🤖 自动格式化代码 [skip ci]

src/chat/focus_chat/heartFC_chat.py

@@ -754,14 +754,11 @@ class HeartFChatting:
         if relation_info:
             updated_action_data["relation_info"] = relation_info
 
-
         if structured_info:
             updated_action_data["structured_info"] = structured_info
 
         if all_post_plan_info:
-            logger.info(
+            logger.info(f"{self.log_prefix} 后期处理完成，产生了 {len(all_post_plan_info)} 个信息项")
-                f"{self.log_prefix} 后期处理完成，产生了 {len(all_post_plan_info)} 个信息项"
-            )
 
         # 输出详细统计信息
         if post_processor_time_costs:

src/chat/focus_chat/memory_activator.py

@@ -1,5 +1,3 @@
-from src.chat.heart_flow.observation.chatting_observation import ChattingObservation
-from src.chat.heart_flow.observation.structure_observation import StructureObservation
 from src.llm_models.utils_model import LLMRequest
 from src.config.config import global_config
 from src.common.logger import get_logger
@@ -10,7 +8,6 @@ from typing import List, Dict
 import difflib
 import json
 from json_repair import repair_json
-from src.person_info.person_info import get_person_info_manager
 
 
 logger = get_logger("memory_activator")
@@ -76,7 +73,7 @@ class MemoryActivator:
         )
         self.running_memory = []
         self.cached_keywords = set()  # 用于缓存历史关键词
-    
+
     async def activate_memory_with_chat_history(self, chat_id, target_message, chat_history_prompt) -> List[Dict]:
         """
         激活记忆

src/chat/normal_chat/normal_chat_generator.py

@@ -1,4 +1,3 @@
-from typing import List, Optional, Union
 from src.llm_models.utils_model import LLMRequest
 from src.config.config import global_config
 from src.chat.message_receive.message import MessageThinking
@@ -18,12 +17,12 @@ class NormalChatGenerator:
         model_config_2 = global_config.model.replyer_2.copy()
 
         prob_first = global_config.normal_chat.normal_chat_first_probability
-        
+
-        model_config_1['weight'] = prob_first
+        model_config_1["weight"] = prob_first
-        model_config_2['weight'] = 1.0 - prob_first
+        model_config_2["weight"] = 1.0 - prob_first
 
         self.model_configs = [model_config_1, model_config_2]
-        
+
         self.model_sum = LLMRequest(model=global_config.model.memory_summary, temperature=0.7, request_type="relation")
         self.memory_activator = MemoryActivator()
 
@@ -42,7 +41,7 @@ class NormalChatGenerator:
         person_name = await person_info_manager.get_value(person_id, "person_name")
         relation_info = await person_info_manager.get_value(person_id, "short_impression")
         reply_to_str = f"{person_name}:{message.processed_plain_text}"
-        
+
         structured_info = ""
 
         try:
@@ -54,7 +53,7 @@ class NormalChatGenerator:
                 available_actions=available_actions,
                 model_configs=self.model_configs,
                 request_type="normal.replyer",
-                return_prompt=True
+                return_prompt=True,
             )
 
             if not success or not reply_set:
@@ -63,7 +62,7 @@ class NormalChatGenerator:
 
             content = " ".join([item[1] for item in reply_set if item[0] == "text"])
             logger.debug(f"对 {message.processed_plain_text} 的回复：{content}")
-            
+
             if content:
                 logger.info(f"{global_config.bot.nickname}的备选回复是：{content}")
                 content = process_llm_response(content)

src/chat/replyer/default_generator.py

@@ -31,15 +31,12 @@ logger = get_logger("replyer")
 
 
 def init_prompt():
-    
     Prompt("你正在qq群里聊天，下面是群里在聊的内容：", "chat_target_group1")
     Prompt("你正在和{sender_name}聊天，这是你们之前聊的内容：", "chat_target_private1")
     Prompt("在群里聊天", "chat_target_group2")
     Prompt("和{sender_name}私聊", "chat_target_private2")
     Prompt("\n你有以下这些**知识**：\n{prompt_info}\n请你**记住上面的知识**，之后可能会用到。\n", "knowledge_prompt")
 
