# coding=utf-8 import asyncio import hashlib import json import logging import os import re import sys import threading import time import cv2 # 添加项目根目录到 sys.path project_root = os.path.dirname(os.path.dirname(os.path.abspath(__file__))) if project_root not in sys.path: sys.path.append(project_root) from WeiXin import WxUtil from Util.LlmUtil import get_llm_response from Util import Win32Patch from Util.AlyTtsKit import QwenTTSManager # 配置日志 log_dir = WxUtil.LOG_DIR if not os.path.exists(log_dir): os.makedirs(log_dir) log_file_path = os.path.join(log_dir, "ChatMonitor_Audio.log") # 设置 logger logger = logging.getLogger("ChatMonitor_Audio") logger.setLevel(logging.INFO) if logger.hasHandlers(): logger.handlers.clear() file_handler = logging.FileHandler(log_file_path, encoding='utf-8', mode='w') file_handler.setFormatter(logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')) logger.addHandler(file_handler) stream_handler = logging.StreamHandler() stream_handler.setFormatter(logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')) logger.addHandler(stream_handler) logger.propagate = False logger.info(f"🚀 日志文件路径: {os.path.abspath(log_file_path)}") # 同时将 WxUtil 的日志也输出到同一个文件 wx_logger = logging.getLogger("WxUtil") wx_logger.propagate = False if not any(isinstance(h, logging.FileHandler) and os.path.abspath(h.baseFilename) == os.path.abspath(log_file_path) for h in wx_logger.handlers): wx_logger.addHandler(file_handler) wx_logger.addHandler(stream_handler) class ChatMonitorAudioBot: """ 大张老师自动巡课系统 (语音版) """ def __init__(self, debug_mode=False): self.device = None self.debug_mode = debug_mode # 同步设置 WxUtil 的调试模式 WxUtil.set_debug_mode(debug_mode) self.screenshot_path = os.path.join(WxUtil.OUTPUT_DIR, "AudioMonitor_live_shot.jpg") self.debug_view_path = os.path.join(WxUtil.OUTPUT_DIR, "AudioMonitor_debug_view.jpg") self.dialogue_log = [] self.input_pos = None self.last_screen_hash = None self.last_processed_msg_hash = None self.processed_hashes = set() self.processed_meta = set() self.check_interval = 3 # 检查频率 (秒) # TTS 配置 self.voice_id = "qwen-tts-vc-guanyu-voice-20260131160431051-8e51" self.tts_manager = None self.persona = ( "你是一名1999年毕业、拥有27年一线教学经验的小学高级女教师,名叫‘大张老师’。你目前在‘长春市少惠林作文素养培养中心’工作。" "你不仅是一位作文教学专家,更是一位心思细腻、能与家长共情的教育智者。" "你的回复风格应该是:温柔、知性、亲切,就像一位邻家大姐姐在聊天。" "【关于作文评价的回复逻辑】:\n" "**仅当且仅当**家长明确询问孩子今天的作文情况(例如:“孩子今天写得怎么样?”、“今天作文有进步吗?”)时,才触发此逻辑。触发时,请随机生成具体、专业的反馈。反馈应包含:\n" "1. 孩子今天在写作中遇到的具体小问题(如修辞手法运用不当、开头略显生硬等);\n" "2. 值得表扬的亮点(如观察细致、某个句子写得很传神等);\n" "3. 对字迹的评价(如字迹工整、卷面整洁等);\n" "4. 针对性的课外阅读建议(推荐具体类型的书目)。\n" "此类特定回复字数可放宽,控制在 150 字以内,确保语音回复时长在 30 秒内。\n" "【关于通用咨询的回复逻辑】:\n" "如果家长是咨询学校信息、地址、课程时间、年级设置,或者是刚加好友打招呼,请务必保持礼貌、知性、亲切,直接回答相关问题。严禁在这些情况下提及孩子的作文表现或虚构作文反馈。\n" "【严格约束】:\n" "1. 绝对禁止输出任何括号内的内容(如动作、神态描述、心理活动等)!例如不要输出‘(微笑)’或‘(亲切地注视)’。\n" "2. 你的回复是直接用于语音播放的,请只输出你想说的话,不要包含任何舞台指导文字。\n" "3. 绝对禁止发散!绝对禁止幻觉!\n" "4. 知道什么就说什么,不要乱讲话,不要自己编造内容!\n" "5. 仅针对家长明确表达的内容进行回复。\n" "6. 严禁使用列表格式。严禁使用‘首先、其次’等逻辑词。\n" "7. 普通咨询回复字数严格控制在 50 字以内,专业作文评价回复控制在 150 字以内!\n" "8. 说话要有人情味,不要回答得太生硬或太简短。例如当家长问‘学校叫什么名字’时,不要只说‘少惠林’,而应该说‘我们的学校名字叫少惠林。’或者‘咱们这儿叫少惠林。’,这样才显得亲切、有礼貌。\n" "9. 对方问什么就答什么。