采集器调整

zouyu

2025-12-09 866a3e6cbd2df9841dfbbd733e1128938cef3e00

 src/main/java/com/chinaztt/mes/docx/util/TakeWords.java

@@ -1,8 +1,9 @@
package com.chinaztt.mes.docx.util;

import cn.hutool.core.io.FileUtil;
import cn.hutool.http.HttpRequest;
import cn.hutool.json.JSONUtil;
import com.chinaztt.mes.docx.dto.GetFileDto;
import com.chinaztt.mes.docx.dto.ThicknessData;
import com.opencsv.CSVReader;
import com.opencsv.CSVReaderBuilder;
import com.opencsv.exceptions.CsvValidationException;
@@ -10,20 +11,19 @@
import net.sourceforge.tess4j.TesseractException;
import org.apache.commons.lang3.ObjectUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.poi.POIXMLDocument;
import org.apache.poi.POIXMLTextExtractor;
import org.apache.poi.hssf.usermodel.HSSFPictureData;
import org.apache.poi.hssf.usermodel.HSSFWorkbook;
import org.apache.poi.hwpf.extractor.WordExtractor;
import org.apache.poi.ooxml.POIXMLDocument;
import org.apache.poi.ooxml.extractor.POIXMLTextExtractor;
import org.apache.poi.openxml4j.opc.OPCPackage;
import org.apache.poi.poifs.filesystem.POIFSFileSystem;
import org.apache.poi.ss.usermodel.Sheet;
import org.apache.poi.ss.usermodel.Workbook;
import org.apache.poi.xssf.usermodel.XSSFSheet;
import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import org.apache.poi.ss.usermodel.WorkbookFactory;
import org.apache.poi.xwpf.extractor.XWPFWordExtractor;

import java.io.*;
import java.nio.file.Files;
import java.sql.*;
import java.util.*;
import java.util.regex.Pattern;
@@ -58,29 +58,42 @@
        return result;
    }

    public static Object readExcelFile(File file) throws IOException {
    /**
     * 读取excel文件,兼容.xlsx,.xls格式
     * @param file
     * @return
     */
    public static Object readExcelFile(File file) {
        StringBuilder result = new StringBuilder();
        //创建工作簿对象
        XSSFWorkbook xssfWorkbook = new XSSFWorkbook(Files.newInputStream(file.toPath()));
        //获取工作簿下sheet的个数 只读取第一个sheet
//            int sheetNum = xssfWorkbook.getNumberOfSheets();
        //遍历工作簿中的所有数据
        XSSFSheet sheet = xssfWorkbook.getSheetAt(0);
        //获取最后一行的num，即总行数。此处从0开始
        int maxRow = sheet.getLastRowNum();
        for (int row = 1; row <= maxRow; row++) {
            //获取最后单元格num，即总单元格数 ***注意：此处从1开始计数***
            int maxRol = sheet.getRow(row).getLastCellNum();
            StringBuilder aLine = new StringBuilder();
            for (int rol = 0; rol < maxRol; rol++) {
                aLine.append(sheet.getRow(row).getCell(rol)).append(splitIdentifier);
        try (
                FileInputStream fis = new FileInputStream(file);
                Workbook workbook = WorkbookFactory.create(fis)
        ) {
            //遍历工作簿中的所有数据
            Sheet sheet = workbook.getSheetAt(0);
            //获取最后一行的num，即总行数。此处从0开始
            int maxRow = sheet.getLastRowNum();
            for (int row = 0; row <= maxRow; row++) {
                int maxRol = sheet.getRow(row).getLastCellNum();
                StringBuilder aLine = new StringBuilder();
                for (int rol = 0; rol < maxRol; rol++) {
                    aLine.append(sheet.getRow(row).getCell(rol)).append(splitIdentifier);
                }
                result.append(aLine).append("\n");
            }
            result.append(aLine).append("\n");
        } catch (Exception e) {
            e.printStackTrace();
        }
        return result.toString();
    }

