From 8e34ab885a33d304beb92dcf94ed2eae7f35b133 Mon Sep 17 00:00:00 2001
From: SengokuCola <1026294844@qq.com>
Date: Mon, 14 Jul 2025 23:44:01 +0800
Subject: [PATCH] =?UTF-8?q?feat=EF=BC=9A=E4=B8=BAs4u=E6=B7=BB=E5=8A=A0?=
=?UTF-8?q?=E4=B8=80=E4=B8=AA=E9=80=8F=E6=98=8E=E5=BA=95=E7=9A=84=E8=81=8A?=
=?UTF-8?q?=E5=A4=A9=E8=AE=B0=E5=BD=95=E7=BD=91=E9=A1=B5?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
---
requirements.txt | 1 +
src/mais4u/mais4u_chat/context_web_manager.py | 544 +++++++++++++++++
src/mais4u/mais4u_chat/s4u_chat.py | 4 +-
src/mais4u/mais4u_chat/s4u_mood_manager.py | 573 +++---------------
src/mais4u/mais4u_chat/s4u_msg_processor.py | 29 +
.../mais4u_chat/s4u_watching_manager.py | 3 +-
template/bot_config_template.toml | 4 +-
7 files changed, 651 insertions(+), 507 deletions(-)
create mode 100644 src/mais4u/mais4u_chat/context_web_manager.py
diff --git a/requirements.txt b/requirements.txt
index 32403c966..a09637a91 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,7 @@
APScheduler
Pillow
aiohttp
+aiohttp-cors
colorama
customtkinter
dotenv
diff --git a/src/mais4u/mais4u_chat/context_web_manager.py b/src/mais4u/mais4u_chat/context_web_manager.py
new file mode 100644
index 000000000..844782092
--- /dev/null
+++ b/src/mais4u/mais4u_chat/context_web_manager.py
@@ -0,0 +1,544 @@
+import asyncio
+import json
+from collections import deque
+from datetime import datetime
+from typing import Dict, List, Optional
+from aiohttp import web, WSMsgType
+import aiohttp_cors
+from threading import Thread
+import weakref
+
+from src.chat.message_receive.message import MessageRecv
+from src.common.logger import get_logger
+
+logger = get_logger("context_web")
+
+
+class ContextMessage:
+ """上下文消息类"""
+
+ def __init__(self, message: MessageRecv):
+ self.user_name = message.message_info.user_info.user_nickname
+ self.user_id = message.message_info.user_info.user_id
+ self.content = message.processed_plain_text
+ self.timestamp = datetime.now()
+ self.group_name = message.message_info.group_info.group_name if message.message_info.group_info else "私聊"
+
+ def to_dict(self):
+ return {
+ "user_name": self.user_name,
+ "user_id": self.user_id,
+ "content": self.content,
+ "timestamp": self.timestamp.strftime("%m-%d %H:%M:%S"),
+ "group_name": self.group_name
+ }
+
+
+class ContextWebManager:
+ """上下文网页管理器"""
+
+ def __init__(self, max_messages: int = 10, port: int = 8765):
+ self.max_messages = max_messages
+ self.port = port
+ self.contexts: Dict[str, deque] = {} # chat_id -> deque of ContextMessage
+ self.websockets: List[web.WebSocketResponse] = []
+ self.app = None
+ self.runner = None
+ self.site = None
+ self._server_starting = False # 添加启动标志防止并发
+
+ async def start_server(self):
+ """启动web服务器"""
+ if self.site is not None:
+ logger.debug("Web服务器已经启动，跳过重复启动")
+ return
+
+ if self._server_starting:
+ logger.debug("Web服务器正在启动中，等待启动完成...")
+ # 等待启动完成
+ while self._server_starting and self.site is None:
+ await asyncio.sleep(0.1)
+ return
+
+ self._server_starting = True
+
+ try:
+ self.app = web.Application()
+
+ # 设置CORS
+ cors = aiohttp_cors.setup(self.app, defaults={
+ "*": aiohttp_cors.ResourceOptions(
+ allow_credentials=True,
+ expose_headers="*",
+ allow_headers="*",
+ allow_methods="*"
+ )
+ })
+
+ # 添加路由
+ self.app.router.add_get('/', self.index_handler)
+ self.app.router.add_get('/ws', self.websocket_handler)
+ self.app.router.add_get('/api/contexts', self.get_contexts_handler)
+ self.app.router.add_get('/debug', self.debug_handler)
+
+ # 为所有路由添加CORS
+ for route in list(self.app.router.routes()):
+ cors.add(route)
+
+ self.runner = web.AppRunner(self.app)
+ await self.runner.setup()
+
