null)
{
urlData.Data = String.Empty;
}
outlinks = parser.ExtractLinks(ref contents, ref urlToCrawl);
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
else
{
urlData.Data = parser.ExtractContent(buffer, false);
if(urlData.Data == null)
{
urlData.Data = String.Empty;
}
outlinks = parser.ExtractLinks(buffer, ref urlToCrawl);
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
}
catch
{
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
urlData.OutLinks = new InternetUrlToIndex[outlinks.Count];
for(int i = 0; i< outlinks.Count; i++)
{
urlData.OutLinks[i] = (InternetUrlToIndex)outlinks[i];
}
//finally update the urlData object with the modified UrlToCrawl
urlData.UrlToCrawl = urlToCrawl;
}
}
//lock and update CrawledUrls
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
}
catch(ThreadAbortException tae)
{
//The thread has been asked to abort. Log information and return at once
if(globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo(Thread.CurrentThread.Name + " has been asked to abort: " + tae.Message);
}
return;
}
catch(ThreadInterruptedException tie)
{
//The thread has been asked to join. Log information and return at once
if(globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo(Thread.CurrentThread.Name + " has been interrupted: " + tie.Message);
}
return;
}
catch(Exception ex)
{
if(!(ex is ThreadAbortException)) // the ThreadAbortedException is rethrown
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("CrawlUrl running in " + Thread.CurrentThread.Name + " encountered an unexpected exception: " + ex.ToString());
}
}
throw ex; //PerformCrawling should catch this one ?????
}
finally
{
GC.Collect();
}
}
/// <summary>
/// DetectConnectionSpeed attempts to detect the computer's internet connection speed
/// by measuring how much time it takes to download the contents of a web site.
/// </summary>
/// <returns>A <see cref="CWConnectionSpeed"/> value containing the estimated internet
/// connection speed.</returns>
public static CWConnectionSpeed DetectConnectionSpeed()
{
CWConnectionSpeed retVal = CWConnectionSpeed.Unknown;
try
{
WebClient client = new WebClient();
HiResTimer timer = new HiResTimer();
byte[] data;
timer.Start();
try
{
data = client.DownloadData("http://www.in.gr/");
}
catch
{
data = new byte[0];
}
finally
{
timer.Stop();
client.Dispose();
}
if (data.Length > 0)
{
double Kbps = ((data.Length * 8) / timer.Duration);
//determine which enumeration value fits best
if ((Kbps > 0) && (Kbps <= 56))
retVal = CWConnectionSpeed.Modem56K;
else if (Kbps <= 64)
retVal = CWConnectionSpeed.ISDN64K;
else if (Kbps <= 128)
retVal = CWConnectionSpeed.ISDN128K;
else if (Kbps <= 256)
retVal = CWConnectionSpeed.DSL256K;
else if (Kbps <= 512)
retVal = CWConnectionSpeed.DSL512K;
else if (Kbps <= 1024)
retVal = CWConnectionSpeed.DSL1M;
else if ((Kbps > 1024) && (Kbps <= 1536))
retVal = CWConnectionSpeed.T1;
else if ((Kbps > 1536) && (Kbps <= 46080))
retVal = CWConnectionSpeed.T3;
else if ((Kbps > 46080) && (Kbps < 158720))
retVal = CWConnectionSpeed.Fiber;
else if (Kbps >= 158720)
retVal = CWConnectionSpeed.ATM;
else
retVal = CWConnectionSpeed.Unknown;
}
}
catch
{ }
return retVal;
}
/// <summary>
/// Crawls a Url and creates a <see cref="UrlCrawlData"/> object that is stored in
/// the internal crawledUrls <see cref="ArrayList"/>. Since it runs in one of the
/// crawling threads that may be interrupted or aborted at any time it must be able
/// to handle ThreadAbortException and ThreadInterruptedException.
