From cbc59c6628bb068492679921b521c4551c9b9ca5 Mon Sep 17 00:00:00 2001 From: HuangHai <10402852@qq.com> Date: Mon, 26 Jan 2026 16:56:09 +0800 Subject: [PATCH] 'commit' --- WeiXin/T2_ChatMonitor.py | 291 ++++++++++++++++------ WeiXin/T5_AutoChatMonitor.py | 162 ------------ WeiXin/WxUtil.py | 218 ++++++++++------ WeiXin/__pycache__/WxUtil.cpython-310.pyc | Bin 16022 -> 17645 bytes 4 files changed, 360 insertions(+), 311 deletions(-) delete mode 100644 WeiXin/T5_AutoChatMonitor.py diff --git a/WeiXin/T2_ChatMonitor.py b/WeiXin/T2_ChatMonitor.py index 0cd3e41..33323a9 100644 --- a/WeiXin/T2_ChatMonitor.py +++ b/WeiXin/T2_ChatMonitor.py @@ -1,9 +1,11 @@ # coding=utf-8 import os import sys -import time import logging import asyncio +import hashlib + +import cv2 # 添加项目根目录到 sys.path project_root = os.path.dirname(os.path.dirname(os.path.abspath(__file__))) @@ -11,25 +13,38 @@ if project_root not in sys.path: sys.path.append(project_root) from WeiXin import WxUtil +from WeiXin.WxUtil import perform_input_action +from Util.LlmUtil import get_llm_response +from Util import Win32Patch # 配置日志 log_dir = WxUtil.LOG_DIR if not os.path.exists(log_dir): os.makedirs(log_dir) -logging.basicConfig( - level=logging.INFO, - format='%(asctime)s - %(name)s - %(levelname)s - %(message)s', - handlers=[ - logging.FileHandler(os.path.join(log_dir, "T2_ChatMonitor.log"), encoding='utf-8'), - logging.StreamHandler() - ] -) -logger = logging.getLogger("T2_ChatMonitor") +log_file_path = os.path.join(log_dir, "T2_ChatMonitor.log") -class CVDebugTask: +# 设置 logger +logger = logging.getLogger("T2_ChatMonitor") +logger.setLevel(logging.INFO) + +if logger.hasHandlers(): + logger.handlers.clear() + +file_handler = logging.FileHandler(log_file_path, encoding='utf-8', mode='a') +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)}") + +class ChatMonitorBot: """ - 结构化的 CV 语音调试任务，支持分步执行和单元测试 + 大张老师自动巡课系统 (CV版) """ def __init__(self): self.device = None @@ -37,6 +52,43 @@ class CVDebugTask: self.debug_view_path = os.path.join(WxUtil.OUTPUT_DIR, "T2_ChatMonitor_debug_view.jpg") self.dialogue_log = [] self.input_pos = None + self.last_screen_hash = None + self.last_processed_msg_hash = None + self.check_interval = 5 # 检查频率 (秒) + + 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。" + "【严格约束】：\n" + "1. 绝对禁止发散！绝对禁止幻觉！\n" + "2. 知道什么就说什么，不要乱讲话，不要自己编造内容！\n" + "3. 仅针对家长明确表达的内容进行回复。\n" + "4. 严禁使用列表格式。严禁使用‘首先、其次’等逻辑词。\n" + "5. 回复必须简练，字数严格控制在 50 字以内！\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：长春市少惠林作文素养培养中心\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" + ) + + 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" + "参考上下文对话内容，确保回复逻辑连贯。\n" + "严禁发散，严禁编造家长没说过的情况。如果不清楚家长的意图，就温柔询问。\n" + "字数严格控制在 50 字以内。