POI操做OFFICE轉HTML

因爲項目需求，須要上傳在線預覽生成的office文件，通過一番痛苦的研究，目前已經成功將DOC、DOCX、PPTX不亂碼轉換成功，PPT當字體不是宋體時出現亂碼，至今沒法解決,待往後再研究吧。而EXCEL轉換未能解決圖片問題。能力有限，先上代碼吧

POI jar的maven地址：需注意的時，統一版本jar可能卻會出現內部一些方法不存在，不明白什麼狀況，經測試，選擇了3.15

    <properties>
        <poi.version>3.15</poi.version>
    </properties>
    <dependency>
            <groupId>org.apache.poi</groupId>
            <artifactId>poi</artifactId>
            <version>${poi.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.poi</groupId>
            <artifactId>poi-ooxml</artifactId>
            <version>${poi.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.poi</groupId>
            <artifactId>poi-scratchpad</artifactId>
            <version>${poi.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.poi</groupId>
            <artifactId>poi-excelant</artifactId>
            <version>${poi.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.poi</groupId>
            <artifactId>ooxml-schemas</artifactId>
            <version>1.3</version>
            <exclusions>
                <exclusion>
                    <groupId>stax</groupId>
                    <artifactId>stax-api</artifactId>
                </exclusion>
                <exclusion>
                    <groupId>org.apache.xmlbeans</groupId>
                    <artifactId>xmlbeans</artifactId>
                </exclusion>
            </exclusions>
        </dependency>

java代碼以下：

package com.boco.investment.common;

import org.apache.commons.io.output.ByteArrayOutputStream;
import org.apache.poi.hslf.usermodel.HSLFAutoShape;
import org.apache.poi.hslf.usermodel.HSLFTable;
import org.apache.poi.hslf.usermodel.HSLFTextParagraph;
import org.apache.poi.hslf.usermodel.HSLFTextRun;
import org.apache.poi.hssf.converter.ExcelToHtmlConverter;
import org.apache.poi.hssf.usermodel.HSSFPictureData;
import org.apache.poi.hssf.usermodel.HSSFWorkbook;
import org.apache.poi.hwpf.HWPFDocument;
import org.apache.poi.hwpf.converter.WordToHtmlConverter;
import org.apache.poi.hwpf.usermodel.Picture;
import org.apache.poi.sl.usermodel.Slide;
import org.apache.poi.sl.usermodel.SlideShow;
import org.apache.poi.sl.usermodel.SlideShowFactory;
import org.apache.poi.xslf.usermodel.*;
import org.apache.poi.xwpf.converter.core.BasicURIResolver;
import org.apache.poi.xwpf.converter.core.FileImageExtractor;
import org.apache.poi.xwpf.converter.xhtml.XHTMLConverter;
import org.apache.poi.xwpf.converter.xhtml.XHTMLOptions;
import org.apache.poi.xwpf.usermodel.XWPFDocument;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.w3c.dom.Document;

import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.parsers.ParserConfigurationException;
import javax.xml.transform.OutputKeys;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerException;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.dom.DOMSource;
import javax.xml.transform.stream.StreamResult;
import java.awt.*;
import java.awt.geom.Rectangle2D;
import java.awt.image.BufferedImage;
import java.io.*;
import java.util.List;

/**
 * Created by wdj on 2017/11/14.
 *
 * 經過poi將doc、docx、ppt、pptx轉爲html
 * 保存圖片的位置在文件所在目錄，以文件名建立文件夾下
 */
public class PoiUtils {

    private static Logger logger = LoggerFactory.getLogger(PoiUtils.class);

    private static String rootPath = ConfigUtils.getProperty("upload.rootPath");

    private static final String IMAGE_SERVER = ConfigUtils.getProperty("file.server");

    /**
     * poi office文件轉html，支持doc,docx,ppt,pptx
     * 根據源文件在同一目錄下生成相同名稱的html文件
     */
    public static boolean officeToHtml(String filePath) {
        filePath = rootPath + filePath;
        String htmlFilePath = FileUtils.getFileNameWithoutExtension(filePath)+".html";
        try {
            if(checkFile(filePath,"doc")){
                return wordToHtml03(filePath,htmlFilePath);
            }else if(checkFile(filePath,"docx")){
                return wordToHtml07(filePath,htmlFilePath);
            }else if(checkFile(filePath,"ppt")){
                return pptToHtml03(filePath,htmlFilePath);
            }else if(checkFile(filePath,"pptx")){
                return pptToHtml07(filePath,htmlFilePath);
            }else {
                logger.error("poi OfficeToHtml出錯，不支持的文件格式");
                return false;
            }
        } catch (Exception e) {
            logger.error("poi OfficeToHtml出錯：",e);
            return false;
        }
    }

