data-acquisition.git

			@@ -2,38 +2,39 @@

			import cn.hutool.core.io.FileUtil;
			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;
			import net.sourceforge.tess4j.Tesseract;
			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.HSSFPicture;
			import org.apache.poi.hssf.usermodel.HSSFPictureData;
			import org.apache.poi.hssf.usermodel.HSSFSheet;
			import org.apache.poi.hssf.usermodel.HSSFWorkbook;
			import org.apache.poi.hwpf.extractor.WordExtractor;
			import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
			import org.apache.poi.openxml4j.opc.OPCPackage;
			import org.apache.poi.ss.usermodel.*;
			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.xwpf.extractor.XWPFWordExtractor;
			import java.awt.Color;
			import java.awt.image.BufferedImage;
			import javax.imageio.ImageIO;

			import java.io.*;
			import java.nio.file.Files;
			import java.sql.*;
			import java.util.*;
			import java.util.regex.Matcher;
			import java.util.regex.Pattern;

			public class TakeWords {

			private static final String splitIdentifier = "@-@"; // 自定义唯一标识符

			// 科学计数法正则模式
			private static final Pattern SCIENTIFIC_PATTERN = Pattern.compile(
			"^[+-]?\\d+(\\.\\d+)?[eE][+-]?\\d+$"
			);

			public static Object readWordFile(File file) {
			String result = "";
			@@ -56,32 +57,30 @@
			return result;
			}

			public static Object readExcelFile(File file) throws FileNotFoundException, IOException {
			public static Object readExcelFile(File file) throws IOException {
			StringBuilder result = new StringBuilder();
			//创建工作簿对象
			XSSFWorkbook xssfWorkbook = new XSSFWorkbook(new FileInputStream(file));
			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
			for (int i = 0; i < 1; i++) {
			XSSFSheet sheet = xssfWorkbook.getSheetAt(i);
			//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++) {
			//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);
			}
			result.append(aLine).append("\n");
			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);
			}
			result.append(aLine).append("\n");
			}
			return result.toString();
			}

			public static Object readExcelxlsFile(File file) throws IOException {
			StringBuilder result = new StringBuilder();
			String result = "";
			try (FileInputStream fis = new FileInputStream(file);
			Workbook workbook = new HSSFWorkbook(fis)) {
			// 获取第一个工作表
			@@ -90,41 +89,66 @@
			if (workbook instanceof HSSFWorkbook) {
			HSSFWorkbook hssfWorkbook = (HSSFWorkbook) workbook;
			List<HSSFPictureData> pictures = hssfWorkbook.getAllPictures();
			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) readPngFile1(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(",");
			//处理最后一张图片数据
			HSSFPictureData lastPicture = pictures.get(pictures.size()-1);
			// 获取图片类型
			String pictureType = lastPicture.suggestFileExtension();
			// 获取图片数据
			byte[] pictureData = lastPicture.getData();
			// 创建临时文件
			File tempFile = File.createTempFile(UUID.randomUUID().toString(), "." + pictureType);
			try (FileOutputStream fos = new FileOutputStream(tempFile)) {
			fos.write(pictureData);
			}
			String ocrResult = "";
			try {
			ocrResult = (String) readPngFile(tempFile);
			} catch (TesseractException e) {
			ocrResult = "OCR识别失败: " + e.getMessage();
			} 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
			@@ -200,9 +224,6 @@
			return tesseract.doOCR(file);
			}




			public static Object readTxtFile(File file) throws IOException {
			FileInputStream fin = new FileInputStream(file);
			InputStreamReader reader = new InputStreamReader(fin);
			@@ -220,21 +241,37 @@

			StringBuilder stringBuilder = new StringBuilder();
			// 创建 reader
			try (BufferedReader br = Files.newBufferedReader(file.toPath())) {
			// CSV文件的分隔符
			String DELIMITER = ",";
			// 按行读取
			String line;
			while ((line = br.readLine()) != null) {
			// 分割
			String[] columns = line.split(DELIMITER);
			// 打印行
			stringBuilder.append(String.join(splitIdentifier, columns)).append("\n");
			try (FileReader fileReader = new FileReader(file);
			CSVReader csvReader = new CSVReaderBuilder(fileReader).build()) {

