lucene入门-解析word文档

下载:

http://mirrors.ibiblio.org/pub/mirrors/maven2/org/textmining/tm-extractors/0.4/

 

java代码如下：

 

 package extract;

import java.io.*;
import org.textmining.text.extraction.WordExtractor;

public class ExtractorWord {

 /**
  * @param args
  */
 public static String getText(String file){
  String s="";
  String wordfile=file;
  WordExtractor extractor=null;
  try {
   FileInputStream in=new FileInputStream(new File(wordfile));
   extractor=new WordExtractor();
   s=extractor.extractText(in);
  } catch (IOException e) {   
   // TODO Auto-generated catch block
   e.printStackTrace();
  } catch (Exception e) {
  // TODO Auto-generated catch block
      e.printStackTrace();
     } 
  return s;     
    }
 public static void toTextFile(String doc,String filename) throws Exception{
  String s="";
  String wordfile=doc;
  String txtfile=filename;
  WordExtractor extractor=null;
  try {
            s=getText(wordfile);
   PrintWriter pw=new PrintWriter(new FileWriter(new File(filename)));
   pw.write(s);
   pw.flush();
   pw.close();
   System.out.print("成功写入文件！");
  } catch (IOException e) {   
   // TODO Auto-generated catch block
   e.printStackTrace();
  }  
 }
 /**
  * @param args
  */
 public static void main(String[] args) {
  // TODO Auto-generated method stub
  try {
   String sc=getText("D:/workspace/testsearch2/htmls/ddd.doc");   
   System.out.print(sc);
   toTextFile("D:/workspace/testsearch2/htmls/ddd.doc","D:/workspace/testsearch2/htmls/ddd.txt");
  } catch (Exception e) {
   // TODO Auto-generated catch block
   e.printStackTrace();
  }
  
 }

}
效果如下: