java使用POI实现html和word相互转换

项目后端使用了springboot，maven，前端使用了ckeditor富文本编辑器。目前从html转换的word为doc格式，而图片处理支持的是docx格式，所以需要手动把doc另存为docx，然后才可以进行图片替换。

一.添加maven依赖

主要使用了以下和poi相关的依赖，为了便于获取html的图片元素，还使用了jsoup：

<dependency>
  <groupid>org.apache.poi</groupid>
  <artifactid>poi</artifactid>
  <version>3.14</version>
</dependency>

<dependency>
  <groupid>org.apache.poi</groupid>
  <artifactid>poi-scratchpad</artifactid>
  <version>3.14</version>
</dependency>

<dependency>
  <groupid>org.apache.poi</groupid>
  <artifactid>poi-ooxml</artifactid>
  <version>3.14</version>
</dependency>

<dependency>
  <groupid>fr.opensagres.xdocreport</groupid>
  <artifactid>xdocreport</artifactid>
  <version>1.0.6</version>
</dependency>

<dependency>
  <groupid>org.apache.poi</groupid>
  <artifactid>poi-ooxml-schemas</artifactid>
  <version>3.14</version>
</dependency>

<dependency>
  <groupid>org.apache.poi</groupid>
  <artifactid>ooxml-schemas</artifactid>
  <version>1.3</version>
</dependency>

<dependency>
  <groupid>org.jsoup</groupid>
  <artifactid>jsoup</artifactid>
  <version>1.11.3</version>
</dependency>

二.word转换为html

在springboot项目的resources目录下新建static文件夹，将需要转换的word文件temp.docx粘贴进去，由于static是springboot的默认资源文件，所以不需要在配置文件里面另行配置了，如果改成其他名字，需要在application.yml进行相应配置。

doc格式转换为html：

public static string doctohtml() throws exception {
  file path = new file(resourceutils.geturl("classpath:").getpath());
  string imagepathstr = path.getabsolutepath() + "\\static\\image\\";
  string sourcefilename = path.getabsolutepath() + "\\static\\test.doc";
  string targetfilename = path.getabsolutepath() + "\\static\\test2.html";
  file file = new file(imagepathstr);
  if(!file.exists()) {
    file.mkdirs();
  }
  hwpfdocument worddocument = new hwpfdocument(new fileinputstream(sourcefilename));
  org.w3c.dom.document document = documentbuilderfactory.newinstance().newdocumentbuilder().newdocument();
  wordtohtmlconverter wordtohtmlconverter = new wordtohtmlconverter(document);
  //保存图片，并返回图片的相对路径
  wordtohtmlconverter.setpicturesmanager((content, picturetype, name, width, height) -> {
    try (fileoutputstream out = new fileoutputstream(imagepathstr + name)) {
      out.write(content);
    } catch (exception e) {
      e.printstacktrace();
    }
    return "image/" + name;
  });
  wordtohtmlconverter.processdocument(worddocument);
  org.w3c.dom.document htmldocument = wordtohtmlconverter.getdocument();
  domsource domsource = new domsource(htmldocument);
  streamresult streamresult = new streamresult(new file(targetfilename));
  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);
  return targetfilename;
}

docx格式转换为html

public static string docxtohtml() throws exception {
  file path = new file(resourceutils.geturl("classpath:").getpath());
  string imagepath = path.getabsolutepath() + "\\static\\image";
  string sourcefilename = path.getabsolutepath() + "\\static\\test.docx";
  string targetfilename = path.getabsolutepath() + "\\static\\test.html";

  outputstreamwriter outputstreamwriter = null;
  try {
    xwpfdocument document = new xwpfdocument(new fileinputstream(sourcefilename));
    xhtmloptions options = xhtmloptions.create();
    // 存放图片的文件夹
    options.setextractor(new fileimageextractor(new file(imagepath)));
    // html中图片的路径
    options.uriresolver(new basicuriresolver("image"));
    outputstreamwriter = new outputstreamwriter(new fileoutputstream(targetfilename), "utf-8");
    xhtmlconverter xhtmlconverter = (xhtmlconverter) xhtmlconverter.getinstance();
    xhtmlconverter.convert(document, outputstreamwriter, options);
  } finally {
    if (outputstreamwriter != null) {
      outputstreamwriter.close();
    }
  }
  return targetfilename;
}

