编写所有基础工具，提供word util pdf util

pom.xml

@@ -55,7 +55,7 @@
         <dependency>
             <groupId>org.apache.pdfbox</groupId>
             <artifactId>pdfbox</artifactId>
-            <version>2.0.23</version>
+            <version>2.0.24</version>
             <type>bundle</type>
         </dependency>
         <dependency>
@@ -100,6 +100,32 @@
             <version>1.4.0</version> <!-- 请检查最新版本 -->
         </dependency>
 
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi</artifactId>
+            <version>5.2.2</version> <!-- 请根据你的实际情况选择版本 -->
+        </dependency>
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi-ooxml</artifactId>
+            <version>5.2.2</version> <!-- 请根据你的实际情况选择版本 -->
+        </dependency>
+
+
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi-scratchpad</artifactId>
+            <version>5.2.2</version>
+        </dependency>
+
+        <!-- https://mvnrepository.com/artifact/net.coobird/thumbnailator -->
+        <dependency>
+            <groupId>net.coobird</groupId>
+            <artifactId>thumbnailator</artifactId>
+            <version>0.4.20</version>
+        </dependency>
+
+
 
     </dependencies>
 

src/main/java/com/docus/sw/fenpan/FenpanService.java

@@ -2,6 +2,7 @@ package com.docus.sw.fenpan;
 
 import com.alibaba.excel.util.FileUtils;
 import com.docus.sw.Config;
+import com.docus.sw.word.PdfBoxUtils;
 import lombok.extern.slf4j.Slf4j;
 import org.apache.commons.imaging.ImageInfo;
 import org.apache.commons.imaging.ImageReadException;
@@ -86,6 +87,10 @@ public class FenpanService {
         new FenpanService().fenpan(saveUrl, readUrl);
     }
 
+    public List<Document> getDocument(Pieces piece){
+        return null;
+    }
+
 
     public Map<String, Zong> readFile(String readUrl) {
         //读取文件夹。
@@ -99,54 +104,37 @@ public class FenpanService {
             //根据文件类型
             if (piece.getFileTypeEnum() == FileTypeEnum.WORD) {
                 //从word 直接提取图片
+                //提取图片为document ，然后用于后面判断
 
             } else if (piece.getFileTypeEnum() == FileTypeEnum.PDF) {
-                // 从pdf 提取图片
+                // 从pdf 提取图片,
+                //提取图片为document ，然后用于后面判断
+                File file = new File("temp");
+                if(!file.exists()){
+                    file.mkdirs();
+                }
+
+                List<Document> documentList = new ArrayList<>();
+                try {
+                    PdfBoxUtils.pdf2image(piece.getAbsolutePath(),file.getAbsolutePath());
+                    File[] files = file.listFiles();
+                    for(File pdfImg:files){
+                        getDocumentList(documentList, pdfImg);
+                    }
+                } catch (IOException e) {
+                    log.error(e.getMessage(),e);
+                }
 
+                piece.put(documentList);
+                //删除对应的temp 文件
+                file.delete();
             } else {
                 //是图片，直接从图片提取
                 List<Document> documentList = new ArrayList<>();
                 File sourceFile = new File(piece.getAbsolutePath());
                 File[] files = sourceFile.listFiles();
                 for (File file : files) {
-                    //非图片模式，跳过。
+                    getDocumentList(documentList, file);
-                    if (!(file.getName().endsWith(".jpg") || file.getName().endsWith(".png")
-                            || file.getName().endsWith(".jpeg") || file.getName().endsWith(".tif")
-                            || file.getName().endsWith(".tiff")) || file.getName().endsWith(".jp2") || file.getName().endsWith(".jpm")|| file.getName().endsWith(".gif")) {
-                        continue;
-                    }
-
-                    if (file.getName().endsWith(".jp2") || file.getName().endsWith(".jpm")) {
-                        // 读取 JPEG 2000 图像文件
-
-                        try {
-                            BufferedImage image = ImageIO.read(file);
-                            int height = image.getHeight();
-                            int width = image.getWidth();
-                            Document document = new Document(width, height, 300);
-                            documentList.add(document);
-                        } catch (IOException e) {
-                            throw new RuntimeException(e);
-                        }
-
-                    } else {
-                        try {
-                            ImageInfo imageInfo = Imaging.getImageInfo(file);
-                            int height = imageInfo.getHeight();
-                            int width = imageInfo.getWidth();
-                            int physicalHeightDpi = imageInfo.getPhysicalHeightDpi();
-                            Document document = new Document(width, height, physicalHeightDpi);
-                            documentList.add(document);
-                        } catch (IOException e) {
-                            FileUtils.delete(file);
-                            throw new RuntimeException("非图片格式", e);
-                        } catch (ImageReadException e) {
-                            FileUtils.delete(file);
-                            throw new RuntimeException(e);
-                        } catch (IllegalArgumentException e) {
-                            FileUtils.delete(file);
-                        }
-                    }
                 }
                 piece.put(documentList);
             }
@@ -204,6 +192,47 @@ public class FenpanService {
 
