diff --git a/.gitignore b/.gitignore index e51abc5cc..6e1be60b4 100644 --- a/.gitignore +++ b/.gitignore @@ -1,4 +1,5 @@ data/ +data1/ mongodb/ NapCat.Framework.Windows.Once/ log/ diff --git a/config/auto_update.py b/config/auto_update.py new file mode 100644 index 000000000..28ab108da --- /dev/null +++ b/config/auto_update.py @@ -0,0 +1,59 @@ +import os +import shutil +import tomlkit +from pathlib import Path + +def update_config(): + # 获取根目录路径 + root_dir = Path(__file__).parent.parent + template_dir = root_dir / "template" + config_dir = root_dir / "config" + + # 定义文件路径 + template_path = template_dir / "bot_config_template.toml" + old_config_path = config_dir / "bot_config.toml" + new_config_path = config_dir / "bot_config.toml" + + # 读取旧配置文件 + old_config = {} + if old_config_path.exists(): + with open(old_config_path, "r", encoding="utf-8") as f: + old_config = tomlkit.load(f) + + # 删除旧的配置文件 + if old_config_path.exists(): + os.remove(old_config_path) + + # 复制模板文件到配置目录 + shutil.copy2(template_path, new_config_path) + + # 读取新配置文件 + with open(new_config_path, "r", encoding="utf-8") as f: + new_config = tomlkit.load(f) + + # 递归更新配置 + def update_dict(target, source): + for key, value in source.items(): + # 跳过version字段的更新 + if key == "version": + continue + if key in target: + if isinstance(value, dict) and isinstance(target[key], (dict, tomlkit.items.Table)): + update_dict(target[key], value) + else: + try: + # 直接使用tomlkit的item方法创建新值 + target[key] = tomlkit.item(value) + except (TypeError, ValueError): + # 如果转换失败，直接赋值 + target[key] = value + + # 将旧配置的值更新到新配置中 + update_dict(new_config, old_config) + + # 保存更新后的配置（保留注释和格式） + with open(new_config_path, "w", encoding="utf-8") as f: + f.write(tomlkit.dumps(new_config)) + +if __name__ == "__main__": + update_config() diff --git a/docs/linux_deploy_guide_for_beginners.md b/docs/linux_deploy_guide_for_beginners.md new file mode 100644 index 000000000..04601923f --- /dev/null +++ b/docs/linux_deploy_guide_for_beginners.md @@ -0,0 +1,444 @@ +# 面向纯新手的Linux服务器麦麦部署指南 + +## 你得先有一个服务器 + +为了能使麦麦在你的电脑关机之后还能运行，你需要一台不间断开机的主机，也就是我们常说的服务器。 + +华为云、阿里云、腾讯云等等都是在国内可以选择的选择。 + +你可以去租一台最低配置的就足敷需要了，按月租大概十几块钱就能租到了。 + +我们假设你已经租好了一台Linux架构的云服务器。我用的是阿里云ubuntu24.04，其他的原理相似。 + +## 0.我们就从零开始吧 + +### 网络问题 + +为访问github相关界面，推荐去下一款加速器，新手可以试试watttoolkit。 + +### 安装包下载 + +#### MongoDB + +对于ubuntu24.04 x86来说是这个： + +https://repo.mongodb.org/apt/ubuntu/dists/noble/mongodb-org/8.0/multiverse/binary-amd64/mongodb-org-server_8.0.5_amd64.deb + +如果不是就在这里自行选择对应版本 + +https://www.mongodb.com/try/download/community-kubernetes-operator + +#### Napcat + +在这里选择对应版本。 + +https://github.com/NapNeko/NapCatQQ/releases/tag/v4.6.7 + +对于ubuntu24.04 x86来说是这个： + +https://dldir1.qq.com/qqfile/qq/QQNT/ee4bd910/linuxqq_3.2.16-32793_amd64.deb + +#### 麦麦 + +https://github.com/SengokuCola/MaiMBot/archive/refs/tags/0.5.8-alpha.zip + +下载这个官方压缩包。 + +### 路径 + +我把麦麦相关文件放在了/moi/mai里面，你可以凭喜好更改，记得适当调整下面涉及到的部分即可。 + +文件结构： + +``` +moi +└─ mai + ├─ linuxqq_3.2.16-32793_amd64.deb + ├─ mongodb-org-server_8.0.5_amd64.deb + └─ bot + └─ MaiMBot-0.5.8-alpha.zip +``` + +### 网络 + +你可以在你的服务器控制台网页更改防火墙规则，允许6099，8080，27017这几个端口的出入。 + +## 1.正式开始！ + +远程连接你的服务器，你会看到一个黑框框闪着白方格，这就是我们要进行设置的场所——终端了。以下的bash命令都是在这里输入。 + +## 2. Python的安装 + +- 导入 Python 的稳定版 PPA： + +```bash +sudo add-apt-repository ppa:deadsnakes/ppa +``` + +- 导入 PPA 后，更新 APT 缓存： + +```bash +sudo apt update +``` + +- 在「终端」中执行以下命令来安装 Python 3.12： + +```bash +sudo apt install python3.12 +``` + +- 验证安装是否成功： + +```bash +python3.12 --version +``` + +- 在「终端」中，执行以下命令安装 pip： + +```bash +sudo apt install python3-pip +``` + +- 检查Pip是否安装成功： + +```bash +pip --version +``` + +- 安装必要组件 + +``` bash +sudo apt install python-is-python3 +``` + +## 3.MongoDB的安装 + +``` bash +cd /moi/mai +``` + +``` bash +dpkg -i mongodb-org-server_8.0.5_amd64.deb +``` + +``` bash +mkdir -p /root/data/mongodb/{data,log} +``` + +## 4.MongoDB的运行 + +```bash +service mongod start +``` + +```bash +systemctl status mongod #通过这条指令检查运行状态 +``` + +有需要的话可以把这个服务注册成开机自启 + +```bash +sudo systemctl enable mongod +``` + +## 5.napcat的安装 + +``` bash +curl -o napcat.sh https://nclatest.znin.net/NapNeko/NapCat-Installer/main/script/install.sh && sudo bash napcat.sh +``` + +上面的不行试试下面的 + +``` bash +dpkg -i linuxqq_3.2.16-32793_amd64.deb +apt-get install -f +dpkg -i linuxqq_3.2.16-32793_amd64.deb +``` + +成功的标志是输入``` napcat ```出来炫酷的彩虹色界面 + +## 6.napcat的运行 + +此时你就可以根据提示在```napcat```里面登录你的QQ号了。 + +```bash +napcat start <你的QQ号> +napcat status #检查运行状态 +``` + +然后你就可以登录napcat的webui进行设置了： + +```http://<你服务器的公网IP>:6099/webui?token=napcat``` + +第一次是这个，后续改了密码之后token就会对应修改。你也可以使用```napcat log <你的QQ号>```来查看webui地址。把里面的```127.0.0.1```改成<你服务器的公网IP>即可。 + +登录上之后在网络配置界面添加websocket客户端，名称随便输一个，url改成`ws://127.0.0.1:8080/onebot/v11/ws`保存之后点启用，就大功告成了。 + +## 7.麦麦的安装 + +### step 1 安装解压软件 + +``` +sudo apt-get install unzip +``` + +### step 2 解压文件 + +```bash +cd /moi/mai/bot # 注意：要切换到压缩包的目录中去 +unzip MaiMBot-0.5.8-alpha.zip +``` + +### step 3 进入虚拟环境安装库 + +```bash +cd /moi/mai/bot +python -m venv venv +source venv/bin/activate +pip install -r requirements.txt +``` + +### step 4 试运行 + +```bash +cd /moi/mai/bot +python -m venv venv +source venv/bin/activate +python bot.py +``` + +肯定运行不成功，不过你会发现结束之后多了一些文件 + +``` +bot +├─ .env.prod +└─ config + └─ bot_config.toml +``` + +你要会vim直接在终端里修改也行，不过也可以把它们下到本地改好再传上去： + +### step 5 文件配置 + +本项目需要配置两个主要文件： + +1. `.env.prod` - 配置API服务和系统环境 +2. `bot_config.toml` - 配置机器人行为和模型 + +#### API + +你可以注册一个硅基流动的账号，通过邀请码注册有14块钱的免费额度：https://cloud.siliconflow.cn/i/7Yld7cfg。 + +#### 在.env.prod中定义API凭证： + +``` +# API凭证配置 +SILICONFLOW_KEY=your_key # 硅基流动API密钥 +SILICONFLOW_BASE_URL=https://api.siliconflow.cn/v1/ # 硅基流动API地址 + +DEEP_SEEK_KEY=your_key # DeepSeek API密钥 +DEEP_SEEK_BASE_URL=https://api.deepseek.com/v1 # DeepSeek API地址 + +CHAT_ANY_WHERE_KEY=your_key # ChatAnyWhere API密钥 +CHAT_ANY_WHERE_BASE_URL=https://api.chatanywhere.tech/v1 # ChatAnyWhere API地址 +``` + +#### 在bot_config.toml中引用API凭证： + +``` +[model.llm_reasoning] +name = "Pro/deepseek-ai/DeepSeek-R1" +base_url = "SILICONFLOW_BASE_URL" # 引用.env.prod中定义的地址 +key = "SILICONFLOW_KEY" # 引用.env.prod中定义的密钥 +``` + +如需切换到其他API服务，只需修改引用： + +``` +[model.llm_reasoning] +name = "Pro/deepseek-ai/DeepSeek-R1" +base_url = "DEEP_SEEK_BASE_URL" # 切换为DeepSeek服务 +key = "DEEP_SEEK_KEY" # 使用DeepSeek密钥 +``` + +#### 配置文件详解 + +##### 环境配置文件 (.env.prod) + +``` +# API配置 +SILICONFLOW_KEY=your_key +SILICONFLOW_BASE_URL=https://api.siliconflow.cn/v1/ +DEEP_SEEK_KEY=your_key +DEEP_SEEK_BASE_URL=https://api.deepseek.com/v1 +CHAT_ANY_WHERE_KEY=your_key +CHAT_ANY_WHERE_BASE_URL=https://api.chatanywhere.tech/v1 + +# 服务配置 +HOST=127.0.0.1 # 如果使用Docker部署，需要改成0.0.0.0，否则QQ消息无法传入 +PORT=8080 + +# 数据库配置 +MONGODB_HOST=127.0.0.1 # 如果使用Docker部署，需要改成数据库容器的名字，默认是mongodb +MONGODB_PORT=27017 +DATABASE_NAME=MegBot +MONGODB_USERNAME = "" # 数据库用户名 +MONGODB_PASSWORD = "" # 数据库密码 +MONGODB_AUTH_SOURCE = "" # 认证数据库 + +# 插件配置 +PLUGINS=["src2.plugins.chat"] +``` + +##### 机器人配置文件 (bot_config.toml) + +``` +[bot] +qq = "机器人QQ号" # 必填 +nickname = "麦麦" # 机器人昵称（你希望机器人怎么称呼它自己） + +[personality] +prompt_personality = [ + "曾经是一个学习地质的女大学生，现在学习心理学和脑科学，你会刷贴吧", + "是一个女大学生，你有黑色头发，你会刷小红书" +] +prompt_schedule = "一个曾经学习地质,现在学习心理学和脑科学的女大学生，喜欢刷qq，贴吧，知乎和小红书" + +[message] +min_text_length = 2 # 最小回复长度 +max_context_size = 15 # 上下文记忆条数 +emoji_chance = 0.2 # 表情使用概率 +ban_words = [] # 禁用词列表 + +[emoji] +auto_save = true # 自动保存表情 +enable_check = false # 启用表情审核 +check_prompt = "符合公序良俗" + +[groups] +talk_allowed = [] # 允许对话的群号 +talk_frequency_down = [] # 降低回复频率的群号 +ban_user_id = [] # 禁止回复的用户QQ号 + +[others] +enable_advance_output = true # 启用详细日志 +enable_kuuki_read = true # 启用场景理解 + +# 模型配置 +[model.llm_reasoning] # 推理模型 +name = "Pro/deepseek-ai/DeepSeek-R1" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + +[model.llm_reasoning_minor] # 轻量推理模型 +name = "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + +[model.llm_normal] # 对话模型 +name = "Pro/deepseek-ai/DeepSeek-V3" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + +[model.llm_normal_minor] # 备用对话模型 +name = "deepseek-ai/DeepSeek-V2.5" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + +[model.vlm] # 图像识别模型 +name = "deepseek-ai/deepseek-vl2" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + +[model.embedding] # 文本向量模型 +name = "BAAI/bge-m3" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" + + +[topic.llm_topic] +name = "Pro/deepseek-ai/DeepSeek-V3" +base_url = "SILICONFLOW_BASE_URL" +key = "SILICONFLOW_KEY" +``` + +**step # 6** 运行 + +现在再运行 + +```bash +cd /moi/mai/bot +python -m venv venv +source venv/bin/activate +python bot.py +``` + +应该就能运行成功了。 + +## 8.事后配置 + +可是现在还有个问题：只要你一关闭终端，bot.py就会停止运行。那该怎么办呢？我们可以把bot.py注册成服务。 + +重启服务器，打开MongoDB和napcat服务。 + +新建一个文件，名为`bot.service`，内容如下 + +``` +[Unit] +Description=maimai bot + +[Service] +WorkingDirectory=/moi/mai/bot +ExecStart=/moi/mai/bot/venv/bin/python /moi/mai/bot/bot.py +Restart=on-failure +User=root + +[Install] +WantedBy=multi-user.target +``` + +里面的路径视自己的情况更改。 + +把它放到`/etc/systemd/system`里面。 + +重新加载 `systemd` 配置： + +```bash +sudo systemctl daemon-reload +``` + +启动服务： + +```bash +sudo systemctl start bot.service # 启动服务 +sudo systemctl restart bot.service # 或者重启服务 +``` + +检查服务状态： + +```bash +sudo systemctl status bot.service +``` + +现在再关闭终端，检查麦麦能不能正常回复QQ信息。如果可以的话就大功告成了！ + +## 9.命令速查 + +```bash +service mongod start # 启动mongod服务 +napcat start <你的QQ号> # 登录napcat +cd /moi/mai/bot # 切换路径 +python -m venv venv # 创建虚拟环境 +source venv/bin/activate # 激活虚拟环境 + +sudo systemctl daemon-reload # 重新加载systemd配置 +sudo systemctl start bot.service # 启动bot服务 +sudo systemctl enable bot.service # 启动bot服务 + +sudo systemctl status bot.service # 检查bot服务状态 +``` + +``` +python bot.py +``` + diff --git a/requirements.txt b/requirements.txt index 0acaade5e..8330c8d06 100644 Binary files a/requirements.txt and b/requirements.txt differ diff --git a/run.py b/run.py index 50e312c37..cfd3a5f14 100644 --- a/run.py +++ b/run.py @@ -128,13 +128,17 @@ if __name__ == "__main__": ) os.system("cls") if choice == "1": - install_napcat() - install_mongodb() + confirm = input("首次安装将下载并配置所需组件\n1.确认\n2.取消\n") + if confirm == "1": + install_napcat() + install_mongodb() + else: + print("已取消安装") elif choice == "2": run_maimbot() - choice = input("是否启动推理可视化？（y/N）").upper() + choice = input("是否启动推理可视化？（未完善）（y/N）").upper() if choice == "Y": run_cmd(r"python src\gui\reasoning_gui.py") - choice = input("是否启动记忆可视化？（y/N）").upper() + choice = input("是否启动记忆可视化？（未完善）（y/N）").upper() if choice == "Y": run_cmd(r"python src/plugins/memory_system/memory_manual_build.py") diff --git a/src/common/database.py b/src/common/database.py index 9d9a596d1..c6cead225 100644 --- a/src/common/database.py +++ b/src/common/database.py @@ -1,5 +1,4 @@ from typing import Optional - from pymongo import MongoClient from pymongo.database import Database as MongoDatabase diff --git a/src/plugins/chat/__init__.py b/src/plugins/chat/__init__.py index 8ae525708..f72f02baf 100644 --- a/src/plugins/chat/__init__.py +++ b/src/plugins/chat/__init__.py @@ -4,7 +4,7 @@ import os from loguru import logger from nonebot import get_driver, on_message, require -from nonebot.adapters.onebot.v11 import Bot, GroupMessageEvent, Message, MessageSegment +from nonebot.adapters.onebot.v11 import Bot, GroupMessageEvent, Message, MessageSegment,MessageEvent from nonebot.typing import T_State from ...common.database import Database @@ -38,8 +38,8 @@ emoji_manager.initialize() logger.debug(f"正在唤醒{global_config.BOT_NICKNAME}......") # 创建机器人实例 chat_bot = ChatBot() -# 注册群消息处理器 -group_msg = on_message(priority=5) +# 注册消息处理器 +msg_in = on_message(priority=5) # 创建定时任务 scheduler = require("nonebot_plugin_apscheduler").scheduler @@ -91,8 +91,8 @@ async def _(bot: Bot): asyncio.create_task(chat_manager._auto_save_task()) -@group_msg.handle() -async def _(bot: Bot, event: GroupMessageEvent, state: T_State): +@msg_in.handle() +async def _(bot: Bot, event: MessageEvent, state: T_State): await chat_bot.handle_message(event, bot) diff --git a/src/plugins/chat/bot.py b/src/plugins/chat/bot.py index 9b2ac06f1..f335a2ba3 100644 --- a/src/plugins/chat/bot.py +++ b/src/plugins/chat/bot.py @@ -2,12 +2,16 @@ import re import time from random import random from loguru import logger -from nonebot.adapters.onebot.v11 import Bot, GroupMessageEvent +from nonebot.adapters.onebot.v11 import ( + Bot, + GroupMessageEvent, + MessageEvent, + PrivateMessageEvent, +) from ..memory_system.memory import hippocampus from ..moods.moods import MoodManager # 导入情绪管理器 from .config import global_config -from .cq_code import CQCode, cq_code_tool # 导入CQCode模块 from .emoji_manager import emoji_manager # 导入表情包管理器 from .llm_generator import ResponseGenerator from .message import MessageSending, MessageRecv, MessageThinking, MessageSet @@ -42,39 +46,53 @@ class ChatBot: if not self._started: self._started = True - async def handle_message(self, event: GroupMessageEvent, bot: Bot) -> None: - """处理收到的群消息""" + async def handle_message(self, event: MessageEvent, bot: Bot) -> None: + """处理收到的消息""" self.bot = bot # 更新 bot 实例 - try: - group_info_api = await bot.get_group_info(group_id=event.group_id) - logger.info(f"成功获取群信息: {group_info_api}") - group_name = group_info_api["group_name"] - except Exception as e: - logger.error(f"获取群信息失败: {str(e)}") - group_name = None - - # 白名单设定由nontbot侧完成 - # 消息过滤，涉及到config有待更新 - if event.group_id: - if event.group_id not in global_config.talk_allowed_groups: - return + # 用户屏蔽,不区分私聊/群聊 if event.user_id in global_config.ban_user_id: return - user_info = UserInfo( - user_id=event.user_id, - user_nickname=event.sender.nickname, - user_cardname=event.sender.card or None, - platform="qq", - ) + # 处理私聊消息 + if isinstance(event, PrivateMessageEvent): + if not global_config.enable_friend_chat: # 私聊过滤 + return + else: + try: + user_info = UserInfo( + user_id=event.user_id, + user_nickname=(await bot.get_stranger_info(user_id=event.user_id, no_cache=True))["nickname"], + user_cardname=None, + platform="qq", + ) + except Exception as e: + logger.error(f"获取陌生人信息失败: {e}") + return + logger.debug(user_info) - group_info = GroupInfo( - group_id=event.group_id, - group_name=group_name, # 使用获取到的群名称或None - platform="qq", - ) + # group_info = GroupInfo(group_id=0, group_name="私聊", platform="qq") + group_info = None + + # 处理群聊消息 + else: + # 白名单设定由nontbot侧完成 + if event.group_id: + if event.group_id not in global_config.talk_allowed_groups: + return + + user_info = UserInfo( + user_id=event.user_id, + user_nickname=event.sender.nickname, + user_cardname=event.sender.card or None, + platform="qq", + ) + + group_info = GroupInfo(group_id=event.group_id, group_name=None, platform="qq") + + # group_info = await bot.get_group_info(group_id=event.group_id) + # sender_info = await bot.get_group_member_info(group_id=event.group_id, user_id=event.user_id, no_cache=True) message_cq = MessageRecvCQ( message_id=event.message_id, @@ -88,7 +106,6 @@ class ChatBot: # 进入maimbot message = MessageRecv(message_json) - groupinfo = message.message_info.group_info userinfo = message.message_info.user_info messageinfo = message.message_info @@ -108,7 +125,9 @@ class ChatBot: # 过滤词 for word in global_config.ban_words: if word in