import org.apache.pdfbox.pdmodel.PDDocument;
    import org.apache.pdfbox.text.PDFTextStripper;

    import java.io.File;
    import java.io.IOException;
    public class PDFBoxDemo {
        public static void main(String[] args) throws IOException {
            PDDocument document = PDDocument.load(new File("D:\\pdf\\test.pdf"));
            PDFTextStripper stripper = new PDFTextStripper();
            String text = stripper.getText(document);
            System.out.println(text);
            document.close();
        }
    }

 /**
    * 獲取文檔坐標(biāo)
    * @param  file PDF文件對(duì)象
    * @param sourceTex 匹配的字符
    * @return 坐標(biāo)
    */
   public static Point getPoint(File file,String sourceTex) {
       Point point = new Point();
       //獲取文檔坐標(biāo)
      try {
        PDDocument document =  Loader.loadPDF(file);
        PDFTextStripper textStripper = new PDFTextStripper() {
            @Override
            protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
                if (text.contains(targetText)) {
                    TextPosition textPositionStart = textPositions.get(0);
                    TextPosition textPositionEnd = textPositions.get(textPositions.size()-1);
                    point.setX(textPositionStart.getX());
                    point.setY(textPositionStart.getY()); 
                }
            }
        };

        textStripper.setSortByPosition(true);
        textStripper.setStartPage(1);
        textStripper.setEndPage(document.getNumberOfPages());

        textStripper.getText(document);

        document.close();

        } catch (IOException e) {
            e.printStackTrace();
        }
        return point;
    }

 public static void main(String[] args) {
        String filePath = "D:\\test\\test.pdf";
         try {
            PDDocument document = Loader.loadPDF(file);
            PDFTextStripperByArea  textStripper = new PDFTextStripperByArea ();
            Rectangle rectangle = new Rectangle(80,120, 250,10);
            String regionName = "regionName";
            textStripper.addRegion(regionName, rectangle);
            PDPage page = document.getPage(0);
            textStripper.extractRegions(page);
            String text = textStripper.getTextForRegion(regionName);
  
            System.out.println(text);
          
            textStripper.setSortByPosition(true);
            textStripper.setStartPage(1);
            textStripper.setEndPage(document.getNumberOfPages());
            textStripper.getText(document);
            document.close();
        }catch (IOException e) {
            e.printStackTrace();
        }
      
    }

package com.example.demo;
import cn.hutool.poi.excel.ExcelUtil;
import cn.hutool.poi.excel.ExcelWriter;
import com.alibaba.fastjson2.JSON;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.text.PDFTextStripper;
import org.apache.pdfbox.text.PDFTextStripperByArea;
import org.apache.pdfbox.text.TextPosition;
import org.springframework.boot.test.autoconfigure.data.cassandra.DataCassandraTest;


import java.awt.*;
import java.awt.geom.Rectangle2D;
import java.io.File;
import java.io.IOException;
import java.util.*;
import java.util.List;
import java.util.stream.Collectors;

/**
 * Desc: 驗(yàn)證pdfbox的可行性
 *
 * @author admin
 * @date since 2023/8/8 18:44
 */

public class PdfDemo {
	//要匹配的位置內(nèi)容點(diǎn)
    private  static final String[] target= {"name", "address"};
    public static void main(String[] args) {
       ExcelWriter excelWriter= ExcelUtil.getWriter("D:\\test\\pdf\\test.xls");
       String folderPath = "D:\\test\\pdf";
       File folder = new File(folderPath);
       if (folder.exists() && folder.isDirectory()) {
           List<Map<String,Object>>  mps =  listPdfFiles(folder);
           excelWriter.write(mps, true);
       } else {
           System.out.println("Invalid folder path.");
       }
       excelWriter.close();
    }
	/**
     * 獲取pdf文件列表
     *
     * @param folder 文件夾
     * @return {@code List<Map<String,Object>>}
     */
    private static  List<Map<String,Object>>  listPdfFiles(File folder) {
        List<Map<String,Object>> mps = new ArrayList<>();
        File[] files = folder.listFiles();
        if (files != null) {
            for (File file : files) {
                if (file.isDirectory()) {
                    listPdfFiles(file); // 遞歸調(diào)用，處理子文件夾
                } else {
                    String fileName = file.getName();
                    if (fileName.toLowerCase().endsWith(".pdf")) {
                        mps.add(getLineData(file));
                    }
                }
            }
        }
        return mps;
    }
    /**
     * 行數(shù)據(jù)
     *
     * @param file 文件
     * @return {@code Map<String,Object>}
     */
    public static Map<String,Object> getLineData(File file){
        Map<String,Object> lineData = new HashMap<>(target.length+2);
        List<Point> pointList =  getPoint(file);
        String[]  arr=  getPointValue(file, pointList.stream().map(s -> new Rectangle(s.getX(), s.getY(), 260, 10)).toArray(Rectangle[]::new));
        if(arr.length>=target.length) {
        for(int i=0;i<target.length;i++)
        {
            lineData.put(target[i], arr[i]);
        }
            lineData.put("fileName", file.getName().toLowerCase().replace(".pdf", ""));
        }
      return lineData;
    }
 	/**
     * 獲得PDF指定坐標(biāo)點(diǎn)文本值
     *
     * @param file       文件
     * @param rectangles 矩形坐標(biāo)
     * @return {@code String[]}
     */
    public  static String[] getPointValue( File file,Rectangle... rectangles){
        String[] textArr = new String[rectangles.length];
       // String text="";
        try {
            PDDocument document = Loader.loadPDF(file);
            PDFTextStripperByArea  textStripper = new PDFTextStripperByArea ();

            for(int i = 0; i < rectangles.length;i++   ) {
                Rectangle rectangle =rectangles[i];
                String regionName = "regionName"+rectangle.getX()+rectangle.getY();
                textStripper.addRegion(regionName, rectangle);
                PDPage page = document.getPage(0);
                textStripper.extractRegions(page);
                // 獲取區(qū)域的text
                String text = textStripper.getTextForRegion(regionName);
                text = text.replace("\u0000","-").replace(" ","");
                System.out.println(">>text"+text);
                textArr[i]=text;
            }

            textStripper.setSortByPosition(true);
            textStripper.setStartPage(1);
            textStripper.setEndPage(document.getNumberOfPages());

            textStripper.getText(document);

            document.close();
        }catch (IOException e) {
            e.printStackTrace();
        }

        return  textArr;
    }

    public  static List<Point> getPoint( File file){
        List<Point> pointList=new ArrayList<>();
        try {
        PDDocument document =  Loader.loadPDF(file);
        PDFTextStripper textStripper = new PDFTextStripper() {
            @Override
            protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
                for(String target:target){
                    if (text.contains(target)) {
                        Point point = new Point();
                        TextPosition textPositionEnd = textPositions.get(textPositions.size() - 1);
                        point.setX((int) textPositionEnd.getEndX());
                        point.setY((int) textPositionEnd.getY());
                        pointList.add(point);
                    }
                }
            }
        };

        textStripper.setSortByPosition(true);
        textStripper.setStartPage(1);
        textStripper.setEndPage(document.getNumberOfPages());
        textStripper.getText(document);
        document.close();

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

        System.out.println(">>>>>pointList" + JSON.toJSONString(pointList));
        return pointList;
    }
}