/// </summary>
/// <param name="urlToCrawl">A reference to the <see cref="InternetUrlToCrawl"/>
/// object that encapsulates the url that must be crawled.</param>
internal void CrawlUrl(ref InternetUrlToCrawl urlToCrawl)
{
try
{
UrlCrawlData urlData = new UrlCrawlData();
HiResTimer timer = new HiResTimer();
//create the web request and download the data
HttpWebRequest pageRequest = null;
try
{
pageRequest = (HttpWebRequest)HttpWebRequest.Create(urlToCrawl.Url);
}
catch
{
urlData.HttpStatusCode=HttpStatusCode.BadRequest;//TODO comment
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.OutLinks = new InternetUrlToIndex[0];
urlData.Data = String.Empty;
UpdateStatistics(HttpStatusCode.BadRequest, 0);
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
return;
}
pageRequest.UserAgent = globals.UserAgent;
pageRequest.Timeout=Backoff.DefaultBackoff; //page timeout = 30 seconds
pageRequest.KeepAlive = false;
HttpWebResponse pageResponse=null;
try
{
timer.Start();
pageResponse = (HttpWebResponse)pageRequest.GetResponse();
//the above line might throw either WebException or UriFormatException
}
catch(WebException we)
{
HttpWebResponse response=(HttpWebResponse)we.Response;
if (response!=null)
{
//although an exception occured we're able to get the Status Code
urlData.HttpStatusCode=response.StatusCode;
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
UpdateStatistics(response.StatusCode, response.ContentLength);
response.Close();
}
else
{
urlData.HttpStatusCode=HttpStatusCode.BadRequest;//TODO comment
urlData.Updated=true;
urlData.UrlToCrawl = urlToCrawl;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
UpdateStatistics(HttpStatusCode.BadRequest, 0);
}
}
catch(UriFormatException)
{
//this will occur if the url is not valid
urlData.HttpStatusCode=HttpStatusCode.BadRequest;
urlData.Updated=false;
urlData.Data = String.Empty;
urlData.UrlToCrawl = urlToCrawl;
urlData.OutLinks = new InternetUrlToIndex[0];
}
finally
{
timer.Stop();
urlData.TimeStamp = DateTime.UtcNow;
}
if(pageResponse !=null)
{
//update the fields
urlData.HttpStatusCode = pageResponse.StatusCode;
//download and parse the contents of the url
Stream receiveStream=pageResponse.GetResponseStream();
/*StreamReader receivedBytes=new StreamReader(receiveStream,defaultEncoding);*/
MemoryStream receivedBytes = new MemoryStream();
byte [] buffer = new byte[4096];
int read = 0;
try
{
while((read=receiveStream.Read(buffer,0,4096))>0)
{
receivedBytes.Write(buffer,0,read);
}
}
catch
{
//it should be response timeout not request timeout
urlData.HttpStatusCode = HttpStatusCode.RequestTimeout;
urlData.Updated = true;
urlData.RetrievalTime = (int)timer.Duration;
urlData.Data = String.Empty;
urlData.OutLinks = new InternetUrlToIndex[0];
urlData.UrlToCrawl = urlToCrawl;
try
{
receivedBytes.Close();
receiveStream.Close();
pageResponse.Close();
}
catch
{}
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
UpdateStatistics(HttpStatusCode.RequestTimeout, 0);
return;
}
buffer = receivedBytes.ToArray();
Parser parser = SelectParser(pageResponse.ContentType);
string contents = String.Empty;
if(parser == htmlParser)
{
Encoding encoding = null;
switch(pageResponse.ContentEncoding)
{
case "":
case "none":
contents = this.defaultEncoding.GetString(buffer, 0, buffer.Length);
//re-check the encoding
encoding = DetectContentEncoding(ref contents);
if(encoding != defaultEncoding)
{
contents = encoding.GetString(buffer, 0, buffer.Length);
}
break;
case "gzip":
case "x-gzip":
//first decompress the stream and then re-check the encoding
byte [] decompressed_buffer = new byte [0];
DecompressGzippedContent(buffer, out decompressed_buffer);
contents = this.defaultEncoding.GetString(decompressed_buffer, 0, decompressed_buffer.Length);
//re-check the encoding
encoding = DetectContentEncoding(ref contents);
if(encoding != defaultEncoding)
{
contents = encoding.GetString(decompressed_buffer, 0, decompressed_buffer.Length);
}
break;
default:
try
{
encoding = Encoding.GetEncoding(pageResponse.ContentEncoding);
contents = encoding.GetString(buffer, 0, buffer.Length);
}
catch//(NotSupportedException)
{
encoding = defaultEncoding;
//the encoding specified is unsupported.
contents = String.Empty;
}
break;
}
}
else
{
if(parser == textParser)
{
try
{
contents = this.defaultEncoding.GetString(buffer, 0, buffer.Length);
}
catch
{
//something went seriously wrong here. The crawler got a header that says the server is
//sending back a plain text document but for some reason we can't get the string contents.