转换成功后会生成对应的html文件，如果想在前端展示，直接读取文件转换为string返回给前端即可。

public static string readfile(string filepath) {
  file file = new file(filepath);
  inputstream input = null;
  try {
    input = new fileinputstream(file);
  } catch (filenotfoundexception e) {
    e.printstacktrace();
  }
  stringbuffer buffer = new stringbuffer();
  byte[] bytes = new byte[1024];
  try {
    for (int n; (n = input.read(bytes)) != -1;) {
      buffer.append(new string(bytes, 0, n, "utf8"));
    }
  } catch (ioexception e) {
    e.printstacktrace();
  }
  return buffer.tostring();
}

在富文本编辑器ckeditor中的显示效果：

三.html转换为word

实现思路就是先把html中的所有图片元素提取出来，统一替换为变量字符”${imgreplace}“，如果多张图片，可以依序排列下去，之后生成对应的doc文件（之前试过直接生成docx文件发现打不开，这个问题尚未找到好的解决方法），我们将其另存为docx文件，之后就可以替换变量为图片了：

public static string writewordfile(string content) {
    string path = "d:/wordfile";
    map<string, object> param = new hashmap<string, object>();

    if (!"".equals(path)) {
      file filedir = new file(path);
      if (!filedir.exists()) {
        filedir.mkdirs();
      }
      content = htmlutils.htmlunescape(content);
      list<hashmap<string, string>> imgs = getimgstr(content);
      int count = 0;
      for (hashmap<string, string> img : imgs) {
        count++;
        //处理替换以“/>”结尾的img标签
        content = content.replace(img.get("img"), "${imgreplace" + count + "}");
        //处理替换以“>”结尾的img标签
        content = content.replace(img.get("img1"), "${imgreplace" + count + "}");
        map<string, object> header = new hashmap<string, object>();

        try {
          file filepath = new file(resourceutils.geturl("classpath:").getpath());
          string imagepath = filepath.getabsolutepath() + "\\static\\";
          imagepath += img.get("src").replaceall("/", "\\\\");
          //如果没有宽高属性，默认设置为400*300
          if(img.get("width") == null || img.get("height") == null) {
            header.put("width", 400);
            header.put("height", 300);
          }else {
            header.put("width", (int) (double.parsedouble(img.get("width"))));
            header.put("height", (int) (double.parsedouble(img.get("height"))));
          }
          header.put("type", "jpg");
          header.put("content", officeutil.inputstream2bytearray(new fileinputstream(imagepath), true));
        } catch (filenotfoundexception e) {
          e.printstacktrace();
        }
        param.put("${imgreplace" + count + "}", header);
      }
      try {
        // 生成doc格式的word文档，需要手动改为docx
        byte by[] = content.getbytes("utf-8");
        bytearrayinputstream bais = new bytearrayinputstream(by);
        poifsfilesystem poifs = new poifsfilesystem();
        directoryentry directory = poifs.getroot();
        documententry documententry = directory.createdocument("worddocument", bais);
        fileoutputstream ostream = new fileoutputstream("d:\\wordfile\\temp.doc");
        poifs.writefilesystem(ostream);
        bais.close();
        ostream.close();

        // 临时文件（手动改好的docx文件）
        customxwpfdocument doc = officeutil.generateword(param, "d:\\wordfile\\temp.docx");
        //最终生成的带图片的word文件
        fileoutputstream fopts = new fileoutputstream("d:\\wordfile\\final.docx");
        doc.write(fopts);
        fopts.close();
      } catch (exception e) {
        e.printstacktrace();
      }

