From b88f200f7450e68a9c1275488e0866987dd28b97 Mon Sep 17 00:00:00 2001 From: SengokuCola <1026294844@qq.com> Date: Wed, 26 Feb 2025 22:18:44 +0800 Subject: [PATCH] =?UTF-8?q?=E6=8B=86=E5=88=860.1.3?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 1 --- src/plugins/chat/bot.py | 4 +- src/plugins/chat/bot_config.toml | 6 +- src/plugins/chat/gpt_response.py | 544 ----------------------------- src/plugins/chat/llm_generator.py | 383 ++++++++++++++++---- src/plugins/chat/prompt_builder.py | 14 +- src/plugins/chat/utils.py | 178 ++++++++++ 6 files changed, 512 insertions(+), 617 deletions(-) delete mode 100644 src/plugins/chat/gpt_response.py diff --git a/src/plugins/chat/bot.py b/src/plugins/chat/bot.py index 675a8c43a..48e967d7d 100644 --- a/src/plugins/chat/bot.py +++ b/src/plugins/chat/bot.py @@ -2,7 +2,7 @@ from nonebot.adapters.onebot.v11 import GroupMessageEvent, Message as EventMessa from .message import Message,MessageSet from .config import BotConfig, global_config from .storage import MessageStorage -from .gpt_response import GPTResponseGenerator +from .llm_generator import LLMResponseGenerator from .message_stream import MessageStream, MessageStreamContainer from .topic_identifier import topic_identifier from random import random @@ -23,7 +23,7 @@ class ChatBot: def __init__(self, config: BotConfig): self.config = config self.storage = MessageStorage() - self.gpt = GPTResponseGenerator(config) + self.gpt = LLMResponseGenerator(config) self.group_info_manager = GroupInfoManager() # 初始化群信息管理器 self.bot = None # bot 实例引用 self._started = False diff --git a/src/plugins/chat/bot_config.toml b/src/plugins/chat/bot_config.toml index 0945f34ae..da9e47246 100644 --- a/src/plugins/chat/bot_config.toml +++ b/src/plugins/chat/bot_config.toml @@ -39,7 +39,7 @@ read_allowed = [ 752426484,#nd1 115843978,#nd2 # 168718420 #bh - 912378791 + # 912378791 ] talk_allowed = [ @@ -60,14 +60,14 @@ talk_allowed = [ # 168718420#bh # 752426484,#nd1 # 115843978,#nd2 - 912378791 + # 912378791 ] talk_frequency_down = [ 549292720, #mrfz 435591861, #m43black # 231561425, - 975992476, + # 975992476, 1140700103, # 534940728 # 752426484,#nd1 diff --git a/src/plugins/chat/gpt_response.py b/src/plugins/chat/gpt_response.py deleted file mode 100644 index 7f1f62dda..000000000 --- a/src/plugins/chat/gpt_response.py +++ /dev/null @@ -1,544 +0,0 @@ -from typing import Dict, Any, List, Optional, Union, Tuple -from openai import OpenAI -import asyncio -import requests -from functools import partial -from .message import Message -from .config import BotConfig -from ...common.database import Database -import random -import time -import subprocess -import os -import sys -import threading -import queue -import numpy as np -from dotenv import load_dotenv -from .relationship_manager import relationship_manager -from ..schedule.schedule_generator import bot_schedule -from .prompt_builder import prompt_builder -from .config import llm_config -from .willing_manager import willing_manager -from .utils import get_embedding -import aiohttp - - -# 获取当前文件的绝对路径 -current_dir = os.path.dirname(os.path.abspath(__file__)) -root_dir = os.path.abspath(os.path.join(current_dir, '..', '..', '..')) -load_dotenv(os.path.join(root_dir, '.env')) - -# 常见的错别字映射 -TYPO_DICT = { - '的': '地得', - '了': '咯啦勒', - '吗': '嘛麻', - '吧': '八把罢', - '是': '事', - '在': '再在', - '和': '合', - '有': '又', - '我': '沃窝喔', - '你': '泥尼拟', - '他': '它她塔祂', - '们': '门', - '啊': '阿哇', - '呢': '呐捏', - '都': '豆读毒', - '很': '狠', - '会': '回汇', - '去': '趣取曲', - '做': '作坐', - '想': '相像', - '说': '说税睡', - '看': '砍堪刊', - '来': '来莱赖', - '好': '号毫豪', - '给': '给既继', - '过': '锅果裹', - '能': '嫩', - '为': '位未', - '什': '甚深伸', - '么': '末麽嘛', - '话': '话花划', - '知': '织直值', - '道': '到', - '听': '听停挺', - '见': '见件建', - '觉': '觉脚搅', - '得': '得德锝', - '着': '着找招', - '像': '向象想', - '等': '等灯登', - '谢': '谢写卸', - '对': '对队', - '里': '里理鲤', - '啦': '啦拉喇', - '吃': '吃持迟', - '哦': '哦喔噢', - '呀': '呀压', - '要': '药', - '太': '太抬台', - '快': '块', - '点': '店', - '以': '以已', - '因': '因应', - '啥': '啥沙傻', - '行': '行型形', - '哈': '哈蛤铪', - '嘿': '嘿黑嗨', - '嗯': '嗯恩摁', - '哎': '哎爱埃', - '呜': '呜屋污', - '喂': '喂位未', - '嘛': '嘛麻马', - '嗨': '嗨害亥', - '哇': '哇娃蛙', - '咦': '咦意易', - '嘻': '嘻西希' -} - -def random_remove_punctuation(text: str) -> str: - """随机处理标点符号，模拟人类打字习惯""" - result = '' - text_len = len(text) - - for i, char in enumerate(text): - if char == '。' and i == text_len - 1: # 结尾的句号 - if random.random() > 0.4: # 80%概率删除结尾句号 - continue - elif char == '，': - rand = random.random() - if rand < 0.25: # 5%概率删除逗号 - continue - elif rand < 0.25: # 20%概率把逗号变成空格 - result += ' ' - continue - result += char - return result - -def add_typos(text: str) -> str: - """随机给文本添加错别字""" - TYPO_RATE = 0.02 # 控制错别字出现的概率(1%) - - result = "" - for char in text: - if char in TYPO_DICT and random.random() < TYPO_RATE: - # 从可能的错别字中随机选择一个 - typos = TYPO_DICT[char] - result += random.choice(typos) - else: - result += char - return result - -def open_new_console_window(text: str): - """在新的控制台窗口中显示文本""" - if sys.platform == 'win32': - # 创建一个临时批处理文件 - temp_bat = "temp_output.bat" - with open(temp_bat, "w", encoding="utf-8") as f: - f.write(f'@echo off\n') - f.write(f'echo {text}\n') - f.write('pause\n') - - # 在新窗口中运行批处理文件 - subprocess.Popen(['start', 'cmd', '/c', temp_bat], shell=True) - - # 等待一会儿再删除批处理文件 - import threading - def delete_bat(): - import time - time.sleep(2) - if os.path.exists(temp_bat): - os.remove(temp_bat) - threading.Thread(target=delete_bat).start() - -class ReasoningWindow: - def __init__(self): - self.process = None - self.message_queue = queue.Queue() - self.is_running = False - self.content_file = "reasoning_content.txt" - - def start(self): - if self.process is None: - # 创建用于显示的批处理文件 - with open("reasoning_window.bat", "w", encoding="utf-8") as f: - f.write('@echo off\n') - f.write('chcp 65001\n') # 设置UTF-8编码 - f.write('title Magellan Reasoning Process\n') - f.write('echo Waiting for reasoning content...\n') - f.write(':loop\n') - f.write('if exist "reasoning_update.txt" (\n') - f.write(' type "reasoning_update.txt" >> "reasoning_content.txt"\n') - f.write(' del "reasoning_update.txt"\n') - f.write(' cls\n') - f.write(' type "reasoning_content.txt"\n') - f.write(')\n') - f.write('timeout /t 1 /nobreak >nul\n') - f.write('goto loop\n') - - # 清空内容文件 - with open(self.content_file, "w", encoding="utf-8") as f: - f.write("") - - # 启动新窗口 - startupinfo = subprocess.STARTUPINFO() - startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW - self.process = subprocess.Popen(['cmd', '/c', 'start', 'reasoning_window.bat'], - shell=True, - startupinfo=startupinfo) - self.is_running = True - - # 启动处理线程 - threading.Thread(target=self._process_messages, daemon=True).start() - - def _process_messages(self): - while self.is_running: - try: - # 获取新消息 - text = self.message_queue.get(timeout=1) - # 写入更新文件 - with open("reasoning_update.txt", "w", encoding="utf-8") as f: - f.write(text) - except queue.Empty: - continue - except Exception as e: - print(f"处理推理内容时出错: {e}") - - def update_content(self, text: str): - if self.is_running: - self.message_queue.put(text) - - def stop(self): - self.is_running = False - if self.process: - self.process.terminate() - self.process = None - # 清理文件 - for file in ["reasoning_window.bat", "reasoning_content.txt", "reasoning_update.txt"]: - if os.path.exists(file): - os.remove(file) - -# 创建全局单例 -reasoning_window = ReasoningWindow() - -class GPTResponseGenerator: - def __init__(self, config: BotConfig): - self.config = config - self.client = OpenAI( - api_key=llm_config.SILICONFLOW_API_KEY, - base_url=llm_config.SILICONFLOW_BASE_URL - ) - - self.db = Database.get_instance() - reasoning_window.start() - # 当前使用的模型类型 - self.current_model_type = 'r1' # 默认使用 R1 - - async def generate_response(self, message: Message) -> Optional[Union[str, List[str]]]: - """根据当前模型类型选择对应的生成函数""" - # 使用随机数选择模型 - rand = random.random() - if rand < 0.6: # 40%概率使用 R1 - self.current_model_type = "r1" - elif rand < 0.5: # 30%概率使用 V3 - self.current_model_type = "v3" - else: # 30%概率使用 R1-Distill - self.current_model_type = "r1_distill" - - print(f"+++++++++++++++++麦麦{self.current_model_type}思考中+++++++++++++++++") - if self.current_model_type == 'r1': - model_response = await self._generate_r1_response(message) - elif self.current_model_type == 'v3': - model_response = await self._generate_v3_response(message) - else: - model_response = await self._generate_r1_distill_response(message) - - # 打印情感标签 - print(f'麦麦的回复是：{model_response}') - model_response , emotion = await self._process_response(model_response) - - if model_response: - print(f"为 '{model_response}' 获取到的情感标签为：{emotion}") - - return model_response,emotion - - async def _generate_r1_response(self, message: Message) -> Optional[Tuple[Union[str, List[str]], List[str]]]: - """使用 DeepSeek-R1 模型生成回复""" - # 获取群聊上下文 - group_chat = await self._get_group_chat_context(message) - sender_name = message.user_nickname or f"用户{message.user_id}" - if relationship_manager.get_relationship(message.user_id): - relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value - print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") - else: - relationship_value = 0.0 - - # 构建 prompt - prompt = prompt_builder._build_prompt( - message_txt=message.processed_plain_text, - sender_name=sender_name, - relationship_value=relationship_value, - group_id=message.group_id - ) - - def create_completion(): - return self.client.chat.completions.create( - model="Pro/deepseek-ai/DeepSeek-R1", - messages=[{"role": "user", "content": prompt}], - stream=False, - max_tokens=1024 - ) - - loop = asyncio.get_event_loop() - response = await loop.run_in_executor(None, create_completion) - if response.choices[0].message.content: - # print(response.choices[0].message.content) - # print(response.choices[0].message.reasoning_content) - # 处理 R1 特有的返回格式 - content = response.choices[0].message.content - reasoning_content = response.choices[0].message.reasoning_content - else: - return None - # 更新推理窗口 - self._update_reasoning_window(message, prompt, reasoning_content, content, sender_name) - - return content - - async def _generate_v3_response(self, message: Message) -> Optional[Tuple[Union[str, List[str]], List[str]]]: - """使用 DeepSeek-V3 模型生成回复""" - # 获取群聊上下文 - group_chat = await self._get_group_chat_context(message) - sender_name = message.user_nickname or f"用户{message.user_id}" - - if relationship_manager.get_relationship(message.user_id): - relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value - print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") - else: - relationship_value = 0.0 - - prompt = prompt_builder._build_prompt(message.processed_plain_text, sender_name,relationship_value,group_id=message.group_id) - - messages = [{"role": "user", "content": prompt}] - - loop = asyncio.get_event_loop() - create_completion = partial( - self.client.chat.completions.create, - model="Pro/deepseek-ai/DeepSeek-V3", - messages=messages, - stream=False, - max_tokens=1024, - temperature=0.8 - ) - response = await loop.run_in_executor(None, create_completion) - - if response.choices[0].message.content: - content = response.choices[0].message.content - # V3 模型没有 reasoning_content - self._update_reasoning_window(message, prompt, "V3模型无推理过程", content, sender_name) - return content - else: - print(f"[ERROR] V3 回复发送生成失败: {response}") - - return None, [] # 返回元组 - - async def _generate_r1_distill_response(self, message: Message) -> Optional[Tuple[Union[str, List[str]], List[str]]]: - """使用 DeepSeek-R1-Distill-Qwen-32B 模型生成回复""" - # 获取群聊上下文 - group_chat = await self._get_group_chat_context(message) - sender_name = message.user_nickname or f"用户{message.user_id}" - if relationship_manager.get_relationship(message.user_id): - relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value - print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") - else: - relationship_value = 0.0 - - # 构建 prompt - prompt = prompt_builder._build_prompt( - message_txt=message.processed_plain_text, - sender_name=sender_name, - relationship_value=relationship_value, - group_id=message.group_id - ) - - def create_completion(): - return self.client.chat.completions.create( - model="deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - messages=[{"role": "user", "content": prompt}], - stream=False, - max_tokens=1024 - ) - - loop = asyncio.get_event_loop() - response = await loop.run_in_executor(None, create_completion) - if response.choices[0].message.content: - # print(response.choices[0].message.content) - # print(response.choices[0].message.reasoning_content) - # 处理 R1 特有的返回格式 - content = response.choices[0].message.content - reasoning_content = response.choices[0].message.reasoning_content - else: - return None - # 更新推理窗口 - self._update_reasoning_window(message, prompt, reasoning_content, content, sender_name) - - return content - - async def _get_group_chat_context(self, message: Message) -> str: - """获取群聊上下文""" - recent_messages = self.db.db.messages.find( - {"group_id": message.group_id} - ).sort("time", -1).limit(15) - - messages_list = list(recent_messages)[::-1] - group_chat = "" - - for msg_dict in messages_list: - time_str = time.strftime("%m-%d %H:%M:%S", time.localtime(msg_dict['time'])) - display_name = msg_dict.get('user_nickname', f"用户{msg_dict['user_id']}") - content = msg_dict.get('processed_plain_text', msg_dict['plain_text']) - - group_chat += f"[{time_str}] {display_name}: {content}\n" - - return group_chat - - def _update_reasoning_window(self, message, prompt, reasoning_content, content, sender_name): - """更新推理窗口内容""" - current_time = time.strftime("%Y-%m-%d %H:%M:%S") - - # 获取当前使用的模型名称 - model_name = { - 'r1': 'DeepSeek-R1', - 