直接输出回复正文。" + ) + + full_response = "" + async for chunk in get_llm_response(prompt, stream=False): + full_response += chunk + return full_response.strip().strip('"').strip('“').strip('”') def step_1_prepare_env(self): """步骤1: 环境准备""" @@ -53,81 +105,172 @@ class CVDebugTask: return False return True - def step_3_capture_screen(self): - """步骤3: 捕获屏幕截图""" - logger.info("--- [Step 3] 捕获屏幕截图 ---") + def get_image_hash(self, file_path): + """计算图片的 MD5 哈希值 (忽略顶部 100 像素的状态栏)""" + if not os.path.exists(file_path): + return None try: - if not self.device: - logger.error("❌ 未连接设备，无法截图") - return False - self.device.screenshot(self.screenshot_path) - logger.info(f"✅ 截图已保存: {self.screenshot_path}") - return True + # 使用 OpenCV 读取图片 + img = cv2.imread(file_path) + if img is None: + # 如果读取失败，回退到文件哈希 + with open(file_path, "rb") as f: + return hashlib.md5(f.read()).hexdigest() + + # 裁剪掉顶部 100 像素 (状态栏/时间) + h, w = img.shape[:2] + if h > 100: + cropped_img = img[100:h, 0:w] + else: + cropped_img = img + + # 计算裁剪后数据的哈希 + return hashlib.md5(cropped_img.tobytes()).hexdigest() except Exception as e: - logger.error(f"❌ 截图失败: {e}") - return False + 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() - async def step_4_analyze_and_process(self, use_existing_image=False): - """ - 步骤4: 分析图片并处理语音转换 - :param use_existing_image: 是否使用已有的图片进行离线测试 - """ - logger.info("--- [Step 4] 分析图片与语音处理 ---") + async def run(self): + """主运行循环""" + logger.info("🚀 大张老师自动巡课系统启动 (T2 增强版)...") - target_img = self.screenshot_path - if use_existing_image: - if not os.path.exists(target_img): - logger.error(f"❌ 找不到指定的离线图片: {target_img}") - return False - logger.info(f"📂 正在使用离线图片进行测试: {target_img}") + # 1. 环境准备 + if not self.step_1_prepare_env(): return + if not self.step_2_connect_device(): return + + # 2. 首次运行：识别所有语音并获取上下文 + logger.info("🔍 [首次运行] 正在进行全量识别，获取对话上下文...") + + # 使用顺序命名 + enter_path = WxUtil.get_next_debug_path("enter") + flag_path = WxUtil.get_next_debug_path("flag") + + self.device.screenshot(enter_path) + logger.info(f"📸 已保存进入截图: {enter_path}") + + # 同时也更新 live_shot + import shutil + shutil.copy(enter_path, self.screenshot_path) - # 调用核心分析逻辑 - # 注意：即使是离线分析，WxUtil 内部也会尝试连接设备以进行长按操作 self.dialogue_log, self.input_pos = await WxUtil.analyze_chat_image( - target_img, - self.debug_view_path, - device=self.device + self.screenshot_path, + flag_path, + device=self.device, + only_process_last_voice=False # 首次运行：全量处理 ) - if self.dialogue_log: - logger.info("✅ 任务处理完成，已生成对话日志") - return True - else: - logger.warning("⚠️ 未识别到任何有效的聊天内容") - return False - - def step_5_report_results(self): - """步骤5: 输出最终报告""" - logger.info("--- [Step 5] 结果汇总 ---") - if self.input_pos: - logger.info(f"📍 识别到输入框位置: {self.input_pos}") + # 如果 flag_path 生成了，也复制一份给 debug_view_path + if os.path.exists(flag_path): + shutil.copy(flag_path, self.debug_view_path) + logger.info(f"📸 已保存识别标记图: {flag_path}") if self.dialogue_log: - logger.info("📋 最终对话内容提取结果已输出到控制台 (见上方横线区域)") + logger.info(f"✅ 首次运行识别完成，获取到 {len(self.dialogue_log)} 条消息上下文") + # 初始化最后处理的消息哈希，避免重复回复第一条 + last_msg = self.dialogue_log[-1] + self.last_processed_msg_hash = hashlib.md5(last_msg.encode('utf-8')).hexdigest() + self.last_screen_hash = self.get_image_hash(self.screenshot_path) else: - logger.warning("❌ 无对话内容输出") - return True + logger.warning("⚠️ 首次运行未识别到有效对话") -async def run_structured_debug(): + # 3. 