Python+Selenium自动化考后核验照片分类

import os
import time
import psutil
from selenium.webdriver.chrome.options import Options
from selenium import webdriver
from selenium.webdriver.common.by import By
import openpyxl
import urllib.request

# 检查是否已经存在浏览器实例，如果存在则不再创建
for proc in psutil.process_iter():
    try:
        if 'chrome' in proc.name() and '--remote-debugging-port=9224' in proc.cmdline():
            options = Options()
            options.add_experimental_option('debuggerAddress', '127.0.0.1:9224')
            break
    except:
        pass
else:
    # 创建浏览器实例
    os.system(r'start chrome --remote-debugging-port=9224 --user-data-dir='D:\评阅用'')
    options = Options()
    options.add_experimental_option('debuggerAddress', '127.0.0.1:9224')

# 在已有的浏览器实例中查找标签页
driver = webdriver.Chrome(options=options)
tabs = driver.window_handles
for tab in tabs:
    driver.switch_to.window(tab)
    if driver.title == '考后核验':
        print('登陆成功')
        break

# 打开Excel表格
wb = openpyxl.load_workbook(r'C:\Users\Administrator\Desktop\考后核验.xlsx')
sheet = wb.active

# 创建照片文件夹
photo_folder = r'C:\Users\Administrator\Desktop\照片'
if not os.path.exists(photo_folder):
    os.mkdir(photo_folder)

# 遍历时间单元K列
time_set = set()
for row in sheet.iter_rows(min_row=2, max_row=sheet.max_row, min_col=11, max_col=11):
    for cell in row:
        # 获取时间单元内的数据
        data = cell.value
        if data:
            time_set.add(data)

# 创建时间文件夹
time_folder_dict = {}
for time_data in time_set:
    folder_name = time.strftime('%Y-%m-%d %H-%M-%S', time.strptime(time_data, '%Y-%m-%d %H:%M:%S'))
    folder_path = os.path.join(photo_folder, folder_name)
    if not os.path.exists(folder_path):
        os.mkdir(folder_path)
    time_folder_dict[time_data] = folder_path

# 遍历表格内'A'列数据
for row in sheet.iter_rows(min_row=2, max_row=sheet.max_row, min_col=1, max_col=1):
    for cell in row:
        # 获取学号
        student_id = cell.value
        if student_id:
            # 在网页内查找考生信息
            search_input = driver.find_element(By.XPATH, '//input[@placeholder='姓名/考生号']')
            search_input.clear()
            search_input.send_keys(student_id)
            time.sleep(1)  # 等待页面加载完成
            search_button = driver.find_element(By.XPATH, '//button[@class='ant-btn ant-btn-primary']')
            driver.execute_script('arguments[0].click();', search_button)
            time.sleep(1)  # 等待页面加载完成
            
            # 获取考场编号
            room_number = None
            for room_cell in sheet.iter_rows(min_row=cell.row, max_row=cell.row, min_col=8, max_col=8):
                for room_cell in room_cell:
                    room_number = room_cell.value
                    break
            
            # 获取时间
            time_data = None
            for time_cell in sheet.iter_rows(min_row=cell.row, max_row=cell.row, min_col=11, max_col=11):
                for time_cell in time_cell:
                    time_data = time_cell.value
                    break
            
            if room_number and time_data:
                # 获取照片
                photo_div = driver.find_element(By.XPATH, '//div[@class='img-lazy-load']')
                photo_img = photo_div.find_element(By.TAG_NAME, 'img')
                photo_src = photo_img.get_attribute('src')
                
                # 创建考场编号文件夹
                room_folder_path = os.path.join(time_folder_dict[time_data], str(room_number))
                if not os.path.exists(room_folder_path):
                    os.mkdir(room_folder_path)
                
                # 下载照片
                photo_name = f'{student_id}.jpg'
                photo_path = os.path.join(room_folder_path, photo_name)
                urllib.request.urlretrieve(photo_src, photo_path)
                print(f'下载照片：{photo_path}')