""" 舆情监控 - Sentiment Monitoring Module 功能： - 直接用全量数据做关键词过滤 - 监控特定关键词的舆情 """ from typing import List, Dict, Optional, Any from dataclasses import dataclass, field from enum import Enum from datetime import datetime import asyncio class SentimentType(Enum): """舆情类型""" ALL = "all" # 全部 POSITIVE = "positive" # 正面 NEGATIVE = "negative" # 负面 NEUTRAL = "neutral" # 中性 @dataclass class SentimentConfig: """舆情监控配置""" keywords: List[str] = field(default_factory=list) # 监控关键词 sentiment_type: SentimentType = SentimentType.ALL # 舆情类型 items_per_platform: int = 10 # 每个平台获取的条目数 total_items: int = 50 # 总条目数限制 include_platforms: List[str] = None # 包含的平台 exclude_platforms: List[str] = None # 排除的平台 class SentimentMonitor: """舆情监控类""" def __init__(self, api_client=None, formatter=None): """ 初始化舆情监控 Args: api_client: API客户端实例（可选） formatter: 格式化器实例（可选） """ self.api_client = api_client self.formatter = formatter self.all_platforms = self._get_default_platforms() def _get_default_platforms(self) -> List[str]: """获取默认的全部平台列表""" return [ "weibo", "zhihu", "douban-group", "douban-movie", "ithome", "36kr", "sspai", "csdn", "juejin", "genshin", "miyoushe", "bilibili", "hupu", "sina-news", "netease-news", "qq-news", "sina-money", "eastmoney", "xueqiu", "autohome", "懂车帝", "mafengwo", "ctrip", "dianping", "xiaohongshu", "weibo" ] async def fetch_all_hot_data(self, limit_per_platform: int = 10) -> List[Dict[str, Any]]: """ 一次性获取全部平台的热榜数据 Args: limit_per_platform: 每个平台获取的条目数 Returns: 全部平台的热榜数据列表 """ all_items = [] if self.api_client: for platform in self.all_platforms: try: items = await self.api_client.get_hot榜单(platform, limit=limit_per_platform) if items: for item in items: item["source_platform"] = platform all_items.extend(items) except Exception as e: print(f"Error fetching {platform}: {e}") continue return all_items async def monitor_keywords(self, keywords: List[str], config: Optional[SentimentConfig] = None) -> Dict[str, Any]: """ 监控关键词舆情（新版：先全部获取，再关键词过滤） Args: keywords: 关键词列表 config: 配置对象（可选） Returns: 舆情监控数据 """ if config is None: config = SentimentConfig(keywords=keywords) # 步骤1：全部获取 all_items = await self.fetch_all_hot_data(limit_per_platform=config.items_per_platform) # 步骤2：关键词过滤 filtered_items = self._filter_by_keywords(all_items, config.keywords) # 平台过滤 if config.include_platforms: filtered_items = [item for item in filtered_items if item.get("source_platform") in config.include_platforms] if config.exclude_platforms: filtered_items = [item for item in filtered_items if item.get("source_platform") not in config.exclude_platforms] # 限制总数 filtered_items = filtered_items[:config.total_items] return { "keywords": config.keywords, "total_items": len(filtered_items), "items": filtered_items } def _filter_by_keywords(self, items: List[Dict], keywords: List[str]) -> List[Dict]: """ 按关键词过滤 Args: items: 条目列表 keywords: 关键词列表 Returns: 过滤后的条目列表 """ if not keywords: return items filtered = [] for item in items: title = item.get("title", "").lower() desc = item.get("description", "").lower() # 检查是否匹配任何关键词 matches = False for kw in keywords: if kw.lower() in title or kw.lower() in desc: matches = True break if matches: filtered.append(item) # 按热度排序 filtered.sort(key=lambda x: x.get("hot", 0) or x.get("score", 0), reverse=True) return filtered def format_monitoring_response(self, monitor_data: Dict[str, Any]) -> str: """ 格式化监控响应 Args: monitor_data: 监控数据 Returns: 格式化的响应文本 """ items = monitor_data.get("items", []) keywords = monitor_data.get("keywords", []) if not items: return f"❌ 暂无关于「{', '.join(keywords)}」的舆情数据" response = f"🔍 **舆情监控 - {', '.join(keywords)}**\n" response += f"共 {monitor_data['total_items']} 条相关内容\n" response += "-" * 40 + "\n\n" for i, item in enumerate(items, 1): title = item.get("title", "无标题") hot = item.get("hot", item.get("score", "")) platform = item.get("source_platform", "") response += f"{i}. {title}\n" if hot: response += f" 🔥 热度: {hot}" if platform: response += f" | 📱 {platform}" response += "\n\n" return response def parse_keywords_from_input(self, user_input: str) -> List[str]: """ 从用户输入解析关键词 Args: user_input: 用户输入 Returns: 关键词列表 """ user_input = user_input.strip() # 常用监控关键词 common_keywords = [ "AI", "人工智能", "ChatGPT", "大模型", "特斯拉", "比亚迪", "新能源汽车", "华为", "iPhone", "小米", "直播", "电商", "网红", "房价", "股票", "基金", "高考", "考研", "留学" ] matched_keywords = [] for kw in common_keywords: if kw.lower() in user_input.lower(): matched_keywords.append(kw) # 如果没有匹配到常用关键词，尝试提取用户输入的词 if not matched_keywords: # 按逗号、空格等分隔 import re words = re.split(r'[,，\s]+', user_input) matched_keywords = [w for w in words if w.strip()] return matched_keywords async def process_user_request(self, user_input: str) -> Dict[str, Any]: """ 处理用户请求 Args: user_input: 用户输入 Returns: 处理结果 """ # 解析关键词 keywords = self.parse_keywords_from_input(user_input) if not keywords: # 返回引导信息 return { "action": "ask_keywords", "message": "🔍 **舆情监控**\n\n" "请输入您想监控的关键词，例如：\n" "• AI、人工智能、ChatGPT\n" "• 特斯拉、比亚迪、新能源汽车\n" "• 华为、iPhone、小米\n" "• 直播、电商、网红\n" "• 房价、股票、基金\n\n" "💡 您可以直接输入任意关键词进行监控" } # 获取监控数据 monitor_data = await self.monitor_keywords(keywords) # 格式化响应 response_text = self.format_monitoring_response(monitor_data) return { "action": "show_monitoring", "data": monitor_data, "message": response_text } # 便捷函数 async def create_sentiment_monitor(api_client=None, formatter=None) -> SentimentMonitor: """创建舆情监控实例""" return SentimentMonitor(api_client=api_client, formatter=formatter) if __name__ == "__main__": # 测试代码 async def test(): monitor = await create_sentiment_monitor() print("测试舆情监控...") result = await monitor.process_user_request("监控AI和特斯拉") print(result["message"]) asyncio.run(test())