Word轉html實現線上預覽
阿新 • • 發佈：2018-12-29
word轉html,可以同時支援doc和docx兩種格式，非常好用
開發工具：idea
專案管理工具：maven
不多說，直接擼程式碼
1、首先配置pom.xml檔案,具體配置如下
2、工具類的開發
/** * WORD轉HTML docx格式 * POI版本: 3.10-FINAL * */ import java.io.*; import java.util.List; import org.apache.poi.xwpf.usermodel.IBodyElement; import org.apache.poi.xwpf.usermodel.ParagraphAlignment; import org.apache.poi.xwpf.usermodel.XWPFDocument; import org.apache.poi.xwpf.usermodel.XWPFParagraph; import org.apache.poi.xwpf.usermodel.XWPFPicture; import org.apache.poi.xwpf.usermodel.XWPFPictureData; 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; public class Docx2Html { /** * 解析DOC * * @param fileName 檔名 * @param isAllHtml 全部為HMTL * @param tmpImgDir 臨時目錄,不包含檔名 * @param tmpImgUrl 臨時連結,圖片 連結不包含檔名,通常這個引數可以傳一個相對路徑 * @throws Exception */ public static String analysisDocument(String fileName,boolean isAllHtml,String tmpImgDir,String tmpImgUrl) throws Exception{ InputStream in = new FileInputStream(new File(fileName)); return analysisDocument(in, isAllHtml, tmpImgDir, tmpImgUrl); } /** * 解析DOC */ public static String analysisDocument(InputStream in,boolean isAllHtml,String tmpImgDir,String tmpImgUrl) throws Exception { XWPFDocument doc = new XWPFDocument(in); StringBuffer buffer=new StringBuffer(); List<IBodyElement> eles= doc.getBodyElements(); for(IBodyElement el:eles){ String name=el.getElementType().name(); if(name.equals("表格")){//表格 XWPFTable table=(XWPFTable)el; buffer.append(analysisTable(table)); }else {//文字 XWPFParagraph graph=(XWPFParagraph)el; List<XWPFRun> runs=graph.getRuns(); ParagraphAlignment alignment=graph.getAlignment(); String align="left"; if(alignment.equals(ParagraphAlignment.CENTER)){ align="center"; }else if(alignment.equals(ParagraphAlignment.RIGHT)){ align="right"; } buffer.append("<div style='text-align:").append(align); StringBuffer pBuffer=new StringBuffer(); int index=0; for(XWPFRun run:runs){ index++; List<XWPFPicture> pics=run.getEmbeddedPictures(); if(null!=pics){ for(XWPFPicture pic:pics){ index++; XWPFPictureData pictureData=pic.getPictureData(); byte[] picBytes=pictureData.getData(); String picName="dzpic_"+System.currentTimeMillis()+"_"+index+".jpg"; File file=new File(tmpImgDir+"/"+picName); OutputStream os =new FileOutputStream(file); os.write(picBytes); os.close(); pBuffer.append("<img src='").append(tmpImgUrl).append(picName).append("'border=0'/>"); } } String text=run.getText(run.getTextPosition()); if(null!=text){ String color=null==run.getColor()?"000":run.getColor(); int fontsize=run.getFontSize()==-1?15:run.getFontSize(); pBuffer.append("<span style='color:#").append(color).append(";font-size:").append(fontsize).append("px;'>"); pBuffer.append(text).append("</span>"); } } buffer.append(";'>").append(pBuffer.toString()).append("</div>"); } } return buffer.toString(); } /** * 解析表格 * @param tb 表格 物件 * @return String * @throws Exception */ static String analysisTable(XWPFTable tb) throws Exception { StringBuffer htmlTextTbl = new StringBuffer(); htmlTextTbl .append("<table border=\"1\" style=\"width:100%;font-size:14px;\" class=\"display table table-striped table-bordered table-hover table-checkable dataTable no-footer\">"); List<XWPFTableRow> rows=tb.getRows(); int rowCount=rows.size(); for (int i = 0; i < rowCount; i++) { XWPFTableRow tr = rows.get(i); String trCls=(i%2==1)?"odd":"even"; htmlTextTbl .append("<tr class='").append(trCls).append("'>"); List<XWPFTableCell> cells=tr.getTableCells(); int cellCount=cells.size(); for (int j = 0; j < cellCount; j++) { XWPFTableCell td = tr.getCell(j); List<XWPFParagraph> cellGraphs= td.getParagraphs(); for (int k = 0; k < cellGraphs.size(); k++) { XWPFParagraph para = cellGraphs.get(k); String s =para.getText()==null?"": para.getText().trim(); if (s == "") { s = " "; } if(i==0){ htmlTextTbl.append("<th class='td-center'>").append(s).append("</th>"); }else{ htmlTextTbl.append("<td class='td-center'>").append(s).append("</td>"); } } } htmlTextTbl.append("</tr>"); } htmlTextTbl.append("</table>"); return htmlTextTbl.toString(); } public static void main(String args[]) { try { String htmlDoc=analysisDocument("d:/uploadtemp/mysql.docx", true, "d:/uploadtemp/","d:/uploadtemp/"); File file=new File("d:/uploadtemp/text.html"); if (file.exists()){ file.createNewFile(); } FileWriter fileWriter=new FileWriter(file); BufferedWriter bufferedWriter=new BufferedWriter(fileWriter); bufferedWriter.write(htmlDoc,0,htmlDoc.length()-1); bufferedWriter.close(); } catch (Exception e) { e.printStackTrace(); } } }
/** * WORD轉HTML doc格式 * POI版本: 3.10-FINAL * */ import java.io.*; import org.apache.poi.hpsf.SummaryInformation; import org.apache.poi.hwpf.HWPFDocument; import org.apache.poi.hwpf.model.PicturesTable; import org.apache.poi.hwpf.usermodel.CharacterRun; import org.apache.poi.hwpf.usermodel.Picture; import org.apache.poi.hwpf.usermodel.Range; import org.apache.poi.hwpf.usermodel.Paragraph; import org.apache.poi.hwpf.usermodel.Table; import org.apache.poi.hwpf.usermodel.TableCell; import org.apache.poi.hwpf.usermodel.TableIterator; import org.apache.poi.hwpf.usermodel.TableRow; public class Doc2Html { private static final short ENTER_ASCII = 13; //回車 private static final short SPACE_ASCII = 32; // 空格 private static final short TABULATION_ASCII = 9; // TAB public static int beginPosi = 0; public static int endPosi = 0; public static int beginArray[]; public static int endArray[]; public static String htmlTextArray[]; public static boolean tblExist = false; /** * 解析DOC * * @param fileName 檔名 * @param isAllHtml 全部為HMTL * @param tmpImgDir 臨時目錄,圖片 * @param tmpImgUrl 臨時連結,圖片 * @throws Exception */ public static String analysisDocument(String fileName,boolean isAllHtml,String tmpImgDir,String tmpImgUrl) throws Exception{ InputStream in = new FileInputStream(new File(fileName)); return analysisDocument(in, isAllHtml, tmpImgDir, tmpImgUrl); } /** * 解析DOC */ public static String analysisDocument(InputStream in,boolean isAllHtml,String tmpImgDir,String tmpImgUrl) throws Exception { HWPFDocument doc = new HWPFDocument(in); Range rangetbl = doc.getRange(); TableIterator it = new TableIterator(rangetbl); int num = 100; beginArray = new int[num]; endArray = new int[num]; htmlTextArray = new String[num]; int length = doc.characterLength(); if(length==0)return ""; PicturesTable pTable = doc.getPicturesTable(); SummaryInformation sif = doc.getSummaryInformation(); String title = "DOC檔案預覽"; if (null != sif) { title = doc.getSummaryInformation().getTitle(); } StringBuffer htmlText=new StringBuffer(""); if(isAllHtml){ htmlText.append("<html><head><title>").append(title).append("</title></head><body>"); } if (it.hasNext()) { analysisTables(it, rangetbl); } int cur = 0; String tempString = ""; int index=0; for (int i = 0; i < length - 1; i++) { Range range = new Range(i, i + 1, doc); CharacterRun cr = range.getCharacterRun(0); if (tblExist) { if (i == beginArray[cur]) { htmlText.append(tempString).append(htmlTextArray[cur]); tempString = ""; i = endArray[cur] - 1; cur++; continue; } } if (pTable.hasPicture(cr)) { htmlText.append(tempString); String picFileName=analysisPicture(pTable, cr,tmpImgDir,index++); htmlText.append("<img src=\"").append(tmpImgUrl.endsWith("/")?tmpImgUrl:(tmpImgUrl+"/")).append(picFileName).append("\" style='width:100%;height:100%;border:0px;'/>"); tempString = ""; } else { Range range2 = new Range(i + 1, i + 2, doc); CharacterRun cr2 = range2.getCharacterRun(0); char c = cr.text().charAt(0); if (c == ENTER_ASCII) { tempString += "<br/>"; } else if (c == SPACE_ASCII) tempString += " "; else if (c == TABULATION_ASCII) tempString += " "; boolean flag = compareCharStyle(cr, cr2); if (flag) tempString += cr.text(); else { StringBuffer fontStyle1 = new StringBuffer("<span style=\"font-family:"); fontStyle1.append(cr.getFontName() ).append(";font-size:").append( cr.getFontSize() / 2).append( "pt;"); if (cr.isBold()) fontStyle1.append("font-weight:bold;"); if (cr.isItalic()) fontStyle1.append("font-style:italic;"); htmlText.append(fontStyle1.toString()).append("\">") .append(tempString).append(cr.text()).append("</span>"); fontStyle1.delete(0,fontStyle1.length()); tempString = ""; } } } htmlText .append(tempString); if(isAllHtml){htmlText .append("</body></html>");} return htmlText.toString(); } /** * 解析表格 * @param it * @param rangetbl DOC段 * @throws Exception */ static void analysisTables(TableIterator it, Range rangetbl) throws Exception { int counter=0; while (it.hasNext()) { tblExist = true; StringBuffer htmlTextTbl = new StringBuffer(); Table tb = (Table) it.next(); beginPosi = tb.getStartOffset(); endPosi = tb.getEndOffset(); beginArray[counter] = beginPosi; endArray[counter] = endPosi; htmlTextTbl .append("<table border=\"1\" style=\"width:100%;font-size:14px;\" class=\"display table table-striped table-bordered table-hover table-checkable dataTable no-footer\">"); for (int i = 0; i < tb.numRows(); i++) { TableRow tr = tb.getRow(i); String trCls=(i%2==1)?"odd":"even"; htmlTextTbl .append("<tr class='").append(trCls).append("'>"); for (int j = 0; j < tr.numCells(); j++) { TableCell td = tr.getCell(j); int cellWidth = td.getWidth(); for (int k = 0; k < td.numParagraphs(); k++) { Paragraph para = td.getParagraph(k); String s = para.text().toString().trim(); if (s == "") { s = " "; } if(i==0){ htmlTextTbl.append("<th class='td-center'>").append(s).append("</th>"); }else{ htmlTextTbl.append("<td class='td-center'>").append(s).append("</td>"); } } } htmlTextTbl.append("</tr>"); } htmlTextTbl.append("</table>"); htmlTextArray[counter++] = htmlTextTbl.toString(); } } /** * 圖片解析 * * @param pTable WORD中的圖片域 * @param cr * @param path 臨時路徑 * @return String 圖片檔名 * @throws Exception */ static String analysisPicture(PicturesTable pTable, CharacterRun cr, String path, int index) throws Exception { // 圖片物件 Picture pic = pTable.extractPicture(cr, false); // 