From 3b258a2fd69dd2c4fcf291773672cc154514569a Mon Sep 17 00:00:00 2001
From: zouyu <2723363702@qq.com>
Date: 星期二, 22 七月 2025 17:10:53 +0800
Subject: [PATCH] 采集器.db与.mysql方法调用错误问题修复
---
src/main/java/com/chinaztt/mes/docx/util/TakeWords.java | 160 +++++++++++++++++++---------------------------------
1 files changed, 59 insertions(+), 101 deletions(-)
diff --git a/src/main/java/com/chinaztt/mes/docx/util/TakeWords.java b/src/main/java/com/chinaztt/mes/docx/util/TakeWords.java
index 39ba107..2814371 100644
--- a/src/main/java/com/chinaztt/mes/docx/util/TakeWords.java
+++ b/src/main/java/com/chinaztt/mes/docx/util/TakeWords.java
@@ -3,37 +3,38 @@
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 = "@-@"; // 鑷畾涔夊敮涓�鏍囪瘑绗�
+
+ // 绉戝璁℃暟娉曟鍒欐ā寮�
+ private static final Pattern SCIENTIFIC_PATTERN = Pattern.compile(
+ "^[+-]?\\d+(\\.\\d+)?[eE][+-]?\\d+$"
+ );
public static Object readWordFile(File file) {
String result = "";
@@ -101,20 +102,19 @@
fos.write(pictureData);
}
// 鍥剧墖棰勫鐞�
- File processedFile = preprocessImage(tempFile, pictureType);
-// ocrResult = (String) readPngFile(tempFile);
-
+// File processedFile = preprocessImage(tempFile, pictureType);
// 璋冪敤 readPngFile1 鏂规硶璇诲彇鍥剧墖鏂囧瓧淇℃伅
String ocrResult = "";
try {
ocrResult = (String) readPngFile1(tempFile);
-// ocrResult = (String) readPngFile1(processedFile);
+// ocrResult = (String) readPngFile(tempFile);
+// ocrResult = (String) readPngFile(processedFile);
} catch (TesseractException e) {
ocrResult = "OCR璇嗗埆澶辫触: " + e.getMessage();
} finally {
// 鍒犻櫎涓存椂鏂囦欢
tempFile.delete();
- processedFile.delete();
+// processedFile.delete();
}
// 灏嗗浘鐗囦俊鎭坊鍔犲埌缁撴灉涓�
@@ -173,81 +173,6 @@
return ocrText;
}
- /**
- * 瀵瑰浘鐗囪繘琛岄澶勭悊锛屽寘鎷伆搴﹀寲銆佷簩鍊煎寲鍜岄攼鍖�
- * @param inputFile 杈撳叆鐨勫浘鐗囨枃浠�
- * @param formatName 鍥剧墖鏍煎紡鍚嶇О
- * @return 澶勭悊鍚庣殑鍥剧墖鏂囦欢
- * @throws IOException 璇诲彇鎴栧啓鍏ュ浘鐗囨椂鍙兘鎶涘嚭鐨勫紓甯�
- */
- private static File preprocessImage(File inputFile, String formatName) throws IOException {
- // 璇诲彇鍥剧墖
- BufferedImage image = ImageIO.read(inputFile);
-
- // 鐏板害鍖�
- image = convertToGrayscale(image);
- // 浜屽�煎寲
- image = applyThreshold(image, 128);
- // 閿愬寲
- image = applySharpening(image);
-
- // 鍒涘缓澶勭悊鍚庣殑涓存椂鏂囦欢
- File outputFile = File.createTempFile(UUID.randomUUID().toString(), "." + formatName);
- ImageIO.write(image, formatName, outputFile);
- return outputFile;
- }
-
- /**
- * 灏嗗浘鐗囪浆鎹负鐏板害鍥�
- * @param image 杈撳叆鐨勫浘鐗�
- * @return 鐏板害鍖栧悗鐨勫浘鐗�
- */
- private static BufferedImage convertToGrayscale(BufferedImage image) {
- BufferedImage grayImage = new BufferedImage(
- image.getWidth(), image.getHeight(), BufferedImage.TYPE_BYTE_GRAY);
- grayImage.getGraphics().drawImage(image, 0, 0, null);
- return grayImage;
- }
-
- /**
- * 瀵瑰浘鐗囪繘琛屼簩鍊煎寲澶勭悊
- * @param image 杈撳叆鐨勫浘鐗�
