diff --git a/src/config/config.py b/src/config/config.py index e2d06687..0a81efa7 100644 --- a/src/config/config.py +++ b/src/config/config.py @@ -278,14 +278,18 @@ class BotConfig: talk_allowed_private = set() enable_pfc_chatting: bool = False # 是否启用PFC聊天 enable_pfc_reply_checker: bool = True # 是否开启PFC回复检查 - - # idle_conversation - enable_idle_conversation: bool = False # 是否启用 pfc 主动发言 - idle_check_interval: int = 10 # 检查间隔，10分钟检查一次 - min_idle_time: int = 7200 # 最短无活动时间，2小时 (7200秒) - max_idle_time: int = 18000 # 最长无活动时间，5小时 (18000秒) + pfc_message_buffer_size: int = ( + 2 # PFC 聊天消息缓冲数量，有利于使聊天节奏更加紧凑流畅，请根据实际 LLM 响应速度进行调整，默认2条 + ) api_polling_max_retries: int = 3 # 神秘小功能 + # idle_chat + enable_idle_chat: bool = False # 是否启用 pfc 主动发言 + idle_check_interval: int = 10 # 检查间隔，10分钟检查一次 + min_cooldown: int = 7200 # 最短冷却时间，2小时 (7200秒) + max_cooldown: int = 18000 # 最长冷却时间，5小时 (18000秒) + + # Group Nickname enable_nickname_mapping: bool = False # 绰号映射功能总开关 max_nicknames_in_prompt: int = 10 # Prompt 中最多注入的绰号数量 @@ -721,18 +725,17 @@ class BotConfig: "enable_pfc_reply_checker", config.enable_pfc_reply_checker ) logger.info(f"PFC Reply Checker 状态: {'启用' if config.enable_pfc_reply_checker else '关闭'}") + config.pfc_message_buffer_size = experimental_config.get( + "pfc_message_buffer_size", config.pfc_message_buffer_size + ) - def idle_conversation(parent: dict): - idle_conversation_config = parent["idle_conversation"] + def idle_chat(parent: dict): + idle_chat_config = parent["idle_chat"] if config.INNER_VERSION in SpecifierSet(">=1.6.1.6"): - config.enable_idle_conversation = idle_conversation_config.get( - "enable_idle_conversation", config.enable_idle_conversation - ) - config.idle_check_interval = idle_conversation_config.get( - "idle_check_interval", config.idle_check_interval - ) - config.min_idle_time = idle_conversation_config.get("min_idle_time", config.min_idle_time) - config.max_idle_time = idle_conversation_config.get("max_idle_time", config.max_idle_time) + config.enable_idle_chat = idle_chat_config.get("enable_idle_chat", config.enable_idle_chat) + config.idle_check_interval = idle_chat_config.get("idle_check_interval", config.idle_check_interval) + config.min_cooldown = idle_chat_config.get("min_cooldown", config.min_cooldown) + config.max_cooldown = idle_chat_config.get("max_cooldown", config.max_cooldown) # 版本表达式：>=1.0.0,<2.0.0 # 允许字段：func: method, support: str, notice: str, necessary: bool @@ -768,7 +771,7 @@ class BotConfig: "normal_chat": {"func": normal_chat, "support": ">=1.6.0", "necessary": False}, "focus_chat": {"func": focus_chat, "support": ">=1.6.0", "necessary": False}, "group_nickname": {"func": group_nickname, "support": ">=1.6.1.1", "necessary": False}, - "idle_conversation": {"func": idle_conversation, "support": ">=1.6.1.6", "necessary": False}, + "idle_chat": {"func": idle_chat, "support": ">=1.6.1.6", "necessary": False}, } # 原地修改，将 字符串版本表达式 转换成 版本对象 diff --git a/src/plugins/PFC/PFC_idle/__init__.py b/src/plugins/PFC/PFC_idle/__init__.py index 77d90210..99c1268a 100644 --- a/src/plugins/PFC/PFC_idle/__init__.py +++ b/src/plugins/PFC/PFC_idle/__init__.py @@ -3,18 +3,8 @@ PFC_idle 包 - 用于空闲时主动聊天的功能模块 该包包含以下主要组件： - IdleChat: 根据关系和活跃度进行智能主动聊天 -- IdleChatManager: 管理多个聊天实例的空闲状态 -- IdleConversation: 处理与空闲聊天相关的功能，与主Conversation类解耦 """ from .idle_chat import IdleChat -from .idle_chat_manager import IdleChatManager -from .idle_conversation import IdleConversation, get_idle_conversation_instance, initialize_idle_conversation -__all__ = [ - "IdleChat", - "IdleChatManager", - "IdleConversation", - "get_idle_conversation_instance", - "initialize_idle_conversation", -] +__all__ = ["IdleChat"] diff --git a/src/plugins/PFC/PFC_idle/idle_chat.py b/src/plugins/PFC/PFC_idle/idle_chat.py index 2962d52c..61f78574 100644 --- a/src/plugins/PFC/PFC_idle/idle_chat.py +++ b/src/plugins/PFC/PFC_idle/idle_chat.py @@ -7,10 +7,12 @@ from datetime import datetime from src.common.logger_manager import get_logger from src.config.config import global_config from src.plugins.models.utils_model import LLMRequest -from src.plugins.utils.prompt_builder import global_prompt_manager + +# from src.plugins.utils.prompt_builder import global_prompt_manager from src.plugins.person_info.person_info import person_info_manager from src.plugins.utils.chat_message_builder import build_readable_messages -from ...schedule.schedule_generator import bot_schedule + +# from ...schedule.schedule_generator import bot_schedule from ..chat_observer import ChatObserver from ..message_sender import DirectMessageSender from src.plugins.chat.chat_stream import ChatStream @@ -144,14 +146,11 @@ class IdleChat: self._task: Optional[asyncio.Task] = None # 配置参数 - 从global_config加载 - self.min_cooldown = getattr( - global_config, "MIN_IDLE_TIME", 7200 - ) # 最短冷却时间（默认2小时）建议修改长一点，你也不希望你的bot一直骚扰你吧 - self.max_cooldown = getattr(global_config, "MAX_IDLE_TIME", 14400) # 最长冷却时间（默认4小时） - self.min_idle_time = getattr(global_config, "MIN_IDLE_TIME", 3600) - self.check_interval = getattr(global_config, "IDLE_CHECK_INTERVAL", 600) # 检查间隔（默认10分钟） - self.active_hours_start = 6 # 活动开始时间 - self.active_hours_end = 24 # 活动结束时间 + self.min_cooldown = global_config.min_cooldown # 最短冷却时间（默认2小时） + self.max_cooldown = global_config.max_cooldown # 最长冷却时间（默认5小时） + self.check_interval = global_config.idle_check_interval * 60 # 检查间隔（默认10分钟，转换为秒） + self.active_hours_start = 7 # 活动开始时间 + self.active_hours_end = 23 # 活动结束时间 # 关系值相关 self.base_trigger_probability = 0.3 # 基础触发概率 @@ -160,8 +159,8 @@ class IdleChat: def start(self) -> None: """启动主动聊天检测""" # 检查是否启用了主动聊天功能 - if not getattr(global_config, "ENABLE_IDLE_CONVERSATION", False): - logger.info(f"[私聊][{self.private_name}]主动聊天功能已禁用（配置ENABLE_IDLE_CONVERSATION=False）") + if not global_config.enable_idle_chat: + logger.info(f"[私聊][{self.private_name}]主动聊天功能已禁用（配置ENABLE_IDLE_CHAT=False）") return if self._running: @@ -353,7 +352,7 @@ class IdleChat: try: while self._running: # 检查是否启用了主动聊天功能 - if not getattr(global_config, "ENABLE_IDLE_CONVERSATION", False): + if not global_config.enable_idle_chat: # 如果禁用了功能，等待一段时间后再次检查配置 await asyncio.sleep(60) # 每分钟检查一次配置变更 continue @@ -488,19 +487,20 @@ class IdleChat: if "：" in full_relationship_text: relationship_description = full_relationship_text.split("：")[1].replace("。", "") - if global_config.ENABLE_SCHEDULE_GEN: - schedule_prompt = await global_prompt_manager.format_prompt( - "schedule_prompt", schedule_info=bot_schedule.get_current_num_task(num=1, time_info=False) - ) - else: - schedule_prompt = "" + # 暂不使用 + # if global_config.ENABLE_SCHEDULE_GEN: + # schedule_prompt = await global_prompt_manager.format_prompt( + # "schedule_prompt", schedule_info=bot_schedule.get_current_num_task(num=1, time_info=False) + # ) + # else: + # schedule_prompt = "" - # 构建提示词 + # 构建提示词，暂存废弃部分这是你的日程{schedule_prompt} current_time = datetime.now().strftime("%H:%M") prompt = f"""你是{global_config.BOT_NICKNAME}。 你正在与用户{self.private_name}进行QQ私聊，你们的关系是{relationship_description} 现在时间{current_time} - 这是你的日程{schedule_prompt} + 你想要主动发起对话。 请基于以下之前的对话历史，生成一条自然、友好、符合关系程度的主动对话消息。 这条消息应能够引起用户的兴趣，重新开始对话。 diff --git a/src/plugins/PFC/PFC_idle/idle_chat_manager.py b/src/plugins/PFC/PFC_idle/idle_chat_manager.py deleted file mode 100644 index a8708331..00000000 --- a/src/plugins/PFC/PFC_idle/idle_chat_manager.py +++ /dev/null @@ -1,183 +0,0 @@ -from typing import Dict, Optional -import asyncio -from src.common.logger_manager import get_logger -from .idle_chat import IdleChat -import traceback - -logger = get_logger("pfc_idle_chat_manager") - - -class IdleChatManager: - """空闲聊天管理器 - - 用于管理所有私聊用户的空闲聊天实例。 - 采用单例模式，确保全局只有一个管理器实例。 - """ - - _instance: Optional["IdleChatManager"] = None - _lock: asyncio.Lock = asyncio.Lock() - - def __init__(self): - """初始化空闲聊天管理器""" - self._idle_chats: Dict[str, IdleChat] = {} # stream_id -> IdleChat - self._active_conversations_count: Dict[str, int] = {} # stream_id -> count - - @classmethod - def get_instance(cls) -> "IdleChatManager": - """获取管理器单例 (同步版本) - - Returns: - IdleChatManager: 管理器实例 - """ - if not cls._instance: - # 在同步环境中创建实例 - cls._instance = cls() - return cls._instance - - @classmethod - async def get_instance_async(cls) -> "IdleChatManager": - """获取管理器单例 (异步版本) - - Returns: - IdleChatManager: 管理器实例 - """ - if not cls._instance: - async with cls._lock: - if not cls._instance: - cls._instance = cls() - return cls._instance - - async def get_or_create_idle_chat(self, stream_id: str, private_name: str) -> IdleChat: - """获取或创建空闲聊天实例 - - Args: - stream_id: 聊天流ID - private_name: 私聊用户名称 - - Returns: - IdleChat: 空闲聊天实例 - """ - if stream_id not in self._idle_chats: - idle_chat = IdleChat(stream_id, private_name) - self._idle_chats[stream_id] = idle_chat - # 初始化活跃对话计数 - if stream_id not in self._active_conversations_count: - self._active_conversations_count[stream_id] = 0 - idle_chat.start() # 启动空闲检测 - logger.info(f"[私聊][{private_name}]创建并启动新的空闲聊天实例") - return self._idle_chats[stream_id] - - async def remove_idle_chat(self, stream_id: str) -> None: - """移除空闲聊天实例 - - Args: - stream_id: 聊天流ID - """ - if stream_id in self._idle_chats: - idle_chat = self._idle_chats[stream_id] - idle_chat.stop() # 停止空闲检测 - del self._idle_chats[stream_id] - if stream_id in self._active_conversations_count: - del self._active_conversations_count[stream_id] - logger.info(f"[私聊][{idle_chat.private_name}]移除空闲聊天实例") - - async def notify_conversation_start(self, stream_id: str) -> None: - """通知对话开始 - - Args: - stream_id: 聊天流ID - """ - try: - if stream_id not in self._idle_chats: - logger.warning(f"对话开始通知: {stream_id} 没有对应的IdleChat实例，将创建一个") - # 从stream_id尝试提取private_name - private_name = stream_id - if stream_id.startswith("private_"): - parts = stream_id.split("_") - if len(parts) >= 2: - private_name = parts[1] # 取第二部分作为名称 - await self.get_or_create_idle_chat(stream_id, private_name) - - if stream_id not in self._active_conversations_count: - self._active_conversations_count[stream_id] = 0 - - # 增加计数前记录当前值，用于日志 - old_count = self._active_conversations_count[stream_id] - self._active_conversations_count[stream_id] += 1 - new_count = self._active_conversations_count[stream_id] - - # 确保IdleChat实例存在 - idle_chat = self._idle_chats.get(stream_id) - if idle_chat: - await idle_chat.increment_active_instances() - logger.debug(f"对话开始通知: {stream_id}, 计数从{old_count}增加到{new_count}") - else: - logger.error(f"对话开始通知: {stream_id}, 计数增加但IdleChat不存在! 计数:{old_count}->{new_count}") - except Exception as e: - logger.error(f"对话开始通知处理失败: {stream_id}, 错误: {e}") - logger.error(traceback.format_exc()) - - async def notify_conversation_end(self, stream_id: str) -> None: - """通知对话结束 - - Args: - stream_id: 聊天流ID - """ - try: - # 记录当前计数用于日志 - old_count = self._active_conversations_count.get(stream_id, 0) - - # 安全减少计数，避免负数 - if stream_id in self._active_conversations_count and self._active_conversations_count[stream_id] > 0: - self._active_conversations_count[stream_id] -= 1 - else: - # 如果计数已经为0或不存在，设置为0 - self._active_conversations_count[stream_id] = 0 - - new_count = self._active_conversations_count.get(stream_id, 0) - - # 确保IdleChat实例存在 - idle_chat = self._idle_chats.get(stream_id) - if idle_chat: - await idle_chat.decrement_active_instances() - logger.debug(f"对话结束通知: {stream_id}, 计数从{old_count}减少到{new_count}") - else: - logger.warning(f"对话结束通知: {stream_id}, 计数减少但IdleChat不存在! 