TII_Automation.py

# 導入所需的套件
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.common.exceptions import TimeoutException
import pytesseract
from PIL import Image, ImageEnhance, ImageFilter
import numpy as np
import cv2
import io
import requests
import time
import os
import urllib3

# 忽略 SSL 警告
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

# 設定 Tesseract 執行檔路徑
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'

def setup_image_folder():
    """建立圖片存放資料夾"""
    # 獲取目前腳本的完整路徑
    script_dir = os.path.dirname(os.path.abspath(__file__))
    # 在腳本目錄下建立 images 資料夾
    image_folder = os.path.join(script_dir, 'images')
    
    # 如果資料夾不存在，就建立它
    if not os.path.exists(image_folder):
        os.makedirs(image_folder)
        print(f"已建立圖片資料夾: {image_folder}")
    
    return image_folder

def setup_driver():
    """設定並返回 Chrome WebDriver"""
    options = webdriver.ChromeOptions()
    options.add_argument('--disable-gpu')
    options.add_argument('--no-sandbox')
    options.add_argument('--ignore-certificate-errors')
    return webdriver.Chrome(options=options)

def click_initial_search(driver):
    """點擊初始搜尋按鈕"""
    try:
        # 訪問初始頁面
        driver.get("https://insprod.tii.org.tw/")
        print("已訪問初始頁面")
        
        # 等待並點擊第一個搜尋按鈕 (使用ID優先)
        wait = WebDriverWait(driver, 10)
        search_button = wait.until(EC.element_to_be_clickable((By.ID, "Go22")))
        print("找到搜尋按鈕")
        search_button.click()
        print("已點擊搜尋按鈕")
        
        # 等待頁面加載
        time.sleep(2)
        return True
        
    except TimeoutException:
        print("等待初始搜尋按鈕超時")
        return False
    except Exception as e:
        print(f"點擊初始搜尋按鈕時發生錯誤: {str(e)}")
        return False

def process_image(img_path):
    """最小程度處理圖片，保持原始藍色"""
    try:
        # 讀取圖片
        img = cv2.imread(img_path)
        if img is None:
            print(f"無法讀取圖片: {img_path}")
            return None

        # 稍微增強對比度但保持原色
        lab = cv2.cvtColor(img, cv2.COLOR_BGR2LAB)
        l, a, b = cv2.split(lab)
        
        # 輕微增強亮度通道
        clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
        enhanced_l = clahe.apply(l)
        
        # 合併回去
        enhanced = cv2.merge([enhanced_l, a, b])
        result = cv2.cvtColor(enhanced, cv2.COLOR_LAB2BGR)
        
        # 稍微放大一點點
        scale = 1.2
        result = cv2.resize(result, None, fx=scale, fy=scale, interpolation=cv2.INTER_CUBIC)
        
        return result
        
    except Exception as e:
        print(f"圖片處理過程中發生錯誤: {str(e)}")
        return None

def get_captcha_text(driver, debug_mode=True):
    """處理驗證碼圖片並進行辨識"""
    image_folder = setup_image_folder()
    temp_img_path = os.path.join(image_folder, "captcha_temp.png")
    processed_path = os.path.join(image_folder, "processed_captcha.png")
    
    try:
        # 獲取驗證碼圖片
        full_img_url = "https://insprod.tii.org.tw/bmp.ashx"
        print(f"正在下載驗證碼圖片...")
        
        headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) Chrome/91.0.4472.124',
            'Referer': 'https://insprod.tii.org.tw/Query.aspx'
        }
        
        response = requests.get(
            full_img_url, 
            headers=headers,
            verify=False,
            timeout=10
        )
        
        if response.status_code != 200:
            print(f"下載圖片失敗，狀態碼: {response.status_code}")
            return None
            
        # 儲存原始圖片
        with open(temp_img_path, "wb") as f:
            f.write(response.content)
        
        print(f"原始驗證碼圖片已保存至: {temp_img_path}")
        
        # 處理圖片
        processed_image = process_image(temp_img_path)
        if processed_image is None:
            print("圖片處理失敗")
            return None
            
        cv2.imwrite(processed_path, processed_image)
        print(f"處理後的圖片已保存至: {processed_path}")
        
        # 使用不同的 OCR 配置嘗試識別
        configs = [
            '--psm 7 --oem 3 -c tessedit_char_whitelist=0123456789',
            '--psm 8 --oem 3 -c tessedit_char_whitelist=0123456789',
            '--psm 6 --oem 3 -c tessedit_char_whitelist=0123456789'
        ]
        
        results = []
        for i, config in enumerate(configs, 1):
            try:
                # 直接使用原圖進行識別
                img = Image.open(processed_path)
                text = pytesseract.image_to_string(img, config=config).strip()
                digits = ''.join(filter(str.isdigit, text))
                
                print(f"OCR配置 {i}: {config}")
                print(f"識別結果: {text}")
                print(f"過濾後數字: {digits}")
                
                if len(digits) == 4:
                    results.append(digits)
                    
            except Exception as e:
                print(f"OCR配置 {i} 識別失敗: {str(e)}")
        
        if results:
            most_common = max(set(results), key=results.count)
            print(f"最終選擇的識別結果: {most_common}")
            return most_common
            
        print("沒有有效的識別結果")
        return None
        
    except Exception as e:
        print(f"驗證碼識別過程中發生錯誤: {str(e)}")
        import traceback
        traceback.print_exc()
        return None
    
def fill_and_submit_form(driver, captcha_text):
    """填寫並提交表單"""
    try:
        # 等待輸入框出現
        wait = WebDriverWait(driver, 10)
        input_field = wait.until(
            EC.presence_of_element_located((By.CSS_SELECTOR, "input[name='bmpC3']"))
        )
        
        # 輸入驗證碼
        input_field.clear()
        input_field.send_keys(captcha_text)
        
        print(f"已輸入驗證碼: {captcha_text}")
        
        # 找到並點擊提交按鈕
        submit_button = driver.find_element(By.CSS_SELECTOR, "input[name='Go2222']")
        submit_button.click()
        
        return True
        
    except Exception as e:
        print(f"表單提交錯誤: {str(e)}")
        return False

def main():
    """主程式"""
    driver = None
    # driver = None 的設置有兩個主要目的：
        # 初始化變數，確保 finally 區塊可以檢查 driver 是否存在
        # 在異常情況下安全地關閉瀏覽器：如果不設置 driver = None，在 setup_driver() 失敗時 finally 區塊會拋出 NameError。
    try:
        # 設置驅動
        driver = setup_driver()
        
        # 步驟1: 點擊初始搜尋按鈕
        if not click_initial_search(driver):
            print("初始搜尋按鈕點擊失敗")
            return
            
        print("成功進入查詢頁面")
        time.sleep(2)  # 等待頁面完全加載
        
        # 步驟2: 處理驗證碼和表單
        captcha_text = get_captcha_text(driver, debug_mode=True)
        
        if captcha_text and len(captcha_text) == 4:  # 確保是4位數
            print(f"識別到的驗證碼: {captcha_text}")
            
            # 填寫並提交表單
            if fill_and_submit_form(driver, captcha_text):
                print("表單提交成功")
            else:
                print("表單提交失敗")
        else:
            print(f"驗證碼識別失敗或格式不正確: {captcha_text}")
            
    except Exception as e:
        print(f"執行過程中發生錯誤: {str(e)}")
    finally:
        # 等待一下看結果
        time.sleep(5)
        if driver:
            driver.quit()

if __name__ == "__main__":
    main()