+ self.site = web.TCPSite(self.runner, 'localhost', self.port)
+ await self.site.start()
+
+ logger.info(f"🌐 上下文网页服务器启动成功在 http://localhost:{self.port}")
+
+ except Exception as e:
+ logger.error(f"❌ 启动Web服务器失败: {e}")
+ # 清理部分启动的资源
+ if self.runner:
+ await self.runner.cleanup()
+ self.app = None
+ self.runner = None
+ self.site = None
+ raise
+ finally:
+ self._server_starting = False
+
+ async def stop_server(self):
+ """停止web服务器"""
+ if self.site:
+ await self.site.stop()
+ if self.runner:
+ await self.runner.cleanup()
+ self.app = None
+ self.runner = None
+ self.site = None
+ self._server_starting = False
+
+ async def index_handler(self, request):
+ """主页处理器"""
+ html_content = '''
+
+
+
+
+ 聊天上下文
+
+
+
+
+
+
+
+
+ '''
+ return web.Response(text=html_content, content_type='text/html')
+
+ async def websocket_handler(self, request):
+ """WebSocket处理器"""
+ ws = web.WebSocketResponse()
+ await ws.prepare(request)
+
+ self.websockets.append(ws)
+ logger.debug(f"WebSocket连接建立，当前连接数: {len(self.websockets)}")
+
+ # 发送初始数据
+ await self.send_contexts_to_websocket(ws)
+
+ async for msg in ws:
+ if msg.type == WSMsgType.ERROR:
+ logger.error(f'WebSocket错误: {ws.exception()}')
+ break
+
+ # 清理断开的连接
+ if ws in self.websockets:
+ self.websockets.remove(ws)
+ logger.debug(f"WebSocket连接断开，当前连接数: {len(self.websockets)}")
+
+ return ws
+
+ async def get_contexts_handler(self, request):
+ """获取上下文API"""
+ all_context_msgs = []
+ for chat_id, contexts in self.contexts.items():
+ all_context_msgs.extend(list(contexts))
+
+ # 按时间排序，最新的在最后
+ all_context_msgs.sort(key=lambda x: x.timestamp)
+
+ # 转换为字典格式
+ contexts_data = [msg.to_dict() for msg in all_context_msgs[-self.max_messages:]]
+
+ logger.debug(f"返回上下文数据，共 {len(contexts_data)} 条消息")
+ return web.json_response({"contexts": contexts_data})
+
+ async def debug_handler(self, request):
+ """调试信息处理器"""
+ debug_info = {
+ "server_status": "running",
+ "websocket_connections": len(self.websockets),
+ "total_chats": len(self.contexts),
+ "total_messages": sum(len(contexts) for contexts in self.contexts.values()),
+ }
+
+ # 构建聊天详情HTML
+ chats_html = ""
+ for chat_id, contexts in self.contexts.items():
+ messages_html = ""
+ for msg in contexts:
+ timestamp = msg.timestamp.strftime("%H:%M:%S")
+ content = msg.content[:50] + "..." if len(msg.content) > 50 else msg.content
+ messages_html += f'[{timestamp}] {msg.user_name}: {content}
'
+
+ chats_html += f'''
+
+
聊天 {chat_id} ({len(contexts)} 条消息)
+ {messages_html}
+
+ '''
+
+ html_content = f'''
+
+
+
+
+ 调试信息
+
+
+
+ 上下文网页管理器调试信息
+
+
+
服务器状态
+
状态: {debug_info["server_status"]}
+
WebSocket连接数: {debug_info["websocket_connections"]}
+
聊天总数: {debug_info["total_chats"]}
+
消息总数: {debug_info["total_messages"]}
+
+
聊天详情
+ {chats_html}
+
+
+
+
操作
+
+
+
+
+
+
+
+
+ '''
+
+ return web.Response(text=html_content, content_type='text/html')
+
+ async def add_message(self, chat_id: str, message: MessageRecv):
+ """添加新消息到上下文"""
+ if chat_id not in self.contexts:
+ self.contexts[chat_id] = deque(maxlen=self.max_messages)
+ logger.debug(f"为聊天 {chat_id} 创建新的上下文队列")
+
+ context_msg = ContextMessage(message)
+ self.contexts[chat_id].append(context_msg)
+
+ # 统计当前总消息数
+ total_messages = sum(len(contexts) for contexts in self.contexts.values())
+
+ logger.info(f"✅ 添加消息到上下文 [总数: {total_messages}]: [{context_msg.group_name}] {context_msg.user_name}: {context_msg.content}")
+
+ # 调试：打印当前所有消息
+ logger.info(f"📝 当前上下文中的所有消息：")
+ for cid, contexts in self.contexts.items():
+ logger.info(f" 聊天 {cid}: {len(contexts)} 条消息")
+ for i, msg in enumerate(contexts):
+ logger.info(f" {i+1}. [{msg.timestamp.strftime('%H:%M:%S')}] {msg.user_name}: {msg.content[:30]}...")