message.processed_plain_text: - logger.info(f"[群{groupinfo.group_id}]{userinfo.user_nickname}:{message.processed_plain_text}") + logger.info( + f"[{chat.group_info.group_name if chat.group_info.group_id else '私聊'}]{userinfo.user_nickname}:{message.processed_plain_text}" + ) logger.info(f"[过滤词识别]消息中含有{word}，filtered") return @@ -116,7 +135,7 @@ class ChatBot: for pattern in global_config.ban_msgs_regex: if re.search(pattern, message.raw_message): logger.info( - f"[群{message.message_info.group_info.group_id}]{message.user_nickname}:{message.raw_message}" + f"[{chat.group_info.group_name if chat.group_info.group_id else '私聊'}]{message.user_nickname}:{message.raw_message}" ) logger.info(f"[正则表达式过滤]消息匹配到{pattern}，filtered") return @@ -124,8 +143,8 @@ class ChatBot: current_time = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(messageinfo.time)) # topic=await topic_identifier.identify_topic_llm(message.processed_plain_text) + topic = "" - interested_rate = 0 interested_rate = await hippocampus.memory_activate_value(message.processed_plain_text) / 100 logger.debug(f"对{message.processed_plain_text}的激活度:{interested_rate}") # logger.info(f"\033[1;32m[主题识别]\033[0m 使用{global_config.topic_extract}主题: {topic}") @@ -144,7 +163,7 @@ class ChatBot: current_willing = willing_manager.get_willing(chat_stream=chat) logger.info( - f"[{current_time}][群{chat.group_info.group_id}]{chat.user_info.user_nickname}:" + f"[{current_time}][{chat.group_info.group_name if chat.group_info.group_id else '私聊'}]{chat.user_info.user_nickname}:" f"{message.processed_plain_text}[回复意愿:{current_willing:.2f}][概率:{reply_probability * 100:.1f}%]" ) @@ -152,12 +171,17 @@ class ChatBot: if random() < reply_probability: bot_user_info = UserInfo( - user_id=global_config.BOT_QQ, user_nickname=global_config.BOT_NICKNAME, platform=messageinfo.platform + user_id=global_config.BOT_QQ, + user_nickname=global_config.BOT_NICKNAME, + platform=messageinfo.platform, ) thinking_time_point = round(time.time(), 2) think_id = "mt" + str(thinking_time_point) thinking_message = MessageThinking( - message_id=think_id, chat_stream=chat, bot_user_info=bot_user_info, reply=message + message_id=think_id, + chat_stream=chat, + bot_user_info=bot_user_info, + reply=message, ) message_manager.add_message(thinking_message) @@ -196,15 +220,16 @@ class ChatBot: # print(f"\033[1;32m[回复内容]\033[0m {msg}") # 通过时间改变时间戳 typing_time = calculate_typing_time(msg) - print(f"typing_time: {typing_time}") + logger.debug(f"typing_time: {typing_time}") accu_typing_time += typing_time timepoint = thinking_time_point + accu_typing_time message_segment = Seg(type="text", data=msg) - print(f"message_segment: {message_segment}") + # logger.debug(f"message_segment: {message_segment}") bot_message = MessageSending( message_id=think_id, chat_stream=chat, bot_user_info=bot_user_info, + sender_info=userinfo, message_segment=message_segment, reply=message, is_head=not mark_head, @@ -218,7 +243,9 @@ class ChatBot: # message_set 可以直接加入 message_manager # print(f"\033[1;32m[回复]\033[0m 将回复载入发送容器") - print(f"添加message_set到message_manager") + + logger.debug("添加message_set到message_manager") + message_manager.add_message(message_set) bot_response_time = thinking_time_point @@ -242,6 +269,7 @@ class ChatBot: message_id=think_id, chat_stream=chat, bot_user_info=bot_user_info, + sender_info=userinfo, message_segment=message_segment, reply=message, is_head=False, diff --git a/src/plugins/chat/config.py b/src/plugins/chat/config.py index 596d120f9..a53237e6a 100644 --- a/src/plugins/chat/config.py +++ b/src/plugins/chat/config.py @@ -69,6 +69,7 @@ class BotConfig: enable_advance_output: bool = False # 是否启用高级输出 enable_kuuki_read: bool = True # 是否启用读空气功能 enable_debug_output: bool = False # 是否启用调试输出 + enable_friend_chat: bool = False # 是否启用好友聊天 mood_update_interval: float = 1.0 # 情绪更新间隔 单位秒 mood_decay_rate: float = 0.95 # 情绪衰减率 @@ -327,7 +328,9 @@ class BotConfig: others_config = parent["others"] config.enable_advance_output = others_config.get("enable_advance_output", config.enable_advance_output) config.enable_kuuki_read = others_config.get("enable_kuuki_read", config.enable_kuuki_read) - config.enable_debug_output = others_config.get("enable_debug_output", config.enable_debug_output) + if config.INNER_VERSION in SpecifierSet(">=0.0.7"): + config.enable_debug_output = others_config.get("enable_debug_output", config.enable_debug_output) + config.enable_friend_chat = others_config.get("enable_friend_chat", config.enable_friend_chat) # 版本表达式：>=1.0.0,<2.0.0 # 允许字段：func: method, support: str, notice: str, necessary: bool diff --git a/src/plugins/chat/emoji_manager.py b/src/plugins/chat/emoji_manager.py index 1743571e9..f15251077 100644 --- a/src/plugins/chat/emoji_manager.py +++ b/src/plugins/chat/emoji_manager.py @@ -161,6 +161,7 @@ class EmojiManager: {'_id': selected_emoji['_id']}, {'$inc': {'usage_count': 1}} ) + logger.success( f"找到匹配的表情包: {selected_emoji.get('description', '无描述')} (相似度: {similarity:.4f})") # 稍微改一下文本描述，不然容易产生幻觉，描述已经包含 表情包 了 @@ -176,8 +177,10 @@ class EmojiManager: logger.error(f"获取表情包失败: {str(e)}") return None + async def _get_emoji_discription(self, image_base64: str) -> str: """获取表情包的标签，使用image_manager的描述生成功能""" + try: # 使用image_manager获取描述，去掉前后的方括号和"表情包："前缀 description = await image_manager.get_emoji_description(image_base64) @@ -272,11 +275,14 @@ class EmojiManager: # 获取表情包的描述 description = await self._get_emoji_discription(image_base64) + + if global_config.EMOJI_CHECK: check = await self._check_emoji(image_base64) if '是' not in check: os.remove(image_path) logger.info(f"描述: {description}") + logger.info(f"描述: {description}") logger.info(f"其不满足过滤规则，被剔除 {check}") continue @@ -287,6 +293,7 @@ class EmojiManager: if description is not None: embedding = await get_embedding(description) + # 准备数据库记录 emoji_record = { 'filename': filename, @@ -301,6 +308,7 @@ class EmojiManager: self.db['emoji'].insert_one(emoji_record) logger.success(f"注册新表情包: {filename}") logger.info(f"描述: {description}") + # 保存到images数据库 image_doc = { @@ -389,5 +397,7 @@ class EmojiManager: # 创建全局单例 + emoji_manager = EmojiManager() + diff --git a/src/plugins/chat/llm_generator.py b/src/plugins/chat/llm_generator.py index 285ea59b7..84e1937b0 100644 --- a/src/plugins/chat/llm_generator.py +++ b/src/plugins/chat/llm_generator.py @@ -8,7 +8,7 @@ from loguru import logger from ...common.database import Database from ..models.utils_model import LLM_request from .config import global_config -from .message import MessageRecv, MessageThinking, MessageSending,Message +from .message import MessageRecv, MessageThinking, Message from .prompt_builder import prompt_builder from .relationship_manager import relationship_manager from .utils import process_llm_response diff --git a/src/plugins/chat/message.py b/src/plugins/chat/message.py index 0505c05a6..626e7cf4e 100644 --- a/src/plugins/chat/message.py +++ b/src/plugins/chat/message.py @@ -3,20 +3,23 @@ import html import re import json from dataclasses import dataclass -from typing import Dict, ForwardRef, List, Optional, Union +from typing import Dict, List, Optional import urllib3 from loguru import logger from .utils_image import image_manager + from .message_base import Seg, GroupInfo, UserInfo, BaseMessageInfo, MessageBase from .chat_stream import ChatStream, chat_manager + # 禁用SSL警告 urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) -#这个类是消息数据类，用于存储和管理消息数据。 -#它定义了消息的属性，包括群组ID、用户ID、消息ID、原始消息内容、纯文本内容和时间戳。 -#它还定义了两个辅助属性：keywords用于提取消息的关键词，is_plain_text用于判断消息是否为纯文本。 +# 这个类是消息数据类，用于存储和管理消息数据。 +# 它定义了消息的属性，包括群组ID、用户ID、消息ID、原始消息内容、纯文本内容和时间戳。 +# 它还定义了两个辅助属性：keywords用于提取消息的关键词，is_plain_text用于判断消息是否为纯文本。 + @dataclass class Message(MessageBase): @@ -64,10 +67,10 @@ class Message(MessageBase): @dataclass class MessageRecv(Message): """接收消息类，用于处理从MessageCQ序列化的消息""" - + def __init__(self, message_dict: Dict): """从MessageCQ的字典初始化 - + Args: message_dict: MessageCQ序列化后的字典 """ @@ -100,68 +103,74 @@ class MessageRecv(Message): async def process(self) -> None: """处理消息内容，生成纯文本和详细文本 - + 这个方法必须在创建实例后显式调用，因为它包含异步操作。 """ - self.processed_plain_text = await self._process_message_segments(self.message_segment) + self.processed_plain_text = await self._process_message_segments( + self.message_segment + ) self.detailed_plain_text = self._generate_detailed_text() async def _process_message_segments(self, segment: Seg) -> str: """递归处理消息段，转换为文字描述 - + Args: segment: 要处理的消息段 - + Returns: str: 处理后的文本 """ - if segment.type == 'seglist': + if segment.type == "seglist": # 处理消息段列表 segments_text = [] for seg in segment.data: processed = await self._process_message_segments(seg) if processed: segments_text.append(processed) - return ' '.join(segments_text) + return " ".join(segments_text) else: # 处理单个消息段 return await self._process_single_segment(segment) async def _process_single_segment(self, seg: Seg) -> str: """处理单个消息段 - + Args: seg: 要处理的消息段 - + Returns: str: 处理后的文本 """ try: - if seg.type == 'text': + if seg.type == "text": return seg.data - elif seg.type == 'image': + elif seg.type == "image": # 如果是base64图片数据 if isinstance(seg.data, str): return await image_manager.get_image_description(seg.data) - return '[图片]' - elif seg.type == 'emoji': - self.is_emoji=True + return "[图片]" + elif seg.type == "emoji": + self.is_emoji = True if isinstance(seg.data, str): return await image_manager.get_emoji_description(seg.data) - return '[表情]' + return "[表情]" else: return f"[{seg.type}:{str(seg.data)}]" except Exception as e: - logger.error(f"处理消息段失败: {str(e)}, 类型: {seg.type}, 数据: {seg.data}") + logger.error( + f"处理消息段失败: {str(e)}, 类型: {seg.type}, 数据: {seg.data}" + ) return f"[处理失败的{seg.type}消息]" def _generate_detailed_text(self) -> str: """生成详细文本，包含时间和用户信息""" - time_str = time.strftime("%m-%d %H:%M:%S", time.localtime(self.message_info.time)) + time_str = time.strftime( + "%m-%d %H:%M:%S", time.localtime(self.message_info.time) + ) user_info = self.message_info.user_info name = ( f"{user_info.user_nickname}(ta的昵称:{user_info.user_cardname},ta的id:{user_info.user_id})" - if user_info.user_cardname!='' + if user_info.user_cardname != "" else f"{user_info.user_nickname}(ta的id:{user_info.user_id})" ) return f"[{time_str}] {name}: {self.processed_plain_text}\n" @@ -170,14 +179,14 @@ class MessageRecv(Message): @dataclass class MessageProcessBase(Message): """消息处理基类，用于处理中和发送中的消息""" - + def __init__( self, message_id: str, chat_stream: ChatStream, bot_user_info: UserInfo, message_segment: Optional[Seg] = None, - reply: Optional['MessageRecv'] = None + reply: Optional["MessageRecv"] = None, ): # 调用父类初始化 super().__init__( @@ -186,7 +195,7 @@ class MessageProcessBase(Message): chat_stream=chat_stream, user_info=bot_user_info, message_segment=message_segment, - reply=reply + reply=reply, ) # 处理状态相关属性 @@ -200,78 +209,83 @@ class MessageProcessBase(Message): async def _process_message_segments(self, segment: Seg) -> str: """递归处理消息段，转换为文字描述 - + Args: segment: 要处理的消息段 - + Returns: str: 处理后的文本 """ - if segment.type == 'seglist': + if segment.type == "seglist": # 处理消息段列表 segments_text = [] for seg in segment.data: processed = await self._process_message_segments(seg) if processed: segments_text.append(processed) - return ' '.join(segments_text) + return " ".join(segments_text) else: # 处理单个消息段 return await self._process_single_segment(segment) async def _process_single_segment(self, seg: Seg) -> str: """处理单个消息段 - + Args: seg: 要处理的消息段 - + Returns: str: 处理后的文本 """ try: - if seg.type == 'text': + if seg.type == "text": return seg.data - elif seg.type == 'image': + elif seg.type == "image": # 如果是base64图片数据 if isinstance(seg.data, str): return await image_manager.get_image_description(seg.data) - return '[图片]' - elif seg.type == 'emoji': + return "[图片]" + elif seg.type == "emoji": if isinstance(seg.data, str): return await image_manager.get_emoji_description(seg.data) - return '[表情]' - elif seg.type == 'at': + return "[表情]" + elif seg.type == "at": return f"[@{seg.data}]" - elif seg.type == 'reply': - if self.reply and hasattr(self.reply, 'processed_plain_text'): + elif seg.type == "reply": + if self.reply and hasattr(self.reply, "processed_plain_text"): return f"[回复：{self.reply.processed_plain_text}]" else: return f"[{seg.type}:{str(seg.data)}]" except Exception as e: - logger.error(f"处理消息段失败: {str(e)}, 类型: {seg.type}, 数据: {seg.data}") + logger.error( + f"处理消息段失败: {str(e)}, 类型: {seg.type}, 数据: {seg.data}" + ) return f"[处理失败的{seg.type}消息]" def _generate_detailed_text(self) -> str: """生成详细文本，包含时间和用户信息""" - time_str = time.strftime("%m-%d %H:%M:%S", time.localtime(self.message_info.time)) + time_str = time.strftime( + "%m-%d %H:%M:%S", time.localtime(self.message_info.time) + ) user_info = self.message_info.user_info name = ( f"{user_info.user_nickname}(ta的昵称:{user_info.user_cardname},ta的id:{user_info.user_id})" - if user_info.user_cardname != '' + if user_info.user_cardname != "" else f"{user_info.user_nickname}(ta的id:{user_info.user_id})" ) return f"[{time_str}] {name}: {self.processed_plain_text}\n" + @dataclass class MessageThinking(MessageProcessBase): """思考状态的消息类""" - + def __init__( self, message_id: str, chat_stream: ChatStream, bot_user_info: UserInfo, - reply: Optional['MessageRecv'] = None + reply: Optional["MessageRecv"] = None, ): # 调用父类初始化 super().__init__( @@ -279,25 +293,27 @@ class MessageThinking(MessageProcessBase): chat_stream=chat_stream, bot_user_info=bot_user_info, message_segment=None, # 思考状态不需要消息段 - reply=reply + reply=reply, ) - + # 思考状态特有属性 self.interrupt = False + @dataclass class MessageSending(MessageProcessBase): """发送状态的消息类""" - + def __init__( self, message_id: str, chat_stream: ChatStream, bot_user_info: UserInfo, + sender_info: UserInfo, # 用来记录发送者信息,用于私聊回复 message_segment: Seg, - reply: Optional['MessageRecv'] = None, + reply: Optional["MessageRecv"] = None, is_head: bool = False, - is_emoji: bool = False + is_emoji: bool = False, ): # 调用父类初始化 super().__init__( @@ -305,28 +321,34 @@ class MessageSending(MessageProcessBase): chat_stream=chat_stream, bot_user_info=bot_user_info, message_segment=message_segment, - reply=reply + reply=reply, ) - + # 发送状态特有属性 + self.sender_info = sender_info self.reply_to_message_id = reply.message_info.message_id if reply else None self.is_head = is_head self.is_emoji = is_emoji - - def set_reply(self, reply: Optional['MessageRecv']) -> None: + + def set_reply(self, reply: Optional["MessageRecv"]) -> None: """设置回复消息""" if reply: self.reply = reply self.reply_to_message_id = self.reply.message_info.message_id - self.message_segment = Seg(type='seglist', data=[ - Seg(type='reply', data=reply.message_info.message_id), - self.message_segment - ]) + self.message_segment = Seg( + type="seglist", + data=[ + Seg(type="reply", data=reply.message_info.message_id), + self.message_segment, + ], + ) async def process(self) -> None: """处理消息内容，生成纯文本和详细文本""" if self.message_segment: - self.processed_plain_text = await self._process_message_segments(self.message_segment) + self.processed_plain_text = await self._process_message_segments( + self.message_segment + ) self.detailed_plain_text = self._generate_detailed_text() @classmethod @@ -335,8 +357,8 @@ class MessageSending(MessageProcessBase): thinking: MessageThinking, message_segment: Seg, is_head: bool = False, - is_emoji: bool = False - ) -> 'MessageSending': + is_emoji: bool = False, + ) -> "MessageSending": """从思考状态消息创建发送状态消息""" return cls( message_id=thinking.message_info.message_id, @@ -345,41 +367,50 @@ class MessageSending(MessageProcessBase): bot_user_info=thinking.message_info.user_info, reply=thinking.reply, is_head=is_head, - is_emoji=is_emoji + is_emoji=is_emoji, ) - + def to_dict(self): - ret= super().to_dict() - ret['message_info']['user_info']=self.chat_stream.user_info.to_dict() + ret = super().to_dict() + ret["message_info"]["user_info"] = self.chat_stream.user_info.to_dict() return ret + def is_private_message(self) -> bool: + """判断是否为私聊消息""" + return ( + self.message_info.group_info is None + or self.message_info.group_info.group_id is None + ) + + @dataclass class MessageSet: """消息集合类，可以存储多个发送消息""" + def __init__(self, chat_stream: ChatStream, message_id: str): self.chat_stream = chat_stream self.message_id = message_id self.messages: List[MessageSending] = [] self.time = round(time.time(), 2) - + def add_message(self, message: MessageSending) -> None: """添加消息到集合""" if not isinstance(message, MessageSending): raise TypeError("MessageSet只能添加MessageSending类型的消息") self.messages.append(message) self.messages.sort(key=lambda x: x.message_info.time) - + def get_message_by_index(self, index: int) -> Optional[MessageSending]: """通过索引获取消息""" if 0 <= index < len(self.messages): return self.messages[index] return None - + def get_message_by_time(self, target_time: float) -> Optional[MessageSending]: """获取最接近指定时间的消息""" if not self.messages: return None - + left, right = 0, len(self.messages) - 1 while left < right: mid = (left + right) // 2 @@ -387,25 +418,22 @@ class MessageSet: left = mid + 1 else: right = mid - + return self.messages[left] - + def clear_messages(self) -> None: """清空所有消息""" self.messages.clear() - + def remove_message(self, message: MessageSending) -> bool: """移除指定消息""" if message in self.messages: self.messages.remove(message) return True return False - + def __str__(self) -> str: return f"MessageSet(id={self.message_id}, count={len(self.messages)})" - + def __len__(self) -> int: return len(self.messages) - - - diff --git a/src/plugins/chat/message_base.py b/src/plugins/chat/message_base.py index d17c2c357..ae7ec3872 100644 --- a/src/plugins/chat/message_base.py +++ b/src/plugins/chat/message_base.py @@ -1,5 +1,5 @@ from dataclasses import dataclass, asdict -from typing import List, Optional, Union, Any, Dict +from typing import List, Optional, Union, Dict @dataclass class Seg: diff --git a/src/plugins/chat/message_cq.py b/src/plugins/chat/message_cq.py index 6bfa47c3f..59d67a455 100644 --- a/src/plugins/chat/message_cq.py +++ b/src/plugins/chat/message_cq.py @@ -1,12 +1,12 @@ import time from dataclasses import dataclass -from typing import Dict, ForwardRef, List, Optional, Union +from typing import Dict, Optional import urllib3 -from .cq_code import CQCode, cq_code_tool +from .cq_code import cq_code_tool from .utils_cq import parse_cq_code -from .utils_user import get_groupname, get_user_cardname, get_user_nickname +from .utils_user import get_groupname from .message_base import Seg, GroupInfo, UserInfo, BaseMessageInfo, MessageBase # 禁用SSL警告 urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) @@ -61,8 +61,12 @@ class MessageRecvCQ(MessageCQ): ): # 调用父类初始化 super().__init__(message_id, user_info, group_info, platform) + + # 私聊消息不携带group_info + if group_info is None: + pass - if group_info.group_name is None: + elif group_info.group_name is None: group_info.group_name = get_groupname(group_info.group_id) # 解析消息段 diff --git a/src/plugins/chat/message_sender.py b/src/plugins/chat/message_sender.py index 9db74633f..584bf9c5d 100644 --- a/src/plugins/chat/message_sender.py +++ b/src/plugins/chat/message_sender.py @@ -5,12 +5,11 @@ from typing import Dict, List, Optional, Union from loguru import logger from nonebot.adapters.onebot.v11 import Bot -from .cq_code import cq_code_tool from .message_cq import MessageSendCQ -from .message import MessageSending, MessageThinking, MessageRecv,MessageSet +from .message import MessageSending, MessageThinking, MessageRecv, MessageSet + from .storage import MessageStorage from .config import global_config -from .chat_stream import chat_manager class Message_Sender: @@ -26,22 +25,24 @@ class Message_Sender: self._current_bot = bot async def send_message( - self, - message: MessageSending, + self, + message: MessageSending, ) -> None: """发送消息""" + if isinstance(message, MessageSending): message_json = message.to_dict() - message_send=MessageSendCQ( - data=message_json - ) - - if message_send.message_info.group_info: + message_send = MessageSendCQ(data=message_json) + # logger.debug(message_send.message_info,message_send.raw_message) + if ( + message_send.message_info.group_info + and message_send.message_info.group_info.group_id + ): try: await self._current_bot.send_group_msg( group_id=message.message_info.group_info.group_id, message=message_send.raw_message, - auto_escape=False + auto_escape=False, ) logger.success(f"[调试] 发送消息{message.processed_plain_text}成功") except Exception as e: @@ -49,10 +50,11 @@ class Message_Sender: logger.error(f"[调试] 发送消息{message.processed_plain_text}失败") else: try: + logger.debug(message.message_info.user_info) await self._current_bot.send_private_msg( - user_id=message.message_info.user_info.user_id, + user_id=message.sender_info.user_id, message=message_send.raw_message, - auto_escape=False + auto_escape=False, ) logger.success(f"[调试] 发送消息{message.processed_plain_text}成功") except Exception as e: @@ -62,13 +64,14 @@ class Message_Sender: class MessageContainer: """单个聊天流的发送/思考消息容器""" + def __init__(self, chat_id: str, max_size: int = 100): self.chat_id = chat_id self.max_size = max_size self.messages = [] self.last_send_time = 0 self.thinking_timeout = 20 # 思考超时时间（秒） - + def get_timeout_messages(self) -> List[MessageSending]: """获取所有超时的Message_Sending对象（思考时间超过30秒），按thinking_start_time排序""" current_time = time.time() @@ -83,12 +86,12 @@ class MessageContainer: timeout_messages.sort(key=lambda x: x.thinking_start_time) return timeout_messages - + def get_earliest_message(self) -> Optional[Union[MessageThinking, MessageSending]]: """获取thinking_start_time最早的消息对象""" if not self.messages: return None - earliest_time = float('inf') + earliest_time = float("inf") earliest_message = None for msg in self.messages: msg_time = msg.thinking_start_time @@ -96,7 +99,7 @@ class MessageContainer: earliest_time = msg_time earliest_message = msg return earliest_message - + def add_message(self, message: Union[MessageThinking, MessageSending]) -> None: """添加消息到队列""" if isinstance(message, MessageSet): @@ -104,7 +107,7 @@ class MessageContainer: self.messages.append(single_message) else: self.messages.append(message) - + def remove_message(self, message: Union[MessageThinking, MessageSending]) -> bool: """移除消息，如果消息存在则返回True，否则返回False""" try: @@ -119,7 +122,7 @@ class MessageContainer: def has_messages(self) -> bool: """检查是否有待发送的消息""" return bool(self.messages) - + def get_all_messages(self) -> List[Union[MessageSending, MessageThinking]]: """获取所有消息""" return list(self.messages) @@ -127,72 +130,91 @@ class MessageContainer: class MessageManager: """管理所有聊天流的消息容器""" + def __init__(self): self.containers: Dict[str, MessageContainer] = {} # chat_id -> MessageContainer self.storage = MessageStorage() self._running = True - + def get_container(self, chat_id: str) -> MessageContainer: """获取或创建聊天流的消息容器""" if