     }
 
+    private static void getDocumentList(List<Document> documentList, File file) {
+        //非图片模式，跳过。
+        if (!(file.getName().endsWith(".jpg") || file.getName().endsWith(".png")
+                || file.getName().endsWith(".jpeg") || file.getName().endsWith(".tif")
+                || file.getName().endsWith(".tiff")) || file.getName().endsWith(".jp2") || file.getName().endsWith(".jpm")|| file.getName().endsWith(".gif")) {
+            return;
+        }
+
+        if (file.getName().endsWith(".jp2") || file.getName().endsWith(".jpm")) {
+            // 读取 JPEG 2000 图像文件
+
+            try {
+                BufferedImage image = ImageIO.read(file);
+                int height = image.getHeight();
+                int width = image.getWidth();
+                Document document = new Document(width, height, 300);
+                documentList.add(document);
+            } catch (IOException e) {
+                throw new RuntimeException(e);
+            }
+
+        } else {
+            try {
+                ImageInfo imageInfo = Imaging.getImageInfo(file);
+                int height = imageInfo.getHeight();
+                int width = imageInfo.getWidth();
+                int physicalHeightDpi = imageInfo.getPhysicalHeightDpi();
+                Document document = new Document(width, height, physicalHeightDpi);
+                documentList.add(document);
+            } catch (IOException e) {
+                FileUtils.delete(file);
+                throw new RuntimeException("非图片格式", e);
+            } catch (ImageReadException e) {
+                FileUtils.delete(file);
+                throw new RuntimeException(e);
+            } catch (IllegalArgumentException e) {
+                FileUtils.delete(file);
+            }
+        }
+    }
+
 
     private void findAllDir(String absolutePath, List<Pieces> allDirectory) {
         File sourceFile = new File(absolutePath);

src/main/java/com/docus/sw/souyin/SuoyinService.java

@@ -70,6 +70,9 @@ public class SuoyinService {
                     List<Pieces> pieceList = new ArrayList<>();
                     File[] piecesFile = rollFile.listFiles();
                     for(File piece:piecesFile){
+
+                        //区分word和pdf
+
                         List<Document> documentList = new ArrayList<>();
                         Pieces pieces = new Pieces(FileTypeEnum.JPG, piece.getAbsolutePath(), piece.getName());
                         for (File docfile : piece.listFiles()) {

src/main/java/com/docus/sw/word/GetPicsDocx.java

@@ -0,0 +1,35 @@
+package com.docus.sw.word;
+
+import java.io.File;
+import java.io.FileInputStream;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.util.List;
+
+import org.apache.poi.xwpf.extractor.XWPFWordExtractor;
+import org.apache.poi.xwpf.usermodel.XWPFDocument;
+import org.apache.poi.xwpf.usermodel.XWPFPictureData;
+
+public class GetPicsDocx {
+    public static void main(String[] args) {
+        String path = "E:\\上海项目测试\\文档\\35.docx";
+        File file = new File(path);
+        try {
+            FileInputStream fis = new FileInputStream(file);
+            XWPFDocument document = new XWPFDocument(fis);
+            XWPFWordExtractor xwpfWordExtractor = new XWPFWordExtractor(document);
+            String text = xwpfWordExtractor.getText();
+            System.out.println(text);
+            List<XWPFPictureData> picList = document.getAllPictures();
+            for (XWPFPictureData pic : picList) {
+                System.out.println(pic.getPictureType() + File.separator + pic.suggestFileExtension() + File.separator + pic.getFileName());
+                byte[] bytev = pic.getData();
+                FileOutputStream fos = new FileOutputStream("E:\\上海项目测试\\docxImage\\" + pic.getFileName());
+                fos.write(bytev);
+            }
+            fis.close();
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
+    }
+}