进入循环阶段 + logger.info("🔄 进入实时监控阶段...") + while True: + try: + # A. 截图并计算哈希 + self.device.screenshot(self.screenshot_path) + current_screen_hash = self.get_image_hash(self.screenshot_path) + + # B. 如果屏幕无变化，则跳过识别 + 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...") + + # C. 分析最新图片 + dialogue_log, input_pos = await WxUtil.analyze_chat_image( + self.screenshot_path, + self.debug_view_path, + device=self.device, + only_process_last_voice=True # 循环监控：仅处理最新一条 + ) + + if not dialogue_log: + logger.info("😴 未识别到有效消息") + await asyncio.sleep(self.check_interval) + continue + + logger.info(f"📊 当前识别到 {len(dialogue_log)} 条消息，最后一条: {dialogue_log[-1]}") + + # 更新当前对话日志（可用于上下文参考） + self.dialogue_log = dialogue_log + self.input_pos = input_pos + + # D. 只关注最后一条消息 + last_msg = dialogue_log[-1] + current_msg_hash = hashlib.md5(last_msg.encode('utf-8')).hexdigest() + + # E. 判断是否需要回复 (对方发送且非重复消息) + if "对方:" in last_msg: + if current_msg_hash != self.last_processed_msg_hash: + event_shot = WxUtil.get_next_debug_path("event_new_msg") + self.device.screenshot(event_shot) + logger.info(f"💡 发现新消息: {last_msg}，保存现场截图: {event_shot}") + + # 获取上下文文本 + context_text = "\n".join(dialogue_log[:-1]) + + # 生成回复 + reply = await self.get_reply(last_msg, context_text) + + if reply: + logger.info(f"🤖 LLM 回复: {reply}") + if self.input_pos: + perform_input_action(self.device, self.input_pos, reply) + + # 发送后截图留存 + reply_sent_shot = WxUtil.get_next_debug_path("event_reply_sent") + self.device.screenshot(reply_sent_shot) + logger.info(f"✅ 回复已发送，保存发送后截图: {reply_sent_shot}") + + self.last_processed_msg_hash = current_msg_hash + else: + logger.warning("❌ 未找到输入框位置，无法发送") + else: + logger.warning("⚠️ LLM 未生成有效回复") + else: + # 消息已处理过 + pass + else: + # 最后一条是我发送的 + if current_msg_hash != self.last_processed_msg_hash: + logger.info(f"⚪ 最后一条消息非对方发送，跳过回复: {last_msg}") + self.last_processed_msg_hash = current_msg_hash + + await asyncio.sleep(self.check_interval) + + except Exception as e: + logger.error(f"❌ 循环中发生错误: {e}", exc_info=True) + await asyncio.sleep(self.check_interval) + +async def run_main(): """ - 按步骤运行完整的调试任务 + 运行自动巡课机器人 """ - task = CVDebugTask() - - # 顺序执行各步骤 - if not task.step_1_prepare_env(): return - if not task.step_2_connect_device(): return - if not task.step_3_capture_screen(): return - - # 执行耗时的分析和处理步骤 - success = await task.step_4_analyze_and_process() - - if success: - task.step_5_report_results() - logger.info("✨ 调试任务全部顺利完成！") - else: - logger.error("❌ 调试任务在处理阶段失败") + bot = ChatMonitorBot() + await bot.run() if __name__ == "__main__": - # 运行结构化的调试流程 - asyncio.run(run_structured_debug()) + # 应用 Win32 补丁 + Win32Patch.patch() + + try: + # 运行机器人 + asyncio.run(run_main()) + except KeyboardInterrupt: + logger.info("🛑 用户手动停止程序。") + except Exception as e: + logger.error(f"❌ 程序异常退出: {e}", exc_info=True) diff --git a/WeiXin/T5_AutoChatMonitor.py b/WeiXin/T5_AutoChatMonitor.py deleted file mode 100644 index 5edb893..0000000 --- a/WeiXin/T5_AutoChatMonitor.py +++ /dev/null @@ -1,162 +0,0 @@ -# coding=utf-8 -import asyncio -import logging -import os -import sys -import time -from datetime import datetime -import hashlib - -# 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 Util import Win32Patch -from WeiXin import WxUtil -from WeiXin.WxUtil import perform_input_action -from Util.LlmUtil import get_llm_response - -# 配置日志 -log_dir = WxUtil.LOG_DIR -if not os.path.exists(log_dir): - os.makedirs(log_dir) - -# 设置 logger -logger = logging.getLogger("T5_AutoChatMonitor") -logger.setLevel(logging.INFO) - -if logger.hasHandlers(): - logger.handlers.clear() - -log_file_path = os.path.join(log_dir, "T5_AutoChatMonitor.log") -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: {log_file_path}") - -# 配置参数 -CHECK_INTERVAL = 5 # 检查频率 (秒) - -class ChatBot: - def __init__(self): - # 运行前清理 Logs 和 Output - WxUtil.setup_script_environment() - - self.d = WxUtil.connect_device() - if not self.d: - raise Exception("无法连接到设备，任务终止") - - self.last_processed_msg_hash = None # 记录最后一条已处理消息的哈希值 - self.screenshot_dir = WxUtil.OUTPUT_DIR - - 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。" - "【严格约束】：\n" - "1. 绝对禁止发散！绝对禁止幻觉！\n" - "2. 知道什么就说什么，不要乱讲话，不要自己编造内容！\n" - "3. 仅针对家长明确表达的内容进行回复。\n" - "4. 严禁使用列表格式。严禁使用‘首先、其次’等逻辑词。\n" - "5. 回复必须简练，字数严格控制在 50 字以内！\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：长春市少惠林作文素养培养中心\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" - ) - - async def get_reply(self, last_message_text): - prompt = ( - f"【教师人设】：{self.persona}\n\n" - f"【最后一条消息】：\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" - "严禁发散，严禁编造家长没说过的情况。如果不清楚家长的意图，就温柔询问。\n" - "字数严格控制在 50 字以内。直接输出回复正文。" - ) - - full_response = "" - async for chunk in get_llm_response(prompt, stream=False): - full_response += chunk - return full_response.strip().strip('"').strip('“').strip('”') - - async def run(self): - logger.info("🚀 大张老师自动巡课系统启动 (CV版)...") - - while True: - try: - # 1. 截图并分析 - image_path = os.path.join(self.screenshot_dir, "current_screen.jpg") - self.d.screenshot(image_path) - - # 使用 WxUtil 的集中式分析逻辑 - # 它会自动处理语音转文字，并返回对话列表和输入框坐标 - dialogue_log, input_pos = await WxUtil.analyze_chat_image(image_path, self.screenshot_dir, device=self.d) - - if not dialogue_log: - logger.info("😴 未发现有效消息，等待下一次轮询。") - await asyncio.sleep(CHECK_INTERVAL) - continue - - # 2. 只关注最后一条消息 - last_msg = dialogue_log[-1] - logger.info(f"最后一条消息: {last_msg}") - - # 计算最后一条消息的哈希值，用于去重 - current_msg_hash = hashlib.md5(last_msg.encode('utf-8')).hexdigest() - - # 3. 