圖片檔名 String afileName = "dzpic_"+System.currentTimeMillis()+"_"+index+".jpg"; OutputStream out = new FileOutputStream(new File((path.endsWith("/")?path:(path+ "/")) + afileName)); pic.writeImageContent(out); out.close(); return afileName; } /** * 切換文字樣式 * @param cr1 * @param cr2 * @return boolean * */ static boolean compareCharStyle(CharacterRun cr1, CharacterRun cr2) { boolean flag = false; if (cr1.isBold() == cr2.isBold() && cr1.isItalic() == cr2.isItalic() && cr1.getFontName().equals(cr2.getFontName()) && cr1.getFontSize() == cr2.getFontSize()) { flag = true; } return flag; } /** * 測試 * */ public static void main(String args[]) { try { String htmlDoc=analysisDocument("d:/uploadtemp/監造流程.doc", true, "d:/uploadtemp/","d:/uploadtemp/"); File file=new File("d:/uploadtemp/監造流程.html"); if (file.exists()){ file.createNewFile(); } FileWriter fileWriter=new FileWriter(file); BufferedWriter bufferedWriter=new BufferedWriter(fileWriter); bufferedWriter.write(htmlDoc,0,htmlDoc.length()-1); bufferedWriter.close(); } catch (Exception e) { e.printStackTrace(); } } }
import javax.servlet.http.HttpServletRequest; import java.io.BufferedWriter; import java.io.File; import java.io.FileWriter; /** * 分析文件的格式是doc還是docx * */ public class AnaliysisDom { public static String analiysisDom(String path,String filename,HttpServletRequest request){ String uri=null; String realPath=null; String inputpath=path+filename; String outputname=(filename.substring(0,filename.lastIndexOf('.')))+".html"; String outputpath=path+outputname; File imgfile=new File(path+filename.substring(0,filename.lastIndexOf('.'))); String string=imgfile.getPath(); if (!imgfile.exists()){ imgfile.mkdirs(); } File outputFile = new File(outputpath); if (outputFile.exists()){ }else { if (inputpath.endsWith(".doc") || inputpath.endsWith(".DOC")){ try { String s = Doc2Html.analysisDocument(inputpath, true, imgfile.getPath(), imgfile.getPath()); File file=new File(outputpath); if (file.exists()){ file.createNewFile(); } FileWriter fileWriter=new FileWriter(file); BufferedWriter bufferedWriter=new BufferedWriter(fileWriter); bufferedWriter.write(s,0,s.length()-1); bufferedWriter.close(); } catch (Exception e) { e.printStackTrace(); } } if (inputpath.endsWith(".docx") || inputpath.endsWith("DOCX")){ try { String s = Docx2Html.analysisDocument(inputpath, true, imgfile.getPath()+"/", imgfile.getPath()+"/"); File file=new File(outputpath); if (file.exists()){ file.createNewFile(); } FileWriter fileWriter=new FileWriter(file); BufferedWriter bufferedWriter=new BufferedWriter(fileWriter); bufferedWriter.write(s,0,s.length()-1); bufferedWriter.close(); } catch (Exception e) { e.printStackTrace(); } } } realPath = request.getSession().getServletContext().getRealPath(outputname); System.out.println(realPath); uri=request.getScheme() + "://" + request.getServerName() + ":" + request.getServerPort() +"/"+ request.getContextPath() +outputname; return uri; } }
controller層程式碼例項
@RequestMapping("/readFile")
@ResponseBody
public String readFile(String templateattachpath,HttpServletRequest request){
String uri="";
try {
uri=AnaliysisDom.analiysisDom(path,templateattachpath,request);
} catch (Exception e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
return uri;
}
這樣word轉html就轉換成功了，超簡單，開擼吧！！！！