    /**
     * Word03 轉爲 HTML
     *
     * @param fileName
     * @param outputFile
     */
    public static boolean wordToHtml03(String fileName, String outputFile){
        if (!(checkFile(fileName,"doc")||checkFile(fileName,"docx"))) {
            logger.error("word03文件轉html出錯，不支持類型爲："+fileName.substring(fileName.lastIndexOf("."))+" 的文件");
            return false;
        }
        HWPFDocument wordDoc = null;
        WordToHtmlConverter wthc = null;
        try {
            wordDoc = new HWPFDocument(new FileInputStream(fileName));
            wthc = new WordToHtmlConverter(
                    DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument());
        } catch (Exception e) {
            logger.error("word03轉html失敗",e);
            return false;
        }
        //html引用圖片位置
        wthc.setPicturesManager((bytes,pt,string,f,f1) ->getImageUrl(fileName)+string);
        wthc.processDocument(wordDoc);
        List<Picture> pics = wordDoc.getPicturesTable().getAllPictures();
        fileExists(getImageSavePath(fileName));
        if (null != pics && pics.size() > 0) {
            for (Picture pic : pics) {
                try {
                    //生成圖片位置
                    pic.writeImageContent(new FileOutputStream(getImageSavePath(fileName)+pic.suggestFullFileName()));
                } catch (IOException e) {
                    logger.error("word03轉html失敗",e);
                    return false;
                }
            }
        }
        Document htmlDocument = wthc.getDocument();
        ByteArrayOutputStream out = new ByteArrayOutputStream();
        DOMSource domSource = new DOMSource(htmlDocument);
        StreamResult streamResult = new StreamResult(out);

        try {
            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);
        } catch (TransformerException e) {
            logger.error("word03轉html失敗",e);
            return false;
        } finally {
            try {
                out.close();
            } catch (IOException e) {
                logger.error("word03轉html文件流關閉失敗",e);
            }
        }
        String htmlStr = new String(out.toByteArray());
        return writeFile(htmlStr, outputFile);
    }

    /**
     * Word07 轉爲 HTML
     *
     * @param fileName
     * @param outputFile
     */
    public static boolean wordToHtml07(String fileName, String outputFile){
        if (!checkFile(fileName,"docx")) {
            logger.error("word07文件轉html出錯，不支持類型爲："+fileName.substring(fileName.lastIndexOf("."))+" 的文件");
            return false;
        }
        //讀取文檔內容
        XWPFDocument document = null;
        try {
            InputStream in = new FileInputStream(fileName);
            document = new XWPFDocument(in);
        } catch (IOException e) {
            logger.error("word07轉html失敗",e);
            return false;
        }
        //加載html頁面時圖片路徑
        XHTMLOptions options = XHTMLOptions.create().URIResolver( new BasicURIResolver(getImageUrl(fileName)));
        //圖片保存文件夾路徑
        fileExists(getImageSavePath(fileName));
        options.setExtractor(new FileImageExtractor(new File(getImageSavePath(fileName))));
        OutputStream out = null;
        try {
            out = new FileOutputStream(new File(outputFile));
            XHTMLConverter.getInstance().convert(document, out, options);
            return true;
        } catch (IOException e) {
            logger.error("word07轉html失敗",e);
            return false;
        } finally {
            try {
                out.close();
            } catch (IOException e) {
                logger.error("word07轉html文件流關閉失敗",e);
            }
        }
    }

    /**
     * excel to html
     * @param path
     * @param file
     */
    // todo 待完成
    public static void testExcel(String path,String file) {
        HSSFWorkbook excelBook= null;
        ExcelToHtmlConverter excelToHtmlConverter = null;
        try {
            InputStream input=new FileInputStream(path+file);
            excelBook = new HSSFWorkbook(input);
            excelToHtmlConverter = new ExcelToHtmlConverter(DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument() );
        } catch (IOException e) {
            e.printStackTrace();
        } catch (ParserConfigurationException e) {
            e.printStackTrace();
        }

