0 条评论

　　1、下载poi的jar包

　　　　下载地址：https://www.apache.org/dyn/closer.lua/poi/release/bin/poi-bin-3.17-20170915.tar.gz

　　　　

　　　　下载解压后用到的jar包，这些包都能在下载的文件夹中找到

　　　　

　　2、代码

//package com.word;
import java.io.*;

import java.io.File;
import java.io.FileInputStream;
import java.io.InputStream;

import org.apache.poi.POIXMLDocument;
import org.apache.poi.POIXMLTextExtractor;
import org.apache.poi.hwpf.extractor.WordExtractor;
import org.apache.poi.openxml4j.opc.OPCPackage;
import org.apache.poi.xwpf.extractor.XWPFWordExtractor;
/**
 * @Author：sks
 * @Description：
 * @Date：Created in 15:49 2018/1/24
 * @Modified by：
 **/
public class poi_doc {
    public static void main(String[] args)  {
        String filePath = "D:/work/Solr/ImportData/test.docx";
        String content = readWord(filePath);
    }

    public static String readWord(String path) {
        String buffer = "";
        try {
            if (path.endsWith(".doc")) {
                InputStream is = new FileInputStream(new File(path));
                WordExtractor ex = new WordExtractor(is);
                buffer = ex.getText();
                ex.close();
            } else if (path.endsWith("docx")) {
                OPCPackage opcPackage = POIXMLDocument.openPackage(path);
                POIXMLTextExtractor extractor = new XWPFWordExtractor(opcPackage);
                buffer = extractor.getText();
                extractor.close();
            } else {
                System.out.println("此文件不是word文件！");
            }

        } catch (Exception e) {
            e.printStackTrace();
        }

        return buffer;
    }


}