/// <summary>
/// Raises an ExtractContentComplete event when the extraction of content is complete
/// </summary>
/// <param name="e">The <see cref="ParserEventArgs"/> related to the event.</param>
private void OnExtractContentComplete(ParserEventArgs e)
{
if (ExtractContentComplete != null)
{
ExtractContentComplete(this, e);
}
}
/// <summary>
/// Extracts links from the contents of a document.
/// </summary>
/// <param name="content">The contents of the document.</param>
/// <param name="contentUrl">The url of the document.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty ArrayList.</remarks>
public override ArrayList ExtractLinks(byte[] content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links = new ArrayList();
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
return(links);
}
/// <summary>
/// Performs the extraction of content from a document.
/// </summary>
/// <param name="content">
/// The contents of the document from which the content must be extracted.
/// </param>
/// <param name="Flag">The parameter is not used in this method.</param>
/// <returns>A string containing the desired extracted content.</returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty string.</remarks>
public override string ExtractContent(byte[] content, bool Flag)
{
string retVal = String.Empty;
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractContentComplete(e);
return(retVal);
}
/// <summary>
/// Extracts text from the contents of a document.
/// </summary>
/// <param name="content">The contents of the document.</param>
/// <returns>The text extracted from the document.</returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty string.</remarks>
public override string ExtractText(ref string content)
{
string retVal = String.Empty;
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return(retVal);
}
/// <summary>
/// Performs the extraction of text from a text document. The text is extracted by
/// compacting consecutive white space characters.
/// </summary>
/// <param name="content">
/// The contents of the document from which the text must be extracted. Passes by
/// reference in order to reduce memory consumption.
/// </param>
/// <returns>A string containing the 'clean' text extracted from the document.</returns>
public override string ExtractText(ref string content)
{
string retVal = String.Empty;
StringBuilder sb = new StringBuilder(content);
sb.Replace("\r\n", " ");
sb.Replace('\n', ' ');
sb.Replace('\t', ' ');
retVal = spacesRegex.Replace(sb.ToString(), " ");
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return(retVal);
}
private void crawler_UrlProcessed(object sender, ParserEventArgs e)
{
log.LogInfo("Crawler processed url " + e.Url);
}
/// <summary>
/// Extracts the desired contents of a document.
/// </summary>
/// <param name="content">The contents of the document.</param>
/// <param name="Flag">The parameter is not used in this method.</param>
/// <returns>The contents extracted from the document.</returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty string.</remarks>
public override string ExtractContent(ref string content, bool Flag)
{
string retVal = String.Empty;
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractContentComplete(e);
return retVal;
}
/// <summary>
/// Extracts text from the contents of a document.
/// </summary>
/// <param name="content">The contents of the document.</param>
/// <returns>The text extracted from the document.</returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty string.</remarks>
public override string ExtractText(byte[] content)
{
string retVal = String.Empty;
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return retVal;
}
/// <summary>
/// Extracts links from the contents of a document.
/// </summary>
/// <param name="content">The contents of the document.</param>
/// <param name="contentUrl">The url of the document.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
/// <remarks>This method <b>ALWAYS</b> returns an empty ArrayList.</remarks>
public override ArrayList ExtractLinks(byte[] content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links=new ArrayList();
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
return links;
}
/// <summary>
/// Extracts text from the contents of a SWF document.