计数:{old_count}->{new_count}") - - # 检查是否所有对话都结束了，帮助调试 - all_counts = sum(self._active_conversations_count.values()) - if all_counts == 0: - logger.info("所有对话实例都已结束，当前总活跃计数为0") - except Exception as e: - logger.error(f"对话结束通知处理失败: {stream_id}, 错误: {e}") - logger.error(traceback.format_exc()) - - def get_idle_chat(self, stream_id: str) -> Optional[IdleChat]: - """获取空闲聊天实例 - - Args: - stream_id: 聊天流ID - - Returns: - Optional[IdleChat]: 空闲聊天实例，如果不存在则返回None - """ - return self._idle_chats.get(stream_id) - - def get_active_conversations_count(self, stream_id: str) -> int: - """获取指定流的活跃对话计数 - - Args: - stream_id: 聊天流ID - - Returns: - int: 活跃对话计数 - """ - return self._active_conversations_count.get(stream_id, 0) - - def get_all_active_conversations_count(self) -> int: - """获取所有活跃对话总计数 - - Returns: - int: 活跃对话总计数 - """ - return sum(self._active_conversations_count.values()) diff --git a/src/plugins/PFC/PFC_idle/idle_conversation.py b/src/plugins/PFC/PFC_idle/idle_conversation.py deleted file mode 100644 index 90036d13..00000000 --- a/src/plugins/PFC/PFC_idle/idle_conversation.py +++ /dev/null @@ -1,513 +0,0 @@ -import traceback -import asyncio -from typing import Optional, Dict -from src.common.logger_manager import get_logger -import time - -logger = get_logger("pfc_idle_conversation") - - -class IdleConversation: - """ - 处理Idle聊天相关的功能，将这些功能从主Conversation类中分离出来， - 以减少代码量并方便维护。 - """ - - def __init__(self): - """初始化IdleConversation实例""" - self._idle_chat_manager = None - self._running = False - self._active_streams: Dict[str, bool] = {} # 跟踪活跃的流 - self._monitor_task = None # 用于后台监控的任务 - self._lock = asyncio.Lock() # 用于线程安全操作 - self._initialization_in_progress = False # 防止并发初始化 - - async def initialize(self): - """初始化Idle聊天管理器""" - # 防止并发初始化 - if self._initialization_in_progress: - logger.debug("IdleConversation正在初始化中，等待完成") - return False - - if self._idle_chat_manager is not None: - logger.debug("IdleConversation已初始化，无需重复操作") - return True - - # 标记开始初始化 - self._initialization_in_progress = True - - try: - # 从PFCManager获取IdleChatManager实例 - from ..pfc_manager import PFCManager - - pfc_manager = PFCManager.get_instance() - self._idle_chat_manager = pfc_manager.get_idle_chat_manager() - logger.debug("IdleConversation初始化完成，已获取IdleChatManager实例") - return True - except Exception as e: - logger.error(f"初始化IdleConversation时出错: {e}") - logger.error(traceback.format_exc()) - return False - finally: - # 无论成功或失败，都清除初始化标志 - self._initialization_in_progress = False - - async def start(self): - """启动IdleConversation，创建后台监控任务""" - if self._running: - logger.debug("IdleConversation已经在运行") - return False - - if not self._idle_chat_manager: - success = await self.initialize() - if not success: - logger.error("无法启动IdleConversation：初始化失败") - return False - - try: - self._running = True - # 创建后台监控任务，使用try-except块来捕获可能的异常 - try: - loop = asyncio.get_running_loop() - if loop.is_running(): - self._monitor_task = asyncio.create_task(self._monitor_loop()) - logger.info("IdleConversation启动成功，后台监控任务已创建") - else: - logger.warning("事件循环不活跃，跳过监控任务创建") - except RuntimeError: - # 如果没有活跃的事件循环，记录警告但继续执行 - logger.warning("没有活跃的事件循环，IdleConversation将不会启动监控任务") - # 尽管没有监控任务，但仍然将running设为True表示IdleConversation已启动 - - return True - except Exception as e: - self._running = False - logger.error(f"启动IdleConversation失败: {e}") - logger.error(traceback.format_exc()) - return False - - async def stop(self): - """停止IdleConversation的后台任务""" - if not self._running: - return - - self._running = False - if self._monitor_task and not self._monitor_task.done(): - try: - self._monitor_task.cancel() - try: - await asyncio.wait_for(self._monitor_task, timeout=2.0) - except asyncio.TimeoutError: - logger.warning("停止IdleConversation监控任务超时") - except asyncio.CancelledError: - pass # 正常取消 - except Exception as e: - logger.error(f"停止IdleConversation监控任务时出错: {e}") - logger.error(traceback.format_exc()) - - self._monitor_task = None - logger.info("IdleConversation已停止") - - async def _monitor_loop(self): - """后台监控循环，定期检查活跃的会话并执行必要的操作""" - try: - while self._running: - try: - # 同步活跃流计数到IdleChatManager - if self._idle_chat_manager: - await self._sync_active_streams_to_manager() - - # 这里可以添加定期检查逻辑，如查询空闲状态等 - active_count = len(self._active_streams) - logger.debug(f"IdleConversation监控中，当前活跃流数量: {active_count}") - - except Exception as e: - logger.error(f"IdleConversation监控循环出错: {e}") - logger.error(traceback.format_exc()) - - # 每30秒执行一次监控 - await asyncio.sleep(30) - except asyncio.CancelledError: - logger.info("IdleConversation监控任务已取消") - except Exception as e: - logger.error(f"IdleConversation监控任务异常退出: {e}") - logger.error(traceback.format_exc()) - self._running = False - - async def _sync_active_streams_to_manager(self): - """同步活跃流计数到IdleChatManager和IdleChat""" - try: - if not self._idle_chat_manager: - return - - # 获取当前的活跃流列表 - async with self._lock: - active_streams = list(self._active_streams.keys()) - - # 对每个活跃流，确保IdleChatManager和IdleChat中的计数是正确的 - for stream_id in active_streams: - # 获取当前IdleChatManager中的计数 - manager_count = self._idle_chat_manager.get_active_conversations_count(stream_id) - - # 由于我们的活跃流字典只记录是否活跃(值为True)，所以计数应该是1 - if manager_count != 1: - # 修正IdleChatManager中的计数 - old_count = manager_count - self._idle_chat_manager._active_conversations_count[stream_id] = 1 - logger.warning(f"同步调整IdleChatManager中的计数: stream_id={stream_id}, {old_count}->1") - - # 同时修正IdleChat中的计数 - idle_chat = self._idle_chat_manager.get_idle_chat(stream_id) - if idle_chat: - if getattr(idle_chat, "active_instances_count", 0) != 1: - old_count = getattr(idle_chat, "active_instances_count", 0) - idle_chat.active_instances_count = 1 - logger.warning(f"同步调整IdleChat中的计数: stream_id={stream_id}, {old_count}->1") - - # 检查IdleChatManager中有没有多余的计数(conversation中已不存在但manager中还有) - for stream_id, count in list(self._idle_chat_manager._active_conversations_count.items()): - if count > 0 and stream_id not in active_streams: - # 重置为0 - self._idle_chat_manager._active_conversations_count[stream_id] = 0 - logger.warning(f"重置IdleChatManager中的多余计数: stream_id={stream_id}, {count}->0") - - # 同时修正IdleChat中的计数 - idle_chat = self._idle_chat_manager.get_idle_chat(stream_id) - if idle_chat and getattr(idle_chat, "active_instances_count", 0) > 0: - old_count = getattr(idle_chat, "active_instances_count", 0) - idle_chat.active_instances_count = 0 - logger.warning(f"同步重置IdleChat中的计数: stream_id={stream_id}, {old_count}->0") - - # 日志记录同步结果 - total_active = len(active_streams) - total_manager = sum(self._idle_chat_manager._active_conversations_count.values()) - logger.debug(f"同步后的计数: IdleConversation活跃流={total_active}, IdleChatManager总计数={total_manager}") - - except Exception as e: - logger.error(f"同步活跃流计数失败: {e}") - logger.error(traceback.format_exc()) - - async def get_or_create_idle_chat(self, stream_id: str, private_name: str): - """ - 获取或创建IdleChat实例 - - Args: - stream_id: 聊天流ID - private_name: 私聊对象名称，用于日志 - - Returns: - bool: 操作是否成功 - """ - # 确保IdleConversation已启动 - if not self._running: - await self.start() - - if not self._idle_chat_manager: - # 如果尚未初始化，尝试初始化 - success = await self.initialize() - if not success: - logger.warning(f"[私聊][{private_name}] 获取或创建IdleChat失败：IdleChatManager未初始化") - return False - - try: - # 创建IdleChat实例 - _idle_chat = await self._idle_chat_manager.get_or_create_idle_chat(stream_id, private_name) - logger.debug(f"[私聊][{private_name}] 已创建或获取IdleChat实例") - return True - except Exception as e: - logger.warning(f"[私聊][{private_name}] 创建或获取IdleChat实例失败: {e}") - logger.warning(traceback.format_exc()) - return False - - async def notify_conversation_start(self, stream_id: str, private_name: str) -> bool: - """ - 通知空闲聊天管理器对话开始 - - Args: - stream_id: 聊天流ID - private_name: 私聊对象名称，用于日志 - - Returns: - bool: 通知是否成功 - """ - try: - # 确保IdleConversation已启动 - if not self._running: - success = await self.start() - if not success: - logger.warning(f"[私聊][{private_name}] 启动IdleConversation失败，无法通知对话开始") - return False - - if not self._idle_chat_manager: - # 如果尚未初始化，尝试初始化 - success = await self.initialize() - if not success: - logger.warning(f"[私聊][{private_name}] 通知对话开始失败：IdleChatManager未初始化") - return False - - try: - # 确保IdleChat实例已创建 - 这是关键步骤，要先创建IdleChat - await self.get_or_create_idle_chat(stream_id, private_name) - - # 先记录活跃状态 - 这是权威源 - async with self._lock: - self._active_streams[stream_id] = True - - # 然后同步到IdleChatManager - if self._idle_chat_manager: - await self._idle_chat_manager.notify_conversation_start(stream_id) - logger.info(f"[私聊][{private_name}] 已通知空闲聊天管理器对话开始") - else: - logger.warning(f"[私聊][{private_name}] IdleChatManager不存在，但已记录活跃状态") - - # 立即进行一次同步，确保数据一致性 - await self._sync_active_streams_to_manager() - - return True - except Exception as e: - logger.warning(f"[私聊][{private_name}] 通知空闲聊天管理器对话开始失败: {e}") - logger.warning(traceback.format_exc()) - # 即使通知失败，也应记录活跃状态 - async with self._lock: - self._active_streams[stream_id] = True - return False - except Exception as outer_e: - logger.error(f"[私聊][{private_name}] 处理对话开始通知时发生严重错误: {outer_e}") - logger.error(traceback.format_exc()) - return False - - async def notify_conversation_end(self, stream_id: str, private_name: str) -> bool: - """ - 通知空闲聊天管理器对话结束 - - Args: - stream_id: 聊天流ID - private_name: 私聊对象名称，用于日志 - - Returns: - bool: 通知是否成功 - """ - try: - # 先从自身的活跃流中移除 - 这是权威源 - was_active = False - async with self._lock: - if stream_id in self._active_streams: - del self._active_streams[stream_id] - was_active = True - logger.debug(f"[私聊][{private_name}] 已从活跃流中移除 {stream_id}") - - if not self._idle_chat_manager: - # 如果尚未初始化，尝试初始化 - success = await self.initialize() - if not success: - logger.warning(f"[私聊][{private_name}] 通知对话结束失败：IdleChatManager未初始化") - return False - - try: - # 然后同步到IdleChatManager - if self._idle_chat_manager: - # 无论如何都尝试通知 - await self._idle_chat_manager.notify_conversation_end(stream_id) - - # 立即进行一次同步，确保数据一致性 - await self._sync_active_streams_to_manager() - - logger.info(f"[私聊][{private_name}] 已通知空闲聊天管理器对话结束") - - # 检查当前活跃流数量 - active_count = len(self._active_streams) - if active_count == 0: - logger.info(f"[私聊][{private_name}] 当前无活跃流，可能会触发主动聊天") - - # 额外调用：如果实例存在且只有在确实移除了活跃流的情况下才触发检查 - if was_active: - idle_chat = self._idle_chat_manager.get_idle_chat(stream_id) - if idle_chat: - # 直接触发IdleChat检查，而不是等待下一个循环 - logger.info(f"[私聊][{private_name}] 对话结束，手动触发一次主动聊天检查") - asyncio.create_task(self._trigger_idle_chat_check(idle_chat, stream_id, private_name)) - - return True - else: - logger.warning(f"[私聊][{private_name}] IdleChatManager不存在，但已更新活跃状态") - return False - except Exception as e: - logger.warning(f"[私聊][{private_name}] 通知空闲聊天管理器对话结束失败: {e}") - logger.warning(traceback.format_exc()) - return False - except Exception as outer_e: - logger.error(f"[私聊][{private_name}] 处理对话结束通知时发生严重错误: {outer_e}") - logger.error(traceback.format_exc()) - return False - - async def _trigger_idle_chat_check(self, idle_chat, stream_id: str, private_name: str): - """在对话结束后，手动触发一次IdleChat的检查""" - try: - # 确保活跃计数与IdleConversation一致 - async with self._lock: - is_active_in_conversation = stream_id in self._active_streams - - # 强制使IdleChat的计数与IdleConversation一致 - if is_active_in_conversation: - # 如果在IdleConversation中是活跃的，IdleChat的计数应该是1 - if idle_chat.active_instances_count != 1: - old_count = idle_chat.active_instances_count - idle_chat.active_instances_count = 1 - logger.warning(f"[私聊][{private_name}] 修正IdleChat计数: {old_count}->1") - else: - # 如果在IdleConversation中不是活跃的，IdleChat的计数应该是0 - if idle_chat.active_instances_count != 0: - old_count = idle_chat.active_instances_count - idle_chat.active_instances_count = 0 - logger.warning(f"[私聊][{private_name}] 修正IdleChat计数: {old_count}->0") - - # 等待1秒，让任何正在进行的处理完成 - await asyncio.sleep(1) - - # 只有当stream不再活跃时才触发检查 - if not is_active_in_conversation: - # 尝试触发一次检查 - if hasattr(idle_chat, "_should_trigger"): - should_trigger = await idle_chat._should_trigger() - logger.info(f"[私聊][{private_name}] 