internal void Stop()
{
UninterruptableMonitor.Enter(this);
try
{
stopped = true;
if (tuple != null)
{
tuple = null;
UninterruptableMonitor.Pulse(this); //Notify();
}
}
finally
{
UninterruptableMonitor.Exit(this);
}
}
public void Run()
{
try
{
Sax.IXMLReader reader = new TagSoup.Parser(); //XMLReaderFactory.createXMLReader();
reader.ContentHandler = this;
reader.ErrorHandler = this;
while (!stopped)
{
Stream localFileIS = outerInstance.@is;
if (localFileIS != null)
{ // null means fileIS was closed on us
try
{
// To work around a bug in XERCES (XERCESJ-1257), we assume the XML is always UTF8, so we simply provide reader.
reader.Parse(new InputSource(IOUtils.GetDecodingReader(localFileIS, Encoding.UTF8)));
}
catch (Exception ioe) when(ioe.IsIOException())
{
UninterruptableMonitor.Enter(outerInstance);
try
{
if (localFileIS != outerInstance.@is)
{
// fileIS was closed on us, so, just fall through
}
else
{
// Exception is real
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
}
}
finally
{
UninterruptableMonitor.Exit(outerInstance);
}
}
}
UninterruptableMonitor.Enter(this);
try
{
if (stopped || !outerInstance.m_forever)
{
nmde = new NoMoreDataException();
UninterruptableMonitor.Pulse(this); //notify();
return;
}
else if (localFileIS == outerInstance.@is)
{
// If file is not already re-opened then re-open it now
outerInstance.@is = outerInstance.OpenInputStream();
}
}
finally
{
UninterruptableMonitor.Exit(this);
}
}
}
catch (SAXException sae)
{
throw RuntimeException.Create(sae);
}
catch (Exception ioe) when(ioe.IsIOException())
{
throw RuntimeException.Create(ioe);
}
finally
{
UninterruptableMonitor.Enter(this);
try
{
threadDone = true;
UninterruptableMonitor.Pulse(this); //Notify();
}
finally
{
UninterruptableMonitor.Exit(this);
}
}
}
public override void EndElement(string @namespace, string simple, string qualified)
{
int elemType = GetElementType(qualified);
switch (elemType)
{
case PAGE:
// the body must be null and we either are keeping image docs or the
// title does not start with Image:
if (body != null && (outerInstance.keepImages || !title.StartsWith("Image:", StringComparison.Ordinal)))
{
string[] tmpTuple = new string[LENGTH];
tmpTuple[TITLE] = title.Replace('\t', ' ');
tmpTuple[DATE] = time.Replace('\t', ' ');
tmpTuple[BODY] = Regex.Replace(body, "[\t\n]", " ");
tmpTuple[ID] = id;
UninterruptableMonitor.Enter(this);
try
{
while (tuple != null && !stopped)
{
try
{
UninterruptableMonitor.Wait(this); //wait();
}
catch (System.Threading.ThreadInterruptedException ie)
{
throw new Util.ThreadInterruptedException(ie);
}
}
tuple = tmpTuple;
UninterruptableMonitor.Pulse(this); //notify();
}
finally
{
UninterruptableMonitor.Exit(this);
}
}
break;
case BODY:
body = contents.ToString();
//workaround that startswith doesn't have an ignore case option, get at least 10 chars.
string startsWith = body.Substring(0, Math.Min(10, contents.Length) - 0).ToLowerInvariant();
if (startsWith.StartsWith("#redirect", StringComparison.Ordinal))
{
body = null;
}
break;
case DATE:
time = Time(contents.ToString());
break;
case TITLE:
title = contents.ToString();
break;
case ID:
//the doc id is the first one in the page. All other ids after that one can be ignored according to the schema
if (id == null)
{
id = contents.ToString();
}
break;
default:
// this element should be discarded.
break;
}
}
