From 216c51d7a27c38fa8b6b4c317a77f88f6fa92f68 Mon Sep 17 00:00:00 2001 From: SengokuCola <1026294844@qq.com> Date: Thu, 18 Dec 2025 17:34:32 +0800 Subject: [PATCH] =?UTF-8?q?feat=EF=BC=9A=E5=8F=AF=E4=BB=A5=E5=85=B3?= =?UTF-8?q?=E9=97=AD=E5=9B=9E=E5=A4=8D=E5=89=8D=E9=BB=91=E8=AF=9D=E8=87=AA?= =?UTF-8?q?=E5=8A=A8=E6=8F=90=E5=8F=96?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- src/bw_learner/expression_learner.py | 12 ++++++++++++ src/chat/replyer/group_generator.py | 18 +++++++++++++----- src/chat/replyer/private_generator.py | 18 +++++++++++++----- src/config/official_configs.py | 3 +++ template/bot_config_template.toml | 3 ++- 5 files changed, 43 insertions(+), 11 deletions(-) diff --git a/src/bw_learner/expression_learner.py b/src/bw_learner/expression_learner.py index b5a4452a..7ff9d789 100644 --- a/src/bw_learner/expression_learner.py +++ b/src/bw_learner/expression_learner.py @@ -188,6 +188,13 @@ class ExpressionLearner: if not context: continue + # 过滤掉包含 SELF 的内容（不学习） + if "SELF" in (situation or "") or "SELF" in (style or "") or "SELF" in context: + logger.info( + f"跳过包含 SELF 的表达方式: situation={situation}, style={style}, source_id={source_id}" + ) + continue + filtered_expressions.append((situation, style, context)) learnt_expressions = filtered_expressions @@ -689,6 +696,11 @@ class ExpressionLearner: if not content: continue + # 过滤掉包含 SELF 的黑话，不学习 + if "SELF" in content: + logger.info(f"跳过包含 SELF 的黑话: {content}") + continue + # 检查是否包含机器人名称 if contains_bot_self_name(content): logger.info(f"跳过包含机器人昵称/别名的黑话: {content}") diff --git a/src/chat/replyer/group_generator.py b/src/chat/replyer/group_generator.py index 9d1c297e..3a82b555 100644 --- a/src/chat/replyer/group_generator.py +++ b/src/chat/replyer/group_generator.py @@ -488,6 +488,10 @@ class DefaultReplyer: duration = end_time - start_time return name, result, duration + async def _build_disabled_jargon_explanation(self) -> str: + """当关闭黑话解释时使用的占位协程，避免额外的LLM调用""" + return "" + def build_chat_history_prompts( self, message_list_before_now: List[DatabaseMessages], target_user_id: str, sender: str ) -> Tuple[str, str]: @@ -819,7 +823,14 @@ class DefaultReplyer: show_actions=True, ) - # 并行执行八个构建任务（包括黑话解释） + # 根据配置决定是否启用黑话解释 + enable_jargon_explanation = getattr(global_config.expression, "enable_jargon_explanation", True) + if enable_jargon_explanation: + jargon_coroutine = explain_jargon_in_context(chat_id, message_list_before_short, chat_talking_prompt_short) + else: + jargon_coroutine = self._build_disabled_jargon_explanation() + + # 并行执行八个构建任务（包括黑话解释，可配置关闭） task_results = await asyncio.gather( self._time_and_run_task( self.build_expression_habits(chat_talking_prompt_short, target, reply_reason, think_level=think_level), @@ -837,10 +848,7 @@ class DefaultReplyer: ), "memory_retrieval", ), - self._time_and_run_task( - explain_jargon_in_context(chat_id, message_list_before_short, chat_talking_prompt_short), - "jargon_explanation", - ), + self._time_and_run_task(jargon_coroutine, "jargon_explanation"), ) # 任务名称中英文映射 diff --git a/src/chat/replyer/private_generator.py b/src/chat/replyer/private_generator.py index a9b78c28..a5cb0e13 100644 --- a/src/chat/replyer/private_generator.py +++ b/src/chat/replyer/private_generator.py @@ -467,6 +467,10 @@ class PrivateReplyer: duration = end_time - start_time return name, result, duration + async def _build_disabled_jargon_explanation(self) -> str: + """当关闭黑话解释时使用的占位协程，避免额外的LLM调用""" + return "" + async def build_actions_prompt( self, available_actions: Dict[str, ActionInfo], chosen_actions_info: Optional[List[ActionPlannerInfo]] = None ) -> str: @@ -702,7 +706,14 @@ class PrivateReplyer: show_actions=True, ) - # 并行执行九个构建任务（包括黑话解释） + # 根据配置决定是否启用黑话解释 + enable_jargon_explanation = getattr(global_config.expression, "enable_jargon_explanation", True) + if enable_jargon_explanation: + jargon_coroutine = explain_jargon_in_context(chat_id, message_list_before_short, chat_talking_prompt_short) + else: + jargon_coroutine = self._build_disabled_jargon_explanation() + + # 并行执行九个构建任务（包括黑话解释，可配置关闭） task_results = await asyncio.gather( self._time_and_run_task( self.build_expression_habits(chat_talking_prompt_short, target, reply_reason), "expression_habits" @@ -720,10 +731,7 @@ class PrivateReplyer: ), "memory_retrieval", ), - self._time_and_run_task( - explain_jargon_in_context(chat_id, message_list_before_short, chat_talking_prompt_short), - "jargon_explanation", - ), + self._time_and_run_task(jargon_coroutine, "jargon_explanation"), ) # 任务名称中英文映射 diff --git a/src/config/official_configs.py b/src/config/official_configs.py index 0a2dc60b..143d5652 100644 --- a/src/config/official_configs.py +++ b/src/config/official_configs.py @@ -325,6 +325,9 @@ class ExpressionConfig(ConfigBase): all_global_jargon: bool = False """是否将所有新增的jargon项目默认为全局（is_global=True），chat_id记录第一次存储时的id。注意，此功能关闭后，已经记录的全局黑话不会改变，需要手动删除""" + enable_jargon_explanation: bool = True + """是否在回复前尝试对上下文中的黑话进行解释（关闭可减少一次LLM调用，仅影响回复前的黑话匹配与解释，不影响黑话学习）""" + def _parse_stream_config_to_chat_id(self, stream_config_str: str) -> Optional[str]: """ 解析流配置字符串并生成对应的 chat_id diff --git a/template/bot_config_template.toml b/template/bot_config_template.toml index a85a241e..2740591a 100644 --- a/template/bot_config_template.toml +++ b/template/bot_config_template.toml @@ -1,5 +1,5 @@ [inner] -version = "7.1.8" +version = "7.2.0" #----以下是给开发人员阅读的，如果你只是部署了麦麦，不需要阅读---- # 如果你想要修改配置文件，请递增version的值 @@ -93,6 +93,7 @@ reflect_operator_id = "" # 表达反思操作员ID，格式：platform:id:type ( allow_reflect = [] # 允许进行表达反思的聊天流ID列表，格式：["qq:123456:private", "qq:654321:group", ...]，只有在此列表中的聊天流才会提出问题并跟踪。如果列表为空，则所有聊天流都可以进行表达反思（前提是 reflect = true） all_global_jargon = true # 是否开启全局黑话模式，注意，此功能关闭后，已经记录的全局黑话不会改变，需要手动删除 +enable_jargon_explanation = true # 是否在回复前尝试对上下文中的黑话进行解释（关闭可减少一次LLM调用，仅影响回复前的黑话匹配与解释，不影响黑话学习） [chat] # 麦麦的聊天设置