YunNanDsBase/src/main/java/com/dsideal/base/Tools/ReadCityDoc.java

package com.dsideal.base.Tools;

import cn.hutool.core.io.FileUtil;
import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
import org.apache.poi.openxml4j.util.ZipSecureFile;
import org.apache.poi.ss.usermodel.CellType;
import org.apache.poi.ss.usermodel.CellValue;
import org.apache.poi.ss.usermodel.FormulaEvaluator;
import org.apache.poi.xssf.usermodel.*;
import org.apache.poi.xwpf.usermodel.XWPFChart;
import org.apache.poi.xwpf.usermodel.XWPFDocument;
import org.apache.poi.xwpf.usermodel.XWPFParagraph;

import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.util.List;

public class ReadCityDoc {
    //https://blog.csdn.net/a346736962/article/details/123037797
    public static void main(String[] args) throws IOException, InvalidFormatException {
        String directoryPath = "D:\\dsWork\\YunNanDsBase\\Doc\\全省及州市县区人口与教育报告集20241023\\16个州市报告2022\\分析报告20240510";
        directoryPath="D:\\dsWork\\YunNanDsBase\\Doc\\全省及州市县区人口与教育报告集20241023\\133个县区报告2022";
        //遍历workingPath下的所有文件,注意，需要递归所有子目录下的所有文件
        System.out.println("开始遍历目录下的所有文件");
        // 调用 Hutool 的 FileUtil.loopFiles 方法递归获取所有文件
        List<File> files = FileUtil.loopFiles(directoryPath, file -> {
            // 这里可以添加你的过滤条件，如果不需要过滤，返回 true 即可
            return true;
        });
        int cnt = 0;
        // 打印所有文件的路径
        for (File file : files) {
            String fileName = file.getName();
            //判断是否为docx文件
            if (fileName.endsWith(".docx") && !fileName.startsWith("~")) {
                cnt++;
                //读取文件
                String inputUrl = file.getAbsolutePath();
                InputStream is = new FileInputStream(inputUrl);
                //读取excel报错 Zip bomb detected! The file would exceed the max.
                //https://blog.csdn.net/baidu_19473529/article/details/109601558
                ZipSecureFile.setMinInflateRatio(-1.0d);
                XWPFDocument doc = new XWPFDocument(is);
                //图表
                List<XWPFChart> charts = doc.getCharts();

                System.out.println("图表数量=" + charts.size());

                for (XWPFChart chart : charts) {
                    XSSFWorkbook workbook = chart.getWorkbook();
                    XSSFSheet sheet = workbook.getSheetAt(0);
                    //遍历一下sheet
                    for (int i = 0; i < sheet.getPhysicalNumberOfRows(); i++) {
                        //遍历行
                        XSSFRow row = sheet.getRow(i);
                        if (row == null) continue;
                        for (int j = 0; j < row.getPhysicalNumberOfCells(); j++) {
                            //遍历列
                            XSSFCell cell = row.getCell(j);
                            // 创建公式计算器
                            FormulaEvaluator evaluator = new XSSFFormulaEvaluator(workbook);

                            // 检查单元格是否包含公式
                            if (cell!=null && cell.getCellType() == CellType.FORMULA) {
                                // 计算公式并获取结果
                                CellValue evaluatedValue = evaluator.evaluate(cell);
                                System.out.println("Calculated Value: " + evaluatedValue.formatAsString());
                            }
                            if (cell != null)
                                System.out.print(cell + " ");
                            else {
                                System.out.print("null ");
                            }
                        }
                        System.out.println();
                    }
                }
                //段落
                List<XWPFParagraph> paragraphs = doc.getParagraphs();
                for (XWPFParagraph paragraph : paragraphs) {
                    String text = paragraph.getText();
                    if (text.startsWith("图")) {
                        System.out.println(text);
                    }

                }
            }
        }
        System.out.println("共读取" + cnt + "个文件");
    }
}