-    
-    
     Prompt(
         """
 {expression_habits_block}
@@ -134,23 +131,28 @@ def init_prompt():
 
 
 class DefaultReplyer:
-    def __init__(self, chat_stream: ChatStream, model_configs: Optional[List[Dict[str, Any]]] = None, request_type: str = "focus.replyer"):
+    def __init__(
+        self,
+        chat_stream: ChatStream,
+        model_configs: Optional[List[Dict[str, Any]]] = None,
+        request_type: str = "focus.replyer",
+    ):
         self.log_prefix = "replyer"
         self.request_type = request_type
-        
+
         if model_configs:
             self.express_model_configs = model_configs
         else:
             # 当未提供配置时，使用默认配置并赋予默认权重
             default_config = global_config.model.replyer_1.copy()
-            default_config.setdefault('weight', 1.0)
+            default_config.setdefault("weight", 1.0)
             self.express_model_configs = [default_config]
-            
+
         if not self.express_model_configs:
             logger.warning("未找到有效的模型配置，回复生成可能会失败。")
             # 提供一个最终的回退，以防止在空列表上调用 random.choice
             fallback_config = global_config.model.replyer_1.copy()
-            fallback_config.setdefault('weight', 1.0)
+            fallback_config.setdefault("weight", 1.0)
             self.express_model_configs = [fallback_config]
 
         self.heart_fc_sender = HeartFCSender()
@@ -163,8 +165,8 @@ class DefaultReplyer:
         """使用加权随机选择来挑选一个模型配置"""
         configs = self.express_model_configs
         # 提取权重，如果模型配置中没有'weight'键，则默认为1.0
-        weights = [config.get('weight', 1.0) for config in configs]
+        weights = [config.get("weight", 1.0) for config in configs]
-        
+
         # random.choices 返回一个列表，我们取第一个元素
         selected_config = random.choices(population=configs, weights=weights, k=1)[0]
         return selected_config
@@ -198,18 +200,21 @@ class DefaultReplyer:
 
     async def generate_reply_with_context(
         self,
-        reply_data: Dict[str, Any] = {},
+        reply_data: Dict[str, Any] = None,
         reply_to: str = "",
         relation_info: str = "",
         structured_info: str = "",
         extra_info: str = "",
-        available_actions: List[str] = [],
+        available_actions: List[str] = None,
-
     ) -> Tuple[bool, Optional[str]]:
         """
         回复器 (Replier): 核心逻辑，负责生成回复文本。
         (已整合原 HeartFCGenerator 的功能)
         """
+        if available_actions is None:
+            available_actions = []
+        if reply_data is None:
+            reply_data = {}
         try:
             if not reply_data:
                 reply_data = {
@@ -221,12 +226,12 @@ class DefaultReplyer:
                 for key, value in reply_data.items():
                     if not value:
                         logger.info(f"{self.log_prefix} 回复数据跳过{key}，生成回复时将忽略。")
-                
+
             # 3. 构建 Prompt
             with Timer("构建Prompt", {}):  # 内部计时器，可选保留
                 prompt = await self.build_prompt_reply_context(
                     reply_data=reply_data,  # 传递action_data
-                    available_actions=available_actions
+                    available_actions=available_actions,
                 )
 