        //加載html頁面時圖片路徑
        XHTMLOptions options = XHTMLOptions.create().URIResolver( new BasicURIResolver(getImageUrl(path)));
        //圖片保存文件夾路徑
        options.setExtractor(new FileImageExtractor(new File(getImageSavePath(path))));
        excelToHtmlConverter.setOutputRowNumbers(false);
        excelToHtmlConverter.setOutputHiddenRows(false);
        excelToHtmlConverter.setOutputColumnHeaders(false);
        excelToHtmlConverter.setOutputHiddenColumns(true);
        excelToHtmlConverter.processWorkbook(excelBook);
        List pics = excelBook.getAllPictures();
        if (pics != null) {
            for (int i = 0; i < pics.size(); i++) {
                HSSFPictureData pic = (HSSFPictureData) pics.get (i);
                    try {
//                        pic.writeImageContent (new FileOutputStream (path + pic.suggestFullFileName() ) );
                        new FileOutputStream (path + "11" ).write(pic.getData());
                    } catch (IOException e) {
                        e.printStackTrace();
                    }

            }
        }
        Document htmlDocument =excelToHtmlConverter.getDocument();
        ByteArrayOutputStream outStream = new ByteArrayOutputStream();
        DOMSource domSource = new DOMSource (htmlDocument);
        StreamResult streamResult = new StreamResult (outStream);
        TransformerFactory tf = TransformerFactory.newInstance();
        try {
            Transformer serializer = tf.newTransformer();
            serializer.setOutputProperty (OutputKeys.ENCODING, "utf-8");
            serializer.setOutputProperty (OutputKeys.INDENT, "yes");
            serializer.setOutputProperty (OutputKeys.METHOD, "html");
            serializer.transform (domSource, streamResult);
        } catch (TransformerException e) {
            e.printStackTrace();
        } finally {
            try {
                outStream.close();
            } catch (IOException e) {
                e.printStackTrace();
            }
        }
        String content = new String (outStream.toByteArray() );

        try {
            FileUtils.writeStringToFile(new File (path, "exportExcel.html"), content, "utf-8");
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    /**
     * ppt03轉html
     * filepath:源文件
     * htmlname:生成html名稱
     * */
    public static boolean pptToHtml03(String filepath, String outputFile){
        File file = new File(filepath);
        // 讀入PPT文件
        if (!checkFile(filepath,"ppt")) {
            logger.error("ppt03文件轉html出錯，不支持類型爲："+FileUtils.getFileExtension(filepath)+" 的文件");
            return false;
        }
        FileInputStream is = null;
        SlideShow ppt;
        try {
            is = new FileInputStream(file);
            ppt = SlideShowFactory.create(is);
        } catch (IOException e) {
            logger.error("ppt03文件轉html出錯：",e);
            return false;
        } finally {
            try {
                is.close();
            } catch (IOException e) {
                logger.error("ppt03文件轉html關閉文件流失敗：",e);
            }
        }

        Dimension pgsize = ppt.getPageSize();
        List<Slide> slide = ppt.getSlides();
        FileOutputStream out =null;
        String imghtml="";
        //保存圖片位置
        fileExists(getImageSavePath(filepath));
        for (int i = 0; i < slide.size(); i++) {
            for (Object o : slide.get(i).getShapes()) {
                if(o instanceof HSLFAutoShape) {
                    HSLFAutoShape shapes = (HSLFAutoShape)o;
                    List<HSLFTextParagraph> list = shapes.getTextParagraphs();
                    for (HSLFTextParagraph hslfTextRuns : list) {
                        for (HSLFTextRun hslfTextRun : hslfTextRuns.getTextRuns()) {
                            hslfTextRun.setFontFamily("宋體");
                        }
                    }
                }else if(o instanceof HSLFTable){
                    HSLFTable hslfTable = (HSLFTable) o;
                    int rowSize = hslfTable.getNumberOfRows();
                    int columnSize = hslfTable.getNumberOfColumns();
                    for (int j = 0; j < rowSize; j++) {
                        for (int k = 0; k < columnSize; k++) {
                            for (int l =0;l <  hslfTable.getCell(j, k).getTextParagraphs().size();l++){
                                HSLFTextParagraph hslfTextRuns = hslfTable.getCell(j, k).getTextParagraphs().get(l);
                                for (int m = 0;m < hslfTextRuns.getTextRuns().size();m++){
                                    HSLFTextRun textRun = hslfTextRuns.getTextRuns().get(m);
                                    //todo 設置字體失敗，輸出html依舊會亂碼
                                    textRun.setFontFamily("宋體");
                                }
                            }
                        }
                    }
                }
            }
            BufferedImage img = new BufferedImage(pgsize.width,pgsize.height, BufferedImage.TYPE_INT_RGB);