/// </summary>
/// <param name="content">The contents of the SWF document.</param>
/// <returns>The text extracted from the SWF document.</returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
/// <remarks>
/// </remarks>
public override string ExtractText(byte[] content)
{
string retVal = String.Empty;
if((content==null)||(content.Length==0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string swfFileName = FileName + ".swf";
string htmFileName = FileName + ".htm";
FileStream swf = null;
StreamReader htm = null;
try
{
//store the swf file
swf = new FileStream(swfFileName,FileMode.Create);
swf.Write(content, 0, content.Length);
swf.Close();
swf = null;
//convert it to html
bool success = converter.ConvertSwfFile(swfFileName, htmFileName);
if(success)
{
htm = new StreamReader(htmFileName, encoding);
string html = htm.ReadToEnd();
htm.Close();
htm = null;
retVal = parser.ExtractText(ref html);
}
}
catch(Exception ex)
{
if(swf!=null)
{
try
{
swf.Close();
}
catch
{}
}
if(htm!=null)
{
try
{
htm.Close();
}
catch
{}
}
if(globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("SwfParser failed to extract text: " + ex.ToString());
}
}
finally
{
File.Delete(swfFileName);
File.Delete(htmFileName);
}
}
catch(Exception ex)
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("SwfParser failed to extract text: " + ex.Message);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return retVal;
}
/// <summary>
/// Performs the extraction of text from a text document. The text is extracted by
/// compacting consecutive white space characters.
/// </summary>
/// <param name="content">
/// The contents of the document from which the text must be extracted. Passes by
/// reference in order to reduce memory consumption.
/// </param>
/// <returns>A string containing the 'clean' text extracted from the document.</returns>
public override string ExtractText(ref string content)
{
string retVal = String.Empty;
StringBuilder sb=new StringBuilder(content);
sb.Replace("\r\n", " ");
sb.Replace('\n', ' ');
sb.Replace('\t', ' ');
retVal = spacesRegex.Replace(sb.ToString()," ");
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return retVal;
}
/// <summary>
/// Performs the extraction of links from a text document. It can extract simple
/// links that are separated from the rest of the text using spaces or line brakes
/// or any other delimiters. The results are returned as an <see cref="ArrayList"/>
/// of <see cref="InternetUrlToIndex"/> objects.
/// </summary>
/// <remarks>
/// Besides the parsing and extraction of Urls, ExtractLinks also performs other
/// tasks as well, such as:<br/>
/// <list type="bullet">
/// <item>
/// <description>Filtering of urls to resources of unsupported content-type, e.g. css, images, etc.</description>
/// </item>
/// <item>
/// <description>Filtering of multimple links to the same url and to the document itself.</description>
/// </item>
/// <item>
/// <description>Filtering of session id variables in dynamic Urls and limiting
/// of the number of GET variables in dynamic Urls.</description>
/// </item>
/// <item>
/// <description>Flagging of Urls according to their country domain.</description>
/// </item>
/// </list>
/// <b>Update History</b>
/// <list type="table">
/// <listheader>
/// <term>Date</term>
/// <description>Description</description>
/// </listheader>
/// <item>
/// <term>15/09/04</term>
/// <description>First release. A lot more needs to be done.</description>
/// </item>
/// </list>
/// </remarks>
/// <param name="content">The text that must be parsed for links. It is passed by
/// reference in order to reduce memory consumption.</param>
/// <param name="contentUrl">The Url from which the content comes.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
public override ArrayList ExtractLinks(ref string content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links = new ArrayList();
// It is important to notice that if the FlagFetchRobots of the contentUrl is
// true then the TextParser must remember this value because during the Robots
// Filtering it will become false so as not to download the robots.txt file
// every time a Url must be filtered.
//bool FlagFetchRobots = contentUrl.FlagFetchRobots;
try
{
//make sure only one thread will parse contents at a time.
