mirror of https://github.com/Mai-with-u/MaiBot.git
106 lines
5.1 KiB
Python
106 lines
5.1 KiB
Python
from typing import Tuple, List, Dict, Any
|
||
from src.common.logger import get_module_logger
|
||
from src.config.config import global_config # 为了获取 BOT_QQ
|
||
from .chat_observer import ChatObserver
|
||
import re
|
||
|
||
logger = get_module_logger("reply_checker")
|
||
|
||
|
||
class ReplyChecker:
|
||
"""回复检查器 - 新版：仅检查机器人自身发言的精确重复"""
|
||
|
||
def __init__(self, stream_id: str, private_name: str):
|
||
self.name = global_config.BOT_NICKNAME
|
||
self.private_name = private_name
|
||
self.chat_observer = ChatObserver.get_instance(stream_id, private_name)
|
||
self.bot_qq_str = str(global_config.BOT_QQ)
|
||
|
||
def _normalize_text(self, text: str) -> str:
|
||
"""
|
||
规范化文本，去除首尾空格，移除末尾的特定标点符号。
|
||
"""
|
||
if not text:
|
||
return ""
|
||
text = text.strip() # 1. 去除首尾空格
|
||
# 2. 移除末尾的一个或多个特定标点符号
|
||
# 可以根据需要调整正则表达式以包含更多或更少的标点
|
||
text = re.sub(r"[~\s,.!?；;，。]+$", "", text)
|
||
# 如果需要忽略大小写，可以取消下面一行的注释
|
||
# text = text.lower()
|
||
return text
|
||
|
||
async def check(
|
||
self,
|
||
reply: str,
|
||
goal: str, # 当前逻辑未使用
|
||
chat_history: List[Dict[str, Any]],
|
||
chat_history_text: str, # 当前逻辑未使用
|
||
current_time_str: str, # 当前逻辑未使用
|
||
retry_count: int = 0, # 当前逻辑未使用
|
||
) -> Tuple[bool, str, bool]:
|
||
"""检查生成的回复是否与机器人之前的发言完全一致（长度大于4）
|
||
|
||
Args:
|
||
reply: 待检查的机器人回复内容
|
||
chat_history: 对话历史记录 (包含用户和机器人的消息字典列表)
|
||
Returns:
|
||
Tuple[bool, str, bool]: (是否合适, 原因, 是否需要重新规划)
|
||
对于重复消息: (False, "机器人尝试发送重复消息", False)
|
||
对于非重复消息: (True, "消息内容未与机器人历史发言重复。", False)
|
||
"""
|
||
if not self.bot_qq_str:
|
||
logger.error(
|
||
f"[私聊][{self.private_name}] ReplyChecker: BOT_QQ 未配置，无法检查{global_config.BOT_NICKNAME}自身消息。"
|
||
)
|
||
return True, "BOT_QQ未配置，跳过重复检查。", False # 无法检查则默认通过
|
||
|
||
# 对当前待发送的回复进行规范化
|
||
normalized_reply = self._normalize_text(reply)
|
||
|
||
if len(normalized_reply) <= 4:
|
||
return True, "消息长度小于等于4字符，跳过重复检查。", False
|
||
|
||
try:
|
||
match_found = False # <--- 用于调试
|
||
for i, msg_dict in enumerate(reversed(chat_history)):
|
||
if not isinstance(msg_dict, dict):
|
||
continue
|
||
|
||
user_info_data = msg_dict.get("user_info")
|
||
if not isinstance(user_info_data, dict):
|
||
continue
|
||
|
||
sender_id = str(user_info_data.get("user_id"))
|
||
|
||
if sender_id == self.bot_qq_str:
|
||
historical_message_text = msg_dict.get("processed_plain_text", "")
|
||
# 对历史消息也进行同样的规范化处理
|
||
normalized_historical_text = self._normalize_text(historical_message_text)
|
||
|
||
logger.debug(
|
||
f"[私聊][{self.private_name}] ReplyChecker: 历史记录 (反向索引 {i}) ({global_config.BOT_NICKNAME}): "
|
||
f"原始='{historical_message_text[:50]}...', 规范化后='{normalized_historical_text[:50]}...'"
|
||
)
|
||
if (
|
||
normalized_reply == normalized_historical_text and len(normalized_reply) > 0
|
||
): # 确保规范化后不为空串才比较
|
||
logger.warning(f"[私聊][{self.private_name}] ReplyChecker: !!! 成功拦截一次复读 !!!")
|
||
logger.warning(
|
||
f"[私聊][{self.private_name}] ReplyChecker 检测到{global_config.BOT_NICKNAME}自身重复消息 (规范化后内容相同): '{normalized_reply[:50]}...'"
|
||
)
|
||
match_found = True
|
||
# 返回: 不合适, 原因, 不需要重规划 (让上层逻辑决定是否重试生成)
|
||
return (False, "机器人尝试发送与历史发言相似的消息 (内容规范化后相同)", False)
|
||
|
||
if not match_found:
|
||
logger.debug(f"[私聊][{self.private_name}] ReplyChecker: 未找到重复内容 (规范化后比较)。")
|
||
return (True, "消息内容未与机器人历史发言重复 (规范化后比较)。", False)
|
||
|
||
except Exception as e:
|
||
import traceback
|
||
|
||
logger.error(f"[私聊][{self.private_name}] ReplyChecker 检查重复时出错: 类型={type(e)}, 值={e}")
|
||
logger.error(f"[私聊][{self.private_name}]{traceback.format_exc()}")
|
||
return (True, f"检查重复时发生内部错误 (规范化检查): {str(e)}", False)
|