src/main/java/com/docus/sw/word/PdfBoxUtils.java

@@ -0,0 +1,192 @@
+package com.docus.sw.word;
+
+import net.coobird.thumbnailator.Thumbnails;
+import net.coobird.thumbnailator.filters.Canvas;
+import net.coobird.thumbnailator.geometry.Positions;
+import org.apache.pdfbox.contentstream.PDFStreamEngine;
+import org.apache.pdfbox.contentstream.operator.Operator;
+import org.apache.pdfbox.cos.COSBase;
+import org.apache.pdfbox.cos.COSName;
+import org.apache.pdfbox.io.MemoryUsageSetting;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDPageTree;
+import org.apache.pdfbox.pdmodel.encryption.AccessPermission;
+import org.apache.pdfbox.pdmodel.encryption.InvalidPasswordException;
+import org.apache.pdfbox.pdmodel.encryption.PDEncryption;
+import org.apache.pdfbox.pdmodel.encryption.SecurityHandler;
+import org.apache.pdfbox.pdmodel.encryption.StandardProtectionPolicy;
+import org.apache.pdfbox.pdmodel.encryption.StandardSecurityHandler;
+import org.apache.pdfbox.pdmodel.graphics.PDXObject;
+import org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject;
+import org.apache.pdfbox.pdmodel.graphics.image.PDImageXObject;
+
+import javax.imageio.ImageIO;
+import java.awt.*;
+import java.awt.image.BufferedImage;
+import java.io.File;
+import java.io.IOException;
+import java.util.List;
+
+/**
+ * pdfbox 加解密
+ */
+public class PdfBoxUtils {
+
+    private static final String vectorKey = "AesVKeyDocus@702";
+
+    /**
+     * 加密
+     *
+     * @param key
+     * @param sourceFilePath
+     * @param destFilePath
+     * @throws Exception
+     */
+    public static void encrypt(String key, String sourceFilePath, String destFilePath) throws Exception {
+
+        File file = new File(sourceFilePath);
+        PDDocument load = PDDocument.load(file);
+
+        AccessPermission permissions = new AccessPermission();
+        // 是否可以插入/删除/旋转页面
+        permissions.setCanAssembleDocument(false);
+        // 是否可以复制和提取内容
+        permissions.setCanExtractContent(false);
+
+        permissions.setCanExtractForAccessibility(false);
+        // 设置用户是否可以填写交互式表单字段（包括签名字段）
+        permissions.setCanFillInForm(false);
+        // 设置用户是否可以修改文档
+        permissions.setCanModify(false);
+        // 设置用户是否可以添加或修改文本注释并填写交互式表单字段，如果canModify()返回true，则创建或修改交互式表单字段（包括签名字段）。
+        permissions.setCanModifyAnnotations(false);
+        // 设置用户是否可以打印。
+        permissions.setCanPrint(false);
+        // 设置用户是否可以降级格式打印文档
+        permissions.setCanPrintDegraded(false);
+
+        StandardProtectionPolicy p = new StandardProtectionPolicy(key, key, permissions);
+        SecurityHandler sh = new StandardSecurityHandler(p);
+        sh.prepareDocumentForEncryption(load);
+        PDEncryption encryptionOptions = new PDEncryption();
+        encryptionOptions.setSecurityHandler(sh);
+
+        load.setEncryptionDictionary(encryptionOptions);
+        load.save(destFilePath);
+        load.close();
+    }
+
+    /**
+     * 解密
+     *
+     * @param key
+     * @param sourceFilePath
+     * @param destFilePath
+     * @throws InvalidPasswordException
+     * @throws IOException
+     */
+    public static void decrypt(String key, String sourceFilePath, String destFilePath) throws InvalidPasswordException, IOException {
+        File file = new File(sourceFilePath);
+        PDDocument load = PDDocument.load(file, key);
+        load.setAllSecurityToBeRemoved(true);
+        load.save(destFilePath);
+        load.close();
+    }
+
+
+    /**
+     * pdfbox pdf to image
+     */
+    public static class DefPdfToImageEngine extends PDFStreamEngine {
+        public int imageNum = 1;
+
+        private final String des;
+        private final String name;
+
+        public DefPdfToImageEngine(String name, String des) {
+            this.des = des;
+            this.name = name;
+        }
+
+        @Override
+        protected void processOperator(Operator operator, List<COSBase> operands) throws IOException {
+            String operatorName = operator.getName();
+            if ("Do".equals(operatorName)) {
+                COSName objName = (COSName) operands.get(0);
+                PDXObject pdxObject = this.getResources().getXObject(objName);
+                if (pdxObject instanceof PDImageXObject) {
+                    PDImageXObject imageXObject = (PDImageXObject) pdxObject;
+                    BufferedImage image = imageXObject.getImage();
+                    ImageIO.write(image,"jpg",new File(this.des + this.name + "-" + this.imageNum + ".jpg"));
+                    ++this.imageNum;
+                } else if (pdxObject instanceof PDFormXObject) {
+                    PDFormXObject form = (PDFormXObject) pdxObject;
+                    this.showForm(form);
+                }
+            } else {
+                super.processOperator(operator, operands);
+            }
+
+        }
+    }
+
+    /**
+     * pdf to image
+     *
+     * @param src
+     * @param des
+     * @throws IOException
+     */
+    public static void pdf2image(String src, String des) throws IOException {
+        File pdfFile = new File(src);
+        PDDocument load = PDDocument.load(pdfFile, MemoryUsageSetting.setupMixed(1024 * 1024));
+        DefPdfToImageEngine imageEngine = new DefPdfToImageEngine(pdfFile.getName().substring(0, pdfFile.getName().lastIndexOf(".")), des);
+        for (PDPage page : load.getPages()) {
+            imageEngine.processPage(page);
+        }
+        load.close();
+    }
+
+
+    /**
+     * 指定提取哪一页
+     *
+     * @param src
+     * @param des
+     * @param pageNum
+     * @throws IOException
+     */
+    public static void pdf2image(String src, String des, int pageNum) throws IOException {
+        File pdfFile = new File(src);
+        PDDocument load = PDDocument.load(pdfFile, MemoryUsageSetting.setupMixed(1024 * 1024));
+        DefPdfToImageEngine imageEngine = new DefPdfToImageEngine(pdfFile.getName().substring(0, pdfFile.getName().lastIndexOf(".")), des);
+
+        PDPageTree pages = load.getPages();
+
+        for (int i = 0; i < pages.getCount(); i++) {
+
+            if (pageNum == i + 1) {
+                imageEngine.processPage(pages.get(i));
+                break;
+            }
+        }
+
+        load.close();
+    }
+
+    public static void main(String[] args) throws Exception {
+        pdf2image("C:\\Users\\zhanghai\\Desktop\\桌面\\test\\a.pdf", "C:\\Users\\zhanghai\\Desktop\\桌面\\test\\");
+
+
+//        long a1 = System.currentTimeMillis();
+//        encrypt(vectorKey, "D:\\workspace\\docus\\业务梳理\\测试数据\\伟淞\\98m.pdf", "D:\\workspace\\docus\\业务梳理\\测试数据\\伟淞\\98m-encrypt.pdf");
+//        long b1 = System.currentTimeMillis();
+//        System.out.println("加密使用时间：" + (b1 - a1));
+//
+//        long a2 = System.currentTimeMillis();
+//        decrypt(vectorKey, "D:\\workspace\\docus\\业务梳理\\测试数据\\伟淞\\98m-encrypt.pdf", "D:\\workspace\\docus\\业务梳理\\测试数据\\伟淞\\98m-decrypt.pdf");
+//        long b2 = System.currentTimeMillis();
+//        System.out.println("解密使用时间：" + (b2 - a2));
+    }
+}