//mutex.WaitOne();
if (contentUrl.FlagDomain != DomainFlagValue.MustVisit)
{
contentUrl.FlagDomain = ExtractDomainFlag(ref content);
if (contentUrl.FlagDomain != DomainFlagValue.MustVisit)
{
if (InternetUtils.HostName(contentUrl).Contains("ebay.com"))
{
contentUrl.FlagDomain = DomainFlagValue.MustVisit;
}
}
}
//perform the hyperlink matching
MatchCollection matches = hrefRegex.Matches(content);
if (matches.Count > 0)
{
string documentUrl = contentUrl.Url;
string baseUrl = BaseUrl(ref documentUrl);
byte priority = 0;
foreach (Match m in matches)
{
try
{
string url = m.Value.Trim();
url = NormalizeUrl(ref url, ref baseUrl);
priority = CleanUrlParams(ref url);
if (FilterUrl(ref url, ref documentUrl))
{
InternetUrlToIndex iurl = new InternetUrlToIndex(url);
iurl.Priority = priority;
iurl.FlagDomain = domainFilter.FilterUrl(ref url);
//[mod 24/2/05] No robots.txt checking is performed for non-greek urls
if (iurl.FlagDomain == DomainFlagValue.MustVisit)
{
iurl.FlagRobots = robotsFilter.FilterUrl(url, contentUrl, RobotsMetaTagValue.NoMeta);
}
else
{
iurl.FlagRobots = false;
}
if (!links.Contains(iurl))
{
links.Add(iurl);
}
}
}
catch
{
if (globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo("TextParser failed to parse " + m.Value);
}
continue;
}
}
}
}
catch (Exception ex)
{
if (globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning(ex.Message);
}
}
finally
{
//mutex.ReleaseMutex();
}
//contentUrl.FlagFetchRobots = FlagFetchRobots;
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
links.TrimToSize();
return(links);
}
/// <summary>
/// Raises the <see cref="UrlProcessed"/> event
/// </summary>
/// <param name="e">The <see cref="ParserEventArgs"/> related to the event.</param>
private void OnUrlProcessed(ParserEventArgs e)
{
if(UrlProcessed != null)
{
UrlProcessed(this, e);
}
}
/// <summary>
/// Extracts text from the contents of a PDF document.
/// </summary>
/// <param name="content">The contents of the PDF document.</param>
/// <returns>The text extracted from the PDF document.</returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
/// <remarks>
/// </remarks>
public override string ExtractText(byte[] content)
{
string retVal = String.Empty;
if((content==null)||(content.Length==0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string pdfFileName = FileName + ".pdf";
string txtFileName = FileName + ".txt";
FileStream pdf = null;
StreamReader txt = null;
try
{
//store the pdf file
pdf = new FileStream(pdfFileName,FileMode.Create);
pdf.Write(content, 0, content.Length);
pdf.Close();
pdf = null;
bool success = false;
//convert it to text
try
{
converter.loadFile(pdfFileName);
converter.convertToTextFile(1, converter.numPages, txtFileName);
success = true;
}
catch
{
success = false;
}
finally
{
converter.closeFile();
}
if(success)
{
txt = new StreamReader(txtFileName, encoding);
string text = txt.ReadToEnd();
txt.Close();
txt = null;
retVal = parser.ExtractText(ref text);
}
else
{
txt = null;
}
}
catch(Exception ex)
{
if(pdf!=null)
{
try
{
pdf.Close();
}
catch
{}
}
if(txt!=null)
{
try
{
txt.Close();
}
catch
{}
}
if(globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("PdfParser failed to extract text: " + ex.ToString());
}
}
finally
{
File.Delete(pdfFileName);
File.Delete(txtFileName);
}
}
catch(Exception ex)
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("PdfParser failed to extract text: " + ex.Message);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractTextComplete(e);
return retVal;
}
/// <summary>
/// Performs the extraction of content from a SWF document. Depending on the value
/// of the Flag provided it simply returns a string containing the HTML format of
/// the input or it returns the text format of the input after performing a white
/// space compaction.
/// </summary>
/// <param name="content">
/// The contents of the document from which the content must be extracted.
/// </param>
/// <param name="Flag">Determines what kind of processing will be performed on the
/// input. If set to false it simply returns a string containing the HTML format of
/// the input. If set to true it returns the text format of the input after performing
/// a white space compaction.
