From 22b579be0271bccd8c6a49ba8b73dab279722828 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=A2=A8=E6=A2=93=E6=9F=92?= <1787882683@qq.com> Date: Thu, 29 May 2025 09:27:06 +0800 Subject: [PATCH 1/3] fix: Ruff --- src/chat/utils/chat_message_builder.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/chat/utils/chat_message_builder.py b/src/chat/utils/chat_message_builder.py index f62c0ad8b..4b9da43ea 100644 --- a/src/chat/utils/chat_message_builder.py +++ b/src/chat/utils/chat_message_builder.py @@ -457,7 +457,7 @@ async def build_anonymous_messages(messages: List[Dict[str, Any]]) -> str: return "SELF" try: person_id = person_info_manager.get_person_id(platform, user_id) - except Exception as e: + except Exception as _e: person_id = None if not person_id: return "?" @@ -472,7 +472,7 @@ async def build_anonymous_messages(messages: List[Dict[str, Any]]) -> str: # user_info = msg.get("user_info", {}) platform = msg.get("chat_info_platform") user_id = msg.get("user_id") - timestamp = msg.get("time") + _timestamp = msg.get("time") # print(f"msg:{msg}") # print(f"platform:{platform}") # print(f"user_id:{user_id}") From 117f426db294992e75dd507af81afd01a82040d0 Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" Date: Thu, 29 May 2025 01:27:37 +0000 Subject: [PATCH 2/3] =?UTF-8?q?=F0=9F=A4=96=20=E8=87=AA=E5=8A=A8=E6=A0=BC?= =?UTF-8?q?=E5=BC=8F=E5=8C=96=E4=BB=A3=E7=A0=81=20[skip=20ci]?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- scripts/mongodb_to_sqlite.py | 2 +- src/chat/focus_chat/expressors/exprssion_learner.py | 2 +- src/chat/focus_chat/heartFC_chat.py | 8 ++++++-- src/chat/utils/chat_message_builder.py | 8 ++++---- src/main.py | 2 +- src/plugins/test_plugin_pic/actions/pic_action.py | 4 +++- 6 files changed, 16 insertions(+), 10 deletions(-) diff --git a/scripts/mongodb_to_sqlite.py b/scripts/mongodb_to_sqlite.py index e3522c6aa..c6d2950fd 100644 --- a/scripts/mongodb_to_sqlite.py +++ b/scripts/mongodb_to_sqlite.py @@ -199,7 +199,7 @@ class MongoToSQLiteMigrator: "timestamp": "timestamp", }, enable_validation=True, # 禁用数据验证" - unique_fields=["user_id", "prompt_tokens","completion_tokens","total_tokens","cost"], # 组合唯一性 + unique_fields=["user_id", "prompt_tokens", "completion_tokens", "total_tokens", "cost"], # 组合唯一性 ), # 消息迁移配置 MigrationConfig( diff --git a/src/chat/focus_chat/expressors/exprssion_learner.py b/src/chat/focus_chat/expressors/exprssion_learner.py index 7b70ce6f6..96d4e231b 100644 --- a/src/chat/focus_chat/expressors/exprssion_learner.py +++ b/src/chat/focus_chat/expressors/exprssion_learner.py @@ -212,7 +212,7 @@ class ExpressionLearner: # random_msg_str: str = await build_readable_messages(random_msg, timestamp_mode="normal") random_msg_str: str = await build_anonymous_messages(random_msg) # print(f"random_msg_str:{random_msg_str}") - + prompt: str = await global_prompt_manager.format_prompt( prompt, chat_str=random_msg_str, diff --git a/src/chat/focus_chat/heartFC_chat.py b/src/chat/focus_chat/heartFC_chat.py index d5e9622f4..abf89479b 100644 --- a/src/chat/focus_chat/heartFC_chat.py +++ b/src/chat/focus_chat/heartFC_chat.py @@ -335,7 +335,9 @@ class HeartFChatting: for pname, ptime in processor_time_costs.items(): formatted_ptime = f"{ptime * 1000:.2f}毫秒" if ptime < 1 else f"{ptime:.2f}秒" processor_time_strings.append(f"{pname}: {formatted_ptime}") - processor_time_log = ("\n各处理器耗时: " + "; ".join(processor_time_strings)) if processor_time_strings else "" + processor_time_log = ( + ("\n各处理器耗时: " + "; ".join(processor_time_strings)) if processor_time_strings else "" + ) logger.info( f"{self.log_prefix} 第{self._current_cycle.cycle_id}次思考," @@ -471,7 +473,9 @@ class HeartFChatting: running_memorys = await self.memory_activator.activate_memory(observations) with Timer("执行 信息处理器", cycle_timers): - all_plan_info, processor_time_costs = await self._process_processors(observations, running_memorys, cycle_timers) + all_plan_info, processor_time_costs = await self._process_processors( + observations, running_memorys, cycle_timers + ) loop_processor_info = { "all_plan_info": all_plan_info, diff --git a/src/chat/utils/chat_message_builder.py b/src/chat/utils/chat_message_builder.py index 4b9da43ea..46d603b5b 100644 --- a/src/chat/utils/chat_message_builder.py +++ b/src/chat/utils/chat_message_builder.py @@ -451,7 +451,7 @@ async def build_anonymous_messages(messages: List[Dict[str, Any]]) -> str: def get_anon_name(platform, user_id): # print(f"get_anon_name: platform:{platform}, user_id:{user_id}") # print(f"global_config.bot.qq_account:{global_config.bot.qq_account}") - + if user_id == global_config.bot.qq_account: # print("SELF11111111111111") return "SELF" @@ -488,11 +488,11 @@ async def build_anonymous_messages(messages: List[Dict[str, Any]]) -> str: content = content.replace("ⁿ", "") # if not all([platform, user_id, timestamp is not None]): - # continue + # continue anon_name = get_anon_name(platform, user_id) # print(f"anon_name:{anon_name}") - + # 处理 回复 reply_pattern = r"回复<([^:<>]+):([^:<>]+)>" match = re.search(reply_pattern, content) @@ -514,7 +514,7 @@ async def build_anonymous_messages(messages: List[Dict[str, Any]]) -> str: new_content = "" last_end = 0 for m in at_matches: - new_content += content[last_end:m.start()] + new_content += content[last_end : m.start()] bbb = m.group(2) try: anon_at = get_anon_name(platform, bbb) diff --git a/src/main.py b/src/main.py index 4846b9134..5680e5525 100644 --- a/src/main.py +++ b/src/main.py @@ -162,7 +162,7 @@ class MainSystem: """学习并存储表达方式任务""" while True: await asyncio.sleep(global_config.expression.learning_interval) - if global_config.expression.enable_expression_learning: + if global_config.expression.enable_expression_learning: logger.info("[表达方式学习] 开始学习表达方式...") await expression_learner.learn_and_store_expression() logger.info("[表达方式学习] 表达方式学习完成") diff --git a/src/plugins/test_plugin_pic/actions/pic_action.py b/src/plugins/test_plugin_pic/actions/pic_action.py index 025ecb9ba..7b0923a9c 100644 --- a/src/plugins/test_plugin_pic/actions/pic_action.py +++ b/src/plugins/test_plugin_pic/actions/pic_action.py @@ -22,7 +22,9 @@ class PicAction(PluginAction): """根据描述使用火山引擎HTTP API生成图片的动作处理类""" action_name = "pic_action" - action_description = "可以根据特定的描述，生成并发送一张图片，如果没提供描述，就根据聊天内容生成,你可以立刻画好，不用等待" + action_description = ( + "可以根据特定的描述，生成并发送一张图片，如果没提供描述，就根据聊天内容生成,你可以立刻画好，不用等待" + ) action_parameters = { "description": "图片描述，输入你想要生成并发送的图片的描述，必填", "size": "图片尺寸，例如 '1024x1024' (可选, 默认从配置或 '1024x1024')", From 72328b890cc489120d3af0c2b2eec61401f10741 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=A2=A8=E6=A2=93=E6=9F=92?= <1787882683@qq.com> Date: Thu, 29 May 2025 09:41:31 +0800 Subject: [PATCH 3/3] =?UTF-8?q?refactor:=20=E4=BC=98=E5=8C=96=E7=9F=A5?= =?UTF-8?q?=E8=AF=86=E5=BA=93=E6=90=9C=E7=B4=A2=E9=80=BB=E8=BE=91=EF=BC=8C?= =?UTF-8?q?=E7=A7=BB=E9=99=A4=E5=86=97=E4=BD=99=E4=BB=A3=E7=A0=81?