手动触发主动聊天检查结果: {should_trigger}") - - # 如果应该触发，直接调用_initiate_chat - if should_trigger and hasattr(idle_chat, "_initiate_chat"): - logger.info(f"[私聊][{private_name}] 手动触发主动聊天") - await idle_chat._initiate_chat() - # 更新最后触发时间 - idle_chat.last_trigger_time = time.time() - else: - logger.warning(f"[私聊][{private_name}] IdleChat没有_should_trigger方法，无法触发检查") - except Exception as e: - logger.error(f"[私聊][{private_name}] 手动触发主动聊天检查时出错: {e}") - logger.error(traceback.format_exc()) - - def is_stream_active(self, stream_id: str) -> bool: - """检查指定的stream是否活跃""" - return stream_id in self._active_streams - - def get_active_streams_count(self) -> int: - """获取当前活跃的stream数量""" - return len(self._active_streams) - - @property - def is_running(self) -> bool: - """检查IdleConversation是否正在运行""" - return self._running - - @property - def idle_chat_manager(self): - """获取IdleChatManager实例""" - return self._idle_chat_manager - - -# 创建单例实例 -_instance: Optional[IdleConversation] = None -_instance_lock = asyncio.Lock() -_initialization_in_progress = False # 防止并发初始化 - - -async def initialize_idle_conversation() -> IdleConversation: - """初始化并启动IdleConversation单例实例""" - global _initialization_in_progress - - # 防止并发初始化 - if _initialization_in_progress: - logger.debug("IdleConversation全局初始化正在进行中，等待完成") - return get_idle_conversation_instance() - - # 标记正在初始化 - _initialization_in_progress = True - - try: - instance = get_idle_conversation_instance() - - # 如果实例已经在运行，避免重复初始化 - if getattr(instance, "_running", False): - logger.debug("IdleConversation已在运行状态，无需重新初始化") - _initialization_in_progress = False - return instance - - # 初始化实例 - success = await instance.initialize() - if not success: - logger.error("IdleConversation初始化失败") - _initialization_in_progress = False - return instance - - # 启动实例 - success = await instance.start() - if not success: - logger.error("IdleConversation启动失败") - else: - # 启动成功，进行初始检查 - logger.info("IdleConversation启动成功，执行初始化后检查") - # 这里可以添加一些启动后的检查，如果需要 - - # 创建一个异步任务，定期检查系统状态 - asyncio.create_task(periodic_system_check(instance)) - - return instance - except Exception as e: - logger.error(f"初始化并启动IdleConversation时出错: {e}") - logger.error(traceback.format_exc()) - # 重置标志，允许下次再试 - _initialization_in_progress = False - return get_idle_conversation_instance() # 返回实例，即使初始化失败 - finally: - # 清除初始化标志 - _initialization_in_progress = False - - -async def periodic_system_check(instance: IdleConversation): - """定期检查系统状态，确保主动聊天功能正常工作""" - try: - # 等待10秒，让系统完全启动 - await asyncio.sleep(10) - - while getattr(instance, "_running", False): - try: - # 检查活跃流数量 - active_streams_count = len(getattr(instance, "_active_streams", {})) - - # 如果IdleChatManager存在，检查其中的活跃对话计数 - idle_chat_manager = getattr(instance, "_idle_chat_manager", None) - if idle_chat_manager and hasattr(idle_chat_manager, "get_all_active_conversations_count"): - manager_count = idle_chat_manager.get_all_active_conversations_count() - - # 如果两者不一致，记录警告 - if active_streams_count != manager_count: - logger.warning( - f"检测到计数不一致: IdleConversation记录的活跃流数量({active_streams_count}) 与 IdleChatManager记录的活跃对话数({manager_count})不匹配" - ) - - # 如果IdleChatManager记录的计数为0但自己的记录不为0，进行修正 - if manager_count == 0 and active_streams_count > 0: - logger.warning("检测到可能的计数错误，尝试修正：清空IdleConversation的活跃流记录") - async with instance._lock: - instance._active_streams.clear() - - # 检查计数如果为0，帮助日志输出 - if active_streams_count == 0: - logger.debug("当前没有活跃的对话流，应该可以触发主动聊天") - - except Exception as check_err: - logger.error(f"执行系统检查时出错: {check_err}") - logger.error(traceback.format_exc()) - - # 每60秒检查一次 - await asyncio.sleep(60) - except asyncio.CancelledError: - logger.debug("系统检查任务被取消") - except Exception as e: - logger.error(f"系统检查任务异常退出: {e}") - logger.error(traceback.format_exc()) - - -def get_idle_conversation_instance() -> IdleConversation: - """获取IdleConversation的单例实例""" - global _instance - if _instance is None: - _instance = IdleConversation() - return _instance diff --git a/src/plugins/PFC/PFC_idle/idle_conversation_starter.py b/src/plugins/PFC/PFC_idle/idle_conversation_starter.py deleted file mode 100644 index 83a58219..00000000 --- a/src/plugins/PFC/PFC_idle/idle_conversation_starter.py +++ /dev/null @@ -1,354 +0,0 @@ -import time -import asyncio -import random -import traceback -from typing import TYPE_CHECKING, Optional - -from src.common.logger_manager import get_logger -from src.plugins.models.utils_model import LLMRequest -from src.config.config import global_config -from src.plugins.chat.chat_stream import chat_manager, ChatStream -from src.individuality.individuality import Individuality -from src.plugins.utils.chat_message_builder import build_readable_messages -from maim_message import UserInfo - -from ..chat_observer import ChatObserver -from ..message_sender import DirectMessageSender - -# 导入富文本回溯，用于更好的错误展示 -from rich.traceback import install - -# 使用TYPE_CHECKING避免循环导入 -if TYPE_CHECKING: - from ..conversation import Conversation - -install(extra_lines=3) - -# 获取当前模块的日志记录器 -logger = get_logger("idle_conversation_starter") - - -class IdleConversationStarter: - """长时间无对话主动发起对话的组件 - - 该组件会在一段时间没有对话后，自动生成一条消息发送给用户，以保持对话的活跃度。 - 时间阈值会在配置的最小和最大值之间随机选择，每次发送消息后都会重置。 - """ - - def __init__(self, stream_id: str, private_name: str): - """初始化空闲对话启动器 - - Args: - stream_id: 聊天流ID - private_name: 私聊用户名称 - """ - self.stream_id: str = stream_id - self.private_name: str = private_name - self.chat_observer = ChatObserver.get_instance(stream_id, private_name) - self.message_sender = DirectMessageSender(private_name) - - # 添加异步锁，保护对共享变量的访问 - self._lock: asyncio.Lock = asyncio.Lock() - - # LLM请求对象，用于生成主动对话内容 - self.llm = LLMRequest( - model=global_config.llm_normal, temperature=0.8, max_tokens=500, request_type="idle_conversation_starter" - ) - - # 个性化信息 - self.personality_info: str = Individuality.get_instance().get_prompt(x_person=2, level=3) - - # 计算实际触发阈值（在min和max之间随机） - self.actual_idle_threshold: int = random.randint(global_config.min_idle_time, global_config.max_idle_time) - - # 工作状态 - self.last_message_time: float = time.time() - self._running: bool = False - self._task: Optional[asyncio.Task] = None - - def start(self) -> None: - """启动空闲对话检测 - - 如果功能被禁用或已经在运行，则不会启动。 - """ - # 如果功能被禁用，则不启动 - if not global_config.enable_idle_conversation: - logger.info(f"[私聊][{self.private_name}]主动发起对话功能已禁用") - return - - if self._running: - logger.debug(f"[私聊][{self.private_name}]主动发起对话功能已在运行中") - return - - self._running = True - self._task = asyncio.create_task(self._check_idle_loop()) - logger.info(f"[私聊][{self.private_name}]启动空闲对话检测，阈值设置为{self.actual_idle_threshold}秒") - - def stop(self) -> None: - """停止空闲对话检测 - - 取消当前运行的任务并重置状态。 - """ - if not self._running: - return - - self._running = False - if self._task: - self._task.cancel() - self._task = None - logger.info(f"[私聊][{self.private_name}]停止空闲对话检测") - - async def update_last_message_time(self, message_time: Optional[float] = None) -> None: - """更新最后一条消息的时间 - - Args: - message_time: 消息时间戳，如果为None则使用当前时间 - """ - async with self._lock: - self.last_message_time = message_time or time.time() - # 重新随机化下一次触发的时间阈值 - self.actual_idle_threshold = random.randint(global_config.min_idle_time, global_config.max_idle_time) - logger.debug( - f"[私聊][{self.private_name}]更新最后消息时间: {self.last_message_time}，新阈值: {self.actual_idle_threshold}秒" - ) - - def reload_config(self) -> None: - """重新加载配置 - - 记录当前配置参数，用于日志输出 - """ - try: - logger.debug( - f"[私聊][{self.private_name}]重新加载主动对话配置: 启用={global_config.enable_idle_conversation}, 检查间隔={global_config.idle_check_interval}秒, 最短间隔={global_config.min_idle_time}秒, 最长间隔={global_config.max_idle_time}秒" - ) - - # 重新计算实际阈值 - async def update_threshold(): - async with self._lock: - self.actual_idle_threshold = random.randint( - global_config.min_idle_time, global_config.max_idle_time - ) - logger.debug(f"[私聊][{self.private_name}]更新空闲检测阈值为: {self.actual_idle_threshold}秒") - - # 创建一个任务来异步更新阈值 - asyncio.create_task(update_threshold()) - - except Exception as e: - logger.error(f"[私聊][{self.private_name}]重新加载配置时出错: {str(e)}") - logger.error(traceback.format_exc()) - - async def _check_idle_loop(self) -> None: - """检查空闲状态的循环 - - 定期检查是否长时间无对话，如果达到阈值则尝试主动发起对话。 - """ - try: - config_reload_counter = 0 - config_reload_interval = 100 # 每100次检查重新加载一次配置 - - while self._running: - # 定期重新加载配置 - config_reload_counter += 1 - if config_reload_counter >= config_reload_interval: - self.reload_config() - config_reload_counter = 0 - - # 检查是否启用了主动对话功能 - if not global_config.enable_idle_conversation: - # 如果禁用了功能，就等待一段时间后再次检查配置 - await asyncio.sleep(global_config.idle_check_interval) - continue - - # 使用锁保护对共享变量的读取 - current_time = time.time() - async with self._lock: - idle_time = current_time - self.last_message_time - threshold = self.actual_idle_threshold - - if idle_time >= threshold: - logger.info(f"[私聊][{self.private_name}]检测到长时间({idle_time:.0f}秒)无对话，尝试主动发起聊天") - await self._initiate_conversation() - # 更新时间，避免连续触发 - await self.update_last_message_time() - - # 等待下一次检查 - await asyncio.sleep(global_config.idle_check_interval) - - except asyncio.CancelledError: - logger.debug(f"[私聊][{self.private_name}]空闲对话检测任务被取消") - except Exception as e: - logger.error(f"[私聊][{self.private_name}]空闲对话检测出错: {str(e)}") - logger.error(traceback.format_exc()) - # 尝试重新启动检测循环 - if self._running: - logger.info(f"[私聊][{self.private_name}]尝试重新启动空闲对话检测") - self._task = asyncio.create_task(self._check_idle_loop()) - - async def _initiate_conversation(self) -> None: - """生成并发送主动对话内容 - - 获取聊天历史记录，使用LLM生成合适的开场白（大概），然后发送消息。 - """ - try: - # 获取聊天历史记录，用于生成更合适的开场白 - messages = self.chat_observer.get_cached_messages(limit=12) # 获取最近12条消息 - chat_history_text = await build_readable_messages( - messages, - replace_bot_name=True, - merge_messages=False, - timestamp_mode="relative", - read_mark=0.0, - ) - - # 构建提示词 - prompt = f"""{self.personality_info}。你的名字是{global_config.BOT_NICKNAME}。 - 你正在与用户{self.private_name}进行QQ私聊, - 但已经有一段时间没有对话了。 - 你想要主动发起一个友好的对话，可以说说自己在做的事情或者询问对方在做什么。 - 请基于以下之前的对话历史，生成一条自然、友好、符合你个性的主动对话消息。 - 这条消息应该能够引起用户的兴趣，重新开始对话。 - 最近的对话历史（可能已经过去了很久）： - {chat_history_text} - 请直接输出一条消息，不要有任何额外的解释或引导文字。消息要简短自然，就像是在日常聊天中的开场白。 - 消息内容尽量简短,不要超过20个字,不要添加任何表情符号。 - """ - - # 尝试生成回复，添加超时处理 - try: - content, _ = await asyncio.wait_for( - self.llm.generate_response_async(prompt), - timeout=30, # 30秒超时 - ) - except asyncio.TimeoutError: - logger.error(f"[私聊][{self.private_name}]生成主动对话内容超时") - return - except Exception as llm_err: - logger.error(f"[私聊][{self.private_name}]生成主动对话内容失败: {str(llm_err)}") - logger.error(traceback.format_exc()) - return - - # 清理结果 - content = content.strip() - content = content.strip("\"'") - - if not content: - logger.error(f"[私聊][{self.private_name}]生成的主动对话内容为空") - return - - # 统一错误处理，从这里开始所有操作都在同一个try-except块中 - logger.debug(f"[私聊][{self.private_name}]成功生成主动对话内容: {content}，准备发送") - - # 在函数内部导入PFCManager，避免循环导入 - from ..pfc_manager import PFCManager - - # 获取当前实例 - 注意这是同步方法，不需要await - pfc_manager = PFCManager.get_instance() - - # 结束当前对话实例（如果存在） - current_conversation = await pfc_manager.get_conversation(self.stream_id) - if current_conversation: - logger.info(f"[私聊][{self.private_name}]结束当前对话实例，准备创建新实例") - try: - await current_conversation.stop() - await pfc_manager.remove_conversation(self.stream_id) - except Exception as e: - logger.warning(f"[私聊][{self.private_name}]结束当前对话实例时出错: {str(e)}，继续创建新实例") - logger.warning(traceback.format_exc()) - - # 创建新的对话实例 - logger.info(f"[私聊][{self.private_name}]创建新的对话实例以发送主动消息") - new_conversation = None - try: - new_conversation = await pfc_manager.get_or_create_conversation(self.stream_id, self.private_name) - except Exception as e: - logger.error(f"[私聊][{self.private_name}]创建新对话实例失败: {str(e)}") - logger.error(traceback.format_exc()) - return - - # 确保新对话实例已初始化完成 - chat_stream = await self._get_chat_stream(new_conversation) - if not chat_stream: - logger.error(f"[私聊][{self.private_name}]无法获取有效的聊天流，取消发送主动消息") - return - - # 发送消息 - try: - await self.message_sender.send_message(chat_stream=chat_stream, content=content, reply_to_message=None) - - # 更新空闲会话启动器的最后消息时间 - await self.update_last_message_time() - - # 如果新对话实例有一个聊天观察者，请触发更新 - if new_conversation and hasattr(new_conversation, "chat_observer"): - logger.info(f"[私聊][{self.private_name}]触发聊天观察者更新") - try: - new_conversation.chat_observer.trigger_update() - except Exception as e: - logger.warning(f"[私聊][{self.private_name}]触发聊天观察者更新失败: {str(e)}") - logger.warning(traceback.format_exc()) - - logger.info(f"[私聊][{self.private_name}]成功主动发起对话: {content}") - except Exception as e: - logger.error(f"[私聊][{self.private_name}]发送主动对话消息失败: {str(e)}") - logger.error(traceback.format_exc()) - - except Exception as e: - # 顶级异常处理，确保任何未捕获的异常都不会导致整个进程崩溃 - logger.error(f"[私聊][{self.private_name}]主动发起对话过程中发生未预期的错误: {str(e)}") - logger.error(traceback.format_exc()) - - async def _get_chat_stream(self, conversation: Optional["Conversation"] = None) -> Optional[ChatStream]: - """获取可用的聊天流 - - 尝试多种方式获取聊天流： - 1. 