src/main/java/com/docus/sw/word/ReadImgDoc.java

@@ -0,0 +1,43 @@
+package com.docus.sw.word;
+
+import java.io.*;
+import java.util.*;
+
+
+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;
+
+public class ReadImgDoc {
+
+    public static void main(String[] args) throws Exception {
+        new ReadImgDoc().readPicture("C:\\Users\\zhanghai\\Desktop\\桌面\\test\\a.doc");
+    }
+
+    private void readPicture(String path)throws Exception{
+        FileInputStream in=new FileInputStream(new File(path));
+        HWPFDocument doc=new HWPFDocument(in);
+        int length=doc.characterLength();
+        PicturesTable pTable=doc.getPicturesTable();
+        // int TitleLength=doc.getSummaryInformation().getTitle().length();
+
+        //  System.out.println(TitleLength);
+        // System.out.println(length);
+        for (int i=0;i<length;i++){
+            Range range=new Range(i, i+1,doc);
+
+            CharacterRun cr=range.getCharacterRun(0);
+            if(pTable.hasPicture(cr)){
+                Picture pic=pTable.extractPicture(cr, false);
+                String afileName=pic.suggestFullFileName();
+                OutputStream out=new FileOutputStream(new File("C:\\Users\\zhanghai\\Desktop\\桌面\\test\\"+UUID.randomUUID()+afileName));
+                pic.writeImageContent(out);
+
+            }
+        }
+
+    }
+
+}