contents = String.Empty;
}
}
}
receiveStream.Close();
receivedBytes.Close();
UpdateStatistics(pageResponse.StatusCode, buffer.Length);
string redirectUrl = string.Empty;
if (pageResponse.ResponseUri.AbsoluteUri!=urlToCrawl.Url)
{
redirectUrl = pageResponse.ResponseUri.AbsoluteUri;
urlData.RedirectedPriority = htmlParser.CleanUrlParams(ref redirectUrl);//CleanupRedirectUrl(ref redirectUrl);
if(urlToCrawl.Url != redirectUrl)//now that was a bloody BUGBUG
{
urlData.Redirected=true;
urlToCrawl.Url=redirectUrl;
}
}
pageResponse.Close();
long CRC = CompressionUtils.BufferCRC(buffer);
if(CRC != urlToCrawl.CRC)
{
urlData.Updated = true;
urlToCrawl.CRC = CRC;
}
if(urlData.Updated)
{
urlData.RetrievalTime = (int)timer.Duration;
//if redirected, calculate robots, domain & priority for redirect url
if(urlData.Redirected)
{
urlData.RedirectedFlagDomain = domainFilter.FilterUrl(ref redirectUrl);
urlData.RedirectedFlagRobots = robotsFilter.FilterUrl(redirectUrl, urlToCrawl, RobotsMetaTagValue.NoMeta);
}
//perform link extraction and content extraction
ArrayList outlinks = null;
try
{
if((parser == htmlParser)||(parser == textParser))
{
string clean = parser.ExtractContent(ref contents, false);
if(clean.Length>1048576)
{
clean = clean.Substring(0,1048576);
}
urlData.Data = InternetUtils.Base64Encode(clean);
if(urlData.Data == null)
{
urlData.Data = String.Empty;
}
outlinks = parser.ExtractLinks(ref contents, ref urlToCrawl);
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
else
{
contents = parser.ExtractContent(buffer, false);
if(contents.Length>1048576)
{
contents = contents.Substring(0,1048576);
}
urlData.Data = InternetUtils.Base64Encode(contents);
if(urlData.Data == null)
{
urlData.Data = String.Empty;
}
if(parser == pdfParser)
{
outlinks = textParser.ExtractLinks(ref contents, ref urlToCrawl);
}
else
{
outlinks = htmlParser.ExtractLinks(ref contents, ref urlToCrawl);
}
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
}
catch
{
if(outlinks == null)
{
outlinks = new ArrayList();
}
}
urlData.OutLinks = new InternetUrlToIndex[outlinks.Count];
for(int i = 0; i< outlinks.Count; i++)
{
urlData.OutLinks[i] = (InternetUrlToIndex)outlinks[i];
}
//finally update the urlData object with the modified UrlToCrawl
urlData.UrlToCrawl = urlToCrawl;
}
}
//lock and update CrawledUrls
lock(crawledUrls.SyncRoot)
{
crawledUrls.Add(urlData);
}
}
catch(ThreadAbortException tae)
{
//The thread has been asked to abort. Log information and return at once
if(globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo(Thread.CurrentThread.Name + " has been asked to abort: " + tae.Message);
}
return;
}
catch(ThreadInterruptedException tie)
{
//The thread has been asked to join. Log information and return at once
if(globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo(Thread.CurrentThread.Name + " has been interrupted: " + tie.Message);
}
return;
}
catch(Exception ex)
{
if(!(ex is ThreadAbortException)) // the ThreadAbortedException is rethrown
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("CrawlUrl running in " + Thread.CurrentThread.Name + " encountered an unexpected exception: " + ex.ToString());
}
}
throw ex; //PerformCrawling should catch this one ?????
}
finally
{
GC.Collect();
}
}
/// <summary>
/// DetectConnectionSpeed attempts to detect the computer's internet connection speed
/// by measuring how much time it takes to download the contents of a web site.
/// </summary>
/// <returns>A <see cref="CWConnectionSpeed"/> value containing the estimated internet
/// connection speed.</returns>
public static CWConnectionSpeed DetectConnectionSpeed()
{
CWConnectionSpeed retVal = CWConnectionSpeed.Unknown;
try
{
WebClient client = new WebClient();
HiResTimer timer = new HiResTimer();
byte[] data;
timer.Start();
try
{
data = client.DownloadData("http://www.in.gr/");
}
catch
{
data = new byte[0];
}
finally
{
timer.Stop();
client.Dispose();
}
if (data.Length > 0)
{
double Kbps = ((data.Length * 8) / timer.Duration);
//determine which enumeration value fits best
if ((Kbps > 0) && (Kbps <= 56))
{
retVal = CWConnectionSpeed.Modem56K;
}
else if (Kbps <= 64)
{
retVal = CWConnectionSpeed.ISDN64K;
}
else if (Kbps <= 128)
{
retVal = CWConnectionSpeed.ISDN128K;
}
else if (Kbps <= 256)
{
retVal = CWConnectionSpeed.DSL256K;
}
else if (Kbps <= 512)
{
retVal = CWConnectionSpeed.DSL512K;
}
else if (Kbps <= 1024)
{
retVal = CWConnectionSpeed.DSL1M;
}
else if ((Kbps > 1024) && (Kbps <= 1536))
{
retVal = CWConnectionSpeed.T1;
}
else if ((Kbps > 1536) && (Kbps <= 46080))
{
retVal = CWConnectionSpeed.T3;
}
else if ((Kbps > 46080) && (Kbps < 158720))
{
retVal = CWConnectionSpeed.Fiber;
}
else if (Kbps >= 158720)
{
retVal = CWConnectionSpeed.ATM;
}
else
{
retVal = CWConnectionSpeed.Unknown;
}
}
}
catch
{ }
return(retVal);
}