chat_id not in self.containers: self.containers[chat_id] = MessageContainer(chat_id) return self.containers[chat_id] - - def add_message(self, message: Union[MessageThinking, MessageSending, MessageSet]) -> None: + + def add_message( + self, message: Union[MessageThinking, MessageSending, MessageSet] + ) -> None: chat_stream = message.chat_stream if not chat_stream: raise ValueError("无法找到对应的聊天流") container = self.get_container(chat_stream.stream_id) container.add_message(message) - + async def process_chat_messages(self, chat_id: str): """处理聊天流消息""" container = self.get_container(chat_id) if container.has_messages(): # print(f"处理有message的容器chat_id: {chat_id}") message_earliest = container.get_earliest_message() - + if isinstance(message_earliest, MessageThinking): message_earliest.update_thinking_time() thinking_time = message_earliest.thinking_time - print(f"消息正在思考中，已思考{int(thinking_time)}秒\r", end='', flush=True) + print( + f"消息正在思考中，已思考{int(thinking_time)}秒\r", + end="", + flush=True, + ) # 检查是否超时 if thinking_time > global_config.thinking_timeout: logger.warning(f"消息思考超时({thinking_time}秒)，移除该消息") container.remove_message(message_earliest) else: - - if message_earliest.is_head and message_earliest.update_thinking_time() > 30: + + if ( + message_earliest.is_head + and message_earliest.update_thinking_time() > 30 + and not message_earliest.is_private_message() # 避免在私聊时插入reply + ): await message_sender.send_message(message_earliest.set_reply()) else: await message_sender.send_message(message_earliest) await message_earliest.process() - - print(f"\033[1;34m[调试]\033[0m 消息'{message_earliest.processed_plain_text}'正在发送中") - - await self.storage.store_message(message_earliest, message_earliest.chat_stream,None) - + + print( + f"\033[1;34m[调试]\033[0m 消息'{message_earliest.processed_plain_text}'正在发送中" + ) + + await self.storage.store_message( + message_earliest, message_earliest.chat_stream, None + ) + container.remove_message(message_earliest) - + message_timeout = container.get_timeout_messages() if message_timeout: logger.warning(f"发现{len(message_timeout)}条超时消息") for msg in message_timeout: if msg == message_earliest: continue - + try: - if msg.is_head and msg.update_thinking_time() > 30: + if ( + msg.is_head + and msg.update_thinking_time() > 30 + and not message_earliest.is_private_message() # 避免在私聊时插入reply + ): await message_sender.send_message(msg.set_reply()) else: await message_sender.send_message(msg) - + # if msg.is_emoji: # msg.processed_plain_text = "[表情包]" await msg.process() - await self.storage.store_message(msg,msg.chat_stream, None) - + await self.storage.store_message(msg, msg.chat_stream, None) + if not container.remove_message(msg): logger.warning("尝试删除不存在的消息") except Exception: @@ -206,7 +228,7 @@ class MessageManager: tasks = [] for chat_id in self.containers.keys(): tasks.append(self.process_chat_messages(chat_id)) - + await asyncio.gather(*tasks) diff --git a/src/plugins/chat/prompt_builder.py b/src/plugins/chat/prompt_builder.py index ea3da777a..c89bf3e07 100644 --- a/src/plugins/chat/prompt_builder.py +++ b/src/plugins/chat/prompt_builder.py @@ -9,7 +9,7 @@ from ..moods.moods import MoodManager from ..schedule.schedule_generator import bot_schedule from .config import global_config from .utils import get_embedding, get_recent_group_detailed_plain_text -from .chat_stream import ChatStream, chat_manager +from .chat_stream import chat_manager class PromptBuilder: diff --git a/src/plugins/chat/relationship_manager.py b/src/plugins/chat/relationship_manager.py index baebb1fe8..fbd8cec59 100644 --- a/src/plugins/chat/relationship_manager.py +++ b/src/plugins/chat/relationship_manager.py @@ -1,6 +1,5 @@ import asyncio -from typing import Optional, Union -from typing import Optional, Union +from typing import Optional from loguru import logger from ...common.database import Database diff --git a/src/plugins/chat/storage.py b/src/plugins/chat/storage.py index dc03e4ced..ec155bbe9 100644 --- a/src/plugins/chat/storage.py +++ b/src/plugins/chat/storage.py @@ -1,8 +1,6 @@ from typing import Optional, Union -from typing import Optional, Union from ...common.database import Database -from .message_base import MessageBase from .message import MessageSending, MessageRecv from .chat_stream import ChatStream from loguru import logger diff --git a/src/plugins/chat/utils.py b/src/plugins/chat/utils.py index 7c658fbf7..186f2ab79 100644 --- a/src/plugins/chat/utils.py +++ b/src/plugins/chat/utils.py @@ -12,8 +12,8 @@ from loguru import logger from ..models.utils_model import LLM_request from ..utils.typo_generator import ChineseTypoGenerator from .config import global_config -from .message import MessageThinking, MessageRecv,MessageSending,MessageProcessBase,Message -from .message_base import MessageBase,BaseMessageInfo,UserInfo,GroupInfo +from .message import MessageRecv,Message +from .message_base import UserInfo from .chat_stream import ChatStream from ..moods.moods import MoodManager diff --git a/src/plugins/chat/utils_image.py b/src/plugins/chat/utils_image.py index ddb3b04cd..8f09a21a0 100644 --- a/src/plugins/chat/utils_image.py +++ b/src/plugins/chat/utils_image.py @@ -1,16 +1,12 @@ import base64 -import io import os import time -import zlib import aiohttp import hashlib -from typing import Optional, Tuple, Union -from urllib.parse import urlparse +from typing import Optional, Union from loguru import logger from nonebot import get_driver -from PIL import Image from ...common.database import Database from ..chat.config import global_config diff --git a/src/plugins/chat/willing_manager.py b/src/plugins/chat/willing_manager.py index 39083f0b8..f34afb746 100644 --- a/src/plugins/chat/willing_manager.py +++ b/src/plugins/chat/willing_manager.py @@ -1,13 +1,9 @@ import asyncio from typing import Dict -from loguru import logger -from typing import Dict -from loguru import logger from .config import global_config -from .message_base import UserInfo, GroupInfo -from .chat_stream import chat_manager,ChatStream +from .chat_stream import ChatStream class WillingManager: diff --git a/src/plugins/knowledege/knowledge_library.py b/src/plugins/knowledege/knowledge_library.py deleted file mode 100644 index 1bebf0930..000000000 --- a/src/plugins/knowledege/knowledge_library.py +++ /dev/null @@ -1,188 +0,0 @@ -import os -import sys -import time - -import requests -from dotenv import load_dotenv - -# 添加项目根目录到 Python 路径 -root_path = os.path.abspath(os.path.join(os.path.dirname(__file__), "../../..")) -sys.path.append(root_path) - -# 加载根目录下的env.edv文件 -env_path = os.path.join(root_path, ".env.dev") -if not os.path.exists(env_path): - raise FileNotFoundError(f"配置文件不存在: {env_path}") -load_dotenv(env_path) - -from src.common.database import Database - -class KnowledgeLibrary: - def __init__(self): - self.db = Database.get_instance() - self.raw_info_dir = "data/raw_info" - self._ensure_dirs() - self.api_key = os.getenv("SILICONFLOW_KEY") - if not self.api_key: - raise ValueError("SILICONFLOW_API_KEY 环境变量未设置") - - def _ensure_dirs(self): - """确保必要的目录存在""" - os.makedirs(self.raw_info_dir, exist_ok=True) - - def get_embedding(self, text: str) -> list: - """获取文本的embedding向量""" - url = "https://api.siliconflow.cn/v1/embeddings" - payload = { - "model": "BAAI/bge-m3", - "input": text, - "encoding_format": "float" - } - headers = { - "Authorization": f"Bearer {self.api_key}", - "Content-Type": "application/json" - } - - response = requests.post(url, json=payload, headers=headers) - if response.status_code != 200: - print(f"获取embedding失败: {response.text}") - return