/// </param>
/// <returns>A string containing the desired extracted content.</returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
public override string ExtractContent(byte[] content, bool Flag)
{
string retVal = null;
if ((content == null) || (content.Length == 0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string swfFileName = FileName + ".swf";
string htmFileName = FileName + ".htm";
FileStream swf = null;
StreamReader htm = null;
try
{
//store the swf file
swf = new FileStream(swfFileName, FileMode.Create);
swf.Write(content, 0, content.Length);
swf.Close();
swf = null;
//convert it to html
bool success = converter.ConvertSwfFile(swfFileName, htmFileName);
if (success)
{
htm = new StreamReader(htmFileName, encoding);
string html = htm.ReadToEnd();
htm.Close();
htm = null;
if (!Flag)
{
retVal = html;
}
else
{
retVal = parser.ExtractText(ref html);
}
}
}
catch (Exception ex)
{
if (swf != null)
{
try
{
swf.Close();
}
catch
{}
}
if (htm != null)
{
try
{
htm.Close();
}
catch
{}
}
if (globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("SwfParser failed to extract text: " + ex.ToString());
}
}
finally
{
File.Delete(swfFileName);
File.Delete(htmFileName);
}
}
catch (Exception ex)
{
if (globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("SwfParser failed to extract content: " + ex.Message);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractContentComplete(e);
return(retVal);
}
/// <summary>
/// Performs the extraction of content from a PDF document. Depending on the value
/// of the Flag provided it simply returns a string same as the text produced from
/// the parsing of the PDF document or it removes consecutive whitespace characters
/// in order to perform a compaction.
/// </summary>
/// <param name="content">
/// The contents of the document from which the content must be extracted.
/// </param>
/// <param name="Flag">Determines what kind of processing will be performed on the
/// input. If set to false it simply returns a string same as the text produced from
/// the parsing of the PDF document. If set to true it removes consecutive white
/// space characters in order to perform a compaction.
/// </param>
/// <returns>A string containing the desired extracted content.</returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
public override string ExtractContent(byte[] content, bool Flag)
{
string retVal = String.Empty;
if ((content == null) || (content.Length == 0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string pdfFileName = FileName + ".pdf";
string txtFileName = FileName + ".txt";
FileStream pdf = null;
StreamReader txt = null;
try
{
//store the pdf file
pdf = new FileStream(pdfFileName, FileMode.Create);
pdf.Write(content, 0, content.Length);
pdf.Close();
pdf = null;
//convert it to text
int success = converter.ConvertPdf2Text(pdfFileName, txtFileName);
if (success == 0)
{
txt = new StreamReader(txtFileName, encoding);
if (!Flag)
{
retVal = txt.ReadToEnd();
}
else
{
string text = txt.ReadToEnd();
retVal = parser.ExtractText(ref text);
}
txt.Close();
txt = null;
}
}
catch (Exception ex)
{
if (pdf != null)
{
try
{
pdf.Close();
}
catch
{}
}
if (txt != null)
{
try
{
txt.Close();
}
catch
{}
}
if (globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("PdfParser failed to extract content: " + ex.ToString());
}
}
finally
{
File.Delete(pdfFileName);
File.Delete(txtFileName);
}
}
catch (Exception ex)
{
if (globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("PdfParser failed to extract content: " + ex.Message);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(String.Empty);
OnExtractContentComplete(e);
return(retVal);
}
/// <summary>
/// Extracts links from the contents of a PDF document.