从传入的对话实例中获取 - 2. 从全局聊天管理器中获取 - 3. 创建一个新的聊天流 - - Args: - conversation: 对话实例，可以为None - - Returns: - Optional[ChatStream]: 如果成功获取则返回聊天流，否则返回None - """ - chat_stream = None - - # 1. 尝试从对话实例获取 - if conversation and hasattr(conversation, "should_continue"): - # 等待一小段时间，确保初始化完成 - retry_count = 0 - max_retries = 10 - while not conversation.should_continue and retry_count < max_retries: - await asyncio.sleep(0.5) - retry_count += 1 - logger.debug(f"[私聊][{self.private_name}]等待新对话实例初始化完成: 尝试 {retry_count}/{max_retries}") - - if not conversation.should_continue: - logger.warning(f"[私聊][{self.private_name}]新对话实例初始化可能未完成，但仍将尝试获取聊天流") - - # 尝试使用对话实例的聊天流 - if hasattr(conversation, "chat_stream") and conversation.chat_stream: - logger.info(f"[私聊][{self.private_name}]使用新对话实例的聊天流") - return conversation.chat_stream - - # 2. 尝试从聊天管理器获取 - try: - logger.info(f"[私聊][{self.private_name}]尝试从chat_manager获取聊天流") - chat_stream = chat_manager.get_stream(self.stream_id) - if chat_stream: - return chat_stream - except Exception as e: - logger.warning(f"[私聊][{self.private_name}]从chat_manager获取聊天流失败: {str(e)}") - logger.warning(traceback.format_exc()) - - # 3. 创建新的聊天流 - try: - logger.warning(f"[私聊][{self.private_name}]无法获取现有聊天流，创建新的聊天流") - # 创建用户信息对象 - user_info = UserInfo(user_id=global_config.BOT_QQ, user_nickname=global_config.BOT_NICKNAME, platform="qq") - # 创建聊天流 - return ChatStream(self.stream_id, "qq", user_info) - except Exception as e: - logger.error(f"[私聊][{self.private_name}]创建新聊天流失败: {str(e)}") - logger.error(traceback.format_exc()) - return None diff --git a/src/plugins/PFC/action_planner.py b/src/plugins/PFC/action_planner.py index 905ee68e..17262075 100644 --- a/src/plugins/PFC/action_planner.py +++ b/src/plugins/PFC/action_planner.py @@ -3,13 +3,11 @@ import traceback from typing import Tuple, Optional, Dict, Any, List from src.common.logger_manager import get_logger -# from src.individuality.individuality import Individuality -from src.plugins.utils.chat_message_builder import build_readable_messages from ..models.utils_model import LLMRequest from src.config.config import global_config # 确保导入路径正确 -from .pfc_utils import get_items_from_json +from .pfc_utils import get_items_from_json, build_chat_history_text from .chat_observer import ChatObserver from .observation_info import ObservationInfo from .conversation_info import ConversationInfo @@ -22,10 +20,10 @@ logger = get_logger("pfc_action_planner") # Prompt(1): 首次回复或非连续回复时的决策 Prompt PROMPT_INITIAL_REPLY = """ 当前时间：{current_time_str} -现在{persona_text}正在与{sender_name}在qq上私聊 +现在[{persona_text}]正在与[{sender_name}]在qq上私聊 他们的关系是：{relationship_text} -{persona_text}现在的心情是是：{current_emotion_text} -你现在需要操控{persona_text}，根据以下【所有信息】灵活，合理的决策{persona_text}的下一步行动，需要符合正常人的社交流程，可以回复，可以倾听，甚至可以屏蔽对方： +[{persona_text}]现在的心情是：{current_emotion_text} +你现在需要操控[{persona_text}]，判断当前氛围和双方的意图，并根据以下【所有信息】灵活，合理的决策{persona_text}的下一步行动，需要符合正常人的社交流程，可以回复，可以倾听，甚至可以屏蔽对方： 【当前对话目标】 {goals_str} @@ -38,15 +36,14 @@ PROMPT_INITIAL_REPLY = """ 【最近的对话记录】(包括你已成功发送的消息 和 新收到的消息) {chat_history_text} -{spam_warning_info} ------ 可选行动类型以及解释： listening: 倾听对方发言，当你认为对方话才说到一半，发言明显未结束时选择 -direct_reply: 直接回复对方 (当有新消息需要处理时，通常应选择此项) +direct_reply: 直接回复对方 rethink_goal: 思考一个对话目标，当你觉得目前对话需要目标，或当前目标不再适用，或话题卡住时选择。注意私聊的环境是灵活的，有可能需要经常选择 end_conversation: 结束对话，对方长时间没回复，繁忙，或者当你觉得对话告一段落时可以选择 -block_and_ignore: 更加极端的结束对话方式，直接结束对话并在一段时间内无视对方所有发言（屏蔽），当对话让你感到十分不适，或你遭到各类骚扰时选择 +block_and_ignore: 更加极端的结束对话方式，直接结束对话并在一段时间内无视对方所有发言（屏蔽），当你觉得对话让[{persona_text}]感到十分不适，或[{persona_text}]遭到各类骚扰时选择 请以JSON格式输出你的决策： {{ @@ -59,10 +56,10 @@ block_and_ignore: 更加极端的结束对话方式，直接结束对话并在 # Prompt(2): 上一次成功回复后，决定继续发言时的决策 Prompt PROMPT_FOLLOW_UP = """ 当前时间：{current_time_str} -现在{persona_text}正在与{sender_name}在qq上私聊，**并且刚刚{persona_text}已经回复了对方** +现在[{persona_text}]正在与[{sender_name}]在qq上私聊，**并且刚刚[{persona_text}]已经回复了对方** 他们的关系是：{relationship_text} -{persona_text}现在的心情是是：{current_emotion_text} -你现在需要操控{persona_text}，根据以下【所有信息】灵活，合理的决策{persona_text}的下一步行动，需要符合正常人的社交流程，可以发送新消息，可以等待，可以倾听，可以结束对话，甚至可以屏蔽对方： +{persona_text}现在的心情是：{current_emotion_text} +你现在需要操控[{persona_text}]，判断当前氛围和双方的意图，并根据以下【所有信息】灵活，合理的决策[{persona_text}]的下一步行动，需要符合正常人的社交流程，可以发送新消息，可以等待，可以倾听，可以结束对话，甚至可以屏蔽对方： 【当前对话目标】 {goals_str} @@ -75,16 +72,14 @@ PROMPT_FOLLOW_UP = """ 【最近的对话记录】(包括你已成功发送的消息 和 新收到的消息) {chat_history_text} -{spam_warning_info} - ------ 可选行动类型以及解释： wait: 暂时不说话，留给对方交互空间，等待对方回复。 listening: 倾听对方发言（虽然你刚发过言，但如果对方立刻回复且明显话没说完，可以选择这个） -send_new_message: 发送一条新消息继续对话，允许适当的追问、补充、深入话题，或开启相关新话题（但是注意看对话记录，如果对方已经没有回复你，end_conversation或wait可能更合适）。 +send_new_message: 发送一条新消息，当你觉得[{persona_text}]还有话要说，或现在适合/需要发送消息时可以选择 rethink_goal: 思考一个对话目标，当你觉得目前对话需要目标，或当前目标不再适用，或话题卡住时选择。注意私聊的环境是灵活的，有可能需要经常选择 -end_conversation: 安全和平的结束对话，对方长时间没回复、繁忙、已经不再回复你消息、明显暗示或表达想结束聊天时，可以果断选择 -block_and_ignore: 更加极端的结束对话方式，直接结束对话并在一段时间内无视对方所有发言（屏蔽），当对话让你感到十分不适，或你遭到各类骚扰时选择 +end_conversation: 安全和平的结束对话，对方长时间没回复、繁忙、或你觉得对话告一段落时可以选择 +block_and_ignore: 更加极端的结束对话方式，直接结束对话并在一段时间内无视对方所有发言（屏蔽），当你觉得对话让[{persona_text}]感到十分不适，或[{persona_text}]遭到各类骚扰时选择 请以JSON格式输出你的决策： {{ @@ -136,7 +131,6 @@ PROMPT_REFLECT_AND_ACT = """ 【最近的对话记录】(包括你已成功发送的消息 和 新收到的消息) {chat_history_text} -{spam_warning_info} ------ 可选行动类型以及解释： @@ -154,6 +148,7 @@ block_and_ignore: 更加极端的结束对话方式，直接结束对话并在 注意：请严格按照JSON格式输出，不要包含任何其他内容。""" + class ActionPlanner: """行动规划器""" @@ -212,16 +207,16 @@ class ActionPlanner: time_since_last_bot_message_info = self._get_bot_last_speak_time_info(observation_info) timeout_context = self._get_timeout_context(conversation_info) goals_str = self._build_goals_string(conversation_info) - chat_history_text = await self._build_chat_history_text(observation_info) + chat_history_text = await build_chat_history_text(observation_info, self.private_name) # 获取 sender_name, relationship_text, current_emotion_text - sender_name_str = getattr(observation_info, 'sender_name', '对方') # 从 observation_info 获取 - if not sender_name_str: sender_name_str = '对方' # 再次确保有默认值 + sender_name_str = self.private_name + if not sender_name_str: + sender_name_str = "对方" # 再次确保有默认值 - relationship_text_str = getattr(conversation_info, 'relationship_text', '你们还不熟悉。') - current_emotion_text_str = getattr(conversation_info, 'current_emotion_text', '心情平静。') + relationship_text_str = getattr(conversation_info, "relationship_text", "你们还不熟悉。") + current_emotion_text_str = getattr(conversation_info, "current_emotion_text", "心情平静。") - - persona_text = f"{self.name}。" + persona_text = f"{self.name}" action_history_summary, last_action_context = self._build_action_history_context(conversation_info) # retrieved_memory_str, retrieved_knowledge_str = await retrieve_contextual_info( # chat_history_text, self.private_name @@ -236,39 +231,41 @@ class ActionPlanner: # --- 2. 选择并格式化 Prompt --- try: - if use_reflect_prompt: # 新增的判断 + if use_reflect_prompt: # 新增的判断 prompt_template = PROMPT_REFLECT_AND_ACT log_msg = "使用 PROMPT_REFLECT_AND_ACT (反思决策)" # 对于 PROMPT_REFLECT_AND_ACT，它不包含 send_new_message 选项，所以 spam_warning_message 中的相关提示可以调整或省略 # 但为了保持占位符填充的一致性，我们仍然计算它 - spam_warning_message = "" - if conversation_info.my_message_count > 5: # 这里的 my_message_count 仍有意义，表示之前连续发送了多少 - spam_warning_message = f"⚠️【警告】**你之前已连续发送{str(conversation_info.my_message_count)}条消息！请谨慎决策。**" - elif conversation_info.my_message_count > 2: - spam_warning_message = f"💬【提示】**你之前已连续发送{str(conversation_info.my_message_count)}条消息。请注意保持对话平衡。**" + # spam_warning_message = "" + # if conversation_info.my_message_count > 5: # 这里的 my_message_count 仍有意义，表示之前连续发送了多少 + # spam_warning_message = ( + # f"⚠️【警告】**你之前已连续发送{str(conversation_info.my_message_count)}条消息！请谨慎决策。**" + # ) + # elif conversation_info.my_message_count > 2: + # spam_warning_message = f"💬【提示】**你之前已连续发送{str(conversation_info.my_message_count)}条消息。请注意保持对话平衡。**" elif last_successful_reply_action in ["direct_reply", "send_new_message"]: prompt_template = PROMPT_FOLLOW_UP log_msg = "使用 PROMPT_FOLLOW_UP (追问决策)" - spam_warning_message = "" - if conversation_info.my_message_count > 5: - spam_warning_message = f"⚠️【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息！请注意不要再选择send_new_message！以免刷屏对造成对方困扰！**" - elif conversation_info.my_message_count > 2: - spam_warning_message = f"💬【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息。请保持理智，如果非必要，请避免选择send_new_message，以免给对方造成困扰。**" + # spam_warning_message = "" + # if conversation_info.my_message_count > 5: + # spam_warning_message = f"⚠️【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息！请注意不要再选择send_new_message！以免刷屏对造成对方困扰！**" + # elif conversation_info.my_message_count > 2: + # spam_warning_message = f"💬【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息。请保持理智，如果非必要，请避免选择send_new_message，以免给对方造成困扰。**" else: prompt_template = PROMPT_INITIAL_REPLY log_msg = "使用 PROMPT_INITIAL_REPLY (首次/非连续回复决策)" - spam_warning_message = "" # 初始回复时通常不需要刷屏警告 + # spam_warning_message = "" # 初始回复时通常不需要刷屏警告 logger.debug(f"[私聊][{self.private_name}] {log_msg}") current_time_value = "获取时间失败" - if observation_info and hasattr(observation_info, 'current_time_str') and observation_info.current_time_str: + if observation_info and hasattr(observation_info, "current_time_str") and observation_info.current_time_str: current_time_value = observation_info.current_time_str - if spam_warning_message: - spam_warning_message = f"\n{spam_warning_message}\n" + # if spam_warning_message: + # spam_warning_message = f"\n{spam_warning_message}\n" prompt = prompt_template.format( persona_text=persona_text, @@ -281,10 +278,10 @@ class ActionPlanner: # retrieved_memory_str=retrieved_memory_str if retrieved_memory_str else "无相关记忆。", # retrieved_knowledge_str=retrieved_knowledge_str if retrieved_knowledge_str else "无相关知识。", current_time_str=current_time_value, - spam_warning_info=spam_warning_message, + # spam_warning_info=spam_warning_message, sender_name=sender_name_str, relationship_text=relationship_text_str, - current_emotion_text=current_emotion_text_str + current_emotion_text=current_emotion_text_str, ) logger.debug(f"[私聊][{self.private_name}] 发送到LLM的最终提示词:\n------\n{prompt}\n------") except KeyError as fmt_key_err: @@ -332,10 +329,11 @@ class ActionPlanner: time_str_for_end_decision = observation_info.current_time_str final_action, final_reason = await self._handle_end_conversation_decision( persona_text, - chat_history_text, initial_reason, - time_str_for_end_decision, + chat_history_text, + initial_reason, + time_str_for_end_decision, sender_name_str=sender_name_str, - relationship_text_str=relationship_text_str + relationship_text_str=relationship_text_str, ) except Exception as end_dec_err: logger.error(f"[私聊][{self.private_name}] 处理结束对话决策时出错: {end_dec_err}") @@ -360,7 +358,7 @@ class ActionPlanner: "block_and_ignore", "say_goodbye", ] - valid_actions_reflect = [ # PROMPT_REFLECT_AND_ACT 的动作 + valid_actions_reflect = [ # PROMPT_REFLECT_AND_ACT 的动作 "wait", "listening", "rethink_goal", @@ -466,52 +464,6 @@ class ActionPlanner: goals_str = "- 构建对话目标时出错。