Python自动化考后核验：提取考生照片并保存

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

# 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
if not os.path.exists(r'C:\Users\Administrator\Desktop\照片'):
    os.mkdir(r'C:\Users\Administrator\Desktop\照片')

# 遍历时间单元K列，并创建子文件夹（重复数据只创建一次）
folder_names = 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:
            # 创建子文件夹
            folder_name = time.strftime('%Y-%m-%d %H-%M-%S', time.strptime(data, '%Y-%m-%d %H:%M:%S'))
            if folder_name not in folder_names:
                folder_names.add(folder_name)
                folder_path = os.path.join(r'C:\Users\Administrator\Desktop\照片', folder_name)
                if not os.path.exists(folder_path):
                    os.mkdir(folder_path)

# 遍历表格内A列数据，在网页内查找考生信息
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
            
            # 遍历时间单元K列数据，提取时间并与网页数据对比
            for row_k in sheet.iter_rows(min_row=2, max_row=sheet.max_row, min_col=11, max_col=11):
                for cell_k in row_k:
                    # 获取时间单元内的数据
                    data_k = cell_k.value
                    if data_k:
                        # 获取网页时间元素
                        time_elements = driver.find_elements(By.XPATH, '//td[@class="ant-table-column-has-actions ant-table-column-has-sorters" and contains(text(), '' + data_k + '')]')
                        if time_elements:
                            # 提取时间元素中的时间
                            for time_element in time_elements:
                                time_text = time_element.text
                                # 提取照片链接
                                photo_element = time_element.find_element(By.XPATH, './following-sibling::td/div/div/img')
                                photo_url = photo_element.get_attribute('src')
                                # 保存照片
                                folder_name = time.strftime('%Y-%m-%d %H-%M-%S', time.strptime(data_k, '%Y-%m-%d %H:%M:%S'))
                                folder_path = os.path.join(r'C:\Users\Administrator\Desktop\照片', folder_name)
                                photo_file_path = os.path.join(folder_path, student_id + '.jpg')
                                try:
                                    # 使用requests库下载图片
                                    import requests
                                    response = requests.get(photo_url)
                                    with open(photo_file_path, 'wb') as f:
                                        f.write(response.content)
                                    print(f'照片保存成功: {photo_file_path}')
                                except Exception as e:
                                    print(f'照片保存失败: {e}')

# 关闭浏览器
driver.quit()