From 3db9fafe651031f87f1e9e06e0728160915b970a Mon Sep 17 00:00:00 2001 From: SengokuCola <1026294844@qq.com> Date: Mon, 15 Dec 2025 00:05:15 +0800 Subject: [PATCH] =?UTF-8?q?feat=EF=BC=9A=E9=BB=91=E5=8C=96=E5=92=8C?= =?UTF-8?q?=E8=A1=A8=E8=BE=BE=E4=B8=8D=E4=BC=9A=E5=B0=86=E5=90=8D=E5=AD=97?= =?UTF-8?q?=E6=80=BB=E7=BB=93=E8=BF=9B=E5=8E=BB?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- src/bw_learner/expression_learner.py | 34 ++++- src/bw_learner/jargon_miner.py | 29 ++++- src/bw_learner/message_recorder.py | 151 +++++++++++++++++++++- src/chat/replyer/prompt/replyer_prompt.py | 2 +- 4 files changed, 201 insertions(+), 15 deletions(-) diff --git a/src/bw_learner/expression_learner.py b/src/bw_learner/expression_learner.py index 71866dea..73522e4a 100644 --- a/src/bw_learner/expression_learner.py +++ b/src/bw_learner/expression_learner.py @@ -3,7 +3,7 @@ import json import os import re import asyncio -from typing import List, Optional, Tuple, Any, Dict +from typing import List, Optional, Tuple, Any, Dict, Callable from src.common.logger import get_logger from src.common.database.database_model import Expression from src.llm_models.utils_model import LLMRequest @@ -97,14 +97,14 @@ class ExpressionLearner: async def learn_and_store( self, messages: List[Any], + person_name_filter: Optional[Callable[[str], bool]] = None, ) -> List[Tuple[str, str, str]]: """ 学习并存储表达方式 Args: messages: 外部传入的消息列表（必需） - num: 学习数量 - timestamp_start: 学习开始的时间戳，如果为None则使用self.last_learning_time + person_name_filter: 可选的过滤函数，用于检查内容是否包含人物名称 """ if not messages: return None @@ -135,6 +135,17 @@ class ExpressionLearner: expressions, jargon_entries = self.parse_expression_response(response) expressions = self._filter_self_reference_styles(expressions) + # 过滤掉包含人物名称的表达方式 + if person_name_filter: + filtered_expressions = [] + for situation, style, source_id in expressions: + # 检查 situation 和 style 是否包含人物名称 + if person_name_filter(situation) or person_name_filter(style): + logger.info(f"跳过包含人物名称的表达方式: situation={situation}, style={style}") + continue + filtered_expressions.append((situation, style, source_id)) + expressions = filtered_expressions + # 检查表达方式数量，如果超过10个则放弃本次表达学习 if len(expressions) > 10: logger.info(f"表达方式提取数量超过10个（实际{len(expressions)}个），放弃本次表达学习") @@ -147,7 +158,7 @@ class ExpressionLearner: # 处理黑话条目，路由到 jargon_miner（即使没有表达方式也要处理黑话） if jargon_entries: - await self._process_jargon_entries(jargon_entries, random_msg) + await self._process_jargon_entries(jargon_entries, random_msg, person_name_filter) # 如果没有表达方式，直接返回 if not expressions: @@ -500,13 +511,19 @@ class ExpressionLearner: logger.error(f"概括表达情境失败: {e}") return None - async def _process_jargon_entries(self, jargon_entries: List[Tuple[str, str]], messages: List[Any]) -> None: + async def _process_jargon_entries( + self, + jargon_entries: List[Tuple[str, str]], + messages: List[Any], + person_name_filter: Optional[Callable[[str], bool]] = None + ) -> None: """ 处理从 expression learner 提取的黑话条目，路由到 jargon_miner Args: jargon_entries: 黑话条目列表，每个元素是 (content, source_id) messages: 消息列表，用于构建上下文 + person_name_filter: 可选的过滤函数，用于检查内容是否包含人物名称 """ if not jargon_entries or not messages: return @@ -527,6 +544,11 @@ class ExpressionLearner: logger.info(f"跳过包含机器人昵称/别名的黑话: {content}") continue + # 检查是否包含人物名称 + if person_name_filter and person_name_filter(content): + logger.info(f"跳过包含人物名称的黑话: {content}") + continue + # 解析 source_id source_id_str = (source_id or "").strip() if not source_id_str.isdigit(): @@ -557,7 +579,7 @@ class ExpressionLearner: return # 调用 jargon_miner 处理这些条目 - await jargon_miner.process_extracted_entries(entries) + await jargon_miner.process_extracted_entries(entries, person_name_filter) init_prompt() diff --git a/src/bw_learner/jargon_miner.py b/src/bw_learner/jargon_miner.py index 275d9cbd..78c5f0dc 100644 --- a/src/bw_learner/jargon_miner.py +++ b/src/bw_learner/jargon_miner.py @@ -3,7 +3,7 @@ import json import asyncio import random from collections import OrderedDict -from typing import List, Dict, Optional, Any +from typing import List, Dict, Optional, Any, Callable from json_repair import repair_json from peewee import fn @@ -478,12 +478,17 @@ class JargonMiner: traceback.print_exc() - async def run_once(self, messages: List[Any]) -> None: + async def run_once( + self, + messages: List[Any], + person_name_filter: Optional[Callable[[str], bool]] = None + ) -> None: """ 运行一次黑话提取 Args: messages: 外部传入的消息列表（必需） + person_name_filter: 可选的过滤函数，用于检查内容是否包含人物名称 """ # 使用异步锁防止并发执行 async with self._extraction_lock: @@ -562,6 +567,11 @@ class JargonMiner: if contains_bot_self_name(content): logger.info(f"解析阶段跳过包含机器人昵称/别名的词条: {content}") continue + + # 检查是否包含人物名称 + if person_name_filter and person_name_filter(content): + logger.info(f"解析阶段跳过包含人物名称的词条: {content}") + continue msg_id_str = str(msg_id_value or "").strip() if not msg_id_str: @@ -723,12 +733,17 @@ class JargonMiner: logger.error(f"JargonMiner 运行失败: {e}") # 即使失败也保持时间戳更新，避免频繁重试 - async def process_extracted_entries(self, entries: List[Dict[str, List[str]]]) -> None: + async def process_extracted_entries( + self, + entries: List[Dict[str, List[str]]], + person_name_filter: Optional[Callable[[str], bool]] = None + ) -> None: """ 处理已提取的黑话条目（从 expression_learner 路由过来的） Args: entries: 黑话条目列表，每个元素格式为 {"content": "...", "raw_content": [...]} + person_name_filter: 可选的过滤函数，用于检查内容是否包含人物名称 """ if not entries: return @@ -738,6 +753,14 @@ class JargonMiner: merged_entries: OrderedDict[str, Dict[str, List[str]]] = OrderedDict() for entry in entries: content_key = entry["content"] + + # 检查是否包含人物名称 + logger.info(f"process_extracted_entries 检查是否包含人物名称: {content_key}") + logger.info(f"person_name_filter: {person_name_filter}") + if person_name_filter and person_name_filter(content_key): + logger.info(f"process_extracted_entries 跳过包含人物名称的黑话: {content_key}") + continue + raw_list = entry.get("raw_content", []) or [] if content_key in merged_entries: merged_entries[content_key]["raw_content"].extend(raw_list) diff --git a/src/bw_learner/message_recorder.py b/src/bw_learner/message_recorder.py index c49da951..8c13013f 100644 --- a/src/bw_learner/message_recorder.py +++ b/src/bw_learner/message_recorder.py @@ -1,16 +1,34 @@ import time import asyncio -from typing import List, Any +from typing import List, Any, Optional +from collections import OrderedDict +from dataclasses import dataclass from src.common.logger import get_logger from src.config.config import global_config from src.chat.message_receive.chat_stream import get_chat_manager from src.chat.utils.chat_message_builder import get_raw_msg_by_timestamp_with_chat_inclusive from src.bw_learner.expression_learner import expression_learner_manager from src.bw_learner.jargon_miner import miner_manager +from src.person_info.person_info import Person logger = get_logger("bw_learner") +@dataclass +class PersonInfo: + """参与聊天的人物信息""" + user_id: str + user_platform: str + user_nickname: str + user_cardname: Optional[str] + person_name: str + last_seen_time: float # 最后发言时间 + + def get_unique_key(self) -> str: + """获取唯一标识（用于去重）""" + return f"{self.user_platform}:{self.user_id}" + + class MessageRecorder: """ 统一的消息记录器，负责管理时间窗口和消息提取，并将消息分发给 expression_learner 和 jargon_miner @@ -27,6 +45,11 @@ class MessageRecorder: # 提取锁，防止并发执行 self._extraction_lock = asyncio.Lock() + # 维护参与该chat_id的人物列表（最多30个，使用OrderedDict保持插入顺序） + # key: f"{platform}:{user_id}", value: PersonInfo + self._person_list: OrderedDict[str, PersonInfo] = OrderedDict() + self._max_person_count = 30 + # 获取 expression 和 jargon 的配置参数 self._init_parameters() @@ -111,6 +134,11 @@ class MessageRecorder: # 按时间排序，确保顺序一致 messages = sorted(messages, key=lambda msg: msg.time or 0) + # 更新参与聊天的人物列表 + self._update_person_list(messages) + + logger.info(f"聊天流 {self.chat_name} 的人物列表: {self._person_list}") + logger.info( f"聊天流 {self.chat_name} 提取到 {len(messages)} 条消息，" f"时间窗口: {extraction_start_time:.2f} - {extraction_end_time:.2f}" @@ -152,8 +180,11 @@ class MessageRecorder: messages: 消息列表 """ try: - # 传递消息给 ExpressionLearner（必需参数） - learnt_style = await self.expression_learner.learn_and_store(messages=messages) + # 传递消息和过滤函数给 ExpressionLearner + learnt_style = await self.expression_learner.learn_and_store( + messages=messages, + person_name_filter=self.contains_person_name + ) if learnt_style: logger.info(f"聊天流 {self.chat_name} 表达学习完成") @@ -179,13 +210,123 @@ class MessageRecorder: messages: 消息列表 """ try: - # 传递消息给 JargonMiner，避免它重复获取 - await self.jargon_miner.run_once(messages=messages) + # 传递消息和过滤函数给 JargonMiner + await self.jargon_miner.run_once( + messages=messages, + person_name_filter=self.contains_person_name + ) except Exception as e: logger.error(f"为聊天流 {self.chat_name} 触发黑话提取失败: {e}") import traceback traceback.print_exc() + + def _update_person_list(self, messages: List[Any]) -> None: + """ + 从消息中提取人物信息并更新人物列表 + + Args: + messages: 消息列表 + """ + for msg in messages: + # 获取消息发送者信息 + # 消息对象可能是 DatabaseMessages，它有 user_info 属性 + if hasattr(msg, 'user_info'): + # DatabaseMessages 类型 + user_info = msg.user_info + user_id = getattr(user_info, 'user_id', None) or '' + user_platform = getattr(user_info, 'platform', None) or '' + user_nickname = getattr(user_info, 'user_nickname', None) or '' + user_cardname = getattr(user_info, 'user_cardname', None) + else: + # 直接属性访问 + user_id = getattr(msg, 'user_id', None) or '' + user_platform = getattr(msg, 'user_platform', None) or '' + user_nickname = getattr(msg, 'user_nickname', None) or '' + user_cardname = getattr(msg, 'user_cardname', None) + + msg_time = getattr(msg, 'time', time.time()) + + # 检查必要信息 + if not user_id or not user_platform: + continue + + # 获取 person_name + try: + person = Person(platform=user_platform, user_id=str(user_id)) + person_name = person.person_name or user_nickname or (user_cardname if user_cardname else "未知用户") + except Exception as e: + logger.info(f"获取person_name失败: {e}, 使用nickname") + person_name = user_nickname or (user_cardname if user_cardname else "未知用户") + + # 生成唯一key + unique_key = f"{user_platform}:{user_id}" + + # 如果已存在，更新最后发言时间 + if unique_key in self._person_list: + self._person_list[unique_key].last_seen_time = msg_time + # 移动到末尾（表示最近活跃） + self._person_list.move_to_end(unique_key) + else: + # 如果超过最大数量，移除最早的（最前面的） + if len(self._person_list) >= self._max_person_count: + oldest_key = next(iter(self._person_list)) + del self._person_list[oldest_key] + logger.info(f"人物列表已满，移除最早的人物: {oldest_key}") + + # 添加新人物 + person_info = PersonInfo( + user_id=str(user_id), + user_platform=user_platform, + user_nickname=user_nickname or "", + user_cardname=user_cardname, + person_name=person_name, + last_seen_time=msg_time + ) + self._person_list[unique_key] = person_info + logger.info(f"添加新人物到列表: {unique_key}, person_name={person_name}") + + def contains_person_name(self, content: str) -> bool: + """ + 检查内容是否包含任何参与聊天的人物的名称或昵称 + + Args: + content: 要检查的内容 + + Returns: + bool: 如果包含任何人物名称或昵称，返回True + """ + if not content or not self._person_list: + return False + + content_lower = content.strip().lower() + if not content_lower: + return False + + # 检查所有人物 + for person_info in self._person_list.values(): + # 检查 person_name + if person_info.person_name: + person_name_lower = person_info.person_name.strip().lower() + if person_name_lower and person_name_lower in content_lower: + logger.debug(f"内容包含person_name: {person_info.person_name} in {content}") + return True + + # 检查 user_nickname + if person_info.user_nickname: + nickname_lower = person_info.user_nickname.strip().lower() + if nickname_lower and nickname_lower in content_lower: + logger.debug(f"内容包含nickname: {person_info.user_nickname} in {content}") + return True + + # 检查 user_cardname（群昵称） + if person_info.user_cardname: + cardname_lower = person_info.user_cardname.strip().lower() + if cardname_lower and cardname_lower in content_lower: + logger.debug(f"内容包含cardname: {person_info.user_cardname} in {content}") + return True + + return False class MessageRecorderManager: diff --git a/src/chat/replyer/prompt/replyer_prompt.py b/src/chat/replyer/prompt/replyer_prompt.py index e17e07f5..858015ac 100644 --- a/src/chat/replyer/prompt/replyer_prompt.py +++ b/src/chat/replyer/prompt/replyer_prompt.py @@ -36,7 +36,7 @@ def init_replyer_prompt(): {reply_target_block}。 {planner_reasoning} {identity} -{chat_prompt}你正在群里聊天,现在请你读读之前的聊天记录，把握当前的话题，然后给出口语化回复， +{chat_prompt}你正在群里聊天,现在请你读读之前的聊天记录，把握当前的话题，然后给出日常且简短的回复，。 {keywords_reaction_prompt} 请注意把握聊天内容。 {reply_style}