/// </summary>
/// <param name="content">The contents of the PDF document.</param>
/// <param name="contentUrl">The url of the PDF document.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
public override ArrayList ExtractLinks(byte[] content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links=null;
if((content==null)||(content.Length==0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string pdfFileName = FileName + ".pdf";
string txtFileName = FileName + ".txt";
FileStream pdf = null;
StreamReader txt = null;
try
{
//store the pdf file
pdf = new FileStream(pdfFileName,FileMode.Create);
pdf.Write(content, 0, content.Length);
pdf.Close();
pdf = null;
//convert it to text
int success = converter.ConvertPdf2Text(pdfFileName, txtFileName);
if(success==0)
{
txt = new StreamReader(txtFileName, encoding);
string text = txt.ReadToEnd();
txt.Close();
txt = null;
links = parser.ExtractLinks(ref text, ref contentUrl);
}
else
{
txt.Close();
txt = null;
}
}
catch(Exception ex)
{
if(pdf!=null)
{
try
{
pdf.Close();
}
catch
{}
}
if(txt!=null)
{
try
{
txt.Close();
}
catch
{}
}
if(globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("PdfParser failed to extract links from " + contentUrl.Url + ": " + ex.ToString());
}
}
finally
{
File.Delete(pdfFileName);
File.Delete(txtFileName);
}
}
catch
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("PdfParser failed to extract links from " + contentUrl.Url);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
return links;
}
/// <summary>
/// Raises an ExtractTextComplete event when the extraction of text is complete
/// </summary>
/// <param name="e">The <see cref="ParserEventArgs"/> related to the event.</param>
private void OnExtractTextComplete(ParserEventArgs e)
{
if(ExtractTextComplete!=null)
{
ExtractTextComplete(this, e);
}
}
/// <summary>
/// Performs the extraction of links from a text document. It can extract simple
/// links that are separated from the rest of the text using spaces or line brakes
/// or any other delimiters. The results are returned as an <see cref="ArrayList"/>
/// of <see cref="InternetUrlToIndex"/> objects.
/// </summary>
/// <remarks>
/// Besides the parsing and extraction of Urls, ExtractLinks also performs other
/// tasks as well, such as:<br/>
/// <list type="bullet">
/// <item>
/// <description>Filtering of urls to resources of unsupported content-type, e.g. css, images, etc.</description>
/// </item>
/// <item>
/// <description>Filtering of multimple links to the same url and to the document itself.</description>
/// </item>
/// <item>
/// <description>Filtering of session id variables in dynamic Urls and limiting
/// of the number of GET variables in dynamic Urls.</description>
/// </item>
/// <item>
/// <description>Flagging of Urls according to their country domain.</description>
/// </item>
/// </list>
/// <b>Update History</b>
/// <list type="table">
/// <listheader>
/// <term>Date</term>
/// <description>Description</description>
/// </listheader>
/// <item>
/// <term>15/09/04</term>
/// <description>First release. A lot more needs to be done.</description>
/// </item>
/// </list>
/// </remarks>
/// <param name="content">The text that must be parsed for links. It is passed by
/// reference in order to reduce memory consumption.</param>
/// <param name="contentUrl">The Url from which the content comes.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
public override ArrayList ExtractLinks(ref string content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links = new ArrayList();
// It is important to notice that if the FlagFetchRobots of the contentUrl is
// true then the TextParser must remember this value because during the Robots
// Filtering it will become false so as not to download the robots.txt file
// every time a Url must be filtered.
//bool FlagFetchRobots = contentUrl.FlagFetchRobots;
try
{
//make sure only one thread will parse contents at a time.