            Graphics2D graphics = img.createGraphics();
            graphics.setPaint(Color.BLUE);
            graphics.fill(new Rectangle2D.Float(0, 0, pgsize.width, pgsize.height));
            slide.get(i).draw(graphics);
            // 這裏設置圖片的存放路徑和圖片的格式(jpeg,png,bmp等等),注意生成文件路徑與源文件同一個目錄
            try {
                out= new FileOutputStream(getImageSavePath(filepath)+(i + 1) + ".jpeg");
                javax.imageio.ImageIO.write(img, "jpeg", out);
            } catch (IOException e) {
                logger.error("ppt03文件轉html出錯：",e);
                try {
                    out.close();
                } catch (IOException e1) {
                    logger.error("ppt03文件轉html關閉文件流失敗：",e);
                }
                return false;
            }
            //圖片在html加載路徑
            String imgs=getImageUrl(filepath)+(i + 1) + ".jpeg";
            imghtml+="<img src=\'"+imgs+"\' style=\'width:1200px;height:830px;vertical-align:text-bottom;\'><br><br><br><br>";

        }
        DOMSource domSource = new DOMSource();
        StreamResult streamResult = new StreamResult(out);
        TransformerFactory tf = TransformerFactory.newInstance();
        try {
            Transformer serializer = tf.newTransformer();
            serializer.setOutputProperty(OutputKeys.ENCODING, "utf-8");
            serializer.setOutputProperty(OutputKeys.INDENT, "yes");
            serializer.setOutputProperty(OutputKeys.METHOD, "html");
            serializer.transform(domSource, streamResult);
        } catch (TransformerException e) {
            logger.error("ppt03文件轉html出錯：",e);
            return false;
        } finally {
            try {
                out.close();
            } catch (IOException e) {
                logger.error("ppt03文件轉html關閉文件流失敗：",e);
            }
        }

        String ppthtml="<html><head><META http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\"></head><body>"+imghtml+"</body></html>";
        try {
            FileUtils.writeStringToFile(new File(outputFile), ppthtml, "utf-8");
        } catch (IOException e) {
            logger.error("ppt03文件轉html出錯：",e);
            return false;
        }
        return true;
    }

    /**
     * ppt07轉html
     * filepath:源文件
     * outputFile:生成html名稱
     * */
    public static boolean pptToHtml07(String filepath,String outputFile) {
        File file = new File(filepath);
        // 讀入PPT文件
        if (!checkFile(filepath,"pptx")) {
            logger.error("ppt07文件轉html出錯，不支持類型爲："+FileUtils.getFileExtension(filepath)+" 的文件");
            return false;
        }
        FileInputStream is = null;
        SlideShow ppt;
        try {
            is = new FileInputStream(file);
            ppt = SlideShowFactory.create(is);
        } catch (IOException e) {
            logger.error("ppt07文件轉html出錯：",e);
            return false;
        } finally {
            try {
                is.close();
            } catch (IOException e) {
                logger.error("ppt07文件轉html關閉文件流失敗：",e);
            }
        }
        Dimension pgsize = ppt.getPageSize();
        List<XSLFSlide> pptPageXSLFSLiseList=ppt.getSlides();
        FileOutputStream out=null;
        String imghtml="";
        //保存圖片位置
        fileExists(getImageSavePath(filepath));
        for (int i = 0; i < pptPageXSLFSLiseList.size(); i++) {
            for(XSLFShape shape : pptPageXSLFSLiseList.get(i).getShapes()){
                //設置文字字體
                if(shape instanceof XSLFTextShape) {
                    XSLFTextShape tsh = (XSLFTextShape)shape;
                    for(XSLFTextParagraph p : tsh){
                        for(XSLFTextRun r : p){
                            r.setFontFamily("宋體");
                        }
                    }
                    //設置表格字體
                }else if(shape instanceof XSLFTable){
                    XSLFTable table = (XSLFTable)shape;
                    int rowSize = table.getNumberOfRows();
                    int columnSize = table.getNumberOfColumns();
                    for (int j = 0; j < rowSize; j++) {
                        for (int k = 0; k < columnSize; k++) {
                            for (int l =0;l <  table.getCell(j, k).getTextParagraphs().size();l++){
                                XSLFTextParagraph xslfTextRuns = table.getCell(j, k).getTextParagraphs().get(l);
                                for (int m = 0;m < xslfTextRuns.getTextRuns().size();m++){
                                    xslfTextRuns.getTextRuns().get(m).setFontFamily("宋體");
                                }
                            }
                        }
                    }
                }
            }
            BufferedImage img = new BufferedImage(pgsize.width, pgsize.height, BufferedImage.TYPE_INT_RGB);
            Graphics2D graphics = img.createGraphics();
            graphics.setPaint(Color.white);
            graphics.fill(new Rectangle2D.Float(0, 0, pgsize.width, pgsize.height));
            pptPageXSLFSLiseList.get(i).draw(graphics);
            //設置圖片存放位置
            String Imgname = getImageSavePath(filepath) + (i+1) + ".jpg";
            try {
                out = new FileOutputStream(Imgname);
                javax.imageio.ImageIO.write(img, "jpg", out);
            } catch (java.io.IOException e) {
                logger.error("ppt07文件轉html出錯：",e);
                try {
                    out.close();
                } catch (IOException e1) {
                    logger.error("ppt07文件轉html關閉流出錯：",e);
                }
                return false;
            }
            //圖片在html加載路徑
            String imgs=getImageUrl(filepath)+(i + 1) + ".jpg";
            imghtml+="<img src=\'"+imgs+"\' style=\'width:1200px;height:830px;vertical-align:text-bottom;\'><br><br><br><br>";
        }