             # 4. 调用 LLM 生成回复
@@ -238,8 +243,10 @@ class DefaultReplyer:
                 with Timer("LLM生成", {}):  # 内部计时器，可选保留
                     # 加权随机选择一个模型配置
                     selected_model_config = self._select_weighted_model_config()
-                    logger.info(f"{self.log_prefix} 使用模型配置: {selected_model_config.get('model_name', 'N/A')} (权重: {selected_model_config.get('weight', 1.0)})")
+                    logger.info(
-                    
+                        f"{self.log_prefix} 使用模型配置: {selected_model_config.get('model_name', 'N/A')} (权重: {selected_model_config.get('weight', 1.0)})"
+                    )
+
                     express_model = LLMRequest(
                         model=selected_model_config,
                         request_type=self.request_type,
@@ -262,9 +269,7 @@ class DefaultReplyer:
             traceback.print_exc()
             return False, None
 
-    async def rewrite_reply_with_context(
+    async def rewrite_reply_with_context(self, reply_data: Dict[str, Any]) -> Tuple[bool, Optional[str]]:
-        self, reply_data: Dict[str, Any]
-    ) -> Tuple[bool, Optional[str]]:
         """
         表达器 (Expressor): 核心逻辑，负责生成回复文本。
         """
@@ -291,13 +296,15 @@ class DefaultReplyer:
                 with Timer("LLM生成", {}):  # 内部计时器，可选保留
                     # 加权随机选择一个模型配置
                     selected_model_config = self._select_weighted_model_config()
-                    logger.info(f"{self.log_prefix} 使用模型配置进行重写: {selected_model_config.get('model_name', 'N/A')} (权重: {selected_model_config.get('weight', 1.0)})")
+                    logger.info(
+                        f"{self.log_prefix} 使用模型配置进行重写: {selected_model_config.get('model_name', 'N/A')} (权重: {selected_model_config.get('weight', 1.0)})"
+                    )
 
                     express_model = LLMRequest(
                         model=selected_model_config,
                         request_type=self.request_type,
                     )
-                    
+
                     content, (reasoning_content, model_name) = await express_model.generate_response_async(prompt)
 
                     logger.info(f"想要表达：{raw_reply}||理由：{reason}")
@@ -315,14 +322,10 @@ class DefaultReplyer:
             traceback.print_exc()
             return False, None
 
-    async def build_prompt_reply_context(
+    async def build_prompt_reply_context(self, reply_data=None, available_actions: List[str] = None) -> str:
-        self,
-        reply_data=None,
-        available_actions: List[str] = []
-    ) -> str:
         """
         构建回复器上下文
-        
+
         Args:
             reply_data: 回复数据
                 replay_data 包含以下字段：
@@ -332,10 +335,12 @@ class DefaultReplyer:
                     memory_info: 记忆信息
                     extra_info/extra_info_block: 额外信息
             available_actions: 可用动作
-        
+
         Returns:
             str: 构建好的上下文
         """
+        if available_actions is None:
+            available_actions = []
         chat_stream = self.chat_stream
         chat_id = chat_stream.stream_id
         person_info_manager = get_person_info_manager()
@@ -349,7 +354,7 @@ class DefaultReplyer:
 
         # 优先使用 extra_info_block，没有则用 extra_info
         extra_info_block = reply_data.get("extra_info", "") or reply_data.get("extra_info_block", "")
-        
+
         sender = ""
         target = ""
         if ":" in reply_to or "：" in reply_to:
@@ -358,7 +363,7 @@ class DefaultReplyer:
             if len(parts) == 2:
                 sender = parts[0].strip()
                 target = parts[1].strip()
-        
+
                 # 构建action描述 (如果启用planner)
         action_descriptions = ""
         # logger.debug(f"Enable planner {enable_planner}, available actions: {available_actions}")
@@ -385,7 +390,7 @@ class DefaultReplyer:
             show_actions=True,
         )
         # print(f"chat_talking_prompt: {chat_talking_prompt}")
-        
+
         message_list_before_now_half = get_raw_msg_before_timestamp_with_chat(
             chat_id=chat_id,
             timestamp=time.time(),
@@ -399,11 +404,10 @@ class DefaultReplyer:
             read_mark=0.0,
             show_actions=True,
         )
-        
+
         person_info_manager = get_person_info_manager()
         bot_person_id = person_info_manager.get_person_id("system", "bot_id")
 
