public WordExtractor(InputStream is) throws IOException 
{
		this( HWPFDocument.verifyAndBuildPOIFS(is) );
}

Create a new Word Extractor

Parameters:

is - InputStream containing the word file

 public WordExtractor(POIFSFileSystem fs) throws IOException 
{
		this(new HWPFDocument(fs));
		this.fs = fs;
}

Create a new Word Extractor

Parameters:

fs - POIFSFileSystem containing the word file

 public WordExtractor(HWPFDocument doc) throws IOException 
{
		super(doc);
		this.doc = doc;
}

Create a new Word Extractor

Parameters:

doc - The HWPFDocument to extract from

 public String[] getParagraphText() 
{
		String[] ret;
		// Extract using the model code
		try {
	    	Range r = doc.getRange();
			ret = new String[r.numParagraphs()];
			for(int i=0; i< ret.length; i++) {
				Paragraph p = r.getParagraph(i);
				ret[i] = p.text();
				// Fix the line ending
				if(ret[i].endsWith("\r")) {
					ret[i] = ret[i] + "\n";
				}
			}
		} catch(Exception e) {
			// Something's up with turning the text pieces into paragraphs
			// Fall back to ripping out the text pieces
			ret = new String[1];
			ret[0] = getTextFromPieces();
		}
		return ret;
}

Get the text from the word file, as an array with one String per paragraph

 public String getText() 
{
		StringBuffer ret = new StringBuffer();
		String[] text = getParagraphText();
		for(int i=0; i< text.length; i++) {
			ret.append(text[i]);
		}
		return ret.toString();
}

Grab the text, based on the paragraphs. Shouldn't include any crud, but slightly slower than getTextFromPieces().

 public String getTextFromPieces() 
{
    	StringBuffer textBuf = new StringBuffer();
    	Iterator textPieces = doc.getTextTable().getTextPieces().iterator();
    	while (textPieces.hasNext()) {
    		TextPiece piece = (TextPiece) textPieces.next();
    		String encoding = "Cp1252";
    		if (piece.usesUnicode()) {
    			encoding = "UTF-16LE";
    		}
    		try {
    			String text = new String(piece.getRawBytes(), encoding);
    			textBuf.append(text);
    		} catch(UnsupportedEncodingException e) {
    			throw new InternalError("Standard Encoding " + encoding + " not found, JVM broken");
    		}
    	}
    	String text = textBuf.toString();
    	// Fix line endings (Note - won't get all of them
    	text = text.replaceAll("\r\r\r", "\r\n\r\n\r\n");
    	text = text.replaceAll("\r\r", "\r\n\r\n");
    	if(text.endsWith("\r")) {
    		text += "\n";
    	}
    	return text;
}

Grab the text out of the text pieces. Might also include various bits of crud, but will work in cases where the text piece -> paragraph mapping is broken. Fast too.

 public static  void main(String[] args) throws IOException 
{
		if(args.length == 0) {
			System.err.println("Use:");
			System.err.println("   java org.apache.poi.hwpf.extractor.WordExtractor < filename >");
			System.exit(1);
		}
		// Process the first argument as a file
		FileInputStream fin = new FileInputStream(args[0]);
		WordExtractor extractor = new WordExtractor(fin);
		System.out.println(extractor.getText());
}

Command line extractor, so people will stop moaning that they can't just run this.