+
+ # 广播更新给所有WebSocket连接
+ await self.broadcast_contexts()
+
+ async def send_contexts_to_websocket(self, ws: web.WebSocketResponse):
+ """向单个WebSocket发送上下文数据"""
+ all_context_msgs = []
+ for chat_id, contexts in self.contexts.items():
+ all_context_msgs.extend(list(contexts))
+
+ # 按时间排序，最新的在最后
+ all_context_msgs.sort(key=lambda x: x.timestamp)
+
+ # 转换为字典格式
+ contexts_data = [msg.to_dict() for msg in all_context_msgs[-self.max_messages:]]
+
+ data = {"contexts": contexts_data}
+ await ws.send_str(json.dumps(data, ensure_ascii=False))
+
+ async def broadcast_contexts(self):
+ """向所有WebSocket连接广播上下文更新"""
+ if not self.websockets:
+ logger.debug("没有WebSocket连接，跳过广播")
+ return
+
+ all_context_msgs = []
+ for chat_id, contexts in self.contexts.items():
+ all_context_msgs.extend(list(contexts))
+
+ # 按时间排序，最新的在最后
+ all_context_msgs.sort(key=lambda x: x.timestamp)
+
+ # 转换为字典格式
+ contexts_data = [msg.to_dict() for msg in all_context_msgs[-self.max_messages:]]
+
+ data = {"contexts": contexts_data}
+ message = json.dumps(data, ensure_ascii=False)
+
+ logger.info(f"广播 {len(contexts_data)} 条消息到 {len(self.websockets)} 个WebSocket连接")
+
+ # 创建WebSocket列表的副本，避免在遍历时修改
+ websockets_copy = self.websockets.copy()
+ removed_count = 0
+
+ for ws in websockets_copy:
+ if ws.closed:
+ if ws in self.websockets:
+ self.websockets.remove(ws)
+ removed_count += 1
+ else:
+ try:
+ await ws.send_str(message)
+ logger.debug("消息发送成功")
+ except Exception as e:
+ logger.error(f"发送WebSocket消息失败: {e}")
+ if ws in self.websockets:
+ self.websockets.remove(ws)
+ removed_count += 1
+
+ if removed_count > 0:
+ logger.debug(f"清理了 {removed_count} 个断开的WebSocket连接")
+
+
+# 全局实例
+_context_web_manager: Optional[ContextWebManager] = None
+
+
+def get_context_web_manager() -> ContextWebManager:
+ """获取上下文网页管理器实例"""
+ global _context_web_manager
+ if _context_web_manager is None:
+ _context_web_manager = ContextWebManager()
+ return _context_web_manager
+
+
+async def init_context_web_manager():
+ """初始化上下文网页管理器"""
+ manager = get_context_web_manager()
+ await manager.start_server()
+ return manager
\ No newline at end of file
diff --git a/src/mais4u/mais4u_chat/s4u_chat.py b/src/mais4u/mais4u_chat/s4u_chat.py
index 8f4d771c6..a1e2efb9e 100644
--- a/src/mais4u/mais4u_chat/s4u_chat.py
+++ b/src/mais4u/mais4u_chat/s4u_chat.py
@@ -142,7 +142,7 @@ def get_s4u_chat_manager() -> S4UChatManager:
class S4UChat:
- _MESSAGE_TIMEOUT_SECONDS = 60 # 普通消息存活时间（秒）
+ _MESSAGE_TIMEOUT_SECONDS = 30 # 普通消息存活时间（秒）
def __init__(self, chat_stream: ChatStream):
"""初始化 S4UChat 实例。"""
@@ -167,7 +167,7 @@ class S4UChat:
self.gpt = S4UStreamGenerator()
self.interest_dict: Dict[str, float] = {} # 用户兴趣分
self.at_bot_priority_bonus = 100.0 # @机器人的优先级加成
- self.normal_queue_max_size = 50 # 普通队列最大容量
+ self.normal_queue_max_size = 5 # 普通队列最大容量
logger.info(f"[{self.stream_name}] S4UChat with two-queue system initialized.")
def _get_priority_info(self, message: MessageRecv) -> dict:
diff --git a/src/mais4u/mais4u_chat/s4u_mood_manager.py b/src/mais4u/mais4u_chat/s4u_mood_manager.py
index 22b1400cb..a394e9429 100644
--- a/src/mais4u/mais4u_chat/s4u_mood_manager.py
+++ b/src/mais4u/mais4u_chat/s4u_mood_manager.py
@@ -1,7 +1,6 @@
import asyncio
import json
import time
-import random
from src.chat.message_receive.message import MessageRecv
from src.llm_models.utils_model import LLMRequest
@@ -13,54 +12,23 @@ from src.manager.async_task_manager import AsyncTask, async_task_manager
from src.plugin_system.apis import send_api
"""
-面部表情系统使用说明：
+情绪管理系统使用说明：
-1. 预定义的面部表情：
- - happy: 高兴表情（眼睛微笑 + 眉毛微笑 + 嘴巴大笑）
- - very_happy: 非常高兴（高兴表情 + 脸红）
- - sad: 悲伤表情（眼睛哭泣 + 眉毛忧伤 + 嘴巴悲伤）
- - angry: 生气表情（眉毛生气 + 嘴巴生气）
- - fear: 恐惧表情（眼睛闭上）
- - shy: 害羞表情（嘴巴嘟起 + 脸红）
- - neutral: 中性表情（无表情）
+1. 情绪数值系统：
+ - 情绪包含四个维度：joy(喜), anger(怒), sorrow(哀), fear(惧)
+ - 每个维度的取值范围为1-10
+ - 当情绪发生变化时，会自动发送到ws端处理
-2. 使用方法：
- # 获取面部表情管理器
- facial_expression = mood_manager.get_facial_expression_by_chat_id(chat_id)
-
- # 发送指定表情
- await facial_expression.send_expression("happy")
-
- # 根据情绪值自动选择表情
- await facial_expression.send_expression_by_mood(mood_values)
-
- # 重置为中性表情
- await facial_expression.reset_expression()
-
- # 执行眨眼动作
- await facial_expression.perform_blink()
+2. 情绪更新机制：
+ - 接收到新消息时会更新情绪状态
+ - 定期进行情绪回归（冷静下来）
+ - 每次情绪变化都会发送到ws端，格式为：
+ type: "emotion"