    public static Object readExcelxlsFile(File file) throws IOException {
    /**
     * 读取excel文件中的图片内容
     * @param file
     * @return
     * @throws IOException
     */
    public static Object readPngContextInExcel(File file) throws IOException {
        String result = "";
        try (FileInputStream fis = new FileInputStream(file);
             Workbook workbook = new HSSFWorkbook(fis)) {
@@ -101,100 +114,18 @@
                try (FileOutputStream fos = new FileOutputStream(tempFile)) {
                    fos.write(pictureData);
                }
                String ocrResult = "";
                String ocrResult;
                try {
                    ocrResult = (String) readPngFile(tempFile);
                } catch (TesseractException e) {
                    ocrResult = "OCR识别失败: " + e.getMessage();
                    ocrResult = ocrImageContext(tempFile.getAbsolutePath());
                } finally {
                    // 删除临时文件
                    tempFile.delete();
                }
                result = ocrResult;
//                String ocrText = fixOcrText(ocrResult);
//                result.append("OCR Result:").append(ocrText).append(",");


//                for (HSSFPictureData picture : pictures) {
//                    // 获取图片类型
//                    String pictureType = picture.suggestFileExtension();
//                    // 获取图片数据
//                    byte[] pictureData = picture.getData();
//                    // 创建临时文件
//                    File tempFile = File.createTempFile(UUID.randomUUID().toString(), "." + pictureType);
//                    try (FileOutputStream fos = new FileOutputStream(tempFile)) {
//                        fos.write(pictureData);
//                    }
//                    // 图片预处理
////                    File processedFile = preprocessImage(tempFile, pictureType);
//                    // 调用 readPngFile1 方法读取图片文字信息
//                    String ocrResult = "";
//                    try {
//                        ocrResult = (String) readPngFile(tempFile);
////                        ocrResult = (String) readPngFile(tempFile);
////                        ocrResult = (String) readPngFile(processedFile);
//                    } catch (TesseractException e) {
//                        ocrResult = "OCR识别失败: " + e.getMessage();
//                    } finally {
//                        // 删除临时文件
////                        tempFile.delete();
////                        processedFile.delete();
//                    }
//
//                    // 将图片信息添加到结果中
////                    result.append("Picture Type: ").append(pictureType)
////                            .append(", Picture Size: ").append(pictureData.length)
////                            .append(" bytes")
////                            .append(", OCR Result: ").append(ocrResult)
////                            .append(",");
//                    String ocrText = fixOcrText(ocrResult);
//                    result.append("OCR Result:").append(ocrText).append(",");
//                }
            }
//
//            // 遍历每一行
//            for (Row row : sheet) {
//                // 遍历每一列
//                for (Cell cell : row) {
//                    CellType cellType = CellType.forInt(cell.getCellType());
//                    switch (cellType) {
//                        case STRING:
//                            result.append(cell.getStringCellValue());
//                            break;
//                        case NUMERIC:
//                            if (DateUtil.isCellDateFormatted(cell)) {
//                                result.append(cell.getDateCellValue());
//                            } else {
//                                result.append(cell.getNumericCellValue());
//                            }
//                            break;
//                        case BOOLEAN:
//                            result.append(cell.getBooleanCellValue());
//                            break;
//                        case FORMULA:
//                            result.append(cell.getCellFormula());
//                            break;
//                        default:
//                            result.append("");
//                    }
//                    result.append("\t");
//                }
//                result.append("\n");
//            }
        }
        return result;

    }

    // 修正 OCR 识别文本中的错误关键词
    public static String fixOcrText(String ocrText) {
        // 定义错误关键词和正确内容的映射，这里处理“击 宇 强 庞”修正为“击穿强度”
        // 考虑到可能有空格分隔，用正则匹配包含这些字的内容
        ocrText = ocrText.replaceAll("击\\s*宇\\s*强\\s*庞", "击穿强度");
        // 还可以继续添加其他错误修正，比如下面假设“电 压 \\(HV\\)”里的空格影响，也修正下
        ocrText = ocrText.replaceAll("电\\s*压\\s*\\(HV\\)", "电压(KV)");
        ocrText = ocrText.replaceAll("电\\s*流\\s*\\(nt\\)", "电流(mA)");
        return ocrText;
    }