'v3': 'DeepSeek-V3', - 'r1_distill': 'DeepSeek-R1-Distill-Qwen-32B' - }.get(self.current_model_type, '未知模型') - - display_text = ( - f"Time: {current_time}\n" - f"Group: {message.group_name}\n" - f"User: {sender_name}\n" - f"Model: {model_name}\n" - f"\033[1;32mMessage:\033[0m {message.processed_plain_text}\n\n" - f"\033[1;32mPrompt:\033[0m \n{prompt}\n" - f"\n-------------------------------------------------------" - f"\n\033[1;32mReasoning Process:\033[0m\n{reasoning_content}\n" - f"\n\033[1;32mResponse Content:\033[0m\n{content}\n" - f"\n{'='*50}\n" - ) - reasoning_window.update_content(display_text) - - async def _get_emotion_tags(self, content: str) -> List[str]: - """提取情感标签""" - try: - prompt = f'''请从以下内容中，从"happy,angry,sad,surprised,disgusted,fearful,neutral"中选出最匹配的1个情感标签并输出 - 只输出标签就好，不要输出其他内容: - 内容：{content} - 输出： - ''' - - messages = [{"role": "user", "content": prompt}] - - loop = asyncio.get_event_loop() - create_completion = partial( - self.client.chat.completions.create, - model="Pro/deepseek-ai/DeepSeek-V3", - messages=messages, - stream=False, - max_tokens=30, - temperature=0.6 - ) - response = await loop.run_in_executor(None, create_completion) - - if response.choices[0].message.content: - # 确保返回的是列表格式 - emotion_tag = response.choices[0].message.content.strip() - return [emotion_tag] # 将单个标签包装成列表返回 - - return ["neutral"] # 如果无法获取情感标签，返回默认值 - - except Exception as e: - print(f"获取情感标签时出错: {e}") - return ["neutral"] # 发生错误时返回默认值 - - async def _process_response(self, content: str) -> Tuple[Union[str, List[str]], List[str]]: - """处理响应内容，返回处理后的内容和情感标签""" - if not content: - return None, [] - - emotion_tags = await self._get_emotion_tags(content) - - # 添加错别字和处理标点符号 - if random.random() < 0.9: # 90%概率进行处理 - processed_response = random_remove_punctuation(add_typos(content)) - else: - processed_response = content - # 处理长消息 - if len(processed_response) > 5: - sentences = self._split_into_sentences(processed_response) - print(f"分割后的句子: {sentences}") - messages = [] - current_message = "" - - for sentence in sentences: - if len(current_message) + len(sentence) <= 5: - current_message += ' ' - current_message += sentence - else: - if current_message: - messages.append(current_message.strip()) - current_message = sentence - - if current_message: - messages.append(current_message.strip()) - - # 翻转消息顺序 - # messages.reverse() - - return messages, emotion_tags - - return processed_response, emotion_tags - - def _split_into_sentences(self, text: str) -> List[str]: - """将文本分割成句子，但保持书名号中的内容完整""" - delimiters = ['。', '！', '，', ',', '？', '…', '!', '?', '\n'] # 添加换行符作为分隔符 - remove_chars = ['，', ','] # 只移除这两种逗号 - sentences = [] - current_sentence = "" - in_book_title = False # 标记是否在书名号内 - - for char in text: - current_sentence += char - - # 检查书名号 - if char == '《': - in_book_title = True - elif char == '》': - in_book_title = False - - # 只有不在书名号内且是分隔符时才分割 - if char in delimiters and not in_book_title: - if current_sentence.strip(): # 确保不是空字符串 - # 只移除逗号 - clean_sentence = current_sentence - if clean_sentence[-1] in remove_chars: - clean_sentence = clean_sentence[:-1] - if clean_sentence.strip(): - sentences.append(clean_sentence.strip()) - current_sentence = "" - - # 处理最后一个句子 - if current_sentence.strip(): - # 如果最后一个字符是逗号，移除它 - if current_sentence[-1] in remove_chars: - current_sentence = current_sentence[:-1] - sentences.append(current_sentence.strip()) - - # 过滤掉空字符串 - sentences = [s for s in sentences if s.strip()] - - return sentences - - -# llm_response = GPTResponseGenerator(config=BotConfig()) \ No newline at end of file diff --git a/src/plugins/chat/llm_generator.py b/src/plugins/chat/llm_generator.py index bf7f1b644..5ab9d1e46 100644 --- a/src/plugins/chat/llm_generator.py +++ b/src/plugins/chat/llm_generator.py @@ -1,108 +1,367 @@ from typing import Dict, Any, List, Optional, Union, Tuple from openai import OpenAI +import asyncio +import requests from functools import partial +from .message import Message from .config import BotConfig from ...common.database import Database import random +import time +import subprocess import os -import aiohttp +import sys +import threading +import queue +import numpy as np from dotenv import load_dotenv from .relationship_manager import relationship_manager +from ..schedule.schedule_generator import bot_schedule +from .prompt_builder import prompt_builder +from .config import llm_config +from .willing_manager import willing_manager +from .utils import get_embedding, split_into_sentences, process_text_with_typos +import aiohttp # 获取当前文件的绝对路径 current_dir = os.path.dirname(os.path.abspath(__file__)) root_dir = os.path.abspath(os.path.join(current_dir, '..', '..', '..')) load_dotenv(os.path.join(root_dir, '.env')) +class ReasoningWindow: + def __init__(self): + self.process = None + self.message_queue = queue.Queue() + self.is_running = False + self.content_file = "reasoning_content.txt" + + def start(self): + if self.process is None: + # 创建用于显示的批处理文件 + with open("reasoning_window.bat", "w", encoding="utf-8") as f: + f.write('@echo off\n') + f.write('chcp 65001\n') # 设置UTF-8编码 + f.write('title Magellan Reasoning Process\n') + f.write('echo Waiting for reasoning content...