如何编写自动抓取基金公告或净值文件的Python代码示例

我来提供一个实用的Python代码框架，用于自动抓取基金公告和净值文件，以满足LOF套利策略的数据需求。

📋 完整抓取方案框架

1. 抓取基金官方净值文件

以下是抓取国投瑞银基金净值的示例代码：

import requests
import pandas as pd
from datetime import datetime, timedelta
import time
import json

class FundDataFetcher:
    def __init__(self, fund_code="161226"):
        self.fund_code = fund_code
        self.headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
        }
    
    def fetch_official_nav(self, date_str=None):
        """
        从官方指定平台获取基金净值
        """
        if not date_str:
            date_str = datetime.now().strftime("%Y-%m-%d")
        
        try:
            # 方法1：尝试从巨潮资讯网获取（官方指定披露平台）
            url = f"http://www.cninfo.com.cn/new/disclosure/stock?stockCode={self.fund_code}"
            response = requests.get(url, headers=self.headers, timeout=10)
            
            # 解析HTML获取净值数据（实际需根据页面结构调整）
            # 这里展示从API获取的示例
            api_url = f"http://api.fund.eastmoney.com/f10/lsjz?fundCode={self.fund_code}&pageIndex=1&pageSize=1"
            api_headers = {
                **self.headers,
                'Referer': 'http://fundf10.eastmoney.com/'
            }
            
            response = requests.get(api_url, headers=api_headers, timeout=10)
            data = response.json()
            
            if data['Data'] and data['Data']['LSJZList']:
                nav_data = data['Data']['LSJZList'][0]
                return {
                    'date': nav_data['FSRQ'],
                    'nav': float(nav_data['DWJZ']),
                    'accumulated_nav': float(nav_data['LJJZ'])
                }
                
        except Exception as e:
            print(f"获取净值失败: {e}")
        
        return None
    
    def fetch_nav_history(self, days=30):
        """获取历史净值数据"""
        url = f"http://api.fund.eastmoney.com/f10/lsjz?fundCode={self.fund_code}&pageIndex=1&pageSize={days}"
        headers = {
            **self.headers,
            'Referer': 'http://fundf10.eastmoney.com/'
        }
        
        try:
            response = requests.get(url, headers=headers, timeout=10)
            data = response.json()
            
            nav_list = []
            if data['Data'] and data['Data']['LSJZList']:
                for item in data['Data']['LSJZList']:
                    nav_list.append({
                        'date': item['FSRQ'],
                        'nav': float(item['DWJZ']),
                        'change': item.get('JZZZL', '0%')
                    })
            
            return pd.DataFrame(nav_list)
            
        except Exception as e:
            print(f"获取历史净值失败: {e}")
            return pd.DataFrame()

　　

2. 抓取基金公告与申购状态

class FundAnnouncementFetcher:
    def __init__(self, fund_code="161226"):
        self.fund_code = fund_code
        self.base_url = "http://www.cninfo.com.cn/new/disclosure"
    
    def fetch_latest_announcements(self, limit=10):
        """
        获取最新基金公告
        重点关注：申购赎回状态、额度限制、基金合同修改等
        """
        try:
            # 构建查询参数
            params = {
                'fundCode': self.fund_code,
                'pageNum': 1,
                'pageSize': limit,
                'category': '基金公告'
            }
            
            response = requests.get(
                f"{self.base_url}/query",
                params=params,
                headers={'User-Agent': 'Mozilla/5.0'},
                timeout=15
            )
            
            announcements = []
            if response.status_code == 200:
                data = response.json()
                for item in data.get('announcements', []):
                    # 筛选关键公告类型
                    title = item.get('title', '')
                    if any(keyword in title for keyword in 
                          ['申购', '赎回', '限购', '额度', '暂停', '恢复']):
                        
                        announcements.append({
                            'title': title,
                            'date': item.get('date'),
                            'url': f"{self.base_url}/download/{item.get('id')}",
                            'content_preview': item.get('content', '')[:200]
                        })
            
            return announcements
            
        except Exception as e:
            print(f"获取公告失败: {e}")
            return []
    
    def check_subscription_status(self):
        """
        专项检查申购限制状态
        返回当前申购额度、是否暂停等信息
        """
        announcements = self.fetch_latest_announcements(limit=20)
        
        # 分析公告内容，提取申购限制信息
        current_status = {
            'can_subscribe': True,
            'daily_limit': None,  # 每日限额
            'last_update': None,
            'reason': '正常'
        }
        
        for ann in announcements:
            title = ann['title'].lower()
            content = ann['content_preview'].lower()
            