			String[] nextLine;
			while ((nextLine = csvReader.readNext()) != null) {
			// 处理每一行数据
			for (String cell : nextLine) {
			if(StringUtils.isNotBlank(cell)){
			stringBuilder.append(scientificToNumber(cell)).append(splitIdentifier);
			}
			}
			stringBuilder.append("\n");
			}
			} catch (IOException ex) {
			ex.printStackTrace();
			} catch (IOException e) {
			e.printStackTrace();
			} catch (CsvValidationException e) {
			throw new RuntimeException(e);
			}
			return stringBuilder.toString();
			}

			/**
			* 将科学计数法转换为数字
			* @param cell
			* @return
			*/
			public static String scientificToNumber(String cell){
			if(SCIENTIFIC_PATTERN.matcher(cell).matches()){
			return String.valueOf(Double.parseDouble(cell));
			}
			return cell;
			}

			public static Object readMdbFile(File file, GetFileDto getFileDto) throws SQLException, ClassNotFoundException, InstantiationException, IllegalAccessException {
			@@ -280,6 +317,7 @@
			}
			tableMap.put("data", list);
			} catch (Exception e) {
			e.printStackTrace();
			} finally {
			closeA1l(conn, preparedStatement, rs);
			}
			@@ -300,6 +338,7 @@
			} catch (Exception ignore) {
			}
			}

			public static Object getMysqlFile(GetFileDto getFileDto){
			Map<String, Object> tableMap = new HashMap<>(16);
			// 从 GetFileDto 获取数据库名，对应【文件名称】字段
			@@ -313,7 +352,7 @@
			return R.failed("数据库名或表名不能为空");
			}
			// 数据库连接信息
			String url = "jdbc:mysql://localhost:3306/"+dbName+"?useSSL=false&serverTimezone=UTC&allowPublicKeyRetrieval=true";
			String url = "jdbc:mysql://localhost:3306/"+dbName+"?useSSL=false&serverTimezone=GMT%2B8&allowPublicKeyRetrieval=true&characterEncoding=utf8";
			Connection connection = null;
			PreparedStatement preparedStatement = null;
			ResultSet resultSet = null;
			@@ -324,9 +363,15 @@
			connection = DriverManager.getConnection(url, user, password);
			// 构建基础 SQL

			StringBuilder sql = new StringBuilder("SELECT * FROM ").append(table);
			String sql = "SELECT * FROM "+table+" WHERE 1=1";
			if(StringUtils.isNotBlank(getFileDto.getMdbEntrustCode())){
			sql+=" AND " + getFileDto.getMdbEntrustCode() + " = '" + getFileDto.getEntrustCode()+ "'";
			}
			if(StringUtils.isNotBlank(getFileDto.getMdbSampleCode())){
			sql+=" AND " + getFileDto.getMdbSampleCode() + " = '" + getFileDto.getSampleCode() + "'";
			}
			// 创建 PreparedStatement 对象执行 SQL
			preparedStatement = connection.prepareStatement(sql.toString());
			preparedStatement = connection.prepareStatement(sql);
			resultSet = preparedStatement.executeQuery();
			ResultSetMetaData metaData = resultSet.getMetaData();
			int columnCount = metaData.getColumnCount();
			@@ -339,11 +384,6 @@
			}
			dataList.add(rowData);
			}
			// while (resultSet.next()) {
			// double thinnestPoint = resultSet.getDouble("ThinnestPoint");
			// double averageThickness = resultSet.getDouble("AverageThickness");
			// dataList.add(new ThicknessData(thinnestPoint, averageThickness));
			// }
			tableMap.put("data", dataList);
			} catch (Exception e) {
			e.printStackTrace();