POI读取Word2003显示到网页

生成HTML:

不生成HTML

doc:

 

读取word2003所需jar包：

poi-scratchpad-3.10-FINAL.jar

poi-3.10-FINAL.jar

commons-io-1.4.jar

 

 

  <servlet>
      <servlet-name>print</servlet-name>
      <servlet-class>com.tpf.PoiWordToHtml</servlet-class>
  </servlet>
  <servlet-mapping>
      <servlet-name>print</servlet-name>
     <url-pattern>/print</url-pattern>  
  </servlet-mapping>

import java.io.ByteArrayOutputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.util.List;
import javax.servlet.ServletException;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.transform.OutputKeys;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.dom.DOMSource;
import javax.xml.transform.stream.StreamResult;
import org.apache.commons.io.FileUtils;
import org.apache.poi.hwpf.HWPFDocument;
import org.apache.poi.hwpf.converter.PicturesManager;
import org.apache.poi.hwpf.converter.WordToHtmlConverter;
import org.apache.poi.hwpf.usermodel.Picture;
import org.apache.poi.hwpf.usermodel.PictureType;
import org.apache.tomcat.util.codec.binary.Base64;
import org.w3c.dom.Document;
public class PoiWordToHtml extends HttpServlet {
     @Override
     protected void doPost(HttpServletRequest req,  HttpServletResponse resp) throws ServletException, IOException {
           doGet(req, resp);
     }
     
     @Override
     protected void doGet(HttpServletRequest req,  HttpServletResponse resp) throws ServletException, IOException {
           PoiWordToHtml poiWordToHtml = new PoiWordToHtml();
           try {
                poiWordToHtml.word2003ToHtml(req,resp);
           } catch (Exception e) {
                e.printStackTrace();
           }
     }
     
     final String path = "E:\\test\\";
     public static void main(String[] args) {
           try {
                PoiWordToHtml poiWordToHtml = new  PoiWordToHtml();
                poiWordToHtml.word2003ToHtml();
                poiWordToHtml.word2007ToHtml();
           } catch (Exception e) {
                e.printStackTrace();
           }
     }
     
    //doc
     public void word2003ToHtml(HttpServletRequest req,  HttpServletResponse resp) throws Exception{
           String file = "签名功能使用手册.doc";
           InputStream input = new FileInputStream(path + file);
           HWPFDocument wordDocument = new HWPFDocument(input);
          
           WordToHtmlConverter wordToHtmlConverter = new  WordToHtmlConverter(
                     DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument());
           //设置图片
           wordToHtmlConverter.setPicturesManager(new  PicturesManager() {
                public String savePicture(byte[] content,  PictureType pictureType, String suggestedName, float  widthInches,
                           float heightInches) {
                     return suggestedName;
                }
           });
           wordToHtmlConverter.processDocument(wordDocument);
           List pics =  wordDocument.getPicturesTable().getAllPictures();
           //图片显示方式1：生成图片文件并显示,图片会保存到指定的目录下
           /*if (pics != null) {
                for (int i = 0; i < pics.size(); i++) {
                     Picture pic = (Picture) pics.get(i);
                     try {
                           pic.writeImageContent(new  FileOutputStream(path + pic.suggestFullFileName()));
                     } catch (FileNotFoundException e) {
                           e.printStackTrace();
                     }
                }
           }*/
           Document htmlDocument =  wordToHtmlConverter.getDocument();
           ByteArrayOutputStream outStream = new  ByteArrayOutputStream();
           DOMSource domSource = new DOMSource(htmlDocument);
           StreamResult streamResult = new  StreamResult(outStream);
           TransformerFactory tf =  TransformerFactory.newInstance();
           Transformer serializer = tf.newTransformer();
           serializer.setOutputProperty(OutputKeys.ENCODING,  "utf-8");
           serializer.setOutputProperty(OutputKeys.INDENT,  "yes");
           serializer.setOutputProperty(OutputKeys.METHOD,  "html");
           serializer.transform(domSource, streamResult);
           outStream.close();
           String content = new String(outStream.toByteArray());
           //图片显示方式2：不生成图片文件，图片以base64形式在页面上展示出来
           if (pics != null) {
                for (int i = 0; i < pics.size(); i++) {
                     Picture pic = (Picture) pics.get(i);
                     byte picContent[] = pic.getContent();
                     byte picBase[] =  Base64.encodeBase64(picContent);
                     content =  content.replace("src=\""+pic.suggestFullFileName(),  "src=\"data:image/png;base64,"+new String(picBase)+"\"");
                }
           }
           System.out.println(content);
           resp.setCharacterEncoding("UTF-8");
           resp.setContentType("text/html;charset=UTF-8");
           //不生成HTML文件，直接返回数据到页面显示
           resp.getWriter().write(content);
           //生成指定的HTML文件
           //FileUtils.writeStringToFile(new File(path, "签名功能使用手册2003.html"), content, "utf-8");
     }     
}