        DOMSource domSource = new DOMSource();
        StreamResult streamResult = new StreamResult(out);
        TransformerFactory tf = TransformerFactory.newInstance();
        try {
            Transformer serializer = tf.newTransformer();
            serializer.setOutputProperty(OutputKeys.ENCODING, "utf-8");
            serializer.setOutputProperty(OutputKeys.INDENT, "yes");
            serializer.setOutputProperty(OutputKeys.METHOD, "html");
            serializer.transform(domSource, streamResult);
        } catch (TransformerException e) {
            logger.error("ppt07文件轉html出錯：",e);
            return false;
        }finally {
            try {
                out.close();
            } catch (java.io.IOException e) {
                logger.error("ppt07文件轉html關閉流出錯：",e);
                return false;
            }
        }

        String ppthtml="<html><head><META http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\"></head><body>"+imghtml+"</body></html>";
        try {
            FileUtils.writeStringToFile(new File(outputFile), ppthtml, "utf-8");
            return true;
        } catch (Exception e) {
            logger.error("ppt07文件轉html關閉流出錯：",e);
            return false;
        }
    }

    /**
     * 輸出文件流
     * @param content
     * @param path
     * @return
     */
    public static boolean writeFile(String content, String path){
        FileOutputStream fos = null;
        BufferedWriter bw = null;

        File file = new File(path);

        try {
            fos = new FileOutputStream(file);
            bw = new BufferedWriter(new OutputStreamWriter(fos, "UTF-8"));
            bw.write(content);
            return true;
        } catch (java.io.IOException e) {
            logger.error("輸出文件出錯",e);
            return false;
        }finally {
            try {
                if (null != bw) {
                    bw.close();
                }
                if (null != fos) {
                    fos.close();
                }
            } catch (java.io.IOException e) {
                logger.error("輸出文件關閉流出錯",e);
            }
        }
    }

    /**
     * 判斷文件夾是否存在，不存在則新建
     * @param path
     */
    private static void fileExists(String path) {
        File file = new File(path);
        if (!file.exists()){
            file.mkdirs();
        }
    }

    /**
     * 檢查文件類型
     * @param fileName
     * @return
     */
    public static boolean checkFile(String fileName,String type) {
        boolean flag = false;
        String suffixname = FileUtils.getFileExtension(fileName);
        if (suffixname != null && suffixname.equalsIgnoreCase(type)) {
            flag = true;
        }
        return flag;
    }

    /**
     * 根據文件全路徑獲取文件所在路徑
     * @param fileFullName
     * @return
     */
    public static String getFilePath(String fileFullName) {
        File file = new File(fileFullName);
        String filePath = fileFullName.replace(file.getName(),"");
        return filePath;
    }

    private static String getImageUrl(String filePath){
        filePath = filePath.replace(rootPath,"");
        //圖片引用地址須要去掉 rootpath
        return IMAGE_SERVER+FileUtils.getFileNameWithoutExtension(filePath)+"/";
    }

    private static String getImageSavePath(String filePath){
        return FileUtils.getFileNameWithoutExtension(filePath)+File.separator;
    }

}