\n') + f.write(':loop\n') + f.write('if exist "reasoning_update.txt" (\n') + f.write(' type "reasoning_update.txt" >> "reasoning_content.txt"\n') + f.write(' del "reasoning_update.txt"\n') + f.write(' cls\n') + f.write(' type "reasoning_content.txt"\n') + f.write(')\n') + f.write('timeout /t 1 /nobreak >nul\n') + f.write('goto loop\n') + + # 清空内容文件 + with open(self.content_file, "w", encoding="utf-8") as f: + f.write("") + + # 启动新窗口 + startupinfo = subprocess.STARTUPINFO() + startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW + self.process = subprocess.Popen(['cmd', '/c', 'start', 'reasoning_window.bat'], + shell=True, + startupinfo=startupinfo) + self.is_running = True + + # 启动处理线程 + threading.Thread(target=self._process_messages, daemon=True).start() + + def _process_messages(self): + while self.is_running: + try: + # 获取新消息 + text = self.message_queue.get(timeout=1) + # 写入更新文件 + with open("reasoning_update.txt", "w", encoding="utf-8") as f: + f.write(text) + except queue.Empty: + continue + except Exception as e: + print(f"处理推理内容时出错: {e}") + + def update_content(self, text: str): + if self.is_running: + self.message_queue.put(text) + + def stop(self): + self.is_running = False + if self.process: + self.process.terminate() + self.process = None + # 清理文件 + for file in ["reasoning_window.bat", "reasoning_content.txt", "reasoning_update.txt"]: + if os.path.exists(file): + os.remove(file) +# 创建全局单例 +reasoning_window = ReasoningWindow() class LLMResponseGenerator: def __init__(self, config: BotConfig): self.config = config - self.API_KEY = os.getenv('SILICONFLOW_KEY') - self.BASE_URL =os.getenv('SILICONFLOW_BASE_URL') self.client = OpenAI( - api_key=self.API_KEY, - base_url=self.BASE_URL + api_key=llm_config.SILICONFLOW_API_KEY, + base_url=llm_config.SILICONFLOW_BASE_URL ) self.db = Database.get_instance() + reasoning_window.start() # 当前使用的模型类型 self.current_model_type = 'r1' # 默认使用 R1 - async def generate_response(self, text: str) -> Optional[str]: + async def generate_response(self, message: Message) -> Optional[Union[str, List[str]]]: """根据当前模型类型选择对应的生成函数""" - if random.random() < self.config.MODEL_R1_PROBABILITY: + # 使用随机数选择模型 + rand = random.random() + if rand < 0.6: # 60%概率使用 R1 self.current_model_type = "r1" - else: + elif rand < 0.5: # 20%概率使用 V3 self.current_model_type = "v3" + else: # 20%概率使用 R1-Distill + self.current_model_type = "r1_distill" print(f"+++++++++++++++++麦麦{self.current_model_type}思考中+++++++++++++++++") if self.current_model_type == 'r1': - model_response = await self._generate_v3_response(text) + model_response = await self._generate_r1_response(message) + elif self.current_model_type == 'v3': + model_response = await self._generate_v3_response(message) else: - model_response = await self._generate_v3_response(text) + model_response = await self._generate_r1_distill_response(message) + # 打印情感标签 - print(f'麦麦的回复------------------------------是：{model_response}') + print(f'麦麦的回复是：{model_response}') + model_response, emotion = await self._process_response(model_response) - return model_response + if model_response: + print(f"为 '{model_response}' 获取到的情感标签为：{emotion}") + + return model_response, emotion - async def _generate_r1_response(self, text: str) -> Optional[str]: + async def _generate_r1_response(self, message: Message) -> Optional[str]: """使用 DeepSeek-R1 模型生成回复""" - messages = [{"role": "user", "content": text}] - async with aiohttp.ClientSession() as session: - headers = { - "Content-Type": "application/json", - "Authorization": f"Bearer {self.API_KEY}" - } - payload = { - "model": "Pro/deepseek-ai/DeepSeek-R1", - "messages": messages, - "stream": False, - "max_tokens": 1024, - "temperature": 0.8 - } - async with session.post(f"{self.BASE_URL}/chat/completions", - headers=headers, - json=payload) as response: - result = await response.json() - if "choices" in result and len(result["choices"]) > 0: - content = result["choices"][0]["message"]["content"] - reasoning_content = result["choices"][0]["message"].get("reasoning_content", "") - print(f"Content: {content}") - print(f"Reasoning: {reasoning_content}") - return content - - return None + # 获取群聊上下文 + group_chat = await self._get_group_chat_context(message) + sender_name = message.user_nickname or f"用户{message.user_id}" + if relationship_manager.get_relationship(message.user_id): + relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value + print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") + else: + relationship_value = 0.0 + + # 构建 prompt + prompt = prompt_builder._build_prompt( + message_txt=message.processed_plain_text, + sender_name=sender_name, + relationship_value=relationship_value, + group_id=message.group_id + ) + + def create_completion(): + return self.client.chat.completions.create( + model="Pro/deepseek-ai/DeepSeek-R1", + messages=[{"role": "user", "content": prompt}], + stream=False, + max_tokens=1024 + ) + + loop = asyncio.get_event_loop() + response = await loop.run_in_executor(None, create_completion) + if response.choices[0].message.content: + content = response.choices[0].message.content + reasoning_content = response.choices[0].message.reasoning_content + else: + return None + # 更新推理窗口 + self._update_reasoning_window(message, prompt, reasoning_content, content, sender_name) + + return content - async def _generate_v3_response(self, text: str) -> Optional[str]: + async def _generate_v3_response(self, message: Message) -> Optional[str]: """使用 DeepSeek-V3 模型生成回复""" - - headers = { - "Content-Type": "application/json", - "Authorization": f"Bearer {self.API_KEY}" - } + # 获取群聊上下文 + group_chat = await self._get_group_chat_context(message) + sender_name = message.user_nickname or f"用户{message.user_id}" - payload = { - "model": "Pro/deepseek-ai/DeepSeek-V3", - "messages": [{"role": "user", "content": text}], - "max_tokens": 1024, - "temperature": 0.8 - } + if relationship_manager.get_relationship(message.user_id): + relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value + print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") + else: + relationship_value = 0.0 - async with aiohttp.ClientSession() as session: - async with session.post(f"{self.BASE_URL}/chat/completions", - headers=headers, - json=payload) as response: - result = await response.json() - - if "choices" in result and len(result["choices"]) > 0: - content = result["choices"][0]["message"]["content"] - return content - else: - print(f"[ERROR] V3 回复发送生成失败: {result}") - + prompt = prompt_builder._build_prompt(message.processed_plain_text, sender_name, relationship_value, group_id=message.group_id) + + messages = [{"role": "user", "content": prompt}] + + loop = asyncio.get_event_loop() + create_completion = partial( + self.client.chat.completions.create, + model="Pro/deepseek-ai/DeepSeek-V3", + messages=messages, + stream=False, + max_tokens=1024, + temperature=0.8 + ) + response = await loop.run_in_executor(None, create_completion) + + if response.choices[0].message.content: + content = response.choices[0].message.content + # V3 模型没有 reasoning_content + self._update_reasoning_window(message, prompt, "V3模型无推理过程", content, sender_name) + return content + else: + print(f"[ERROR] V3 回复发送生成失败: {response}") + return None + async def _generate_r1_distill_response(self, message: Message) -> Optional[str]: + """使用 DeepSeek-R1-Distill-Qwen-32B 模型生成回复""" + # 获取群聊上下文 + group_chat = await self._get_group_chat_context(message) + sender_name = message.user_nickname or f"用户{message.user_id}" + if relationship_manager.get_relationship(message.user_id): + relationship_value = relationship_manager.get_relationship(message.user_id).relationship_value + print(f"\033[1;32m[关系管理]\033[0m 回复中_当前关系值: {relationship_value}") + else: + relationship_value = 0.0 + + # 构建 prompt + prompt = prompt_builder._build_prompt( + message_txt=message.processed_plain_text, + sender_name=sender_name, + relationship_value=relationship_value, + group_id=message.group_id + ) + + def create_completion(): + return self.client.chat.completions.create( + model="deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + messages=[{"role": "user", "content": prompt}], + stream=False, + max_tokens=1024 + ) + + loop = asyncio.get_event_loop() + response = await loop.run_in_executor(None, create_completion) + if response.choices[0].message.content: + content = response.choices[0].message.content + reasoning_content = response.choices[0].message.reasoning_content + else: + return None + # 