    }
    return "d:/wordfile/final.docx";
  }

  //获取html中的图片元素信息
  public static list<hashmap<string, string>> getimgstr(string htmlstr) {
    list<hashmap<string, string>> pics = new arraylist<hashmap<string, string>>();

    document doc = jsoup.parse(htmlstr);
    elements imgs = doc.select("img");
    for (element img : imgs) {
      hashmap<string, string> map = new hashmap<string, string>();
      if(!"".equals(img.attr("width"))) {
        map.put("width", img.attr("width").substring(0, img.attr("width").length() - 2));
      }
      if(!"".equals(img.attr("height"))) {
        map.put("height", img.attr("height").substring(0, img.attr("height").length() - 2));
      }
      map.put("img", img.tostring().substring(0, img.tostring().length() - 1) + "/>");
      map.put("img1", img.tostring());
      map.put("src", img.attr("src"));
      pics.add(map);
    }
    return pics;
  }

officeutil工具类，之前发现网上的写法只支持一张图片的修改，多张图片就会报错，是因为添加了图片，processparagraphs方法中的runs的大小改变了，会报arraylist的异常，就和我们循环list中删除元素会报异常道理一样，解决方法就是复制一个新的arraylist进行循环即可：

package com.example.demo.util; 

import java.io.bytearrayinputstream;
import java.io.fileinputstream;
import java.io.ioexception;
import java.io.inputstream;
import java.util.arraylist;
import java.util.iterator;
import java.util.list;
import java.util.map;
import java.util.map.entry;

import org.apache.poi.poixmldocument;
import org.apache.poi.hwpf.extractor.wordextractor;
import org.apache.poi.openxml4j.opc.opcpackage;
import org.apache.poi.xwpf.usermodel.xwpfparagraph;
import org.apache.poi.xwpf.usermodel.xwpfrun;
import org.apache.poi.xwpf.usermodel.xwpftable;
import org.apache.poi.xwpf.usermodel.xwpftablecell;
import org.apache.poi.xwpf.usermodel.xwpftablerow; 

/** 
 * 适用于word 2007
 */ 
public class officeutil { 

  /** 
   * 根据指定的参数值、模板，生成 word 文档 
   * @param param 需要替换的变量 
   * @param template 模板 
   */ 
  public static customxwpfdocument generateword(map<string, object> param, string template) { 
    customxwpfdocument doc = null;
    try { 
      opcpackage pack = poixmldocument.openpackage(template); 
      doc = new customxwpfdocument(pack); 
      if (param != null && param.size() > 0) { 

        //处理段落 
        list<xwpfparagraph> paragraphlist = doc.getparagraphs(); 
        processparagraphs(paragraphlist, param, doc); 