None - - return response.json()['data'][0]['embedding'] - - def process_files(self): - """处理raw_info目录下的所有txt文件""" - for filename in os.listdir(self.raw_info_dir): - if filename.endswith('.txt'): - file_path = os.path.join(self.raw_info_dir, filename) - self.process_single_file(file_path) - - def process_single_file(self, file_path: str): - """处理单个文件""" - try: - # 检查文件是否已处理 - if self.db.processed_files.find_one({"file_path": file_path}): - print(f"文件已处理过，跳过: {file_path}") - return - - with open(file_path, 'r', encoding='utf-8') as f: - content = f.read() - - # 按1024字符分段 - segments = [content[i:i+600] for i in range(0, len(content), 300)] - - # 处理每个分段 - for segment in segments: - if not segment.strip(): # 跳过空段 - continue - - # 获取embedding - embedding = self.get_embedding(segment) - if not embedding: - continue - - # 存储到数据库 - doc = { - "content": segment, - "embedding": embedding, - "file_path": file_path, - "segment_length": len(segment) - } - - # 使用文本内容的哈希值作为唯一标识 - content_hash = hash(segment) - - # 更新或插入文档 - self.db.knowledges.update_one( - {"content_hash": content_hash}, - {"$set": doc}, - upsert=True - ) - - # 记录文件已处理 - self.db.processed_files.insert_one({ - "file_path": file_path, - "processed_time": time.time() - }) - - print(f"成功处理文件: {file_path}") - - except Exception as e: - print(f"处理文件 {file_path} 时出错: {str(e)}") - - def search_similar_segments(self, query: str, limit: int = 5) -> list: - """搜索与查询文本相似的片段""" - query_embedding = self.get_embedding(query) - if not query_embedding: - return [] - - # 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"]}] - } - } - }, - {"$sort": {"similarity": -1}}, - {"$limit": limit}, - {"$project": {"content": 1, "similarity": 1, "file_path": 1}} - ] - - results = list(self.db.knowledges.aggregate(pipeline)) - return results - -# 创建单例实例 -knowledge_library = KnowledgeLibrary() - -if __name__ == "__main__": - # 测试知识库功能 - print("开始处理知识库文件...") - knowledge_library.process_files() - - # 测试搜索功能 - test_query = "麦麦评价一下僕と花" - print(f"\n搜索与'{test_query}'相似的内容:") - results = knowledge_library.search_similar_segments(test_query) - for result in results: - print(f"相似度: {result['similarity']:.4f}") - print(f"内容: {result['content'][:100]}...") - print("-" * 50) diff --git a/src/plugins/memory_system/memory_manual_build.py b/src/plugins/memory_system/memory_manual_build.py index 9c1d43ce9..736a50e97 100644 --- a/src/plugins/memory_system/memory_manual_build.py +++ b/src/plugins/memory_system/memory_manual_build.py @@ -10,7 +10,6 @@ from pathlib import Path import matplotlib.pyplot as plt import networkx as nx -import pymongo from dotenv import load_dotenv from loguru import logger import jieba diff --git a/src/plugins/zhishi/knowledge_library.py b/src/plugins/zhishi/knowledge_library.py new file mode 100644 index 000000000..2411e3112 --- /dev/null +++ b/src/plugins/zhishi/knowledge_library.py @@ -0,0 +1,383 @@ +import os +import sys +import time +import requests +from dotenv import load_dotenv +import hashlib +from datetime import datetime +from tqdm import tqdm +from rich.console import Console +from rich.table import Table + +# 添加项目根目录到 Python 路径 +root_path = os.path.abspath(os.path.join(os.path.dirname(__file__), "../../..")) +sys.path.append(root_path) + +# 现在可以导入src模块 +from src.common.database import Database + +# 加载根目录下的env.edv文件 +env_path = os.path.join(root_path, ".env.prod") +if not os.path.exists(env_path): + raise FileNotFoundError(f"配置文件不存在: {env_path}") +load_dotenv(env_path) + +class KnowledgeLibrary: + def __init__(self): + # 初始化数据库连接 + if Database._instance is None: + Database.initialize( + uri=os.getenv("MONGODB_URI"), + host=os.getenv("MONGODB_HOST", "127.0.0.1"), + port=int(os.getenv("MONGODB_PORT", "27017")), + db_name=os.getenv("DATABASE_NAME", "MegBot"), + username=os.getenv("MONGODB_USERNAME"), + password=os.getenv("MONGODB_PASSWORD"), + auth_source=os.getenv("MONGODB_AUTH_SOURCE"), + ) + self.db = Database.get_instance() + self.raw_info_dir = "data/raw_info" + self._ensure_dirs() + self.api_key = os.getenv("SILICONFLOW_KEY") + if not self.api_key: + raise ValueError("SILICONFLOW_API_KEY 环境变量未设置") + self.console = Console() + + def _ensure_dirs(self): + """确保必要的目录存在""" + os.makedirs(self.raw_info_dir, exist_ok=True) + + def read_file(self, file_path: str) -> str: + """读取文件内容""" + with open(file_path, 'r', encoding='utf-8') as f: + return f.read() + + def split_content(self, content: str, max_length: int = 512) -> list: + """将内容分割成适当大小的块，保持段落完整性 + + Args: + content: 要分割的文本内容 + max_length: 每个块的最大长度 + + Returns: + list: 分割后的文本块列表 + """ + # 首先按段落分割 + paragraphs = [p.strip() for p in content.split('\n\n') if p.strip()] + chunks = [] + current_chunk = [] + current_length = 0 + + for para in paragraphs: + para_length = len(para) + + # 如果单个段落就超过最大长度 + if para_length > max_length: + # 如果当前chunk不为空，先保存 + if current_chunk: + chunks.append('\n'.join(current_chunk)) + current_chunk = [] + current_length = 0 + + # 将长段落按句子分割 + sentences = [s.strip() for s in para.replace('。', '。\n').replace('！', '！\n').replace('？', '？\n').split('\n') if s.strip()] + temp_chunk = [] + temp_length = 0 + + for sentence in sentences: + sentence_length = len(sentence) + if sentence_length > max_length: + # 如果单个句子超长，强制按长度分割 + if temp_chunk: + chunks.append('\n'.join(temp_chunk)) + temp_chunk = [] + temp_length = 0 + for i in range(0, len(sentence), max_length): + chunks.append(sentence[i:i + max_length]) + elif temp_length + sentence_length + 1 <= max_length: + temp_chunk.append(sentence) + temp_length += sentence_length + 1 + else: + chunks.append('\n'.join(temp_chunk)) + temp_chunk = [sentence] + temp_length = sentence_length + + if temp_chunk: + chunks.append('\n'.join(temp_chunk)) + + # 如果当前段落加上现有chunk不超过最大长度 + elif current_length + para_length + 1 <= max_length: + current_chunk.append(para) + current_length += para_length + 1 + else: + # 保存当前chunk并开始新的chunk + chunks.append('\n'.join(current_chunk)) + current_chunk = [para] + current_length = para_length + + # 添加最后一个chunk + if current_chunk: + chunks.append('\n'.join(current_chunk)) + + return chunks + + def get_embedding(self, text: str) -> list: + """获取文本的embedding向量""" + url = "https://api.siliconflow.cn/v1/embeddings" + payload = { + "model": "BAAI/bge-m3", + "input": text, + "encoding_format": "float" + } + headers = { + "Authorization": f"Bearer {self.api_key}", + "Content-Type": "application/json" + } + + response = requests.post(url, json=payload, headers=headers) + if response.status_code != 200: + print(f"获取embedding失败: {response.text}") + return None + + return response.json()['data'][0]['embedding'] + + def process_files(self, knowledge_length:int=512): + """处理raw_info目录下的所有txt文件""" + txt_files = [f for f in os.listdir(self.raw_info_dir) if f.endswith('.txt')] + + if not txt_files: + self.console.print("[red]警告：在 {} 目录下没有找到任何txt文件[/red]".format(self.raw_info_dir)) + self.console.print("[yellow]请将需要处理的文本文件放入该目录后再运行程序[/yellow]") + return + + total_stats = { + "processed_files": 0, + "total_chunks": 0, + "failed_files": [], + "skipped_files": [] + } + + self.console.print(f"\n[bold blue]开始处理知识库文件 - 共{len(txt_files)}个文件[/bold blue]") + + for filename in tqdm(txt_files, desc="处理文件进度"): + file_path = os.path.join(self.raw_info_dir, filename) + result = self.process_single_file(file_path, knowledge_length) + self._update_stats(total_stats, result, filename) + + self._display_processing_results(total_stats) + + def process_single_file(self, file_path: str, knowledge_length: int = 512): + """处理单个文件""" + result = { + "status": "success", + "chunks_processed": 0, + "error": None + } + + try: + current_hash = self.calculate_file_hash(file_path) + processed_record = self.db.db.processed_files.find_one({"file_path": file_path}) + + if processed_record: + if processed_record.get("hash") == current_hash: + if knowledge_length in processed_record.get("split_by", []): + result["status"] = "skipped" + return result + + content = self.read_file(file_path) + chunks = self.split_content(content, knowledge_length) + + for chunk in tqdm(chunks, desc=f"处理 {os.path.basename(file_path)} 的文本块", leave=False): + embedding = self.get_embedding(chunk) + if embedding: + knowledge = { + "content": chunk, + "embedding": embedding, + "source_file": file_path, + "split_length": knowledge_length, + "created_at": datetime.now() + } + self.db.db.knowledges.insert_one(knowledge) + result["chunks_processed"] += 1 + + split_by = processed_record.get("split_by", []) if processed_record else [] + if knowledge_length not in split_by: + split_by.append(knowledge_length) + + self.db.db.processed_files.update_one( + {"file_path": file_path}, + { + "$set": { + "hash": current_hash, + "last_processed": datetime.now(), + "split_by": split_by + } + }, + upsert=True + ) + + except Exception as e: + result["status"] = "failed" + result["error"] = str(e) + + return result + + def _update_stats(self, total_stats, result, filename): + """更新总体统计信息""" + if result["status"] == "success": + total_stats["processed_files"] += 1 + total_stats["total_chunks"] += result["chunks_processed"] + elif result["status"] == "failed": + total_stats["failed_files"].append((filename, result["error"])) + elif result["status"] == "skipped": + total_stats["skipped_files"].append(filename) + + def _display_processing_results(self, stats): + """显示处理结果统计""" + self.console.print("\n[bold green]处理完成！统计信息如下：[/bold green]") + + table = Table(show_header=True, header_style="bold magenta") + table.add_column("统计项", style="dim") + table.add_column("数值") + + table.add_row("成功处理文件数", str(stats["processed_files"])) + table.add_row("处理的知识块总数", str(stats["total_chunks"])) + table.add_row("跳过的文件数", str(len(stats["skipped_files"]))) + table.add_row("失败的文件数", str(len(stats["failed_files"]))) + + self.console.print(table) + + if stats["failed_files"]: + self.console.print("\n[bold red]处理失败的文件：[/bold red]") + for filename, error in stats["failed_files"]: + self.console.print(f"[red]- {filename}: {error}[/red]") + + if stats["skipped_files"]: + self.console.print("\n[bold yellow]跳过的文件（已处理）：[/bold yellow]") + for filename in stats["skipped_files"]: + self.console.print(f"[yellow]- {filename}[/yellow]") + + def calculate_file_hash(self, file_path): + """计算文件的MD5哈希值""" + hash_md5 = hashlib.md5() + with open(file_path, "rb") as f: + for chunk in iter(lambda: f.read(4096), b""): + hash_md5.update(chunk) + return hash_md5.hexdigest() + + def search_similar_segments(self, query: str, limit: int = 5) -> list: + """搜索与查询文本相似的片段""" + query_embedding = self.get_embedding(query) + if not query_embedding: + return [] + + # 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"]}] + } + } + }, + {"$sort": {"similarity": -1}}, + {"$limit": limit}, + {"$project": {"content": 1, "similarity": 1, "file_path": 1}} + ] + + results = list(self.db.db.knowledges.aggregate(pipeline)) + return results + +# 创建单例实例 +knowledge_library = KnowledgeLibrary() + +if __name__ == "__main__": + console = Console() + console.print("[bold green]知识库处理工具[/bold green]") + + while True: + console.print("\n请选择要执行的操作：") + console.print("[1] 麦麦开始学习") + console.print("[2] 麦麦全部忘光光（仅知识）") + console.print("[q] 退出程序") + + choice = input("\n请输入选项: ").strip() + + if choice.lower() == 'q': + console.print("[yellow]程序退出[/yellow]") + sys.exit(0) + elif choice == '2': + confirm = input("确定要删除所有知识吗？这个操作不可撤销！(y/n): ").strip().lower() + if confirm == 'y': + knowledge_library.db.db.knowledges.delete_many({}) + console.print("[green]已清空所有知识！[/green]") + continue + elif choice == '1': + if not os.path.exists(knowledge_library.raw_info_dir): + console.print(f"[yellow]创建目录：{knowledge_library.raw_info_dir}[/yellow]") + os.makedirs(knowledge_library.raw_info_dir, exist_ok=True) + + # 询问分割长度 + while True: + try: + length_input = input("请输入知识分割长度（默认512，输入q退出，回车使用默认值）: ").strip() + if length_input.lower() == 'q': + break + if not length_input: # 如果直接回车，使用默认值 + knowledge_length = 512 + break + knowledge_length = int(length_input) + if knowledge_length <= 0: + print("分割长度必须大于0，请重新输入") + continue + break + except ValueError: + print("请输入有效的数字") + continue + + if length_input.lower() == 'q': + continue + + # 测试知识库功能 + print(f"开始处理知识库文件，使用分割长度: {knowledge_length}...") + knowledge_library.process_files(knowledge_length=knowledge_length) + else: + console.print("[red]无效的选项，请重新选择[/red]") + continue diff --git a/template/bot_config_template.toml b/template/bot_config_template.toml index bea6ab7b7..eb0323cec 100644 --- a/template/bot_config_template.toml +++ b/template/bot_config_template.toml @@ -1,5 +1,5 @@ [inner] -version = "0.0.6" +version = "0.0.7" #如果你想要修改配置文件，请在修改后将version的值进行变更 #如果新增项目，请在BotConfig类下新增相应的变量 @@ -101,6 +101,7 @@ word_replace_rate=0.006 # 整词替换概率 enable_advance_output = true # 是否启用高级输出 enable_kuuki_read = true # 是否启用读空气功能 enable_debug_output = false # 是否启用调试输出 +enable_friend_chat = false # 是否启用好友聊天 [groups] talk_allowed = [ diff --git a/如果你更新了版本，点我.txt b/如果你更新了版本，点我.txt new file mode 100644 index 000000000..400e8ae0c --- /dev/null +++ b/如果你更新了版本，点我.txt @@ -0,0 +1,4 @@ +更新版本后，建议删除数据库messages中所有内容，不然会出现报错 +该操作不会影响你的记忆 + +如果显示配置文件版本过低，运行根目录的bat \ No newline at end of file diff --git a/如果你的配置文件版本太老就点我.bat b/如果你的配置文件版本太老就点我.bat new file mode 100644 index 000000000..fec1f4cdb --- /dev/null +++ b/如果你的配置文件版本太老就点我.bat @@ -0,0 +1,45 @@ +@echo off +setlocal enabledelayedexpansion +chcp 65001 +cd /d %~dp0 + +echo ===================================== +echo 选择Python环境: +echo 1 - venv (推荐) +echo 2 - conda +echo ===================================== +choice /c 12 /n /m "输入数字(1或2): " + +if errorlevel 2 ( + echo ===================================== + set "CONDA_ENV=" + set /p CONDA_ENV="请输入要激活的 conda 环境名称: " + + :: 检查输入是否为空 + if "!CONDA_ENV!"=="" ( + echo 错误：环境名称不能为空 + pause + exit /b 1 + ) + + call conda activate !CONDA_ENV! + if errorlevel 1 ( + echo 激活 conda 环境失败 + pause + exit /b 1 + ) + + echo Conda 环境 "!CONDA_ENV!" 激活成功 + python config/auto_update.py +) else ( + if exist "venv\Scripts\python.exe" ( + venv\Scripts\python config/auto_update.py + ) else ( + echo ===================================== + echo 错误: venv环境不存在，请先创建虚拟环境 + pause + exit /b 1 + ) +) +endlocal +pause diff --git a/麦麦开始学习.bat b/麦麦开始学习.bat new file mode 100644 index 000000000..f7391150f --- /dev/null +++ b/麦麦开始学习.bat @@ -0,0 +1,45 @@ +@echo off +setlocal enabledelayedexpansion +chcp 65001 +cd /d %~dp0 + +echo ===================================== +echo 选择Python环境: +echo 1 - venv (推荐) +echo 2 - conda +echo ===================================== +choice /c 12 /n /m "输入数字(1或2): " + +if errorlevel 2 ( + echo ===================================== + set "CONDA_ENV=" + set /p CONDA_ENV="请输入要激活的 conda 环境名称: " + + :: 检查输入是否为空 + if "!CONDA_ENV!"=="" ( + echo 错误：环境名称不能为空 + pause + exit /b 1 + ) + + call conda activate !CONDA_ENV! + if errorlevel 1 ( + echo 激活 conda 环境失败 + pause + exit /b 1 + ) + + echo Conda 环境 "!CONDA_ENV!" 激活成功 + python src/plugins/zhishi/knowledge_library.py +) else ( + if exist "venv\Scripts\python.exe" ( + venv\Scripts\python src/plugins/zhishi/knowledge_library.py + ) else ( + echo ===================================== + echo 错误: venv环境不存在，请先创建虚拟环境 + pause + exit /b 1 + ) +) +endlocal +pause