更新推理窗口 + self._update_reasoning_window(message, prompt, reasoning_content, content, sender_name) + + return content + + async def _get_group_chat_context(self, message: Message) -> str: + """获取群聊上下文""" + recent_messages = self.db.db.messages.find( + {"group_id": message.group_id} + ).sort("time", -1).limit(15) + + messages_list = list(recent_messages)[::-1] + group_chat = "" + + for msg_dict in messages_list: + time_str = time.strftime("%m-%d %H:%M:%S", time.localtime(msg_dict['time'])) + display_name = msg_dict.get('user_nickname', f"用户{msg_dict['user_id']}") + content = msg_dict.get('processed_plain_text', msg_dict['plain_text']) + + group_chat += f"[{time_str}] {display_name}: {content}\n" + + return group_chat + + def _update_reasoning_window(self, message, prompt, reasoning_content, content, sender_name): + """更新推理窗口内容""" + current_time = time.strftime("%Y-%m-%d %H:%M:%S") + + # 获取当前使用的模型名称 + model_name = { + 'r1': 'DeepSeek-R1', + 'v3': 'DeepSeek-V3', + 'r1_distill': 'DeepSeek-R1-Distill-Qwen-32B' + }.get(self.current_model_type, '未知模型') + + display_text = ( + f"Time: {current_time}\n" + f"Group: {message.group_name}\n" + f"User: {sender_name}\n" + f"Model: {model_name}\n" + f"\033[1;32mMessage:\033[0m {message.processed_plain_text}\n\n" + f"\033[1;32mPrompt:\033[0m \n{prompt}\n" + f"\n-------------------------------------------------------" + f"\n\033[1;32mReasoning Process:\033[0m\n{reasoning_content}\n" + f"\n\033[1;32mResponse Content:\033[0m\n{content}\n" + f"\n{'='*50}\n" + ) + reasoning_window.update_content(display_text) + + async def _get_emotion_tags(self, content: str) -> List[str]: + """提取情感标签""" + try: + prompt = f'''请从以下内容中，从"happy,angry,sad,surprised,disgusted,fearful,neutral"中选出最匹配的1个情感标签并输出 + 只输出标签就好，不要输出其他内容: + 内容：{content} + 输出： + ''' + + messages = [{"role": "user", "content": prompt}] + + loop = asyncio.get_event_loop() + create_completion = partial( + self.client.chat.completions.create, + model="Pro/deepseek-ai/DeepSeek-V3", + messages=messages, + stream=False, + max_tokens=30, + temperature=0.6 + ) + response = await loop.run_in_executor(None, create_completion) + + if response.choices[0].message.content: + # 确保返回的是列表格式 + emotion_tag = response.choices[0].message.content.strip() + return [emotion_tag] # 将单个标签包装成列表返回 + + return ["neutral"] # 如果无法获取情感标签，返回默认值 + + except Exception as e: + print(f"获取情感标签时出错: {e}") + return ["neutral"] # 发生错误时返回默认值 + async def _process_response(self, content: str) -> Tuple[Union[str, List[str]], List[str]]: + """处理响应内容，返回处理后的内容和情感标签""" + if not content: + return None, [] + + emotion_tags = await self._get_emotion_tags(content) + + # 添加错别字和处理标点符号 + processed_response = process_text_with_typos(content) + + # 处理长消息 + if len(processed_response) > 5: + sentences = split_into_sentences(processed_response) + print(f"分割后的句子: {sentences}") + messages = [] + current_message = "" + + for sentence in sentences: + if len(current_message) + len(sentence) <= 5: + current_message += ' ' + current_message += sentence + else: + if current_message: + messages.append(current_message.strip()) + current_message = sentence + + if current_message: + messages.append(current_message.strip()) + + return messages, emotion_tags + + return processed_response, emotion_tags + +# 创建全局实例 llm_response = LLMResponseGenerator(config=BotConfig()) \ No newline at end of file diff --git a/src/plugins/chat/prompt_builder.py b/src/plugins/chat/prompt_builder.py index 90dee41df..c0f3a8aed 100644 --- a/src/plugins/chat/prompt_builder.py +++ b/src/plugins/chat/prompt_builder.py @@ -51,7 +51,7 @@ class PromptBuilder: current_date = time.strftime("%Y-%m-%d", time.localtime()) current_time = time.strftime("%H:%M:%S", time.localtime()) bot_schedule_now_time,bot_schedule_now_activity = bot_schedule.get_current_task() - prompt_date = f'''今天是{current_date}，现在是{current_time}，你今天的日程是：\n{bot_schedule.today_schedule}\n''' + prompt_date = f'''今天是{current_date}，现在是{current_time}，你今天的日程是：\n{bot_schedule.today_schedule}\n你现在正在{bot_schedule_now_activity}\n''' #知识构建 prompt_info = '' @@ -74,7 +74,7 @@ class PromptBuilder: #激活prompt构建 activate_prompt = '' - activate_prompt = f"以上是群里正在进行的聊天，昵称为 '{sender_name}' 的用户说的:{message_txt}。引起了你的注意,你和他{relation_prompt}，你想要{relation_prompt_2} \n" + activate_prompt = f"以上是群里正在进行的聊天，昵称为 '{sender_name}' 的用户说的:{message_txt}。