            # 检查暂停申购公告
            if '暂停申购' in title or '暂停申购' in content:
                current_status['can_subscribe'] = False
                current_status['reason'] = '已暂停申购'
                current_status['last_update'] = ann['date']
                break
                
            # 检查限购公告
            elif '限购' in title or '限额' in title:
                # 提取限额金额（简化示例，实际需更复杂的文本解析）
                import re
                amount_pattern = r'(\d+,?\d*)元'
                matches = re.search(amount_pattern, ann['title'] + content)
                
                if matches:
                    current_status['daily_limit'] = matches.group(1)
                    current_status['reason'] = f"每日限购{current_status['daily_limit']}元"
                    current_status['last_update'] = ann['date']
        
        return current_status

　　

3. 与QMT集成的数据服务模块

class QMTDataService:
"""将抓取的数据集成到QMT策略中"""

def __init__(self, fund_code="161226"):
self.fund_code = fund_code
self.nav_fetcher = FundDataFetcher(fund_code)
self.ann_fetcher = FundAnnouncementFetcher(fund_code)
self.cache_file = f"{fund_code}_nav_cache.json"

def get_trading_nav_data(self):
"""
交易日获取最新净值数据，供QMT策略调用
"""
# 1. 检查缓存
cached_data = self._load_cached_data()
today = datetime.now().strftime("%Y-%m-%d")

if cached_data and cached_data.get('date') == today:
return cached_data

# 2. 获取最新数据
nav_data = self.nav_fetcher.fetch_official_nav()

# 3. 获取申购状态
sub_status = self.ann_fetcher.check_subscription_status()

# 4. 合并数据并缓存
trading_data = {
**nav_data,
'subscription_status': sub_status,
'update_time': datetime.now().isoformat(),
'fund_code': self.fund_code
}

self._save_to_cache(trading_data)
return trading_data

def _load_cached_data(self):
"""加载缓存数据"""
try:
with open(self.cache_file, 'r', encoding='utf-8') as f:
return json.load(f)
except:
return None

def _save_to_cache(self, data):
"""保存数据到缓存"""
with open(self.cache_file, 'w', encoding='utf-8') as f:
json.dump(data, f, ensure_ascii=False, indent=2)

def should_trigger_arbitrage(self, market_price):
"""
综合判断是否触发套利
"""
data = self.get_trading_nav_data()

if not data or 'nav' not in data:
return {'should_trigger': False, 'reason': '净值数据缺失'}

# 检查申购状态
sub_status = data['subscription_status']
if not sub_status['can_subscribe']:
return {'should_trigger': False, 'reason': f"无法申购: {sub_status['reason']}"}

# 计算溢价率
premium_rate = (market_price - data['nav']) / data['nav']

# 考虑交易成本（示例值）
total_cost_rate = 0.006 # 0.6%

# 判断逻辑
if premium_rate > total_cost_rate:
return {
'should_trigger': True,
'direction': 'premium_arb',
'premium_rate': premium_rate,
'nav': data['nav'],
'market_price': market_price,
'daily_limit': sub_status.get('daily_limit')
}
elif premium_rate < -total_cost_rate:
return {
'should_trigger': True,
'direction': 'discount_arb',
'premium_rate': premium_rate
}

return {'should_trigger': False, 'reason': '溢价率未达阈值'}

 

 

# 在QMT策略中的使用方式
def initialize(context):
# 初始化数据服务
context.data_service = QMTDataService(fund_code="161226")

# 设置套利参数
context.min_premium = 0.015 # 1.5%最小溢价率
context.check_interval = 60 # 每60秒检查一次

def handle_data(context, data):
# 获取当前市价（从QMT）
market_price = get_market_data(
['close'],
stock_code=['161226.SZ'],
period='1m',
count=1
)['close'].iloc[-1]

# 获取净值并判断套利机会
decision = context.data_service.should_trigger_arbitrage(market_price)

if decision['should_trigger']:
log.info(f"发现套利机会: {decision}")

# 检查申购限额
if decision.get('direction') == 'premium_arb':
daily_limit = decision.get('daily_limit')
if daily_limit and daily_limit < 1000:
log.warning(f"申购限额过低: {daily_limit}元，可能无法有效套利")
return

# 执行套利逻辑（调用之前的套利函数）
execute_arbitrage(context, decision)

　　

4. 实际使用示例

 