- * @param threshold 浜屽�煎寲闃堝��
- * @return 浜屽�煎寲鍚庣殑鍥剧墖
- */
- private static BufferedImage applyThreshold(BufferedImage image, int threshold) {
- BufferedImage binaryImage = new BufferedImage(
- image.getWidth(), image.getHeight(), BufferedImage.TYPE_BYTE_BINARY);
- for (int y = 0; y < image.getHeight(); y++) {
- for (int x = 0; x < image.getWidth(); x++) {
- int rgb = image.getRGB(x, y);
- int gray = (rgb >> 16) & 0xff;
- if (gray < threshold) {
- binaryImage.setRGB(x, y, Color.BLACK.getRGB());
- } else {
- binaryImage.setRGB(x, y, Color.WHITE.getRGB());
- }
- }
- }
- return binaryImage;
- }
-
- /**
- * 瀵瑰浘鐗囪繘琛岄攼鍖栧鐞�
- * @param image 杈撳叆鐨勫浘鐗�
- * @return 閿愬寲鍚庣殑鍥剧墖
- */
- private static BufferedImage applySharpening(BufferedImage image) {
- float[] sharpenMatrix = {
- 0f, -1f, 0f,
- -1f, 5f, -1f,
- 0f, -1f, 0f
- };
- java.awt.image.Kernel kernel = new java.awt.image.Kernel(3, 3, sharpenMatrix);
- java.awt.image.ConvolveOp op = new java.awt.image.ConvolveOp(kernel, java.awt.image.ConvolveOp.EDGE_NO_OP, null);
- return op.filter(image, null);
- }
-
public static Object readPngFile1(File file) throws IOException, TesseractException {
// 鑾峰彇 tessdata 鐩綍鐨勭粷瀵硅矾寰�
String arch = System.getProperty("sun.arch.data.model");
@@ -296,21 +221,52 @@
StringBuilder stringBuilder = new StringBuilder();
// 鍒涘缓 reader
- try (BufferedReader br = Files.newBufferedReader(file.toPath())) {
- // CSV鏂囦欢鐨勫垎闅旂
- String DELIMITER = ",";
- // 鎸夎璇诲彇
- String line;
- while ((line = br.readLine()) != null) {
- // 鍒嗗壊
- String[] columns = line.split(DELIMITER);
- // 鎵撳嵃琛�
- stringBuilder.append(String.join(splitIdentifier, columns)).append("\n");
+// try (BufferedReader br = Files.newBufferedReader(file.toPath())) {
+// // CSV鏂囦欢鐨勫垎闅旂
+// String DELIMITER = ",";
+// // 鎸夎璇诲彇
+// String line;
+// System.out.println(br.readLine());
+// while ((line = br.readLine()) != null) {
+// // 鍒嗗壊
+// String[] columns = line.split(DELIMITER);
+// // 鎵撳嵃琛�
+// stringBuilder.append(String.join(splitIdentifier, columns)).append("\n");
+// }
+// } catch (IOException ex) {
+// ex.printStackTrace();
+// }
+ try (FileReader fileReader = new FileReader(file);
+ CSVReader csvReader = new CSVReaderBuilder(fileReader).build()) {
+
+ String[] nextLine;
+ while ((nextLine = csvReader.readNext()) != null) {
+ // 澶勭悊姣忎竴琛屾暟鎹�
+ 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();
+ }
+
+ /**
+ * 灏嗙瀛﹁鏁版硶杞崲涓烘暟瀛�
+ * @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 {
@@ -466,6 +422,8 @@
} else {
path = canonicalPath64.replaceAll("/chi_sim.traineddata", "").replaceAll("\\\\", "/");
}
+ // 璁剧疆 TESSDATA_PREFIX 鐜鍙橀噺
+// System.setProperty("TESSDATA_PREFIX", path);
//璁剧疆閰嶇疆鏂囦欢澶瑰井瑙嗐�佽瘑鍒瑷�銆佽瘑鍒ā寮�
Tesseract tesseract = new Tesseract();
tesseract.setDatapath(path);
--
Gitblit v1.9.3