\n" return goals_str - async def _build_chat_history_text(self, observation_info: ObservationInfo) -> str: - """构建聊天历史记录文本 (包含未处理消息)""" - - chat_history_text = "" - try: - if hasattr(observation_info, "chat_history_str") and observation_info.chat_history_str: - chat_history_text = observation_info.chat_history_str - elif hasattr(observation_info, "chat_history") and observation_info.chat_history: - history_slice = observation_info.chat_history[-20:] - chat_history_text = await build_readable_messages( - history_slice, replace_bot_name=True, merge_messages=False, timestamp_mode="relative", read_mark=0.0 - ) - else: - chat_history_text = "还没有聊天记录。\n" - unread_count = getattr(observation_info, "new_messages_count", 0) - unread_messages = getattr(observation_info, "unprocessed_messages", []) - if unread_count > 0 and unread_messages: - bot_qq_str = str(global_config.BOT_QQ) - other_unread_messages = [ - msg for msg in unread_messages if msg.get("user_info", {}).get("user_id") != bot_qq_str - ] - other_unread_count = len(other_unread_messages) - if other_unread_count > 0: - new_messages_str = await build_readable_messages( - other_unread_messages, - replace_bot_name=True, - merge_messages=False, - timestamp_mode="relative", - read_mark=0.0, - ) - chat_history_text += ( - f"\n--- 以下是 {other_unread_count} 条你需要处理的新消息 ---\n{new_messages_str}\n------\n" - ) - logger.debug(f"[私聊][{self.private_name}] 向 LLM 追加了 {other_unread_count} 条未读消息。") - else: - chat_history_text += ( - f"\n--- 以上均为已读消息，未读消息均已处理完毕 ---\n" - ) - except AttributeError as e: - logger.warning(f"[私聊][{self.private_name}] 构建聊天记录文本时属性错误: {e}") - chat_history_text = "[获取聊天记录时出错]\n" - except Exception as e: - logger.error(f"[私聊][{self.private_name}] 处理聊天记录时发生未知错误: {e}") - chat_history_text = "[处理聊天记录时出错]\n" - return chat_history_text - def _build_action_history_context(self, conversation_info: ConversationInfo) -> Tuple[str, str]: """构建行动历史概要和上一次行动详细情况""" @@ -561,11 +513,23 @@ class ActionPlanner: # --- Helper method for handling end_conversation decision --- async def _handle_end_conversation_decision( - self, persona_text: str, chat_history_text: str, initial_reason: str, current_time_str: str, sender_name_str: str, relationship_text_str: str + self, + persona_text: str, + chat_history_text: str, + initial_reason: str, + current_time_str: str, + sender_name_str: str, + relationship_text_str: str, ) -> Tuple[str, str]: """处理结束对话前的告别决策""" logger.info(f"[私聊][{self.private_name}] 初步规划结束对话，进入告别决策...") - end_decision_prompt = PROMPT_END_DECISION.format(persona_text=persona_text, chat_history_text=chat_history_text,current_time_str=current_time_str,sender_name = sender_name_str, relationship_text = relationship_text_str) + end_decision_prompt = PROMPT_END_DECISION.format( + persona_text=persona_text, + chat_history_text=chat_history_text, + current_time_str=current_time_str, + sender_name=sender_name_str, + relationship_text=relationship_text_str, + ) logger.debug(f"[私聊][{self.private_name}] 发送到LLM的结束决策提示词:\n------\n{end_decision_prompt}\n------") llm_start_time = time.time() end_content, _ = await self.llm.generate_response_async(end_decision_prompt) diff --git a/src/plugins/PFC/actions.py b/src/plugins/PFC/actions.py index 8d8a9e0f..9f966097 100644 --- a/src/plugins/PFC/actions.py +++ b/src/plugins/PFC/actions.py @@ -295,7 +295,9 @@ async def handle_action( # 后续的 plan 循环会检测到这个 "done_no_reply" 状态并使用反思 prompt elif is_suitable: # 适用于 direct_reply 或 (send_new_message 且 RG决定发送并通过检查) - logger.debug(f"[私聊][{conversation_instance.private_name}] 动作 '{action}': 找到合适的回复，准备发送。") + logger.debug( + f"[私聊][{conversation_instance.private_name}] 动作 '{action}': 找到合适的回复，准备发送。" + ) # conversation_info.last_reply_rejection_reason = None # 已在循环内清除 # conversation_info.last_rejected_reply_content = None conversation_instance.generated_reply = generated_content_for_check_or_send # 使用检查通过的内容 @@ -361,7 +363,7 @@ async def handle_action( observation_info.chat_history_str = "[构建聊天记录出错]" # --- 新增结束 --- - # 更新 idle_conversation_starter 的最后消息时间 + # 更新 idle_chat 的最后消息时间 # (避免在发送消息后很快触发主动聊天) if conversation_instance.idle_chat: await conversation_instance.idle_chat.update_last_message_time(send_end_time) @@ -506,7 +508,7 @@ async def handle_action( action_successful = True # 标记成功 # final_status 和 final_reason 会在 finally 中设置 logger.info(f"[私聊][{conversation_instance.private_name}] 成功发送告别语，即将停止对话实例。") - # 更新 idle_conversation_starter 的最后消息时间 + # 更新 idle_chat 的最后消息时间 # (避免在发送消息后很快触发主动聊天) if conversation_instance.idle_chat: await conversation_instance.idle_chat.update_last_message_time(send_end_time) diff --git a/src/plugins/PFC/conversation.py b/src/plugins/PFC/conversation.py index da613317..6132ae13 100644 --- a/src/plugins/PFC/conversation.py +++ b/src/plugins/PFC/conversation.py @@ -25,12 +25,10 @@ from .observation_info import ObservationInfo from .conversation_info import ConversationInfo from .reply_generator import ReplyGenerator from .PFC_idle.idle_chat import IdleChat -from .pfc_KnowledgeFetcher import KnowledgeFetcher from .waiter import Waiter from .reply_checker import ReplyChecker -# >>> 新增导入 <<< -from .conversation_loop import run_conversation_loop # 导入新的循环函数 +from .conversation_loop import run_conversation_loop from rich.traceback import install @@ -38,13 +36,6 @@ install(extra_lines=3) logger = get_logger("pfc_conversation") -# 时区配置移到 loop 文件或更全局的位置，这里不再需要 -# configured_tz = getattr(global_config, 'TIME_ZONE', 'Asia/Shanghai') -# TIME_ZONE = tz.gettz(configured_tz) -# if TIME_ZONE is None: -# logger.error(f"配置的时区 '{configured_tz}' 无效，将使用默认时区 'Asia/Shanghai'") -# TIME_ZONE = tz.gettz('Asia/Shanghai') - class Conversation: """ @@ -59,7 +50,7 @@ class Conversation: self.stream_id: str = stream_id self.private_name: str = private_name self.state: ConversationState = ConversationState.INIT - self.should_continue: bool = False # Manager 会在初始化后设置 + self.should_continue: bool = False self.ignore_until_timestamp: Optional[float] = None self.generated_reply: str = "" self.chat_stream: Optional[ChatStream] = None @@ -74,7 +65,6 @@ class Conversation: self.action_planner: Optional[ActionPlanner] = None self.goal_analyzer: Optional[GoalAnalyzer] = None self.reply_generator: Optional[ReplyGenerator] = None - self.knowledge_fetcher: Optional[KnowledgeFetcher] = None self.waiter: Optional[Waiter] = None self.direct_sender: Optional[DirectMessageSender] = None self.idle_chat: Optional[IdleChat] = None @@ -83,13 +73,14 @@ class Conversation: self.conversation_info: Optional[ConversationInfo] = None self.reply_checker: Optional[ReplyChecker] = None - self._initialized: bool = False # Manager 会在初始化成功后设为 True + self._initialized: bool = False self.bot_qq_str: Optional[str] = str(global_config.BOT_QQ) if global_config.BOT_QQ else None if not self.bot_qq_str: logger.error(f"[私聊][{self.private_name}] 严重错误：未能从配置中获取 BOT_QQ ID！") - # _initialize 和 _load_initial_history 方法已被移除 + # 确保这个属性被正确初始化 + self.consecutive_llm_action_failures: int = 0 # LLM相关动作连续失败的计数器 async def start(self): """ @@ -105,27 +96,27 @@ class Conversation: logger.info(f"[私聊][{self.private_name}] 对话系统启动，准备创建规划循环任务...") try: - # >>> 修改后的调用 <<< + # 创建PFC主循环任务 _loop_task = asyncio.create_task(run_conversation_loop(self)) logger.info(f"[私聊][{self.private_name}] 规划循环任务已创建。") except Exception as task_err: logger.error(f"[私聊][{self.private_name}] 创建规划循环任务时出错: {task_err}") - await self.stop() + await self.stop() # 发生错误时尝试停止 async def stop(self): """ 停止对话实例并清理相关资源。 """ logger.info(f"[私聊][{self.private_name}] 正在停止对话实例: {self.stream_id}") - self.should_continue = False + self.should_continue = False # 设置标志以退出循环 # 最终关系评估 if ( - self._initialized + self._initialized # 确保已初始化 and self.relationship_updater and self.conversation_info and self.observation_info - and self.chat_observer + and self.chat_observer # 确保所有需要的组件都存在 ): try: logger.info(f"[私聊][{self.private_name}] 准备执行最终关系评估...") @@ -143,11 +134,10 @@ class Conversation: # 停止其他组件 if self.idle_chat: - # 减少活跃实例计数，而不是停止IdleChat - await self.idle_chat.decrement_active_instances() + await self.idle_chat.decrement_active_instances() # 减少活跃实例计数 logger.debug(f"[私聊][{self.private_name}] 已减少IdleChat活跃实例计数") - if self.observation_info and self.chat_observer: - self.observation_info.unbind_from_chat_observer() + if self.observation_info and self.chat_observer: # 确保二者都存在 + self.observation_info.unbind_from_chat_observer() # 解绑 if self.mood_mng and hasattr(self.mood_mng, "stop_mood_update") and self.mood_mng._running: # type: ignore self.mood_mng.stop_mood_update() # type: ignore logger.debug(f"[私聊][{self.private_name}] MoodManager 后台更新已停止。") @@ -155,12 +145,11 @@ class Conversation: self._initialized = False # 标记为未初始化 logger.info(f"[私聊][{self.private_name}] 对话实例 {self.stream_id} 已停止。") - # _plan_and_action_loop 方法已被移除 - def _convert_to_message(self, msg_dict: Dict[str, Any]) -> Optional[Message]: """将从数据库或其他来源获取的消息字典转换为内部使用的 Message 对象""" # 这个方法似乎没有被其他内部方法调用，但为了完整性暂时保留 try: + # 尝试获取与此对话实例关联的 ChatStream chat_stream_to_use = self.chat_stream or chat_manager.get_stream(self.stream_id) if not chat_stream_to_use: logger.error( @@ -168,6 +157,7 @@ class Conversation: ) return None + # 解析 UserInfo user_info_dict = msg_dict.get("user_info", {}) user_info: Optional[UserInfo] = None if isinstance(user_info_dict, dict): @@ -177,21 +167,22 @@ class Conversation: logger.warning( f"[私聊][{self.private_name}] 从字典创建 UserInfo 时出错: {e}, dict: {user_info_dict}" ) - if not user_info: + if not user_info: # 如果没有有效的 UserInfo，则无法创建 Message 对象 logger.warning( f"[私聊][{self.private_name}] 消息缺少有效的 UserInfo，无法转换。 