回答要完整、体面,但不要过度发散地址和电话(除非被问到)。\n" "如果涉及到校区信息,必须且只能使用以下真实数据:\n" "- 单位/学校名称:长春市少惠林作文素养培养中心(简称:少惠林)\n" "- 地址:南环城路与临河街交汇,TOUCH12街3楼325号\n" "- 联系人:小张老师(电话:18686619970)\n" "- 每学期开学招收小学三年级至六年级,初中七年级的学生入学,其它年段不招生。\n" ) def _record_processed_hash(self, msg, msg_hash): """记录已处理的消息哈希和元数据 (仅内存)""" self.processed_hashes.add(msg_hash) if msg: meta = (msg.get("sender", ""), msg.get("time_display", ""), msg.get("type", "")) self.processed_meta.add(meta) if len(self.processed_hashes) > 100: temp = list(self.processed_hashes)[-100:] self.processed_hashes = set(temp) if len(self.processed_meta) > 100: temp_meta = list(self.processed_meta)[-100:] self.processed_meta = set(temp_meta) async def get_reply(self, last_message_text, context_text=""): prompt = ( f"【教师人设】:{self.persona}\n\n" f"【上下文对话内容】:\n{context_text}\n\n" f"【最后一条待回复消息】:\n{last_message_text}\n\n" "【任务要求】:\n" "请作为大张老师回复家长。**必须且只能针对最后一条消息进行回复!**\n" "1. **意图识别**:判断家长是否在询问“孩子今天的作文写得怎么样”。\n" "2. **条件回复**:\n" " - 如果是询问作文,请严格按【关于作文评价的回复逻辑】生成 150 字以内的详细专业反馈。\n" " - 如果是咨询学校地址、课程、年级、打招呼等通用问题,请按【关于通用咨询的回复逻辑】简洁回答,严禁提到任何关于孩子作文的内容,字数控制在 50 字以内。\n" "3. **格式要求**:严禁发散,直接输出回复正文,不包含任何括号内的动作描述。\n" "4. **口语化处理**:由于回复将用于语音播放,请确保称呼自然。如果称呼家长,请直接使用‘糖豆爸爸’、‘糖豆妈妈’等格式,不要在中间加标点符号,也不要使用‘某某的爸爸’这种书面语,直接称呼‘糖豆爸爸’会更亲切自然。" ) full_response = "" async for chunk in get_llm_response(prompt, stream=False): full_response += chunk reply = full_response.strip().strip('"').strip('“').strip('”') # [User Requested] 过滤掉括号内容 (如动作描述),防止 TTS 播放 reply = re.sub(r'\(.*?\)', '', reply) reply = re.sub(r'(.*?)', '', reply) # [Optimization] 口语化称呼预处理,确保 TTS 衔接自然 # 针对“XX爸爸/妈妈”中间可能出现的空格或异常停顿符号进行清理 reply = re.sub(r'(爸爸|妈妈|爷爷|奶奶|姥姥|姥爷)\s+', r'\1', reply) # 移除称谓后的空格 return reply.strip() def step_1_prepare_env(self): """步骤1: 环境准备""" logger.info("--- [Step 1] 环境准备 ---") WxUtil.setup_script_environment() try: self.tts_manager = QwenTTSManager() logger.info("✅ TTS 引擎初始化成功") except Exception as e: logger.error(f"❌ TTS 引擎初始化失败: {e}") return False return True def step_2_connect_device(self): """步骤2: 连接设备""" logger.info("--- [Step 2] 连接设备 ---") self.device = WxUtil.connect_device() if not self.device: logger.error("❌ 设备连接失败,请检查手机是否连接且开启了调试模式") return False return True def get_image_hash(self, file_path): """计算图片的 MD5 哈希值 (忽略顶部 150 像素的状态栏)""" if not os.path.exists(file_path): return None try: img = cv2.imread(file_path) if img is None: with open(file_path, "rb") as f: return hashlib.md5(f.read()).hexdigest() h, w = img.shape[:2] if h > 150: cropped_img = img[150:h, 0:w] else: cropped_img = img return hashlib.md5(cropped_img.tobytes()).hexdigest() except Exception as e: logger.error(f"计算哈希出错: {e}, 回退到文件哈希") with open(file_path, "rb") as f: return hashlib.md5(f.read()).hexdigest() def get_stable_message_hash(self, msg): """计算消息的稳定哈希值""" if not msg: return "" stable_data = { "sender": msg.get("sender", ""), "content": msg.get("content") or "", "time_display": msg.get("time_display", ""), "type": msg.get("type", "") } msg_str = json.dumps(stable_data, sort_keys=True, ensure_ascii=False) return hashlib.md5(msg_str.encode('utf-8')).hexdigest() def _tts_worker(self, text, finished_event, start_event): """TTS 