+ data: {"joy": 5, "anger": 1, "sorrow": 1, "fear": 1}
-3. 自动表情系统：
- - 当情绪值更新时，系统会自动根据mood_values选择合适的面部表情
- - 只有当新表情与当前表情不同时才会发送，避免重复发送
- - 支持joy >= 8时显示very_happy，joy >= 6时显示happy等梯度表情
-
-4. amadus表情更新系统：
- - 每1秒检查一次表情是否有变化，如有变化则发送到amadus
- - 每次mood更新后立即发送表情更新
- - 发送消息类型为"amadus_expression_update"，格式为{"action": "表情名", "data": 1.0}
-
-5. 眨眼系统：
- - 每4-6秒随机执行一次眨眼动作
- - 眨眼包含两个阶段：先闭眼（eye_close=1.0），保持0.1-0.15秒，然后睁眼（eye_close=0.0）
- - 眨眼使用override_values参数临时覆盖eye_close值，不修改原始表情状态
- - 眨眼时会发送完整的表情状态，包含当前表情的所有动作
- - 当eye部位已经是eye_happy_weak时，跳过眨眼动作
-
-6. 表情选择逻辑：
- - 系统会找出最强的情绪（joy, anger, sorrow, fear）
- - 根据情绪强度选择相应的表情组合
- - 默认情况下返回neutral表情
+3. ws端处理：
+ - 本地只负责情绪计算和发送情绪数值
+ - 表情渲染和动作由ws端根据情绪数值处理
"""
logger = get_logger("mood")
@@ -137,300 +105,6 @@ def init_prompt():
)
-class FacialExpression:
- def __init__(self, chat_id: str):
- self.chat_id: str = chat_id
-
- # 预定义面部表情动作（根据用户定义的表情动作）
- self.expressions = {
- # 眼睛表情
- "eye_happy_weak": {"action": "eye_happy_weak", "data": 1.0},
- "eye_close": {"action": "eye_close", "data": 1.0},
- "eye_shift_left": {"action": "eye_shift_left", "data": 1.0},
- "eye_shift_right": {"action": "eye_shift_right", "data": 1.0},
- # "eye_smile": {"action": "eye_smile", "data": 1.0}, # 未定义，占位
- # "eye_cry": {"action": "eye_cry", "data": 1.0}, # 未定义，占位
- # "eye_normal": {"action": "eye_normal", "data": 1.0}, # 未定义，占位
-
- # 眉毛表情
- "eyebrow_happy_weak": {"action": "eyebrow_happy_weak", "data": 1.0},
- "eyebrow_happy_strong": {"action": "eyebrow_happy_strong", "data": 1.0},
- "eyebrow_angry_weak": {"action": "eyebrow_angry_weak", "data": 1.0},
- "eyebrow_angry_strong": {"action": "eyebrow_angry_strong", "data": 1.0},
- "eyebrow_sad_weak": {"action": "eyebrow_sad_weak", "data": 1.0},
- "eyebrow_sad_strong": {"action": "eyebrow_sad_strong", "data": 1.0},
- # "eyebrow_smile": {"action": "eyebrow_smile", "data": 1.0}, # 未定义，占位
- # "eyebrow_angry": {"action": "eyebrow_angry", "data": 1.0}, # 未定义，占位
- # "eyebrow_sad": {"action": "eyebrow_sad", "data": 1.0}, # 未定义，占位
- # "eyebrow_normal": {"action": "eyebrow_normal", "data": 1.0}, # 未定义，占位
-
- # 嘴巴表情（注意：用户定义的是mouth，可能是mouth的拼写错误）
- "mouth_default": {"action": "mouth_default", "data": 1.0},
- "mouth_happy_strong": {"action": "mouth_happy_strong", "data": 1.0}, # 保持用户原始拼写
- "mouth_angry_weak": {"action": "mouth_angry_weak", "data": 1.0},
- # "mouth_sad": {"action": "mouth_sad", "data": 1.0}, # 未定义，占位
- # "mouth_angry": {"action": "mouth_angry", "data": 1.0}, # 未定义，占位
- # "mouth_laugh": {"action": "mouth_laugh", "data": 1.0}, # 未定义，占位
- # "mouth_pout": {"action": "mouth_pout", "data": 1.0}, # 未定义，占位
- # "mouth_normal": {"action": "mouth_normal", "data": 1.0}, # 未定义，占位
-
- # 脸部表情
- # "face_blush": {"action": "face_blush", "data": 1.0}, # 未定义，占位
- # "face_normal": {"action": "face_normal", "data": 1.0}, # 未定义，占位
- }
-
- # 表情组合模板（根据新的表情动作调整）
- self.expression_combinations = {
- "happy": {
- "eye": "eye_happy_weak",
- "eyebrow": "eyebrow_happy_weak",
- "mouth": "mouth_default",
- },
- "very_happy": {
- "eye": "eye_happy_weak",
- "eyebrow": "eyebrow_happy_strong",
- "mouth": "mouth_happy_strong",
- },
- "sad": {
- "eyebrow": "eyebrow_sad_strong",
- "mouth": "mouth_default",
- },
- "angry": {
- "eyebrow": "eyebrow_angry_strong",
- "mouth": "mouth_angry_weak",
- },
- "fear": {
- "eyebrow": "eyebrow_sad_weak",
- "mouth": "mouth_default",
- },
- "shy": {
- "eyebrow": "eyebrow_happy_weak",
- "mouth": "mouth_default",
- },
- "neutral": {
- "eyebrow": "eyebrow_happy_weak",
- "mouth": "mouth_default",
- }
- }
-
- # 未定义的表情部位（保留备用）：
- # 眼睛：eye_smile, eye_cry, eye_close, eye_normal
- # 眉毛：eyebrow_smile, eyebrow_angry, eyebrow_sad, eyebrow_normal
- # 嘴巴：mouth_sad, mouth_angry, mouth_laugh, mouth_pout, mouth_normal
- # 脸部：face_blush, face_normal
-
- # 初始化当前表情状态
- self.last_expression = "neutral"
-
- def select_expression_by_mood(self, mood_values: dict[str, int]) -> str:
- """根据情绪值选择合适的表情组合"""
- joy = mood_values.get("joy", 5)
- anger = mood_values.get("anger", 1)
- sorrow = mood_values.get("sorrow", 1)
- fear = mood_values.get("fear", 1)
-
- # 找出最强的情绪
- emotions = {
- "joy": joy,
- "anger": anger,
- "sorrow": sorrow,
- "fear": fear
- }
-
- # 获取最强情绪
- dominant_emotion = max(emotions, key=emotions.get)
- _dominant_value = emotions[dominant_emotion]
-
- # 根据情绪强度和类型选择表情
- if dominant_emotion == "joy":
- if joy >= 8:
- return "very_happy"
- elif joy >= 6:
- return "happy"
- elif joy >= 4:
- return "shy"
- else:
- return "neutral"
- elif dominant_emotion == "anger" and anger >= 6:
- return "angry"
- elif dominant_emotion == "sorrow" and sorrow >= 6:
- return "sad"
- elif dominant_emotion == "fear" and fear >= 6:
- return "fear"
- else:
- return "neutral"
-
- async def _send_expression_actions(self, expression_name: str, log_prefix: str = "发送面部表情", override_values: dict = None):
- """统一的表情动作发送函数 - 发送完整的表情状态
-
- Args:
- expression_name: 表情名称
- log_prefix: 日志前缀
- override_values: 需要覆盖的动作值，格式为 {"action_name": value}
- """
- if expression_name not in self.expression_combinations:
- logger.warning(f"[{self.chat_id}] 未知表情: {expression_name}")
- return
-
- combination = self.expression_combinations[expression_name]
-
- # 按部位分组所有已定义的表情动作
- expressions_by_part = {
- "eye": {},
- "eyebrow": {},
- "mouth": {}
- }
-
- # 将所有已定义的表情按部位分组
- for expression_key, expression_data in self.expressions.items():
- if expression_key.startswith("eye_"):
- expressions_by_part["eye"][expression_key] = expression_data
- elif expression_key.startswith("eyebrow_"):
- expressions_by_part["eyebrow"][expression_key] = expression_data
- elif expression_key.startswith("mouth_"):
- expressions_by_part["mouth"][expression_key] = expression_data
-
- # 构建完整的表情状态
- complete_expression_state = {}
-
- # 为每个部位构建完整的表情动作状态
- for part in expressions_by_part.keys():
- if expressions_by_part[part]: # 如果该部位有已定义的表情
- part_actions = {}
- active_expression = combination.get(part) # 当前激活的表情
-
- # 添加该部位所有已定义的表情动作
- for expression_key, expression_data in expressions_by_part[part].items():
- # 复制表情数据并设置激活状态
- action_data = expression_data.copy()
-
- # 检查是否有覆盖值
- if override_values and expression_key in override_values:
- action_data["data"] = override_values[expression_key]
- else:
- action_data["data"] = 1.0 if expression_key == active_expression else 0.0
-
- part_actions[expression_key] = action_data
-
- complete_expression_state[part] = part_actions
- logger.debug(f"[{self.chat_id}] 部位 {part}: 激活 {active_expression}, 总共 {len(part_actions)} 个动作")
-
- # 发送完整的表情状态
- if complete_expression_state:
- package_data = {
- "expression_name": expression_name,
- "actions": complete_expression_state
- }
-
- await send_api.custom_to_stream(
- message_type="face_emotion",
- content=package_data,
- stream_id=self.chat_id,
- storage_message=False,
- show_log=False,
- )
-
- # 统计信息
- total_actions = sum(len(part_actions) for part_actions in complete_expression_state.values())
- active_actions = [f"{part}:{combination.get(part, 'none')}" for part in complete_expression_state.keys()]
- logger.info(f"[{self.chat_id}] {log_prefix}: {expression_name} - 发送{total_actions}个动作，激活: {', '.join(active_actions)}")
- else:
- logger.warning(f"[{self.chat_id}] 表情 {expression_name} 没有有效的动作可发送")
-
- async def send_expression(self, expression_name: str):
- """发送表情组合"""
- await self._send_expression_actions(expression_name, "发送面部表情")
-
- # 通知ChatMood需要更新amadus
- # 这里需要从mood_manager获取ChatMood实例并标记
- chat_mood = mood_manager.get_mood_by_chat_id(self.chat_id)
- if chat_mood.last_expression != expression_name:
- chat_mood.last_expression = expression_name
- chat_mood.expression_needs_update = True
-
- async def send_expression_by_mood(self, mood_values: dict[str, int]):
- """根据情绪值发送相应的面部表情"""
- expression_name = self.select_expression_by_mood(mood_values)
- logger.info(f"[{self.chat_id}] 根据情绪值选择表情: {expression_name}, 情绪值: {mood_values}")
- await self.send_expression(expression_name)
-
- async def reset_expression(self):
- """重置为中性表情"""
- await self.send_expression("neutral")
-
- async def perform_blink(self):
- """执行眨眼动作"""
- # 检查当前表情组合中eye部位是否为eye_happy_weak
- current_combination = self.expression_combinations.get(self.last_expression, {})
- current_eye_expression = current_combination.get("eye")
-
- if current_eye_expression == "eye_happy_weak":
- logger.debug(f"[{self.chat_id}] 当前eye表情为{current_eye_expression}，跳过眨眼动作")
- return
-
- logger.debug(f"[{self.chat_id}] 执行眨眼动作")
-
- # 第一阶段：闭眼
- await self._send_expression_actions(
- self.last_expression,
- "眨眼-闭眼",
- override_values={"eye_close": 1.0}
- )
-
- # 等待0.1-0.15秒
- blink_duration = random.uniform(0.7, 0.12)
- await asyncio.sleep(blink_duration)
-
- # 第二阶段：睁眼
- await self._send_expression_actions(