msg_id: {msg_dict.get('message_id')}" ) return None + # 创建并返回 Message 对象 return Message( - message_id=msg_dict.get("message_id", f"gen_{time.time()}"), + message_id=msg_dict.get("message_id", f"gen_{time.time()}"), # 提供默认 message_id chat_stream=chat_stream_to_use, - time=msg_dict.get("time", time.time()), + time=msg_dict.get("time", time.time()), # 提供默认时间 user_info=user_info, - processed_plain_text=msg_dict.get("processed_plain_text", ""), - detailed_plain_text=msg_dict.get("detailed_plain_text", ""), + processed_plain_text=msg_dict.get("processed_plain_text", ""), # 提供默认文本 + detailed_plain_text=msg_dict.get("detailed_plain_text", ""), # 提供默认详细文本 ) except Exception as e: logger.error(f"[私聊][{self.private_name}] 转换消息时出错: {e}") logger.error(f"[私聊][{self.private_name}] {traceback.format_exc()}") - return None + return None # 出错时返回 None diff --git a/src/plugins/PFC/conversation_info.py b/src/plugins/PFC/conversation_info.py index 0e7a5138..8bbf5ef5 100644 --- a/src/plugins/PFC/conversation_info.py +++ b/src/plugins/PFC/conversation_info.py @@ -17,4 +17,5 @@ class ConversationInfo: self.relationship_text: Optional[str] = "你们还不熟悉。" # 与当前对话者的关系描述文本 self.current_emotion_text: Optional[str] = "心情平静。" # 机器人当前的情绪描述文本 self.current_instance_message_count: int = 0 # 当前私聊实例中的消息计数 + self.other_new_messages_during_planning_count: int = 0 # 在计划阶段期间收到的其他新消息计数 # --- 新增字段结束 --- diff --git a/src/plugins/PFC/conversation_initializer.py b/src/plugins/PFC/conversation_initializer.py index 932f74f5..118bc171 100644 --- a/src/plugins/PFC/conversation_initializer.py +++ b/src/plugins/PFC/conversation_initializer.py @@ -18,7 +18,6 @@ from .observation_info import ObservationInfo from .conversation_info import ConversationInfo from .reply_generator import ReplyGenerator from .PFC_idle.idle_chat import IdleChat -from .pfc_KnowledgeFetcher import KnowledgeFetcher # 修正大小写 from .waiter import Waiter from .pfc_utils import get_person_id from .reply_checker import ReplyChecker @@ -166,9 +165,6 @@ async def initialize_core_components(conversation_instance: "Conversation"): conversation_instance.stream_id, conversation_instance.private_name ) - logger.debug(f"[私聊][{conversation_instance.private_name}] (Initializer) 初始化 KnowledgeFetcher...") - conversation_instance.knowledge_fetcher = KnowledgeFetcher(conversation_instance.private_name) - logger.debug(f"[私聊][{conversation_instance.private_name}] (Initializer) 初始化 Waiter...") conversation_instance.waiter = Waiter(conversation_instance.stream_id, conversation_instance.private_name) diff --git a/src/plugins/PFC/conversation_loop.py b/src/plugins/PFC/conversation_loop.py index d0cd287d..aab031cc 100644 --- a/src/plugins/PFC/conversation_loop.py +++ b/src/plugins/PFC/conversation_loop.py @@ -15,18 +15,19 @@ if TYPE_CHECKING: logger = get_logger("pfc_loop") -# 时区配置 (从 conversation.py 移过来，或者考虑放到更全局的配置模块) +# 时区配置 configured_tz = getattr(global_config, "TIME_ZONE", "Asia/Shanghai") TIME_ZONE = tz.gettz(configured_tz) if TIME_ZONE is None: logger.error(f"配置的时区 '{configured_tz}' 无效，将使用默认时区 'Asia/Shanghai'") TIME_ZONE = tz.gettz("Asia/Shanghai") +MAX_CONSECUTIVE_LLM_ACTION_FAILURES = 3 # 可配置的最大LLM连续失败次数 + async def run_conversation_loop(conversation_instance: "Conversation"): """ 核心的规划与行动循环 (PFC Loop)。 - 之前是 Conversation 类中的 _plan_and_action_loop 方法。 """ logger.debug(f"[私聊][{conversation_instance.private_name}] 进入 run_conversation_loop 循环。") @@ -34,16 +35,20 @@ async def run_conversation_loop(conversation_instance: "Conversation"): logger.error(f"[私聊][{conversation_instance.private_name}] 尝试在未初始化状态下运行规划循环，退出。") return - force_reflect_and_act = False # 用于强制使用反思 prompt 的标志 + _force_reflect_and_act_next_iter = False while conversation_instance.should_continue: loop_iter_start_time = time.time() - logger.debug(f"[私聊][{conversation_instance.private_name}] 开始新一轮循环迭代 ({loop_iter_start_time:.2f})") + current_force_reflect_and_act = _force_reflect_and_act_next_iter + _force_reflect_and_act_next_iter = False + + logger.debug( + f"[私聊][{conversation_instance.private_name}] 开始新一轮循环迭代 ({loop_iter_start_time:.2f}), force_reflect_next_iter: {current_force_reflect_and_act}, consecutive_llm_failures: {conversation_instance.consecutive_llm_action_failures}" + ) - # 更新当前时间 try: - global TIME_ZONE # 引用全局 TIME_ZONE - if TIME_ZONE is None: # 如果还未加载成功 + global TIME_ZONE + if TIME_ZONE is None: configured_tz_loop = getattr(global_config, "TIME_ZONE", "Asia/Shanghai") TIME_ZONE = tz.gettz(configured_tz_loop) if TIME_ZONE is None: @@ -54,7 +59,6 @@ async def run_conversation_loop(conversation_instance: "Conversation"): if conversation_instance.observation_info: time_str = current_time_dt.strftime("%Y-%m-%d %H:%M:%S %Z%z") conversation_instance.observation_info.current_time_str = time_str - logger.debug(f"[私聊][{conversation_instance.private_name}] 更新 ObservationInfo 当前时间: {time_str}") else: logger.warning( f"[私聊][{conversation_instance.private_name}] ObservationInfo 未初始化，无法更新当前时间。" @@ -64,15 +68,11 @@ async def run_conversation_loop(conversation_instance: "Conversation"): f"[私聊][{conversation_instance.private_name}] 更新 ObservationInfo 当前时间时出错: {time_update_err}" ) - # 处理忽略状态 if ( conversation_instance.ignore_until_timestamp and loop_iter_start_time < conversation_instance.ignore_until_timestamp ): if conversation_instance.idle_chat and conversation_instance.idle_chat._running: - # 不直接停止服务，改为暂时忽略此用户 - # 虽然我们仍然可以通过active_instances_count来决定是否触发主动聊天 - # 但为了安全起见，我们只记录一个日志 logger.debug(f"[私聊][{conversation_instance.private_name}] 对话被暂时忽略，暂停对该用户的主动聊天") sleep_duration = min(30, conversation_instance.ignore_until_timestamp - loop_iter_start_time) await asyncio.sleep(sleep_duration) @@ -85,18 +85,13 @@ async def run_conversation_loop(conversation_instance: "Conversation"): f"[私聊][{conversation_instance.private_name}] 忽略时间已到 {conversation_instance.stream_id}，准备结束对话。" ) conversation_instance.ignore_until_timestamp = None - await conversation_instance.stop() # 调用 Conversation 实例的 stop 方法 + await conversation_instance.stop() continue else: - # 忽略状态结束，这里不需要任何特殊处理 - # IdleChat会通过active_instances_count自动决定是否触发 pass - # 核心规划与行动逻辑 try: - # 更新关系和情绪文本 (在每次循环开始时进行) if conversation_instance.conversation_info and conversation_instance._initialized: - # 更新关系 if ( conversation_instance.conversation_info.person_id and conversation_instance.relationship_translator @@ -121,13 +116,11 @@ async def run_conversation_loop(conversation_instance: "Conversation"): except Exception as e_rel: logger.error(f"[私聊][{conversation_instance.private_name}] (Loop) 更新关系文本时出错: {e_rel}") conversation_instance.conversation_info.relationship_text = "你们的关系是：普通。" - # 更新情绪 if conversation_instance.mood_mng: conversation_instance.conversation_info.current_emotion_text = ( conversation_instance.mood_mng.get_prompt() - ) # type: ignore + ) - # 检查核心组件 if not all( [ conversation_instance.action_planner, @@ -141,7 +134,6 @@ async def run_conversation_loop(conversation_instance: "Conversation"): await asyncio.sleep(5) continue - # 规划 planning_start_time = time.time() logger.debug( f"[私聊][{conversation_instance.private_name}] --- (Loop) 开始规划 ({planning_start_time:.2f}) ---" @@ -155,79 +147,55 @@ async def run_conversation_loop(conversation_instance: "Conversation"): conversation_instance.conversation_info.last_successful_reply_action if conversation_instance.conversation_info else None, - use_reflect_prompt=force_reflect_and_act, + use_reflect_prompt=current_force_reflect_and_act, ) - force_reflect_and_act = False + logger.debug( f"[私聊][{conversation_instance.private_name}] (Loop) ActionPlanner.plan 完成，初步规划动作: {action}" ) - # 检查中断 - current_unprocessed_messages = getattr(conversation_instance.observation_info, "unprocessed_messages", []) - new_messages_during_planning: List[Dict[str, Any]] = [] - other_new_messages_during_planning: List[Dict[str, Any]] = [] + current_unprocessed_messages_after_plan = getattr( + conversation_instance.observation_info, "unprocessed_messages", [] + ) + new_messages_during_action_planning: List[Dict[str, Any]] = [] + other_new_messages_during_action_planning: List[Dict[str, Any]] = [] - for msg in current_unprocessed_messages: - msg_time = msg.get("time") - sender_id_info = msg.get("user_info", {}) - sender_id = str(sender_id_info.get("user_id")) if sender_id_info else None - if msg_time and msg_time >= planning_start_time: - new_messages_during_planning.append(msg) - if sender_id != conversation_instance.bot_qq_str: - other_new_messages_during_planning.append(msg) + for msg_ap in current_unprocessed_messages_after_plan: + msg_time_ap = msg_ap.get("time") + sender_id_info_ap = msg_ap.get("user_info", {}) + sender_id_ap = str(sender_id_info_ap.get("user_id")) if sender_id_info_ap else None + if msg_time_ap and msg_time_ap >= planning_start_time: + new_messages_during_action_planning.append(msg_ap) + if sender_id_ap != conversation_instance.bot_qq_str: + other_new_messages_during_action_planning.append(msg_ap) - new_msg_count = len(new_messages_during_planning) - other_new_msg_count = len(other_new_messages_during_planning) + new_msg_count_action_planning = len(new_messages_during_action_planning) + other_new_msg_count_action_planning = len(other_new_messages_during_action_planning) - if conversation_instance.conversation_info and other_new_msg_count > 0: - conversation_instance.conversation_info.current_instance_message_count += other_new_msg_count - # 触发关系和情绪更新（如果需要） - if ( - conversation_instance.relationship_updater - and conversation_instance.observation_info - and conversation_instance.chat_observer - ): - await conversation_instance.relationship_updater.update_relationship_incremental( - conversation_info=conversation_instance.conversation_info, - observation_info=conversation_instance.observation_info, - chat_observer_for_history=conversation_instance.chat_observer, - ) - if ( - conversation_instance.emotion_updater - and other_new_messages_during_planning - and conversation_instance.observation_info - and conversation_instance.chat_observer - ): - last_user_msg = other_new_messages_during_planning[-1] - last_user_msg_text = last_user_msg.get("processed_plain_text", "用户发了新消息") - sender_name_for_event = getattr(conversation_instance.observation_info, "sender_name", "对方") - event_desc = f"用户【{sender_name_for_event}】发送了新消息: '{last_user_msg_text[:30]}...'" - await conversation_instance.emotion_updater.update_emotion_based_on_context( - conversation_info=conversation_instance.conversation_info, - observation_info=conversation_instance.observation_info, - chat_observer_for_history=conversation_instance.chat_observer, - event_description=event_desc, - ) + if conversation_instance.conversation_info and other_new_msg_count_action_planning > 0: + pass - should_interrupt: bool = False - interrupt_reason: str = "" - if action in ["wait", "listening"] and new_msg_count > 0: - should_interrupt = True - interrupt_reason = f"规划 {action} 期间收到 {new_msg_count} 条新消息" - elif other_new_msg_count > 2: # Threshold for other actions - should_interrupt = True - interrupt_reason = f"规划 {action} 期间收到 {other_new_msg_count} 条来自他人的新消息" - - if should_interrupt: - logger.info( - f"[私聊][{conversation_instance.private_name}] (Loop) 中断 '{action}'，原因: {interrupt_reason}。重新规划..." + should_interrupt_action_planning: bool = False + interrupt_reason_action_planning: str = "" + if action in ["wait", "listening"] and new_msg_count_action_planning > 0: + should_interrupt_action_planning = True + interrupt_reason_action_planning = f"规划 {action} 期间收到 {new_msg_count_action_planning} 条新消息" + elif other_new_msg_count_action_planning > 2: + should_interrupt_action_planning = True + interrupt_reason_action_planning = ( + f"规划 {action} 期间收到 {other_new_msg_count_action_planning} 条来自他人的新消息" ) - cancel_record = { + + if should_interrupt_action_planning: + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) 中断 '{action}' (在ActionPlanner.plan后)，原因: {interrupt_reason_action_planning}。