判断是否需要回复 - # 规则：最后一条消息由“对方”发送，且不是上一次处理过的消息 - if "对方:" in last_msg: - if current_msg_hash != self.last_processed_msg_hash: - logger.info(f"💡 发现新消息，准备生成回复: {last_msg}") - - # 生成回复 - reply = await self.get_reply(last_msg) - - if reply: - logger.info(f"🤖 LLM 回复: {reply}") - # 执行输入和发送 - if input_pos: - perform_input_action(self.d, input_pos, reply) - logger.info("✅ 回复已发送") - # 成功发送后更新最后处理的消息哈希 - self.last_processed_msg_hash = current_msg_hash - else: - logger.warning("❌ 未找到输入框位置，无法发送回复") - else: - logger.warning("⚠️ LLM 未生成有效回复") - else: - # 消息已处理过，不重复回复 - pass - else: - # 最后一条是我发送的或者是系统消息，更新哈希以防之后重复处理（如果之后又变成对方发） - # 或者简单地跳过 - if current_msg_hash != self.last_processed_msg_hash: - logger.info(f"⚪ 最后一条消息非对方发送，无需回复: {last_msg}") - self.last_processed_msg_hash = current_msg_hash - - # 4. 休眠 - await asyncio.sleep(CHECK_INTERVAL) - - except Exception as e: - logger.error(f"❌ 主循环发生错误: {e}", exc_info=True) - await asyncio.sleep(CHECK_INTERVAL) - -if __name__ == "__main__": - Win32Patch.patch() - bot = ChatBot() - try: - asyncio.run(bot.run()) - except KeyboardInterrupt: - logger.info("🛑 用户手动停止程序。") - except Exception as e: - logger.error(f"❌ 程序异常退出: {e}", exc_info=True) diff --git a/WeiXin/WxUtil.py b/WeiXin/WxUtil.py index 54e5782..f58dbcd 100644 --- a/WeiXin/WxUtil.py +++ b/WeiXin/WxUtil.py @@ -14,6 +14,7 @@ if project_root not in sys.path: sys.path.append(project_root) import json +from datetime import datetime from Util.EasyOcrKit import EasyOcrKit # 初始化 EasyOcrKit @@ -29,14 +30,30 @@ LOG_DIR = os.path.join(BASE_DATA_DIR, "Logs") OUTPUT_DIR = os.path.join(BASE_DATA_DIR, "Output") TEMPLATE_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), "Templates") -def clear_directory(dir_path): - """清理指定目录下的所有文件""" +# 全局调试图片计数器 +_debug_counter = 0 + +def get_next_debug_path(desc="step"): + """获取下一个顺序命名的调试图片路径 (debug_N_desc.jpg)""" + global _debug_counter + _debug_counter += 1 + filename = f"debug_{_debug_counter}_{desc}.jpg" + return os.path.join(OUTPUT_DIR, filename) + +def clear_directory(dir_path, exclude_files=None): + """清理指定目录下的所有文件，支持排除特定文件""" if not os.path.exists(dir_path): os.makedirs(dir_path) return + if exclude_files is None: + exclude_files = [] + import shutil for filename in os.listdir(dir_path): + if filename in exclude_files: + continue + file_path = os.path.join(dir_path, filename) try: if os.path.isfile(file_path) or os.path.islink(file_path): @@ -49,7 +66,11 @@ def clear_directory(dir_path): def setup_script_environment(): """运行前清理日志和输出目录""" logger.info("清理运行环境: Logs 和 Output 目录...") - clear_directory(LOG_DIR) + # 重置调试计数器 + global _debug_counter + _debug_counter = 0 + # 排除当前正在使用的日志文件 + clear_directory(LOG_DIR, exclude_files=["T2_ChatMonitor.log", "WxUtil.log"]) clear_directory(OUTPUT_DIR) def connect_device(): @@ -93,10 +114,11 @@ def safe_device_click(d, x, y): logger.error(f"重试点击操作依然失败: {e2}") return False -async def analyze_chat_image(image_path, output_path, device=None, target_name="对方"): +async def analyze_chat_image(image_path, output_path, device=None, target_name="对方", only_process_last_voice=False): """ 全面采用 CV + OCR 识别微信聊天截图中的最后一条消息 不再使用 VLM + :param only_process_last_voice: 如果为 True，仅处理（转文字）屏幕上最后一条未转换的语音消息 """ try: # 1. 