⚠️ 关键注意事项

1. 法律合规性

   • 检查目标网站的robots.txt文件，尊重爬取频率限制

   • 添加适当的请求间隔（如time.sleep(1)）

   • 仅用于个人研究，避免商业用途

2. 数据源稳定性

   • 官方平台（如巨潮资讯）是最可靠的数据源

   • 东方财富等第三方API可能有频率限制

   • 实现多数据源备选机制

3. 错误处理

   • 增加重试机制和超时设置

   • 实现数据验证（检查净值范围合理性）

   • 记录日志以便调试

4. 性能优化

   • 缓存频繁访问的数据

   • 使用异步请求提高效率

   • 定期清理历史数据

🔄 完整工作流程

class QMTDataService:
    """将抓取的数据集成到QMT策略中"""
    
    def __init__(self, fund_code="161226"):
        self.fund_code = fund_code
        self.nav_fetcher = FundDataFetcher(fund_code)
        self.ann_fetcher = FundAnnouncementFetcher(fund_code)
        self.cache_file = f"{fund_code}_nav_cache.json"
        
    def get_trading_nav_data(self):
        """
        交易日获取最新净值数据，供QMT策略调用
        """
        # 1. 检查缓存
        cached_data = self._load_cached_data()
        today = datetime.now().strftime("%Y-%m-%d")
        
        if cached_data and cached_data.get('date') == today:
            return cached_data
        
        # 2. 获取最新数据
        nav_data = self.nav_fetcher.fetch_official_nav()
        
        # 3. 获取申购状态
        sub_status = self.ann_fetcher.check_subscription_status()
        
        # 4. 合并数据并缓存
        trading_data = {
            **nav_data,
            'subscription_status': sub_status,
            'update_time': datetime.now().isoformat(),
            'fund_code': self.fund_code
        }
        
        self._save_to_cache(trading_data)
        return trading_data
    
    def _load_cached_data(self):
        """加载缓存数据"""
        try:
            with open(self.cache_file, 'r', encoding='utf-8') as f:
                return json.load(f)
        except:
            return None
    
    def _save_to_cache(self, data):
        """保存数据到缓存"""
        with open(self.cache_file, 'w', encoding='utf-8') as f:
            json.dump(data, f, ensure_ascii=False, indent=2)
    
    def should_trigger_arbitrage(self, market_price):
        """
        综合判断是否触发套利
        """
        data = self.get_trading_nav_data()
        
        if not data or 'nav' not in data:
            return {'should_trigger': False, 'reason': '净值数据缺失'}
        
        # 检查申购状态
        sub_status = data['subscription_status']
        if not sub_status['can_subscribe']:
            return {'should_trigger': False, 'reason': f"无法申购: {sub_status['reason']}"}
        
        # 计算溢价率
        premium_rate = (market_price - data['nav']) / data['nav']
        
        # 考虑交易成本（示例值）
        total_cost_rate = 0.006  # 0.6%
        
        # 判断逻辑
        if premium_rate > total_cost_rate:
            return {
                'should_trigger': True,
                'direction': 'premium_arb',
                'premium_rate': premium_rate,
                'nav': data['nav'],
                'market_price': market_price,
                'daily_limit': sub_status.get('daily_limit')
            }
        elif premium_rate < -total_cost_rate:
            return {
                'should_trigger': True,
                'direction': 'discount_arb',
                'premium_rate': premium_rate
            }
        
        return {'should_trigger': False, 'reason': '溢价率未达阈值'}

# 主程序示例
def main():
    # 1. 初始化
    fetcher = QMTDataService("161226")
    
    # 2. 检查申购状态
    status = fetcher.ann_fetcher.check_subscription_status()
    print(f"当前申购状态: {status}")
    
    # 3. 获取最新净值
    nav_data = fetcher.get_trading_nav_data()
    print(f"最新净值: {nav_data}")
    
    # 4. 判断套利机会（假设当前市价2.1元）
    decision = fetcher.should_trigger_arbitrage(market_price=2.1)
    print(f"套利决策: {decision}")

if __name__ == "__main__":
    main()

# 主程序示例
def main():
# 1. 初始化
fetcher = QMTDataService("161226")

# 2. 检查申购状态
status = fetcher.ann_fetcher.check_subscription_status()
print(f"当前申购状态: {status}")

# 3. 获取最新净值
nav_data = fetcher.get_trading_nav_data()
print(f"最新净值: {nav_data}")

# 4. 判断套利机会（假设当前市价2.1元）
decision = fetcher.should_trigger_arbitrage(market_price=2.1)
print(f"套利决策: {decision}")

if __name__ == "__main__":
main()

　　

📊 后续优化建议

1. 增加数据源：同时监控多个数据源，提高可靠性

2. 实时监控：使用WebSocket或定时轮询，及时发现公告更新

3. 机器学习：对公告文本进行自然语言处理，自动识别限制条件

4. 集成到QMT：将数据服务作为QMT策略的常驻模块

这个框架提供了从数据抓取到决策判断的完整流程。实际使用时，你需要根据目标网站的实际HTML结构调整解析逻辑，并充分测试各种异常情况。