播音工作线程""" try: logger.info(f"[TTS] 启动语音合成,音色ID: {self.voice_id}") self.tts_manager.start_synthesis(self.voice_id, [text], wait_finished=False, buffer_seconds=0.3) self.tts_manager.wait_for_playback_start() start_event.set() if self.tts_manager.callback: self.tts_manager.callback.wait_for_finished() logger.info("[TTS] 语音播放完成") except Exception as e: logger.error(f"[TTS] 播放过程中发生异常: {e}") finally: finished_event.set() async def send_voice_reply(self, reply_text): """发送语音回复""" logger.info(f"🎤 准备发送语音回复: {reply_text}") # 1. 确保处于语音输入模式 tmp_screen = os.path.join(WxUtil.OUTPUT_DIR, "audio_reply_check.jpg") self.device.screenshot(tmp_screen) press_say_template = os.path.join(WxUtil.TEMPLATE_DIR, "press_say.jpg") pos = WxUtil.match_template_center(tmp_screen, press_say_template, threshold=0.8) if not pos: logger.info(">>> 未发现 '按住说话' 按钮,尝试切换模式...") audio_reply_template = os.path.join(WxUtil.TEMPLATE_DIR, "audio_reply.jpg") switch_pos = WxUtil.match_template_center(tmp_screen, audio_reply_template, threshold=0.8) if switch_pos: self.device.click(switch_pos[0], switch_pos[1]) await asyncio.sleep(1.5) self.device.screenshot(tmp_screen) pos = WxUtil.match_template_center(tmp_screen, press_say_template, threshold=0.8) if not pos: logger.error("❌ 无法定位到 '按住说话' 按钮") return False # 2. 启动 TTS 线程 finished_event = threading.Event() playback_start_event = threading.Event() tts_thread = threading.Thread(target=self._tts_worker, args=(reply_text, finished_event, playback_start_event)) # 3. 执行录音动作 logger.info(">>> 立即按住发送语音按钮...") self.device.touch.down(pos[0], pos[1]) logger.info(">>> 等待 0.5 秒确保微信进入录音状态...") await asyncio.sleep(0.5) logger.info(">>> 启动 TTS 播音...") tts_thread.start() # 等待播放开始 playback_start_event.wait(timeout=10) # 等待播放结束 start_time = time.time() while not finished_event.is_set(): await asyncio.sleep(0.1) if time.time() - start_time > 60: logger.warning("录音超时,强制结束") break # 4. 释放按钮 self.device.touch.up(pos[0], pos[1]) logger.info(">>> 录音结束,语音已发送") tts_thread.join() return True async def run(self): """主运行循环""" logger.info("🚀 正在启动 ChatMonitor_Audio (Voice-Reply)...") if not self.step_1_prepare_env(): return if not self.step_2_connect_device(): return logger.info("🚀 启动完成,进入实时监控阶段...") while True: try: self.device.screenshot(self.screenshot_path) # [User Requested] 检查是否在聊天界面 if not WxUtil.check_is_chat_interface(self.screenshot_path): logger.info("当前不在聊天界面,跳过本次循环,等待中...") await asyncio.sleep(self.check_interval) continue current_screen_hash = self.get_image_hash(self.screenshot_path) if current_screen_hash == self.last_screen_hash: await asyncio.sleep(self.check_interval) continue self.last_screen_hash = current_screen_hash logger.info("📸 屏幕发生变化,正在分析...") dialogue_log, input_pos = await WxUtil.analyze_chat_image( self.screenshot_path, self.debug_view_path, device=self.device, process_strategy="UNREAD", restore_processed_voice=False ) if not dialogue_log: await asyncio.sleep(self.check_interval) continue self.dialogue_log = dialogue_log self.input_pos = input_pos # [Optimization] 核心逻辑:在 Voice-Reply 模式下,只有当存在"对方"发来的、带红点的未读消息时,才进行处理 unread_voices = [m for m in dialogue_log if m.get('type') == 'voice' and m.get('is_unread') and m.get('sender') != "我"] if not unread_voices: if self.debug_mode: logger.debug("当前屏幕无带红点的未读语音消息,跳过分析") await asyncio.sleep(self.check_interval) continue # 进一步确认最后一条消息不是"我"发的(双重保险) chat_msgs = [m for m in dialogue_log if m.get('type') in ['text', 'voice']] if not chat_msgs: await asyncio.sleep(self.check_interval) continue absolute_last_msg = chat_msgs[-1] if absolute_last_msg.get('sender') == "我": if self.debug_mode: logger.debug("最后一条消息是'我'发送的,跳过回复") await asyncio.sleep(self.check_interval) continue # 准备处理最新的未读语音 last_msg = unread_voices[-1] last_message_text = last_msg.get('content') or "" msg_type = last_msg.get('type') current_msg_hash = self.get_stable_message_hash(last_msg) is_processed = current_msg_hash in self.processed_hashes if is_processed: if current_msg_hash != self.last_processed_msg_hash: self.last_processed_msg_hash = current_msg_hash await asyncio.sleep(self.check_interval) continue if not is_processed and current_msg_hash != self.last_processed_msg_hash: logger.info(f"💡 发现新消息 [{last_msg.get('type')}]: {last_msg.get('content')}") msg_shot_path = os.path.join(WxUtil.OUTPUT_DIR, f"NewMsg_{int(time.time())}.jpg") self.device.screenshot(msg_shot_path) context_text = "\n".join([f"{m.get('time_display', '') + ' ' if m.get('time_display') else ''}{m.get('sender')}: {m.get('content')}" for m in dialogue_log if m != last_msg]) last_content = last_msg.get('content') or "" if last_msg.get('type') == 'voice' and not last_content.strip(): logger.info("检测到未成功转换的语音消息,尝试强制重试 OCR 转换...") dialogue_log_retry, _ = await WxUtil.analyze_chat_image( self.screenshot_path, self.debug_view_path, device=self.device, process_strategy="LAST", restore_processed_voice=False ) if dialogue_log_retry: retry_other_msgs = [m for m in dialogue_log_retry if m.get('sender') != "我"] if retry_other_msgs: last_msg = retry_other_msgs[-1] last_content = last_msg.get('content') or "" current_msg_hash = self.get_stable_message_hash(last_msg) if current_msg_hash in self.processed_hashes: self.last_processed_msg_hash = current_msg_hash continue if last_msg.get('type') == 'voice' and not last_content.strip(): logger.warning("语音消息内容为空,暂不生成回复") await asyncio.sleep(self.check_interval) continue reply = await self.get_reply(last_content, context_text) if reply: logger.info(f"LLM 建议回复: {reply}") # 发送语音回复 success = await self.send_voice_reply(reply) if success: logger.info(">>> 语音回复发送成功 <<<") self._record_processed_hash(last_msg, current_msg_hash) self.last_processed_msg_hash = current_msg_hash else: logger.error("语音回复动作执行失败") else: logger.info("LLM 认为无需回复") self._record_processed_hash(last_msg, current_msg_hash) self.last_processed_msg_hash = current_msg_hash await asyncio.sleep(self.check_interval) except Exception as e: logger.error(f"Error in monitoring loop: {e}", exc_info=True) await asyncio.sleep(self.check_interval) async def run_main(): # 默认关闭调试模式以提高响应速度,如需调试可设为 True bot = ChatMonitorAudioBot(debug_mode=False) await bot.run() if __name__ == "__main__": Win32Patch.patch() asyncio.run(run_main())