        //处理表格 
        iterator<xwpftable> it = doc.gettablesiterator(); 
        while (it.hasnext()) {
          xwpftable table = it.next(); 
          list<xwpftablerow> rows = table.getrows(); 
          for (xwpftablerow row : rows) { 
            list<xwpftablecell> cells = row.gettablecells(); 
            for (xwpftablecell cell : cells) { 
              list<xwpfparagraph> paragraphlisttable = cell.getparagraphs(); 
              processparagraphs(paragraphlisttable, param, doc); 
            } 
          } 
        } 
      } 
    } catch (exception e) { 
      e.printstacktrace(); 
    } 
    return doc; 
  } 
  /** 
   * 处理段落 
   * @param paragraphlist 
   */ 
  public static void processparagraphs(list<xwpfparagraph> paragraphlist,map<string, object> param,customxwpfdocument doc){ 
    if(paragraphlist != null && paragraphlist.size() > 0){ 
      for(xwpfparagraph paragraph:paragraphlist){
        //poi转换过来的行间距过大，需要手动调整
        if(paragraph.getspacingbefore() >= 1000 || paragraph.getspacingafter() > 1000) {
          paragraph.setspacingbefore(0);
          paragraph.setspacingafter(0);
        }
        //设置word中左右间距
        paragraph.setindentationleft(0);
        paragraph.setindentationright(0);
        list<xwpfrun> runs = paragraph.getruns();
        //加了图片，修改了paragraph的runs的size，所以循环不能使用runs
        list<xwpfrun> allruns = new arraylist<xwpfrun>(runs);
        for (xwpfrun run : allruns) {
          string text = run.gettext(0); 
          if(text != null){
            boolean issettext = false; 
            for (entry<string, object> entry : param.entryset()) { 
              string key = entry.getkey(); 
              if(text.indexof(key) != -1){ 
                issettext = true; 
                object value = entry.getvalue(); 
                if (value instanceof string) {//文本替换 
                  text = text.replace(key, value.tostring()); 
                } else if (value instanceof map) {//图片替换 
                  text = text.replace(key, ""); 
                  map pic = (map)value; 
                  int width = integer.parseint(pic.get("width").tostring()); 
                  int height = integer.parseint(pic.get("height").tostring()); 
                  int pictype = getpicturetype(pic.get("type").tostring()); 
                  byte[] bytearray = (byte[]) pic.get("content"); 
                  bytearrayinputstream byteinputstream = new bytearrayinputstream(bytearray); 
                  try { 
                    string blipid = doc.addpicturedata(byteinputstream,pictype); 
                    doc.createpicture(blipid,doc.getnextpicnamenumber(pictype), width, height,paragraph);
                  } catch (exception e) { 
                    e.printstacktrace(); 
                  } 
                } 
              } 
            } 
            if(issettext){ 
              run.settext(text,0); 
            } 
          } 
        } 
      } 
    } 
  } 
  /** 
   * 根据图片类型，取得对应的图片类型代码 
   * @param pictype 
   * @return int 
   */ 
  private static int getpicturetype(string pictype){ 
    int res = customxwpfdocument.picture_type_pict; 
    if(pictype != null){ 
      if(pictype.equalsignorecase("png")){ 
        res = customxwpfdocument.picture_type_png; 
      }else if(pictype.equalsignorecase("dib")){ 
        res = customxwpfdocument.picture_type_dib; 
      }else if(pictype.equalsignorecase("emf")){ 
        res = customxwpfdocument.picture_type_emf; 
      }else if(pictype.equalsignorecase("jpg") || pictype.equalsignorecase("jpeg")){ 
        res = customxwpfdocument.picture_type_jpeg; 
      }else if(pictype.equalsignorecase("wmf")){ 
        res = customxwpfdocument.picture_type_wmf; 
      } 
    } 
    return res; 
  } 
  /** 
   * 将输入流中的数据写入字节数组 
   * @param in 
   * @return 
   */ 
  public static byte[] inputstream2bytearray(inputstream in,boolean isclose){ 
    byte[] bytearray = null; 
    try { 
      int total = in.available(); 
      bytearray = new byte[total]; 
      in.read(bytearray); 
    } catch (ioexception e) { 
      e.printstacktrace(); 
    }finally{ 
      if(isclose){ 
        try { 
          in.close(); 
        } catch (exception e2) { 
          system.out.println("关闭流失败"); 
        } 
      } 
    } 
    return bytearray; 
  } 
}

