为实现上海二手房数据的爬取并保存为CSV文件，可参考以下思路与代码示例： 思路上，编写`process_region`函数遍历每个区域页面，提取标题、单价、总价等信息并以字典格式存储。在`main`函数中设置异步任务，利用`asyncio.Semaphore`控制并发数，避免对服务器造成过大压力。最后将收集的数据整合到`pandas DataFrame`并保存为CSV文件，使用`if __name__ == "__main__":`作为程序入口启动。 以下是示例代码： ```python import asyncio import aiohttp from bs4 import BeautifulSoup import pandas as pd headers = { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.138 Safari/537.36' } async def fetch(session, url): async with session.get(url, headers=headers) as response: return await response.text() async def process_region(session, region_url): html = await fetch(session, region_url) soup = BeautifulSoup(html, 'html.parser') # 假设这里提取标题、单价、总价等信息 house_list = [] # 这里需要根据实际页面结构修改选择器 for house in soup.select('.house-item'): title = house.select_one('.title').text.strip() unit_price = house.select_one('.unit-price').text.strip() total_price = house.select_one('.total-price').text.strip() house_info = { 'title': title, 'unit_price': unit_price, 'total_price': total_price } house_list.append(house_info) return house_list async def main(): semaphore = asyncio.Semaphore(5) # 控制并发数 async with aiohttp.ClientSession() as session: tasks = [] # 这里假设需要爬取多个区域的二手房信息，需要替换为实际区域URL region_urls = ['https://example.com/region1', 'https://example.com/region2'] for url in region_urls: async with semaphore: task = asyncio.create_task(process_region(session, url)) tasks.append(task) results = await asyncio.gather(*tasks) all_houses = [house for sublist in results for house in sublist] df = pd.DataFrame(all_houses) df.to_csv('shanghai_second_hand_houses.csv', index=False) if __name__ == "__main__": asyncio.run(main()) ```