- self.last_expression,
- "眨眼-睁眼",
- override_values={"eye_close": 0.0}
- )
-
-
- async def perform_shift(self):
- """执行眨眼动作"""
- # 检查当前表情组合中eye部位是否为eye_happy_weak
- current_combination = self.expression_combinations.get(self.last_expression, {})
- current_eye_expression = current_combination.get("eye")
-
- direction = random.choice(["left", "right"])
- strength = random.randint(6, 9) / 10
- time_duration = random.randint(5, 15) / 10
-
- if current_eye_expression == "eye_happy_weak" or current_eye_expression == "eye_close":
- logger.debug(f"[{self.chat_id}] 当前eye表情为{current_eye_expression}，跳过漂移动作")
- return
-
- logger.debug(f"[{self.chat_id}] 执行漂移动作，方向：{direction}，强度：{strength}，时间：{time_duration}")
-
- if direction == "left":
- override_values = {"eye_shift_left": strength}
- back_values = {"eye_shift_left": 0.0}
- else:
- override_values = {"eye_shift_right": strength}
- back_values = {"eye_shift_right": 0.0}
-
- # 第一阶段：闭眼
- await self._send_expression_actions(
- self.last_expression,
- "漂移",
- override_values=override_values
- )
-
- # 等待0.1-0.15秒
- await asyncio.sleep(time_duration)
-
- # 第二阶段：睁眼
- await self._send_expression_actions(
- self.last_expression,
- "回归",
- override_values=back_values
- )
-
-
-
class ChatMood:
def __init__(self, chat_id: str):
self.chat_id: str = chat_id
@@ -452,14 +126,8 @@ class ChatMood:
self.last_change_time = 0
- # 添加面部表情系统
- self.facial_expression = FacialExpression(chat_id)
- self.last_expression = "neutral" # 记录上一次的表情
- self.expression_needs_update = False # 标记表情是否需要更新
-
- # 设置初始中性表情
- asyncio.create_task(self.facial_expression.reset_expression())
- self.expression_needs_update = True # 初始化时也标记需要更新
+ # 发送初始情绪状态到ws端
+ asyncio.create_task(self.send_emotion_update(self.mood_values))
def _parse_numerical_mood(self, response: str) -> dict[str, int] | None:
try:
@@ -564,13 +232,10 @@ class ChatMood:
_old_mood_values = self.mood_values.copy()
self.mood_values = numerical_mood_response
- # 发送面部表情
- new_expression = self.facial_expression.select_expression_by_mood(self.mood_values)
- if new_expression != self.last_expression:
- # 立即发送表情
- asyncio.create_task(self.facial_expression.send_expression(new_expression))
- self.last_expression = new_expression
- self.expression_needs_update = True # 标记表情已更新
+ # 发送情绪更新到ws端
+ await self.send_emotion_update(self.mood_values)
+
+ logger.info(f"[{self.chat_id}] 情绪变化: {_old_mood_values} -> {self.mood_values}")
self.last_change_time = message_time
@@ -644,33 +309,31 @@ class ChatMood:
_old_mood_values = self.mood_values.copy()
self.mood_values = numerical_mood_response
- # 发送面部表情
- new_expression = self.facial_expression.select_expression_by_mood(self.mood_values)
- if new_expression != self.last_expression:
- # 立即发送表情
- asyncio.create_task(self.facial_expression.send_expression(new_expression))
- self.last_expression = new_expression
- self.expression_needs_update = True # 标记表情已更新
+ # 发送情绪更新到ws端
+ await self.send_emotion_update(self.mood_values)
+
+ logger.info(f"[{self.chat_id}] 情绪回归: {_old_mood_values} -> {self.mood_values}")
self.regression_count += 1
- async def send_expression_update_if_needed(self):
- """如果表情有变化，发送更新到amadus"""
- if self.expression_needs_update:
- # 使用统一的表情发送函数
- await self.facial_expression._send_expression_actions(
- self.last_expression,
- "发送表情更新到amadus"
- )
- self.expression_needs_update = False # 重置标记
-
- async def perform_blink(self):
- """执行眨眼动作"""
- await self.facial_expression.perform_blink()
+ async def send_emotion_update(self, mood_values: dict[str, int]):
+ """发送情绪更新到ws端"""
+ emotion_data = {
+ "joy": mood_values.get("joy", 5),
+ "anger": mood_values.get("anger", 1),
+ "sorrow": mood_values.get("sorrow", 1),
+ "fear": mood_values.get("fear", 1)
+ }
- async def perform_shift(self):
- """执行漂移动作"""
- await self.facial_expression.perform_shift()
+ await send_api.custom_to_stream(
+ message_type="emotion",
+ content=emotion_data,
+ stream_id=self.chat_id,
+ storage_message=False,
+ show_log=True,
+ )
+
+ logger.info(f"[{self.chat_id}] 发送情绪更新: {emotion_data}")
class MoodRegressionTask(AsyncTask):
@@ -695,17 +358,38 @@ class MoodRegressionTask(AsyncTask):
time_since_last_change = now - mood.last_change_time
- if time_since_last_change > 120: # 2分钟
+ # 检查是否有极端情绪需要快速回归
+ high_emotions = {k: v for k, v in mood.mood_values.items() if v >= 8}