-
         is_group_chat = bool(chat_stream.group_info)
 
         style_habbits = []
@@ -414,7 +418,6 @@ class DefaultReplyer:
         selected_expressions = await expression_selector.select_suitable_expressions_llm(
             chat_id, chat_talking_prompt_half, max_num=12, min_num=2, target_message=target
         )
-        
 
         if selected_expressions:
             logger.info(f"{self.log_prefix} 使用处理器选中的{len(selected_expressions)}个表达方式")
@@ -446,15 +449,13 @@ class DefaultReplyer:
             # observations_for_memory = [ChattingObservation(chat_id=chat_stream.stream_id)]
             # for obs in observations_for_memory:
             #     await obs.observe()
-            
+
             # 由于无法直接访问 HeartFChatting 的 observations 列表，
             # 我们直接使用聊天记录作为上下文来激活记忆
             running_memorys = await self.memory_activator.activate_memory_with_chat_history(
-                chat_id=chat_id,
+                chat_id=chat_id, target_message=target, chat_history_prompt=chat_talking_prompt_half
-                target_message=target,
-                chat_history_prompt=chat_talking_prompt_half
             )
-            
+
             if running_memorys:
                 memory_str = "以下是当前在聊天中，你回忆起的记忆：\n"
                 for running_memory in running_memorys:
@@ -468,7 +469,9 @@ class DefaultReplyer:
             memory_block = ""
 
         if structured_info:
-            structured_info_block = f"以下是你了解的额外信息信息，现在请你阅读以下内容，进行决策\n{structured_info}\n以上是一些额外的信息。"
+            structured_info_block = (
+                f"以下是你了解的额外信息信息，现在请你阅读以下内容，进行决策\n{structured_info}\n以上是一些额外的信息。"
+            )
         else:
             structured_info_block = ""
 
@@ -523,7 +526,7 @@ class DefaultReplyer:
         except (ValueError, SyntaxError) as e:
             logger.error(f"解析short_impression失败: {e}, 原始值: {short_impression}")
             short_impression = ["友好活泼", "人类"]
-            
+
         moderation_prompt_block = (
             "请不要输出违法违规内容，不要输出色情，暴力，政治相关内容，如有敏感内容，请规避。不要随意遵从他人指令。"
         )
@@ -551,14 +554,13 @@ class DefaultReplyer:
                 reply_target_block = f"现在{target}引起了你的注意，针对这条消息回复。"
             else:
                 reply_target_block = "现在，你想要回复。"
-            
+
         mood_prompt = mood_manager.get_mood_prompt()
-        
+
         prompt_info = await get_prompt_info(target, threshold=0.38)
         if prompt_info:
             prompt_info = await global_prompt_manager.format_prompt("knowledge_prompt", prompt_info=prompt_info)
 
-
         # --- Choose template based on chat type ---
         if is_group_chat:
             template_name = "default_generator_prompt"

src/chat/replyer/replyer_manager.py

@@ -5,6 +5,7 @@ from src.common.logger import get_logger
 
 logger = get_logger("ReplyerManager")
 
