Python实现自动爬取学生照片并分类保存

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
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)

for data in time_set:
    # 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'))
    folder_name = folder_name.replace(':', '_')  # 替换冒号字符
    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
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

            # 获取考场编号和姓名
            try:
                room_element = driver.find_element(By.XPATH, '//td[contains(text(),'{}')]'.format(data))
                room_number = room_element.text
                name_element = driver.find_element(By.XPATH, '//td[contains(text(),'{}')]/following-sibling::td'.format(data))
                name = name_element.text
            except:
                continue

            # 创建考场编号文件夹
            room_folder_path = os.path.join(folder_path, room_number)
            if not os.path.exists(room_folder_path):
                os.mkdir(room_folder_path)

            # 保存照片
            photo_element = driver.find_element(By.XPATH, '//td[contains(text(),'{}')]/following-sibling::td/img'.format(data))
            photo_url = photo_element.get_attribute('src')
            photo_extension = os.path.splitext(photo_url)[1]
            photo_name = '{}_{}'.format(student_id, name) + photo_extension
            photo_path = os.path.join(room_folder_path, photo_name)
            driver.get(photo_url)
            with open(photo_path, 'wb') as f:
                f.write(driver.page_source.encode('utf-8'))