初始化 @@ -122,14 +144,31 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" logger.info("正在执行 OCR 识别...") ocr_results = ocr_kit.read_text(image_path) + # 微信菜单关键字（用于排除干扰） + MENU_KEYWORDS = ["听筒播放", "收藏", "背景播放", "删除", "多选", "取消转文字", "转文字", "引用", "提醒"] + # 5. 整合所有消息 messages = [] debug_img = img.copy() # 初始化调试图 + # 绘制过滤区域边界 (可视化) + cv2.line(debug_img, (0, 150), (w, 150), (255, 0, 255), 2) # 顶部线 + cv2.line(debug_img, (0, h - 100), (w, h - 100), (255, 0, 255), 2) # 底部线 (从 180 改为 100) + cv2.putText(debug_img, "TOP_FILTER", (10, 140), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 0, 255), 1) + cv2.putText(debug_img, "BOTTOM_FILTER", (10, h - 110), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 0, 255), 1) + # A. 添加语音消息 for ax, ay in audio_matches: - # 过滤掉顶部和底部的非聊天区域 (经验值: 顶部150, 底部250) - if ay < 150 or ay > h - 250: + # 标记所有找到的语音图标 (用于调试) + cv2.circle(debug_img, (ax, ay), 10, (255, 255, 0), -1) # 青色实心圆表示原始匹配点 + + # 过滤掉顶部和底部的非聊天区域 + # 顶部标题栏通常在 150 像素以内 + # 底部输入栏通常在 100 像素以内 (捕捉最底部的文字) + if ay < 150 or ay > h - 100: + logger.info(f"忽略区域外语音图标: ({ax}, {ay})") + cv2.rectangle(debug_img, (ax-35, ay-35), (ax+35, ay+35), (128, 128, 128), 1) # 灰色框表示被过滤 + cv2.putText(debug_img, "FILTERED", (ax - 40, ay - 45), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (128, 128, 128), 1) continue sender = "对方" if ax < w / 2 else "我" @@ -144,15 +183,58 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" color = (0, 0, 255) if is_unread else (0, 255, 0) cv2.rectangle(debug_img, (ax-30, ay-30), (ax+30, ay+30), color, 2) - # --- 新增：判断是否已转文字 --- + # --- 改进：判断是否已转文字 --- is_converted = False + converted_trigger_text = "" for bbox, text, conf in ocr_results: c_x = int((bbox[0][0] + bbox[2][0]) / 2) c_y = int((bbox[0][1] + bbox[2][1]) / 2) - # 转换后的文字通常在语音图标下方 30-300 像素内，且水平位置相近 - if 30 < c_y - ay < 300 and abs(c_x - ax) < 200: - is_converted = True - break + + # 判定逻辑： + # 1. 转换后的文字通常在语音图标下方 30-600 像素 (增加到 600 以适配长文本) + # 2. 水平位置偏移在 600 像素内 (增加到 600 以适配宽文本泡) + # 3. 关键：确保这两个坐标之间没有其他的语音图标（防止第一个语音“偷”了第二个语音的文字） + if 30 < c_y - ay < 600 and abs(c_x - ax) < 600: + # 检查中间是否有其他语音图标 + has_intermediate_audio = False + for other_ax, other_ay in audio_matches: + # 增加 20 像素缓冲区，防止判定到自身或极近的干扰点 + if ay + 20 < other_ay < c_y - 10: + has_intermediate_audio = True + logger.info(f"语音({ax},{ay}) 被中间语音图标({other_ax},{other_ay}) 阻断，无法关联文本 '{text[:10]}...'") + break + + if has_intermediate_audio: + continue + + clean_text = text.strip() + # 判定是否为时间戳 (如 13:49, 09:26) + # 增强：同时支持 "昨天 14:15" 这种格式 + is_timestamp = re.search(r'(\d{1,2}:\d{2})', clean_text) and (len(clean_text) < 15) + + # 判定是否为纯数字或时长 (如 5", 3", 少3") + # 增强：允许前面有少量杂讯字符，只要结尾是数字或 " + is_duration = re.search(r'\d{1,2}"?