+ has_extreme_emotion = len(high_emotions) > 0
+
+ # 回归条件：1. 正常时间间隔(120s) 或 2. 有极端情绪且距上次变化>=30s
+ should_regress = False
+ regress_reason = ""
+
+ if time_since_last_change > 120:
+ should_regress = True
+ regress_reason = f"常规回归(距上次变化{int(time_since_last_change)}秒)"
+ elif has_extreme_emotion and time_since_last_change > 30:
+ should_regress = True
+ high_emotion_str = ", ".join([f"{k}={v}" for k, v in high_emotions.items()])
+ regress_reason = f"极端情绪快速回归({high_emotion_str}, 距上次变化{int(time_since_last_change)}秒)"
+
+ if should_regress:
if mood.regression_count >= 3:
logger.debug(f"[回归任务] {chat_info} 已达到最大回归次数(3次)，停止回归")
continue
- logger.info(f"[回归任务] {chat_info} 开始情绪回归 (距上次变化{int(time_since_last_change)}秒，第{mood.regression_count + 1}次回归)")
+ logger.info(f"[回归任务] {chat_info} 开始情绪回归 ({regress_reason}，第{mood.regression_count + 1}次回归)")
await mood.regress_mood()
regression_executed += 1
else:
- remaining_time = 120 - time_since_last_change
- logger.debug(f"[回归任务] {chat_info} 距离回归还需等待{int(remaining_time)}秒")
+ if has_extreme_emotion:
+ remaining_time = 5 - time_since_last_change
+ high_emotion_str = ", ".join([f"{k}={v}" for k, v in high_emotions.items()])
+ logger.debug(f"[回归任务] {chat_info} 存在极端情绪({high_emotion_str})，距离快速回归还需等待{int(remaining_time)}秒")
+ else:
+ remaining_time = 120 - time_since_last_change
+ logger.debug(f"[回归任务] {chat_info} 距离回归还需等待{int(remaining_time)}秒")
if regression_executed > 0:
logger.info(f"[回归任务] 本次执行了{regression_executed}个聊天的情绪回归")
@@ -713,96 +397,6 @@ class MoodRegressionTask(AsyncTask):
logger.debug(f"[回归任务] 本次没有符合回归条件的聊天")
-class ExpressionUpdateTask(AsyncTask):
- def __init__(self, mood_manager: "MoodManager"):
- super().__init__(task_name="ExpressionUpdateTask", run_interval=0.3)
- self.mood_manager = mood_manager
- self.run_count = 0
- self.last_log_time = 0
-
- async def run(self):
- self.run_count += 1
- now = time.time()
-
- # 每60秒输出一次状态信息（避免日志太频繁）
- if now - self.last_log_time > 60:
- logger.info(f"[表情任务] 已运行{self.run_count}次，当前管理{len(self.mood_manager.mood_list)}个聊天的表情状态")
- self.last_log_time = now
-
- updates_sent = 0
- for mood in self.mood_manager.mood_list:
- if mood.expression_needs_update:
- logger.debug(f"[表情任务] chat {mood.chat_id} 检测到表情变化，发送更新")
- await mood.send_expression_update_if_needed()
- updates_sent += 1
-
- if updates_sent > 0:
- logger.info(f"[表情任务] 发送了{updates_sent}个表情更新")
-
-
-class BlinkTask(AsyncTask):
- def __init__(self, mood_manager: "MoodManager"):
- # 初始随机间隔4-6秒
- super().__init__(task_name="BlinkTask", run_interval=4)
- self.mood_manager = mood_manager
- self.run_count = 0
- self.last_log_time = 0
-
- async def run(self):
- self.run_count += 1
- now = time.time()
-
- # 每60秒输出一次状态信息（避免日志太频繁）
- if now - self.last_log_time > 20:
- logger.debug(f"[眨眼任务] 已运行{self.run_count}次，当前管理{len(self.mood_manager.mood_list)}个聊天的眨眼状态")
- self.last_log_time = now
-
- interval_add = random.randint(0, 2)
- await asyncio.sleep(interval_add)
-
- blinks_executed = 0
- for mood in self.mood_manager.mood_list:
- try:
- await mood.perform_blink()
- blinks_executed += 1
- except Exception as e:
- logger.error(f"[眨眼任务] 处理chat {mood.chat_id}时出错: {e}")
-
- if blinks_executed > 0:
- logger.debug(f"[眨眼任务] 本次执行了{blinks_executed}个聊天的眨眼动作")
-
-class ShiftTask(AsyncTask):
- def __init__(self, mood_manager: "MoodManager"):
- # 初始随机间隔4-6秒
- super().__init__(task_name="ShiftTask", run_interval=8)
- self.mood_manager = mood_manager
- self.run_count = 0
- self.last_log_time = 0
-
- async def run(self):
- self.run_count += 1
- now = time.time()
-
- # 每60秒输出一次状态信息（避免日志太频繁）
- if now - self.last_log_time > 20:
- logger.debug(f"[漂移任务] 已运行{self.run_count}次，当前管理{len(self.mood_manager.mood_list)}个聊天的漂移状态")
- self.last_log_time = now
-
- interval_add = random.randint(0, 3)
- await asyncio.sleep(interval_add)
-
- blinks_executed = 0
- for mood in self.mood_manager.mood_list:
- try:
- await mood.perform_shift()
- blinks_executed += 1
- except Exception as e:
- logger.error(f"[漂移任务] 处理chat {mood.chat_id}时出错: {e}")
-
- if blinks_executed > 0:
- logger.debug(f"[漂移任务] 本次执行了{blinks_executed}个聊天的漂移动作")
-
-
class MoodManager:
def __init__(self):
self.mood_list: list[ChatMood] = []
@@ -820,20 +414,8 @@ class MoodManager:
regression_task = MoodRegressionTask(self)
await async_task_manager.add_task(regression_task)
- # 启动表情更新任务