重新规划..." + ) + cancel_record_ap = { "action": action, "plan_reason": reason, - "status": "cancelled_due_to_new_messages", + "status": "cancelled_due_to_new_messages_during_action_plan", "time": datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S"), - "final_reason": interrupt_reason, + "final_reason": interrupt_reason_action_planning, } if conversation_instance.conversation_info: if ( @@ -235,40 +203,208 @@ async def run_conversation_loop(conversation_instance: "Conversation"): or conversation_instance.conversation_info.done_action is None ): conversation_instance.conversation_info.done_action = [] - conversation_instance.conversation_info.done_action.append(cancel_record) + conversation_instance.conversation_info.done_action.append(cancel_record_ap) conversation_instance.conversation_info.last_successful_reply_action = None conversation_instance.state = ConversationState.ANALYZING await asyncio.sleep(0.1) continue - # 执行动作 (调用 actions 模块的函数) - logger.debug( - f"[私聊][{conversation_instance.private_name}] (Loop) 未中断，调用 actions.handle_action 执行动作 '{action}'..." - ) - if conversation_instance.conversation_info: - conversation_instance.conversation_info.other_new_messages_during_planning_count = other_new_msg_count + # --- LLM Action Handling with Shield and Failure Count --- + if action in ["direct_reply", "send_new_message"]: + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) 动作 '{action}' 需要LLM生成，进入监控执行模式..." + ) + llm_call_start_time = time.time() - await actions.handle_action( - conversation_instance, - action, - reason, - conversation_instance.observation_info, - conversation_instance.conversation_info, - ) - logger.debug(f"[私聊][{conversation_instance.private_name}] (Loop) actions.handle_action 完成。") + if conversation_instance.conversation_info: + conversation_instance.conversation_info.other_new_messages_during_planning_count = ( + other_new_msg_count_action_planning + ) + + llm_action_task = asyncio.create_task( + actions.handle_action( + conversation_instance, + action, + reason, + conversation_instance.observation_info, + conversation_instance.conversation_info, + ) + ) + + interrupted_by_new_messages = False + llm_action_completed_successfully = False + action_outcome_processed = False # Flag to ensure we process outcome only once + + while not llm_action_task.done() and not action_outcome_processed: + try: + # Shield the task so wait_for timeout doesn't cancel it directly + await asyncio.wait_for(asyncio.shield(llm_action_task), timeout=1.5) + # If wait_for completes without timeout, the shielded task is done (or errored/cancelled by itself) + action_outcome_processed = True # Outcome will be processed outside this loop + except asyncio.TimeoutError: + # Shielded task didn't finish in 1.5s. This is our chance to check messages. + current_time_for_check = time.time() + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM Monitor polling. llm_call_start_time: {llm_call_start_time:.2f}, current_check_time: {current_time_for_check:.2f}. Task still running, checking for new messages." + ) + + current_unprocessed_messages_during_llm = getattr( + conversation_instance.observation_info, "unprocessed_messages", [] + ) + other_new_messages_this_check: List[Dict[str, Any]] = [] + + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) Checking unprocessed_messages (count: {len(current_unprocessed_messages_during_llm)}):" + ) + for msg_llm in current_unprocessed_messages_during_llm: + msg_time_llm = msg_llm.get("time") + sender_id_info_llm = msg_llm.get("user_info", {}) + sender_id_llm = str(sender_id_info_llm.get("user_id")) if sender_id_info_llm else None + is_new_enough = msg_time_llm and msg_time_llm >= llm_call_start_time + is_other_sender = sender_id_llm != conversation_instance.bot_qq_str + + time_str_for_log = f"{msg_time_llm:.2f}" if msg_time_llm is not None else "N/A" + logger.debug( + f" - Msg ID: {msg_llm.get('message_id')}, Time: {time_str_for_log}, Sender: {sender_id_llm}. New enough? {is_new_enough}. Other sender? {is_other_sender}." + ) + + if is_new_enough and is_other_sender: + other_new_messages_this_check.append(msg_llm) + + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) Found {len(other_new_messages_this_check)} 'other_new_messages_this_check'." + ) + + if len(other_new_messages_this_check) > global_config.pfc_message_buffer_size: + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作 '{action}' 执行期间收到 {len(other_new_messages_this_check)} 条来自他人的新消息，将取消LLM任务。" + ) + if not llm_action_task.done(): # Check again before cancelling + llm_action_task.cancel() # Now we explicitly cancel the original task + interrupted_by_new_messages = True + action_outcome_processed = True # We've made a decision, exit monitoring + # else: continue polling if not enough new messages + # Shield ensures CancelledError from llm_action_task itself is caught by the outer try/except + + # After the monitoring loop (either task finished, or we decided to cancel it) + # Await the task properly to get its result or handle its exception/cancellation + action_final_status_in_history = "unknown" + try: + await llm_action_task # This will re-raise CancelledError if we cancelled it, or other exceptions + + # If no exception, it means the task completed. + # actions.handle_action updates done_action, so we check its status. + if conversation_instance.conversation_info and conversation_instance.conversation_info.done_action: + # Check if done_action is not empty + if conversation_instance.conversation_info.done_action: + action_final_status_in_history = conversation_instance.conversation_info.done_action[ + -1 + ].get("status", "unknown") + + if action_final_status_in_history in ["done", "done_no_reply"]: + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作 '{action}' 任务最终成功完成 (status: {action_final_status_in_history})。" + ) + conversation_instance.consecutive_llm_action_failures = 0 + llm_action_completed_successfully = True + else: + logger.warning( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作 '{action}' 任务完成但未成功 (status: {action_final_status_in_history})。" + ) + if not interrupted_by_new_messages: + conversation_instance.consecutive_llm_action_failures += 1 + + except asyncio.CancelledError: + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作 '{action}' 任务最终确认被取消。" + ) + if not interrupted_by_new_messages: + conversation_instance.consecutive_llm_action_failures += 1 + logger.warning( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM任务因外部原因取消，连续失败次数: {conversation_instance.consecutive_llm_action_failures}" + ) + else: # interrupted_by_new_messages is True + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM任务因新消息被内部逻辑取消，不计为LLM失败。" + ) + + except Exception as e_llm_final: + logger.error( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作 '{action}' 任务执行时发生最终错误: {e_llm_final}" + ) + logger.error(traceback.format_exc()) + conversation_instance.state = ConversationState.ERROR + if not interrupted_by_new_messages: + conversation_instance.consecutive_llm_action_failures += 1 + + # --- Post LLM Action Task Handling --- + if not llm_action_completed_successfully: + if conversation_instance.consecutive_llm_action_failures >= MAX_CONSECUTIVE_LLM_ACTION_FAILURES: + logger.error( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM相关动作连续失败或被取消 {conversation_instance.consecutive_llm_action_failures} 次。将强制等待并重置计数器。" + ) + + action = "wait" # Force action to wait + reason = f"LLM连续失败{conversation_instance.consecutive_llm_action_failures}次，强制等待" + conversation_instance.consecutive_llm_action_failures = 0 + + if conversation_instance.conversation_info: + conversation_instance.conversation_info.last_successful_reply_action = None + + logger.info(f"[私聊][{conversation_instance.private_name}] (Loop) 执行强制等待动作...") + await actions.handle_action( + conversation_instance, + action, + reason, + conversation_instance.observation_info, + conversation_instance.conversation_info, + ) + _force_reflect_and_act_next_iter = False + conversation_instance.state = ConversationState.ANALYZING + await asyncio.sleep(1) + continue + else: + conversation_instance.state = ConversationState.ANALYZING + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) LLM动作中断/失败，准备重新规划。Interrupted by new msgs: {interrupted_by_new_messages}, Consecutive LLM Failures: {conversation_instance.consecutive_llm_action_failures}" + ) + await asyncio.sleep(0.1) + continue + else: + logger.debug(f"[私聊][{conversation_instance.private_name}] (Loop) 执行非LLM类动作 '{action}'...") + conversation_instance.consecutive_llm_action_failures = 0 + logger.debug( + f"[私聊][{conversation_instance.private_name}] (Loop) 重置 consecutive_llm_action_failures due to non-LLM action." + ) + + if conversation_instance.conversation_info: + conversation_instance.conversation_info.other_new_messages_during_planning_count = ( + other_new_msg_count_action_planning + ) + + await actions.handle_action( + conversation_instance, + action, + reason, + conversation_instance.observation_info, + conversation_instance.conversation_info, + ) + logger.debug(f"[私聊][{conversation_instance.private_name}] (Loop) 非LLM类动作 '{action}' 完成。") - # 检查是否需要反思 last_action_record = {} if conversation_instance.conversation_info and conversation_instance.conversation_info.done_action: - last_action_record = conversation_instance.conversation_info.done_action[-1] + if conversation_instance.conversation_info.done_action: + last_action_record = conversation_instance.conversation_info.done_action[-1] + if ( last_action_record.get("action") == "send_new_message" and last_action_record.get("status") == "done_no_reply" ): - logger.info(f"[私聊][{conversation_instance.private_name}] (Loop) 检测到需反思，设置标志。") - force_reflect_and_act = True + logger.info( + f"[私聊][{conversation_instance.private_name}] (Loop) 检测到 ReplyGenerator 决定不发送消息，下一轮将强制反思。" + ) + _force_reflect_and_act_next_iter = True - # 检查结束条件 goal_ended: bool = False if ( conversation_instance.conversation_info @@ -286,7 +422,9 @@ async def run_conversation_loop(conversation_instance: "Conversation"): last_action_record_for_end_check = {} if conversation_instance.conversation_info and conversation_instance.conversation_info.done_action: - last_action_record_for_end_check = conversation_instance.conversation_info.done_action[-1] + if conversation_instance.conversation_info.done_action: + last_action_record_for_end_check = conversation_instance.conversation_info.done_action[-1] + action_ended: bool = ( last_action_record_for_end_check.get("action") in ["end_conversation", "say_goodbye"] and last_action_record_for_end_check.get("status") == "done" @@ -294,12 +432,12 @@ async def run_conversation_loop(conversation_instance: "Conversation"): if goal_ended or action_ended: logger.info(f"[私聊][{conversation_instance.private_name}] (Loop) 检测到结束条件，停止循环。") - await conversation_instance.stop() # 调用 Conversation 的 stop - continue # 虽然会 break，但 continue 更明确 + await conversation_instance.stop() + continue except asyncio.CancelledError: logger.info(f"[私聊][{conversation_instance.private_name}] (Loop) PFC 主循环任务被取消。") - await conversation_instance.stop() # 调用 Conversation 的 stop + await conversation_instance.stop() break except Exception as loop_err: logger.error(f"[私聊][{conversation_instance.private_name}] (Loop) PFC 主循环出错: {loop_err}") @@ -307,7 +445,6 @@ async def run_conversation_loop(conversation_instance: "Conversation"): conversation_instance.state = ConversationState.ERROR await asyncio.sleep(5) - # 控制循环频率 loop_duration = time.time() - loop_iter_start_time min_loop_interval = 0.1 logger.debug(f"[私聊][{conversation_instance.private_name}] (Loop) 循环迭代耗时: {loop_duration:.3f} 秒。") diff --git a/src/plugins/PFC/observation_info.py b/src/plugins/PFC/observation_info.py index 7b3d6279..a372b3dc 100644 --- a/src/plugins/PFC/observation_info.py +++ b/src/plugins/PFC/observation_info.py @@ -114,8 +114,6 @@ class ObservationInfo: """初始化 ObservationInfo""" self.private_name: str = private_name - # 新增：发信人信息 - self.sender_name: Optional[str] = None self.sender_user_id: Optional[str] = None # 存储为字符串 self.sender_platform: Optional[str] = None @@ -232,23 +230,20 @@ class ObservationInfo: if user_info: try: self.sender_user_id = str(user_info.user_id) # 确保是字符串 - self.sender_name = user_info.user_nickname # 或者 user_info.card 如果私聊时card更准 self.sender_platform = user_info.platform current_message_sender_id = self.sender_user_id # 用于后续逻辑 logger.debug( - f"[私聊][{self.private_name}] 更新发信人信息: ID={self.sender_user_id}, Name={self.sender_name}, Platform={self.sender_platform}" + f"[私聊][{self.private_name}] 更新发信人信息: ID={self.sender_user_id}, Name={self.private_name}, Platform={self.sender_platform}" ) except AttributeError as e: logger.error(f"[私聊][{self.private_name}] 从 UserInfo 对象提取信息时出错: {e}, UserInfo: {user_info}") # 