//mutex.WaitOne();
if(contentUrl.FlagDomain!=DomainFlagValue.MustVisit)
{
contentUrl.FlagDomain = ExtractDomainFlag(ref content);
if (contentUrl.FlagDomain != DomainFlagValue.MustVisit)
if (InternetUtils.HostName(contentUrl).Contains("ebay.com"))
contentUrl.FlagDomain = DomainFlagValue.MustVisit;
}
//perform the hyperlink matching
MatchCollection matches = hrefRegex.Matches(content);
if(matches.Count>0)
{
string documentUrl = contentUrl.Url;
string baseUrl = BaseUrl(ref documentUrl);
byte priority = 0;
foreach(Match m in matches)
{
try
{
string url = m.Value.Trim();
url = NormalizeUrl(ref url, ref baseUrl);
priority = CleanUrlParams(ref url);
if(FilterUrl(ref url, ref documentUrl))
{
InternetUrlToIndex iurl = new InternetUrlToIndex(url);
iurl.Priority = priority;
iurl.FlagDomain = domainFilter.FilterUrl(ref url);
//[mod 24/2/05] No robots.txt checking is performed for non-greek urls
if(iurl.FlagDomain == DomainFlagValue.MustVisit)
{
iurl.FlagRobots = robotsFilter.FilterUrl(url, contentUrl, RobotsMetaTagValue.NoMeta);
}
else
{
iurl.FlagRobots = false;
}
if(!links.Contains(iurl))
{
links.Add(iurl);
}
}
}
catch
{
if(globals.Settings.LogLevel == CWLogLevel.LogInfo)
{
globals.FileLog.LogInfo("TextParser failed to parse " + m.Value);
}
continue;
}
}
}
}
catch(Exception ex)
{
if(globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning(ex.Message);
}
}
finally
{
//mutex.ReleaseMutex();
}
//contentUrl.FlagFetchRobots = FlagFetchRobots;
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
links.TrimToSize();
return links;
}
private void crawler_UrlProcessed(object sender, ParserEventArgs e)
{
log.LogInfo("Crawler processed url " + e.Url);
}
/// <summary>
/// Extracts links from the contents of a PDF document.
/// </summary>
/// <param name="content">The contents of the PDF document.</param>
/// <param name="contentUrl">The url of the PDF document.</param>
/// <returns>
/// An <see cref="ArrayList"/> of <see cref="InternetUrlToIndex"/> objects, one for
/// each link found in the content.
/// </returns>
/// <exception cref="ArgumentNullException">If the input buffer is null or empty.</exception>
public override ArrayList ExtractLinks(byte[] content, ref InternetUrlToCrawl contentUrl)
{
ArrayList links = null;
if ((content == null) || (content.Length == 0))
{
throw new ArgumentNullException("content", "The input buffer cannot be empty or null.");
}
try
{
mutex.WaitOne();
string FileName = globals.AppWorkPath + Guid.NewGuid().ToString();
string pdfFileName = FileName + ".pdf";
string txtFileName = FileName + ".txt";
FileStream pdf = null;
StreamReader txt = null;
try
{
//store the pdf file
pdf = new FileStream(pdfFileName, FileMode.Create);
pdf.Write(content, 0, content.Length);
pdf.Close();
pdf = null;
bool success = false;
//convert it to text
try
{
converter.loadFile(pdfFileName);
converter.convertToTextFile(1, converter.numPages, txtFileName);
success = true;
}
catch
{
success = false;
}
finally
{
converter.closeFile();
}
if (success)
{
txt = new StreamReader(txtFileName, encoding);
string text = txt.ReadToEnd();
txt.Close();
txt = null;
links = parser.ExtractLinks(ref text, ref contentUrl);
}
else
{
txt = null;
}
}
catch (Exception ex)
{
if (pdf != null)
{
try
{
pdf.Close();
}
catch
{}
}
if (txt != null)
{
try
{
txt.Close();
}
catch
{}
}
if (globals.Settings.LogLevel <= CWLogLevel.LogInfo)
{
globals.FileLog.LogWarning("PdfParser failed to extract links from " + contentUrl.Url + ": " + ex.ToString());
}
}
finally
{
File.Delete(pdfFileName);
File.Delete(txtFileName);
}
}
catch
{
if (globals.Settings.LogLevel <= CWLogLevel.LogWarning)
{
globals.FileLog.LogWarning("PdfParser failed to extract links from " + contentUrl.Url);
}
}
finally
{
GC.Collect();
mutex.ReleaseMutex();
}
ParserEventArgs e = new ParserEventArgs(contentUrl.Url);
OnExtractLinksComplete(e);
return(links);
}