- expression_task = ExpressionUpdateTask(self)
- await async_task_manager.add_task(expression_task)
-
- # 启动眨眼任务
- blink_task = BlinkTask(self)
- await async_task_manager.add_task(blink_task)
-
- # 启动漂移任务
- shift_task = ShiftTask(self)
- await async_task_manager.add_task(shift_task)
-
self.task_started = True
- logger.info("情绪管理任务已启动（包含情绪回归、表情更新和眨眼动作）")
+ logger.info("情绪管理任务已启动（情绪回归）")
def get_mood_by_chat_id(self, chat_id: str) -> ChatMood:
for mood in self.mood_list:
@@ -850,28 +432,15 @@ class MoodManager:
mood.mood_state = "感觉很平静"
mood.mood_values = {"joy": 5, "anger": 1, "sorrow": 1, "fear": 1}
mood.regression_count = 0
- # 重置面部表情为中性
- asyncio.create_task(mood.facial_expression.reset_expression())
- mood.last_expression = "neutral"
- mood.expression_needs_update = True # 标记表情需要更新
+ # 发送重置后的情绪状态到ws端
+ asyncio.create_task(mood.send_emotion_update(mood.mood_values))
return
# 如果没有找到现有的mood，创建新的
new_mood = ChatMood(chat_id)
self.mood_list.append(new_mood)
- asyncio.create_task(new_mood.facial_expression.reset_expression())
- new_mood.expression_needs_update = True # 标记表情需要更新
-
- def get_facial_expression_by_chat_id(self, chat_id: str) -> FacialExpression:
- """获取聊天对应的面部表情管理器"""
- for mood in self.mood_list:
- if mood.chat_id == chat_id:
- return mood.facial_expression
-
- # 如果没有找到，创建新的
- new_mood = ChatMood(chat_id)
- self.mood_list.append(new_mood)
- return new_mood.facial_expression
+ # 发送初始情绪状态到ws端
+ asyncio.create_task(new_mood.send_emotion_update(new_mood.mood_values))
init_prompt()
diff --git a/src/mais4u/mais4u_chat/s4u_msg_processor.py b/src/mais4u/mais4u_chat/s4u_msg_processor.py
index 6a6cf25f0..86ea90275 100644
--- a/src/mais4u/mais4u_chat/s4u_msg_processor.py
+++ b/src/mais4u/mais4u_chat/s4u_msg_processor.py
@@ -13,6 +13,7 @@ from src.config.config import global_config
from src.mais4u.mais4u_chat.body_emotion_action_manager import action_manager
from src.mais4u.mais4u_chat.s4u_mood_manager import mood_manager
from src.mais4u.mais4u_chat.s4u_watching_manager import watching_manager
+from src.mais4u.mais4u_chat.context_web_manager import get_context_web_manager
from .s4u_chat import get_s4u_chat_manager
@@ -115,5 +116,33 @@ class S4UMessageProcessor:
chat_watching = watching_manager.get_watching_by_chat_id(chat.stream_id)
asyncio.create_task(chat_watching.on_message_received())
+ # 上下文网页管理：启动独立task处理消息上下文
+ asyncio.create_task(self._handle_context_web_update(chat.stream_id, message))
+
# 7. 日志记录
logger.info(f"[S4U]{userinfo.user_nickname}:{message.processed_plain_text}")
+
+ async def _handle_context_web_update(self, chat_id: str, message: MessageRecv):
+ """处理上下文网页更新的独立task
+
+ Args:
+ chat_id: 聊天ID
+ message: 消息对象
+ """
+ try:
+ logger.debug(f"🔄 开始处理上下文网页更新: {message.message_info.user_info.user_nickname}")
+
+ context_manager = get_context_web_manager()
+
+ # 只在服务器未启动时启动（避免重复启动）
+ if context_manager.site is None:
+ logger.info("🚀 首次启动上下文网页服务器...")
+ await context_manager.start_server()
+
+ # 添加消息到上下文并更新网页
+ await context_manager.add_message(chat_id, message)
+
+ logger.debug(f"✅ 上下文网页更新完成: {message.message_info.user_info.user_nickname}")
+
+ except Exception as e:
+ logger.error(f"❌ 处理上下文网页更新失败: {e}", exc_info=True)
diff --git a/src/mais4u/mais4u_chat/s4u_watching_manager.py b/src/mais4u/mais4u_chat/s4u_watching_manager.py
index afa1421f1..897ef7f70 100644
--- a/src/mais4u/mais4u_chat/s4u_watching_manager.py
+++ b/src/mais4u/mais4u_chat/s4u_watching_manager.py
@@ -121,7 +121,8 @@ class ChatWatching:
await send_api.custom_to_stream(
message_type="watching",
content=self.current_state.value,
- stream_id=self.chat_id
+ stream_id=self.chat_id,
+ storage_message=False
)
logger.info(f"[{self.chat_id}] 发送视线状态更新: {self.current_state.value}")
diff --git a/template/bot_config_template.toml b/template/bot_config_template.toml
index d6d13017c..41fc80d9b 100644
--- a/template/bot_config_template.toml
+++ b/template/bot_config_template.toml
@@ -20,10 +20,10 @@ alias_names = ["麦叠", "牢麦"] # 麦麦的别名
[personality]
# 建议50字以内，描述人格的核心特质
personality_core = "是一个积极向上的女大学生"
-# 人格的细节，可以描述人格的一些侧面，条数任意，不能为0，不宜太多
+# 人格的细节，描述人格的一些侧面
personality_side = "用一句话或几句话描述人格的侧面特质"
#アイデンティティがない 生まれないらららら
-# 可以描述外貌，性别，身高，职业，属性等等描述,条数任意，不能为0
+# 可以描述外貌，性别，身高，职业，属性等等描述
identity = "年龄为19岁,是女孩子,身高为160cm,有黑色的短发"
compress_personality = false # 是否压缩人格，压缩后会精简人格信息，节省token消耗并提高回复性能，但是会丢失一些信息，如果人设不长，可以关闭