package com.atlassian.bonnie.search.extractor;
   
   import org.apache.poi.hslf.model.TextRun;
   import org.apache.poi.hslf.record.*;
   import org.apache.poi.poifs.eventfilesystem.POIFSReaderEvent;
   import org.apache.poi.poifs.eventfilesystem.POIFSReaderListener;
   import org.apache.poi.poifs.filesystem.DocumentInputStream;
   import org.apache.poi.util.LittleEndian;
   import org.slf4j.Logger;
  import org.slf4j.LoggerFactory;
  
  import java.io.ByteArrayOutputStream;
  import java.util.Collections;
  import java.util.HashSet;
  import java.util.Iterator;
  import java.util.Set;
  
  /**
   * Listener for responding to read events thrown when reading a powerpoint document. This implementation is largely
   * based on {@link org.apache.poi.hslf.extractor.QuickButCruddyTextExtractor} (the {@link #findTextRecords(int,byte[])}
   * has been copied and adapted). This listener is able to extract unicode text.
   */
  @Deprecated
  public class PowerPointListener implements POIFSReaderListener
  {
  	private static Logger log = LoggerFactory.getLogger(PowerPointListener.class);
  	private final StringBuffer buffer;
  
  	public static final Set/*<String>*/ IGNORED_TEXT;
  	public static final Set/*<String>*/ IGNORED_PREFIXES;
  
  	static
  	{
  		Set/*<String>*/ tempIgnoredText = new HashSet/*<String>*/(9);
  		tempIgnoredText.add("*");
  		tempIgnoredText.add("Microsoft PowerPoint Presentation");
  		tempIgnoredText.add("Click to edit Master title style");
  		tempIgnoredText.add("Click to edit Master subtitle style");
  		tempIgnoredText.add("Click to edit Master text styles\nSecond level\nThird level\nFourth level\nFifth level");
  		tempIgnoredText.add("Default Design");
  		tempIgnoredText.add("Microsoft Excel Worksheet");
  		tempIgnoredText.add("Worksheet");
  		tempIgnoredText.add("MS Org Chart");
  
  		IGNORED_TEXT = Collections.unmodifiableSet(tempIgnoredText);
  
  		Set/*<String>*/ tempIgnoredPrefixes = new HashSet/*<String>*/(9);
  		tempIgnoredPrefixes.add("PowerPoint.Show.");
  		tempIgnoredPrefixes.add("Excel.Sheet.");
  		tempIgnoredPrefixes.add("___PPT");
  		tempIgnoredPrefixes.add("MS Organization Chart ");
  		tempIgnoredPrefixes.add("WordArt ");
  		tempIgnoredPrefixes.add("MSWordArt.");
  		tempIgnoredPrefixes.add("Microsoft WordArt ");
  		tempIgnoredPrefixes.add("Microsoft WordArt ");
  		tempIgnoredPrefixes.add("OrgPlusWOPX.");
  
  		IGNORED_PREFIXES = Collections.unmodifiableSet(tempIgnoredPrefixes);
  	}
  
  	public PowerPointListener(StringBuffer buff)
  	{
  		this.buffer = buff;
  	}
  
  	public void processPOIFSReaderEvent(POIFSReaderEvent event)
  	{
  		try
  		{
  			if ("PowerPoint Document".equals(event.getName()))
  			{
  				DocumentInputStream dis = event.getStream();
  				ByteArrayOutputStream bos = new ByteArrayOutputStream();
  				try
  				{
  					byte[] documentBytes = new byte[dis.available()];
  					int bytesRead = dis.read(documentBytes, 0, dis.available());
  
  					if (log.isDebugEnabled() && bytesRead != documentBytes.length)
  						log.debug("Bytes read from powerpoint file != bytes available");
  
  					int pos = 0;
  
  					do
  					{
  						pos = findTextRecords(pos, documentBytes);
  					}
  					while (pos != -1);
  				}
  
  
  				finally
  				{
  					dis.close();
  					bos.close();
  				}
  			}
  		}
  		catch (Exception e)
 		{
 			log.error("Error extracting PowerPoint text: " + e, e);
 		}
 	}
 
 
 	public int findTextRecords(int startPos, byte[] documentBytes)
 	{
 		// Grab the length, and the first option byte
 		// Note that the length doesn't include the 8 byte atom header
 		int len = (int) LittleEndian.getUInt(documentBytes, startPos + 4);
 		byte opt = documentBytes[startPos];
 
 		// If it's a container, step into it and return
 		// (If it's a container, option byte 1 BINARY_AND 0x0f will be 0x0f)
 		int container = (int) opt & 0x0f;
 		if (container == 0x0f)
 		{
 			return (startPos + 8);
 		}
 
 		// Otherwise, check the type to see if it's text
 		long type = LittleEndian.getUShort(documentBytes, startPos + 2);
 		TextRun textRun = null;
 
 		// TextBytesAtom
 		if (type == RecordTypes.TextBytesAtom.typeID)
 		{
 			TextBytesAtom tba = (TextBytesAtom) Record.createRecordForType(type, documentBytes, startPos, len + 8);
 			textRun = new TextRun((TextHeaderAtom) null, tba, (StyleTextPropAtom) null);
 		}
 		// TextCharsAtom
 		if (type == RecordTypes.TextCharsAtom.typeID)
 		{
 			TextCharsAtom tca = (TextCharsAtom) Record.createRecordForType(type, documentBytes, startPos, len + 8);
 			textRun = new TextRun((TextHeaderAtom) null, tca, (StyleTextPropAtom) null);
 		}
 
 		// CString (doesn't go via a TextRun)
 		if (type == RecordTypes.CString.typeID)
 		{
 			CString cs = (CString) Record.createRecordForType(type, documentBytes, startPos, len + 8);
 			appendText(cs.getText());
 		}
 
 		// If we found text via a TextRun, save it in the vector
 		if (textRun != null)
 		{
 			String textRunAsText = textRun.getText();
 			appendText(textRunAsText);
 		}
 
 		// Wind on by the atom length, and check we're not at the end
 		int newPos = (startPos + 8 + len);
 		if (newPos > (documentBytes.length - 8))
 		{
 			newPos = -1;
 		}
 		return newPos;
 	}
 
 	private void appendText(String textRunAsText)
 	{
 		if (log.isDebugEnabled()) log.debug(textRunAsText);
 
 		if (IGNORED_TEXT.contains(textRunAsText))
 			return;
 
 		for (Iterator iterator = IGNORED_PREFIXES.iterator(); iterator.hasNext();)
 		{
 			String prefix = (String) iterator.next();
 			if (textRunAsText.startsWith(prefix))
 				return;
 		}
 
 		buffer.append(textRunAsText);
 		buffer.append(" ");
 	}
 
 }