    public static Object readPngFile1(File file) throws IOException, TesseractException {
@@ -232,6 +163,7 @@
        StringBuilder stringBuilder = new StringBuilder();
        String strTmp = "";
        while ((strTmp = buffReader.readLine()) != null) {
            strTmp = strTmp.replaceAll("\t",",");
            stringBuilder.append(strTmp).append("\n");
        }
        buffReader.close();
@@ -280,8 +212,8 @@
        Properties prop = new Properties();
        //设置编码
        prop.put("charSet", "UTF-8");
        prop.put("user", "");
        prop.put("password", "");
        prop.put("user",  StringUtils.isNotBlank(getFileDto.getDbUserName())?getFileDto.getDbUserName():"");
        prop.put("password", StringUtils.isNotBlank(getFileDto.getDbPassword())?getFileDto.getDbPassword():"");
        //数据地址
        String dbUrl = "jdbc:ucanaccess://" + file.getPath();
        //引入驱动
@@ -295,8 +227,6 @@
        try {
            List<Object> list = new ArrayList<>();
            //遍历获取多张表数据
//            String s = "select * from " + getFileDto.getDbFileName() + " where 1=1" + getFileDto.getMdbEntrustCode() + " = '" + getFileDto.getEntrustCode() +
//                    "' and " + getFileDto.getMdbSampleCode() + " = '" + getFileDto.getSampleCode() + "'";
            String s = "select * from " + getFileDto.getDbFileName() + " where 1=1";
            if(StringUtils.isNotBlank(getFileDto.getMdbEntrustCode())){
                s+=" and " + getFileDto.getMdbEntrustCode() + " = '" + getFileDto.getEntrustCode()+ "'";
@@ -304,6 +234,10 @@
            if(StringUtils.isNotBlank(getFileDto.getMdbSampleCode())){
                s+=" and " + getFileDto.getMdbSampleCode() + " = '" + getFileDto.getSampleCode() + "'";
            }
            if(StringUtils.isNotBlank(getFileDto.getBushingColorField())){
                s+=" and " + getFileDto.getBushingColorField() + " = '" + getFileDto.getBushingColor() + "'";
            }

            preparedStatement = conn.prepareStatement(s);
            rs = preparedStatement.executeQuery();
            ResultSetMetaData data = rs.getMetaData();
@@ -365,11 +299,18 @@
            // 构建基础 SQL

            String sql = "SELECT * FROM "+table+" WHERE 1=1";
            if(StringUtils.isNotBlank(getFileDto.getMdbEntrustCode())){
                sql+=" AND " + getFileDto.getMdbEntrustCode() + " = '" + getFileDto.getEntrustCode()+ "'";
            if(StringUtils.isNotBlank(getFileDto.getMdbEntrustCode()) ){
                sql+=" AND (" + getFileDto.getMdbEntrustCode() + " = TRIM('" + getFileDto.getEntrustCode()+ "')";
                if(StringUtils.isNotBlank(getFileDto.getLotBatchNo())){
                    sql+=" OR "+ getFileDto.getMdbEntrustCode() + " = TRIM('" + getFileDto.getLotBatchNo()+ "')";
                }
                sql+=")";
            }
            if(StringUtils.isNotBlank(getFileDto.getMdbSampleCode())){
                sql+=" AND " + getFileDto.getMdbSampleCode() + " = '" + getFileDto.getSampleCode() + "'";
            }
            if(StringUtils.isNotBlank(getFileDto.getCableTag())){
                sql+=" AND Color = '" + getFileDto.getCableTag() + "'";
            }
            // 创建 PreparedStatement 对象执行 SQL
            preparedStatement = connection.prepareStatement(sql);
@@ -456,6 +397,19 @@
        return tableMap;
    }

    /**
     * ocr识别图片内容
     * @param imagePath 图片路径
     * @return
     */
    public static String ocrImageContext(String imagePath){
        //调用ocr识别服务
        Map<String,Object> jsonMap = new HashMap<>();
        jsonMap.put("imagePath",imagePath);
        String requestBody = JSONUtil.toJsonStr(jsonMap);
        return HttpRequest.post("localhost:8080/ocr/recognize").body(requestBody).execute().body();
    }

    public static Object readPngFile(File file) throws IOException, TesseractException {
        String canonicalPath32 = FileUtil.file(".", "/jre_32/tessdata").getCanonicalPath();
        String canonicalPath64 = FileUtil.file(".", "/jre_64/tessdata").getCanonicalPath();