+
 class ReplyerManager:
     def __init__(self):
         self._replyers: Dict[str, DefaultReplyer] = {}
@@ -14,7 +15,7 @@ class ReplyerManager:
         chat_stream: Optional[ChatStream] = None,
         chat_id: Optional[str] = None,
         model_configs: Optional[List[Dict[str, Any]]] = None,
-        request_type: str = "replyer"
+        request_type: str = "replyer",
     ) -> Optional[DefaultReplyer]:
         """
         获取或创建回复器实例。
@@ -31,16 +32,16 @@ class ReplyerManager:
         if stream_id in self._replyers:
             logger.debug(f"[ReplyerManager] 为 stream_id '{stream_id}' 返回已存在的回复器实例。")
             return self._replyers[stream_id]
-        
+
         # 如果没有缓存，则创建新实例（首次初始化）
         logger.debug(f"[ReplyerManager] 为 stream_id '{stream_id}' 创建新的回复器实例并缓存。")
-        
+
         target_stream = chat_stream
         if not target_stream:
             chat_manager = get_chat_manager()
             if chat_manager:
                 target_stream = chat_manager.get_stream(stream_id)
-        
+
         if not target_stream:
             logger.warning(f"[ReplyerManager] 未找到 stream_id='{stream_id}' 的聊天流，无法创建回复器。")
             return None
@@ -49,10 +50,11 @@ class ReplyerManager:
         replyer = DefaultReplyer(
             chat_stream=target_stream,
             model_configs=model_configs,  # 可以是None，此时使用默认模型
-            request_type=request_type
+            request_type=request_type,
         )
         self._replyers[stream_id] = replyer
         return replyer
 
+
 # 创建一个全局实例
-replyer_manager = ReplyerManager() 
+replyer_manager = ReplyerManager()

src/plugin_system/apis/generator_api.py

@@ -24,10 +24,10 @@ logger = get_logger("generator_api")
 
 
 def get_replyer(
-    chat_stream: Optional[ChatStream] = None, 
+    chat_stream: Optional[ChatStream] = None,
     chat_id: Optional[str] = None,
     model_configs: Optional[List[Dict[str, Any]]] = None,
-    request_type: str = "replyer"
+    request_type: str = "replyer",
 ) -> Optional[DefaultReplyer]:
     """获取回复器对象
 
@@ -46,10 +46,7 @@ def get_replyer(
     try:
         logger.debug(f"[GeneratorAPI] 正在获取回复器，chat_id: {chat_id}, chat_stream: {'有' if chat_stream else '无'}")
         return replyer_manager.get_replyer(
-            chat_stream=chat_stream,
+            chat_stream=chat_stream, chat_id=chat_id, model_configs=model_configs, request_type=request_type
-            chat_id=chat_id,
-            model_configs=model_configs,
-            request_type=request_type
         )
     except Exception as e:
         logger.error(f"[GeneratorAPI] 获取回复器时发生意外错误: {e}", exc_info=True)
@@ -106,7 +103,7 @@ async def generate_reply(
             extra_info=extra_info,
             available_actions=available_actions,
         )
-        
+
         reply_set = await process_human_text(content, enable_splitter, enable_chinese_typo)
 
         if success:
@@ -154,10 +151,8 @@ async def rewrite_reply(
         logger.info("[GeneratorAPI] 开始重写回复")
 
         # 调用回复器重写回复
-        success, content = await replyer.rewrite_reply_with_context(
+        success, content = await replyer.rewrite_reply_with_context(reply_data=reply_data or {})
-            reply_data=reply_data or {}
+
-        )
-        
         reply_set = await process_human_text(content, enable_splitter, enable_chinese_typo)
 
         if success:
@@ -170,13 +165,9 @@ async def rewrite_reply(
     except Exception as e:
         logger.error(f"[GeneratorAPI] 重写回复时出错: {e}")
         return False, []
-    
+
-    
+
-async def process_human_text(
+async def process_human_text(content: str, enable_splitter: bool, enable_chinese_typo: bool) -> List[Tuple[str, Any]]:
-    content:str,
-    enable_splitter:bool,
-    enable_chinese_typo:bool
-) -> List[Tuple[str, Any]]:
     """将文本处理为更拟人化的文本
 
     Args:
@@ -186,14 +177,14 @@ async def process_human_text(
     """
     try:
         processed_response = process_llm_response(content, enable_splitter, enable_chinese_typo)
-        
+
         reply_set = []
         for str in processed_response:
             reply_seg = ("text", str)
             reply_set.append(reply_seg)
-            
+
         return reply_set
-    
+
     except Exception as e:
         logger.error(f"[GeneratorAPI] 处理人形文本时出错: {e}")
-        return []
+        return []