引起了你的注意,你和他{relation_prompt}，你想要{relation_prompt_2}。" # 检测机器人相关词汇 bot_keywords = ['人机', 'bot', '机器', '入机', 'robot', '机器人'] @@ -88,11 +88,13 @@ class PromptBuilder: prompt_personality = '' personality_choice = random.random() if personality_choice < 4/6: # 第一种人格 - prompt_personality = f'''你的网名叫麦麦，是一个学习地质的女大学生，喜欢摄影，你会刷贴吧，你现在正在{bot_schedule_now_activity},并且正在浏览qq群,{promt_info_prompt} + prompt_personality = f'''你的网名叫麦麦，是一个学习地质的女大学生，喜欢摄影，你会刷贴吧，你正在浏览qq群,{promt_info_prompt}, + {activate_prompt} 现在请你给出日常且口语化的回复，平淡一些，尽量简短一些,不要输出感叹号。{is_bot_prompt} - 不要回复的太有逻辑或者太有条理，可以有个性。''' + 请注意把握群里的聊天内容，不要回复的太有条理，可以有个性。''' elif personality_choice < 1: # 第二种人格 - prompt_personality = f'''你的网名叫麦麦，是一个女大学生，你有黑色头发，你会刷小红书，你现在正在{bot_schedule_now_activity},并且正在浏览qq群,{promt_info_prompt} + prompt_personality = f'''你的网名叫麦麦，是一个女大学生，你有黑色头发，你会刷小红书, 你正在浏览qq群,{promt_info_prompt}, + {activate_prompt} 现在请你给出日常且口语化的回复，请表现你自己的见解，不要一昧迎合，尽量简短一些,不要输出感叹号。{is_bot_prompt} 请你表达自己的见解和观点。可以有个性。''' @@ -104,7 +106,7 @@ class PromptBuilder: # prompt += f"{prompt_info}\n" prompt += f"{prompt_date}\n" prompt += f"{chat_talking_prompt}\n" - prompt += f"{activate_prompt}\n" + # prompt += f"{activate_prompt}\n" prompt += f"{prompt_personality}\n" prompt += f"{extra_info}\n" diff --git a/src/plugins/chat/utils.py b/src/plugins/chat/utils.py index 9d3abd0f7..96e842114 100644 --- a/src/plugins/chat/utils.py +++ b/src/plugins/chat/utils.py @@ -1,4 +1,5 @@ import time +import random from typing import List from .message import Message import requests @@ -113,3 +114,180 @@ def get_recent_group_messages(db, group_id: int, limit: int = 12) -> list: # 按时间正序排列 message_objects.reverse() return message_objects + +def split_into_sentences(text: str) -> List[str]: + """将文本分割成句子，但保持书名号中的内容完整 + + Args: + text: 要分割的文本字符串 + + Returns: + List[str]: 分割后的句子列表 + """ + delimiters = ['。', '！', '，', ',', '？', '…', '!', '?', '\n'] # 添加换行符作为分隔符 + remove_chars = ['，', ','] # 只移除这两种逗号 + sentences = [] + current_sentence = "" + in_book_title = False # 标记是否在书名号内 + + for char in text: + current_sentence += char + + # 检查书名号 + if char == '《': + in_book_title = True + elif char == '》': + in_book_title = False + + # 只有不在书名号内且是分隔符时才分割 + if char in delimiters and not in_book_title: + if current_sentence.strip(): # 确保不是空字符串 + # 只移除逗号 + clean_sentence = current_sentence + if clean_sentence[-1] in remove_chars: + clean_sentence = clean_sentence[:-1] + if clean_sentence.strip(): + sentences.append(clean_sentence.strip()) + current_sentence = "" + + # 处理最后一个句子 + if current_sentence.strip(): + # 如果最后一个字符是逗号，移除它 + if current_sentence[-1] in remove_chars: + current_sentence = current_sentence[:-1] + sentences.append(current_sentence.strip()) + + # 过滤掉空字符串 + sentences = [s for s in sentences if s.strip()] + + return sentences + +# 常见的错别字映射 +TYPO_DICT = { + '的': '地得', + '了': '咯啦勒', + '吗': '嘛麻', + '吧': '八把罢', + '是': '事', + '在': '再在', + '和': '合', + '有': '又', + '我': '沃窝喔', + '你': '泥尼拟', + '他': '它她塔祂', + '们': '门', + '啊': '阿哇', + '呢': '呐捏', + '都': '豆读毒', + '很': '狠', + '会': '回汇', + '去': '趣取曲', + '做': '作坐', + '想': '相像', + '说': '说税睡', + '看': '砍堪刊', + '来': '来莱赖', + '好': '号毫豪', + '给': '给既继', + '过': '锅果裹', + '能': '嫩', + '为': '位未', + '什': '甚深伸', + '么': '末麽嘛', + '话': '话花划', + '知': '织直值', + '道': '到', + '听': '听停挺', + '见': '见件建', + '觉': '觉脚搅', + '得': '得德锝', + '着': '着找招', + '像': '向象想', + '等': '等灯登', + '谢': '谢写卸', + '对': '对队', + '里': '里理鲤', + '啦': '啦拉喇', + '吃': '吃持迟', + '哦': '哦喔噢', + '呀': '呀压', + '要': '药', + '太': '太抬台', + '快': '块', + '点': '店', + '以': '以已', + '因': '因应', + '啥': '啥沙傻', + '行': '行型形', + '哈': '哈蛤铪', + '嘿': '嘿黑嗨', + '嗯': '嗯恩摁', + '哎': '哎爱埃', + '呜': '呜屋污', + '喂': '喂位未', + '嘛': '嘛麻马', + '嗨': '嗨害亥', + '哇': '哇娃蛙', + '咦': '咦意易', + '嘻': '嘻西希' +} + +def random_remove_punctuation(text: str) -> str: + """随机处理标点符号，模拟人类打字习惯 + + Args: + text: 要处理的文本 + + Returns: + str: 处理后的文本 + """ + result = '' + text_len = len(text) + + for i, char in enumerate(text): + if char == '。' and i == text_len - 1: # 结尾的句号 + if random.random() > 0.4: # 80%概率删除结尾句号 + continue + elif char == '，': + rand = random.random() + if rand < 0.25: # 5%概率删除逗号 + continue + elif rand < 0.25: # 20%概率把逗号变成空格 + result += ' ' + continue + result += char + return result + +def add_typos(text: str) -> str: + """随机给文本添加错别字 + + Args: + text: 要处理的文本 + + Returns: + str: 添加错别字后的文本 + """ + TYPO_RATE = 0.02 # 控制错别字出现的概率(2%) + + result = "" + for char in text: + if char in TYPO_DICT and random.random() < TYPO_RATE: + # 从可能的错别字中随机选择一个 + typos = TYPO_DICT[char] + result += random.choice(typos) + else: + result += char + return result + +def process_text_with_typos(text: str) -> str: + """处理文本，添加错别字和处理标点符号 + + Args: + text: 要处理的文本 + + Returns: + str: 处理后的文本 + """ + if random.random() < 0.9: # 90%概率进行处理 + return random_remove_punctuation(add_typos(text)) + return text