如果提取失败，将这些新字段设为 None，避免使用旧数据 self.sender_user_id = None - self.sender_name = None self.sender_platform = None else: logger.warning(f"[私聊][{self.private_name}] 处理消息更新时缺少有效的 UserInfo, message_id: {message_id}") # 如果没有 UserInfo，也将这些新字段设为 None self.sender_user_id = None - self.sender_name = None self.sender_platform = None # --- 新增/修改结束 --- diff --git a/src/plugins/PFC/pfc.py b/src/plugins/PFC/pfc.py index 77a8f4b7..741f20ea 100644 --- a/src/plugins/PFC/pfc.py +++ b/src/plugins/PFC/pfc.py @@ -3,7 +3,7 @@ from src.common.logger import get_module_logger from ..models.utils_model import LLMRequest from ...config.config import global_config from .chat_observer import ChatObserver -from .pfc_utils import get_items_from_json +from .pfc_utils import get_items_from_json, build_chat_history_text from src.individuality.individuality import Individuality from .conversation_info import ConversationInfo from .observation_info import ObservationInfo @@ -86,21 +86,7 @@ class GoalAnalyzer: goals_str = f"目标：{goal}，产生该对话目标的原因：{reasoning}\n" # 获取聊天历史记录 - chat_history_text = observation_info.chat_history_str - - if observation_info.new_messages_count > 0: - new_messages_list = observation_info.unprocessed_messages - new_messages_str = await build_readable_messages( - new_messages_list, - replace_bot_name=True, - merge_messages=False, - timestamp_mode="relative", - read_mark=0.0, - ) - chat_history_text += f"\n--- 以下是 {observation_info.new_messages_count} 条新消息 ---\n{new_messages_str}" - else: - chat_history_text += "\n--- 以上均为已读消息，未读消息均已处理完毕 ---\n" - # await observation_info.clear_unprocessed_messages() + chat_history_text = await build_chat_history_text(observation_info, self.private_name) persona_text = f"你的名字是{self.name}，{self.personality_info}。" # 构建action历史文本 diff --git a/src/plugins/PFC/pfc_KnowledgeFetcher.py b/src/plugins/PFC/pfc_KnowledgeFetcher.py deleted file mode 100644 index 0989339d..00000000 --- a/src/plugins/PFC/pfc_KnowledgeFetcher.py +++ /dev/null @@ -1,85 +0,0 @@ -from typing import List, Tuple -from src.common.logger import get_module_logger -from src.plugins.memory_system.Hippocampus import HippocampusManager -from ..models.utils_model import LLMRequest -from ...config.config import global_config -from ..chat.message import Message -from ..knowledge.knowledge_lib import qa_manager -from ..utils.chat_message_builder import build_readable_messages - -logger = get_module_logger("knowledge_fetcher") - - -class KnowledgeFetcher: - """知识调取器""" - - def __init__(self, private_name: str): - self.llm = LLMRequest( - model=global_config.llm_normal, - temperature=global_config.llm_normal["temp"], - max_tokens=1000, - request_type="knowledge_fetch", - ) - self.private_name = private_name - - def _lpmm_get_knowledge(self, query: str) -> str: - """获取相关知识 - - Args: - query: 查询内容 - - Returns: - str: 构造好的,带相关度的知识 - """ - - logger.debug(f"[私聊][{self.private_name}]正在从LPMM知识库中获取知识") - try: - knowledge_info = qa_manager.get_knowledge(query) - logger.debug(f"[私聊][{self.private_name}]LPMM知识库查询结果: {knowledge_info:150}") - return knowledge_info - except Exception as e: - logger.error(f"[私聊][{self.private_name}]LPMM知识库搜索工具执行失败: {str(e)}") - return "未找到匹配的知识" - - async def fetch(self, query: str, chat_history: List[Message]) -> Tuple[str, str]: - """获取相关知识 - - Args: - query: 查询内容 - chat_history: 聊天历史 - - Returns: - Tuple[str, str]: (获取的知识, 知识来源) - """ - # 构建查询上下文 - chat_history_text = await build_readable_messages( - chat_history, - replace_bot_name=True, - merge_messages=False, - timestamp_mode="relative", - read_mark=0.0, - ) - - # 从记忆中获取相关知识 - related_memory = await HippocampusManager.get_instance().get_memory_from_text( - text=f"{query}\n{chat_history_text}", - max_memory_num=3, - max_memory_length=2, - max_depth=3, - fast_retrieval=False, - ) - knowledge_text = "" - sources_text = "无记忆匹配" # 默认值 - if related_memory: - sources = [] - for memory in related_memory: - knowledge_text += memory[1] + "\n" - sources.append(f"记忆片段{memory[0]}") - knowledge_text = knowledge_text.strip() - sources_text = "，".join(sources) - - knowledge_text += "\n现在有以下**知识**可供参考：\n " - knowledge_text += self._lpmm_get_knowledge(query) - knowledge_text += "\n请记住这些**知识**，并根据**知识**回答问题。\n" - - return knowledge_text or "未找到相关知识", sources_text or "无记忆匹配" diff --git a/src/plugins/PFC/pfc_emotion.py b/src/plugins/PFC/pfc_emotion.py index d2496520..c202d04a 100644 --- a/src/plugins/PFC/pfc_emotion.py +++ b/src/plugins/PFC/pfc_emotion.py @@ -71,9 +71,7 @@ class PfcEmotionUpdater: ) current_mood_text_from_manager = self.mood_mng.current_mood.text # 从 MoodManager 获取当前情绪文本 - sender_name_for_prompt = getattr(observation_info, "sender_name", "对方") - if not sender_name_for_prompt: - sender_name_for_prompt = "对方" + sender_name_for_prompt = self.private_name relationship_text_for_prompt = getattr( conversation_info, "relationship_text", "关系一般。" ) # 从 ConversationInfo 获取关系文本 diff --git a/src/plugins/PFC/pfc_utils.py b/src/plugins/PFC/pfc_utils.py index 38ead120..fc5437ab 100644 --- a/src/plugins/PFC/pfc_utils.py +++ b/src/plugins/PFC/pfc_utils.py @@ -8,6 +8,9 @@ from src.plugins.heartFC_chat.heartflow_prompt_builder import prompt_builder # from src.plugins.chat.chat_stream import ChatStream from ..person_info.person_info import person_info_manager import math +from src.plugins.utils.chat_message_builder import build_readable_messages +from .observation_info import ObservationInfo +from src.config.config import global_config logger = get_logger("pfc_utils") @@ -339,3 +342,43 @@ async def adjust_relationship_value_nonlinear(old_value: float, raw_adjustment: value = 0 return value + + +async def build_chat_history_text(observation_info: ObservationInfo, private_name: str) -> str: + """构建聊天历史记录文本 (包含未处理消息)""" + + chat_history_text = "" + try: + if hasattr(observation_info, "chat_history_str") and observation_info.chat_history_str: + chat_history_text = observation_info.chat_history_str + elif hasattr(observation_info, "chat_history") and observation_info.chat_history: + history_slice = observation_info.chat_history[-20:] + chat_history_text = await build_readable_messages( + history_slice, replace_bot_name=True, merge_messages=False, timestamp_mode="relative", read_mark=0.0 + ) + else: + chat_history_text = "还没有聊天记录。\n" + unread_count = getattr(observation_info, "new_messages_count", 0) + unread_messages = getattr(observation_info, "unprocessed_messages", []) + if unread_count > 0 and unread_messages: + bot_qq_str = str(global_config.BOT_QQ) + other_unread_messages = [ + msg for msg in unread_messages if msg.get("user_info", {}).get("user_id") != bot_qq_str + ] + other_unread_count = len(other_unread_messages) + if other_unread_count > 0: + new_messages_str = await build_readable_messages( + other_unread_messages, + replace_bot_name=True, + merge_messages=False, + timestamp_mode="relative", + read_mark=0.0, + ) + chat_history_text += f"\n{new_messages_str}\n------\n" + except AttributeError as e: + logger.warning(f"[私聊][{private_name}] 构建聊天记录文本时属性错误: {e}") + chat_history_text = "[获取聊天记录时出错]\n" + except Exception as e: + logger.error(f"[私聊][{private_name}] 处理聊天记录时发生未知错误: {e}") + chat_history_text = "[处理聊天记录时出错]\n" + return chat_history_text diff --git a/src/plugins/PFC/reply_generator.py b/src/plugins/PFC/reply_generator.py index 5f3b3591..174e3ba0 100644 --- a/src/plugins/PFC/reply_generator.py +++ b/src/plugins/PFC/reply_generator.py @@ -1,4 +1,5 @@ import random + from .pfc_utils import retrieve_contextual_info from src.common.logger_manager import get_logger @@ -9,7 +10,7 @@ from .reply_checker import ReplyChecker from src.individuality.individuality import Individuality from .observation_info import ObservationInfo from .conversation_info import ConversationInfo -from src.plugins.utils.chat_message_builder import build_readable_messages +from .pfc_utils import build_chat_history_text logger = get_logger("reply_generator") @@ -103,8 +104,6 @@ PROMPT_SEND_NEW_MESSAGE = """ {last_rejection_info} -{spam_warning_info} - 请根据上述信息，判断你是否要继续发一条新消息（例如对之前消息的补充，深入话题，或追问等等）。如果你觉得要发送，该消息应该： 1. 符合对话目标，以"你"的角度发言（不要自己与自己对话！） 2. 符合你的性格特征和身份细节 @@ -216,25 +215,9 @@ class ReplyGenerator: else: goals_str = "- 目前没有明确对话目标\n" - chat_history_text = observation_info.chat_history_str - if observation_info.new_messages_count > 0 and observation_info.unprocessed_messages: - new_messages_list = observation_info.unprocessed_messages - new_messages_str = await build_readable_messages( - new_messages_list, - replace_bot_name=True, - merge_messages=False, - timestamp_mode="relative", - read_mark=0.0, - ) - chat_history_text += f"\n--- 以下是 {observation_info.new_messages_count} 条新消息 ---\n{new_messages_str}" - elif not chat_history_text: - chat_history_text = "还没有聊天记录。" - else: - chat_history_text += "\n--- 以上均为已读消息，未读消息均已处理完毕 ---\n" + chat_history_text = await build_chat_history_text(observation_info, self.private_name) - sender_name_str = getattr(observation_info, "sender_name", "对方") - if not sender_name_str: - sender_name_str = "对方" + sender_name_str = self.private_name relationship_text_str = getattr(conversation_info, "relationship_text", "你们还不熟悉。") current_emotion_text_str = getattr(conversation_info, "current_emotion_text", "心情平静。") @@ -280,14 +263,14 @@ class ReplyGenerator: ) # 新增：构建刷屏警告信息 for PROMPT_SEND_NEW_MESSAGE - spam_warning_message = "" - if action_type == "send_new_message": # 只在 send_new_message 时构建刷屏警告 - if conversation_info.my_message_count > 5: - spam_warning_message = f"⚠️【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息！请谨慎考虑是否继续发送！以免刷屏对造成对方困扰！**" - elif conversation_info.my_message_count > 2: - spam_warning_message = f"💬【提示】**你已连续发送{str(conversation_info.my_message_count)}条消息。如果非必要，请避免连续发送，以免给对方造成困扰。**" - if spam_warning_message: - spam_warning_message = f"\n{spam_warning_message}\n" + # spam_warning_message = "" + # if action_type == "send_new_message": # 只在 send_new_message 时构建刷屏警告 + # if conversation_info.my_message_count > 5: + # spam_warning_message = f"⚠️【警告】**你已连续发送{str(conversation_info.my_message_count)}条消息！请谨慎考虑是否继续发送！以免刷屏对造成对方困扰！**" + # elif conversation_info.my_message_count > 2: + # spam_warning_message = f"💬【提示】**你已连续发送{str(conversation_info.my_message_count)}条消息。如果非必要，请避免连续发送，以免给对方造成困扰。**" + # if spam_warning_message: + # spam_warning_message = f"\n{spam_warning_message}\n" # --- 选择 Prompt --- if action_type == "send_new_message": @@ -326,7 +309,7 @@ class ReplyGenerator: if action_type == "send_new_message": current_format_params = base_format_params.copy() - current_format_params["spam_warning_info"] = spam_warning_message + # current_format_params["spam_warning_info"] = spam_warning_message prompt = prompt_template.format(**current_format_params) elif action_type == "say_goodbye": farewell_params = { diff --git a/template/bot_config_template.toml b/template/bot_config_template.toml index f0b220a8..29c23d86 100644 --- a/template/bot_config_template.toml +++ b/template/bot_config_template.toml @@ -202,12 +202,13 @@ talk_allowed_private = [] # 可以回复消息的QQ号 pfc_chatting = false # 是否启用PFC聊天，该功能仅作用于私聊，与回复模式独立 api_polling_max_retries = 3 enable_pfc_reply_checker = true # 是否启用 PFC 的回复检查器 +pfc_message_buffer_size = 2 # PFC 聊天消息缓冲数量，有利于使聊天节奏更加紧凑流畅，请根据实际 LLM 响应速度进行调整，默认2条 -[idle_conversation] -enable_idle_conversation = false # 是否启用 pfc 主动发言 +[idle_chat] +enable_idle_chat = false # 是否启用 pfc 主动发言 idle_check_interval = 10 # 检查间隔，10分钟检查一次 -min_idle_time = 7200 # 最短无活动时间，2小时 (7200秒) -max_idle_time = 18000 # 最长无活动时间，5小时 (18000秒) +min_cooldown = 7200 # 最短冷却时间，2小时 (7200秒) +max_cooldown = 18000 # 最长冷却时间，5小时 (18000秒) #下面的模型若使用硅基流动则不需要更改，使用ds官方则改成.env自定义的宏，使用自定义模型则选择定位相似的模型自己填写 #推理模型