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- src/tools/tool_can_use/lpmm_get_knowledge.py | 125 ++----------------- 1 file changed, 11 insertions(+), 114 deletions(-) diff --git a/src/tools/tool_can_use/lpmm_get_knowledge.py b/src/tools/tool_can_use/lpmm_get_knowledge.py index f7c0bd753..f8b9e95dc 100644 --- a/src/tools/tool_can_use/lpmm_get_knowledge.py +++ b/src/tools/tool_can_use/lpmm_get_knowledge.py @@ -38,125 +38,22 @@ class SearchKnowledgeFromLPMMTool(BaseTool): # threshold = function_args.get("threshold", 0.4) # 调用知识库搜索 - embedding = await get_embedding(query, request_type="info_retrieval") - if embedding: - knowledge_info = qa_manager.get_knowledge(query) - logger.debug(f"知识库查询结果: {knowledge_info}") - if knowledge_info: - content = f"你知道这些知识: {knowledge_info}" - else: - content = f"你不太了解有关{query}的知识" - return {"type": "lpmm_knowledge", "id": query, "content": content} - # 如果获取嵌入失败 - return {"type": "info", "id": query, "content": f"无法获取关于'{query}'的嵌入向量，你lpmm知识库炸了"} + + knowledge_info = qa_manager.get_knowledge(query) + + logger.debug(f"知识库查询结果: {knowledge_info}") + + if knowledge_info: + content = f"你知道这些知识: {knowledge_info}" + else: + content = f"你不太了解有关{query}的知识" + return {"type": "lpmm_knowledge", "id": query, "content": content} except Exception as e: + # 捕获异常并记录错误 logger.error(f"知识库搜索工具执行失败: {str(e)}") # 在其他异常情况下，确保 id 仍然是 query (如果它被定义了) query_id = query if "query" in locals() else "unknown_query" return {"type": "info", "id": query_id, "content": f"lpmm知识库搜索失败，炸了: {str(e)}"} - # def get_info_from_db( - # self, query_embedding: list, limit: int = 1, threshold: float = 0.5, return_raw: bool = False - # ) -> Union[str, list]: - # """从数据库中获取相关信息 - - # Args: - # query_embedding: 查询的嵌入向量 - # limit: 最大返回结果数 - # threshold: 相似度阈值 - # return_raw: 是否返回原始结果 - - # Returns: - # Union[str, list]: 格式化的信息字符串或原始结果列表 - # """ - # if not query_embedding: - # return "" if not return_raw else [] - - # # 使用余弦相似度计算 - # pipeline = [ - # { - # "$addFields": { - # "dotProduct": { - # "$reduce": { - # "input": {"$range": [0, {"$size": "$embedding"}]}, - # "initialValue": 0, - # "in": { - # "$add": [ - # "$$value", - # { - # "$multiply": [ - # {"$arrayElemAt": ["$embedding", "$$this"]}, - # {"$arrayElemAt": [query_embedding, "$$this"]}, - # ] - # }, - # ] - # }, - # } - # }, - # "magnitude1": { - # "$sqrt": { - # "$reduce": { - # "input": "$embedding", - # "initialValue": 0, - # "in": {"$add": ["$$value", {"$multiply": ["$$this", "$$this"]}]}, - # } - # } - # }, - # "magnitude2": { - # "$sqrt": { - # "$reduce": { - # "input": query_embedding, - # "initialValue": 0, - # "in": {"$add": ["$$value", {"$multiply": ["$$this", "$$this"]}]}, - # } - # } - # }, - # } - # }, - # {"$addFields": {"similarity": {"$divide": ["$dotProduct", {"$multiply": ["$magnitude1", "$magnitude2"]}]}}}, - # { - # "$match": { - # "similarity": {"$gte": threshold} # 只保留相似度大于等于阈值的结果 - # } - # }, - # {"$sort": {"similarity": -1}}, - # {"$limit": limit}, - # {"$project": {"content": 1, "similarity": 1}}, - # ] - - # results = list(db.knowledges.aggregate(pipeline)) - # logger.debug(f"知识库查询结果数量: {len(results)}") - - # if not results: - # return "" if not return_raw else [] - - # if return_raw: - # return results - # else: - # # 返回所有找到的内容，用换行分隔 - # return "\n".join(str(result["content"]) for result in results) - - def _format_results(self, results: list) -> str: - """格式化结果""" - if not results: - return "未找到相关知识。" - - formatted_string = "我找到了一些相关知识：\n" - for i, result in enumerate(results): - # chunk_id = result.get("chunk_id") - text = result.get("text", "") - source = result.get("source", "未知来源") - source_type = result.get("source_type", "未知类型") - similarity = result.get("similarity", 0.0) - - formatted_string += ( - f"{i + 1}. (相似度: {similarity:.2f}) 类型: {source_type}, 来源: {source} \n内容片段: {text}\n\n" - ) - # 暂时去掉chunk_id - # formatted_string += f"{i + 1}. (相似度: {similarity:.2f}) 类型: {source_type}, 来源: {source}, Chunk ID: {chunk_id} \n内容片段: {text}\n\n" - - return formatted_string -# 注册工具 -# register_tool(SearchKnowledgeTool)