我认为之所以word2003不支持图片替换，主要是处理2003版本的hwpfdocument对象被声明为了final，我们就无法重写他的方法了。而处理2007版本的类为xwpfdocument，是可以继承的，通过继承xwpfdocument，重写createpicture方法即可实现图片替换，以下为对应的customxwpfdocument类：

package com.example.demo.util;  

import java.io.ioexception; 
import java.io.inputstream; 
import org.apache.poi.openxml4j.opc.opcpackage; 
import org.apache.poi.xwpf.usermodel.xwpfdocument; 
import org.apache.poi.xwpf.usermodel.xwpfparagraph; 
import org.apache.xmlbeans.xmlexception; 
import org.apache.xmlbeans.xmltoken; 
import org.openxmlformats.schemas.drawingml.x2006.main.ctnonvisualdrawingprops; 
import org.openxmlformats.schemas.drawingml.x2006.main.ctpositivesize2d; 
import org.openxmlformats.schemas.drawingml.x2006.wordprocessingdrawing.ctinline; 

/** 
 * 自定义 xwpfdocument，并重写 createpicture()方法 
 */ 
public class customxwpfdocument extends xwpfdocument {  
  public customxwpfdocument(inputstream in) throws ioexception {  
    super(in);  
  }  

  public customxwpfdocument() {  
    super();  
  }  

  public customxwpfdocument(opcpackage pkg) throws ioexception {  
    super(pkg);  
  }  

  /** 
   * @param ind 
   * @param width 宽 
   * @param height 高 
   * @param paragraph 段落 
   */ 
  public void createpicture(string blipid, int ind, int width, int height,xwpfparagraph paragraph) {  
    final int emu = 9525;  
    width *= emu;  
    height *= emu;  
    ctinline inline = paragraph.createrun().getctr().addnewdrawing().addnewinline();  
    string picxml = ""  
        + "<a:graphic xmlns:a=\"http://schemas.openxmlformats.org/drawingml/2006/main\">"  
        + "  <a:graphicdata uri=\"http://schemas.openxmlformats.org/drawingml/2006/picture\">"  
        + "   <pic:pic xmlns:pic=\"http://schemas.openxmlformats.org/drawingml/2006/picture\">"  
        + "     <pic:nvpicpr>" + "      <pic:cnvpr id=\""  
        + ind  
        + "\" name=\"generated\"/>"  
        + "      <pic:cnvpicpr/>"  
        + "     </pic:nvpicpr>"  
        + "     <pic:blipfill>"  
        + "      <a:blip r:embed=\""  
        + blipid  
        + "\" xmlns:r=\"http://schemas.openxmlformats.org/officedocument/2006/relationships\"/>"  
        + "      <a:stretch>"  
        + "        <a:fillrect/>"  
        + "      </a:stretch>"  
        + "     </pic:blipfill>"  
        + "     <pic:sppr>"  
        + "      <a:xfrm>"  
        + "        <a:off x=\"0\" y=\"0\"/>"  
        + "        <a:ext cx=\""  
        + width  
        + "\" cy=\""  
        + height  
        + "\"/>"  
        + "      </a:xfrm>"  
        + "      <a:prstgeom prst=\"rect\">"  
        + "        <a:avlst/>"  
        + "      </a:prstgeom>"  
        + "     </pic:sppr>"  
        + "   </pic:pic>"  
        + "  </a:graphicdata>" + "</a:graphic>";  

    inline.addnewgraphic().addnewgraphicdata();  
    xmltoken xmltoken = null;  
    try {  
      xmltoken = xmltoken.factory.parse(picxml);  
    } catch (xmlexception xe) {  
      xe.printstacktrace();  
    }  
    inline.set(xmltoken);  

    inline.setdistt(0);   
    inline.setdistb(0);   
    inline.setdistl(0);   
    inline.setdistr(0);   

    ctpositivesize2d extent = inline.addnewextent();  
    extent.setcx(width);  
    extent.setcy(height);  

    ctnonvisualdrawingprops docpr = inline.addnewdocpr();   
    docpr.setid(ind);   
    docpr.setname("图片" + ind);   
    docpr.setdescr("测试");  
  }  
}

以上就是通过poi实现html和word的相互转换，对于html无法转换为可读的docx这个问题尚未解决，如果大家有好的解决方法可以交流一下。