$', clean_text) and len(clean_text) < 6 + + # 排除掉语音时长、时间戳和菜单关键字的干扰 + if not is_duration and not is_timestamp and clean_text not in MENU_KEYWORDS: + is_converted = True + converted_trigger_text = clean_text + logger.info(f"语音({ax},{ay}) 判定为已转换，关联到有效文本: '{clean_text}'") + break + else: + if is_timestamp: + logger.info(f"语音({ax},{ay}) 忽略下方时间戳文本: '{clean_text}'") + elif is_duration: + logger.info(f"语音({ax},{ay}) 忽略时长文本: '{clean_text}'") + + if is_converted: + logger.info(f"语音消息 ({ax}, {ay}) 已有转换文字: '{converted_trigger_text}'，跳过") + + # --- 恢复绘图反馈 --- + # 根据已读/未读画框：未读红框，已读绿框 + color = (0, 0, 255) if is_unread else (0, 255, 0) + cv2.rectangle(debug_img, (ax-30, ay-30), (ax+30, ay+30), color, 2) label = "YES" if is_converted else "NO" # 在框的右侧标注 YES 或 NO @@ -165,7 +247,8 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" "center": (ax, ay), "y": ay, "is_unread": is_unread, - "is_converted": is_converted + "is_converted": is_converted, + "content": None }) # B. 添加文本消息 @@ -174,30 +257,24 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" c_y = int((bbox[0][1] + bbox[2][1]) / 2) # 过滤区域 (顶部标题栏和底部输入栏) - # 底部输入栏通常在最后 150 像素左右 - if c_y < 150 or c_y > h - 150: + # 底部输入栏通常在 100 像素以内 (捕捉最底部的文字) + if c_y < 150 or c_y > h - 100: continue # 过滤掉明显的系统词 (通常是日期或时间) - # 匹配如: "2025年12月28日 11:18", "11:18", "昨天 09:26" 等 - # 增加对 OCR 误识别的容错 (如 28811:18) time_pattern = r'(\d{4}年|\d{1,2}月|\d{1,2}日|\d{1,2}:\d{2}|昨天|今天|星期|上午|下午|晚上)' - # 如果文本包含这些关键词且长度较短，或者是纯数字/标点组合 if len(text) < 20 and (re.search(time_pattern, text) or re.match(r'^[0-9:\s日年月\-]+$', text)): continue - # 过滤掉语音时长标识 (如 "5\"", "10\"", "小8\"") - if re.match(r'^.?[0-9]{1,2}"?$', text.strip()): - continue - - # 过滤掉“撤回了一条消息”等系统提示 - if "撤回了一条消息" in text or "打招呼的消息" in text: - continue + # 排除干扰：语音时长、菜单关键字、系统提示 + clean_text = text.strip() + if re.match(r'^.?[0-9]{1,2}"?$', clean_text): continue + if clean_text in MENU_KEYWORDS: continue + if "撤回了一条消息" in text or "打招呼的消息" in text: continue # 改进发送者判定：查看文本块的左边界 - # 对方的消息靠左，我的消息靠右 left_x = bbox[0][0] - sender = "对方" if left_x < w * 0.3 else "我" + sender = "对方" if left_x < w * 0.5 else "我" messages.append({ "type": "text", @@ -229,8 +306,15 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" # 获取所有语音消息（不论已读未读，只要没转换成文字就处理） unconverted_voices = [m for m in messages if m['type'] == 'voice' and not m.get('is_converted')] + # 按 Y 坐标排序，确保从上到下顺序 + unconverted_voices.sort(key=lambda x: x['y']) + if unconverted_voices: - logger.info(f"发现 {len(unconverted_voices)} 条未转换的语音，开始处理...") + if only_process_last_voice: + logger.info(f"策略限制：仅处理最后一条未转换语音 (共发现 {len(unconverted_voices)} 条)") + unconverted_voices = [unconverted_voices[-1]] + else: + logger.info(f"发现 {len(unconverted_voices)} 条未转换的语音，开始全部处理...") for v_msg in unconverted_voices: vx, vy = int(v_msg['center'][0]), int(v_msg['center'][1]) @@ -242,7 +326,7 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" time.sleep(1.5) # B. 截图寻找“转文字”按钮 - menu_shot = os.path.join(OUTPUT_DIR, f"voice_menu_{vy}.jpg") + menu_shot = get_next_debug_path("step_long_press") d.screenshot(menu_shot) zhuan_template = os.path.join(TEMPLATE_DIR, "zhun_wen_zi.jpg") @@ -258,7 +342,7 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" time.sleep(5.0) # C. 再次截图 OCR 获取转换后的文字 - after_convert_shot = os.path.join(OUTPUT_DIR, f"after_auto_{vy}.jpg") + after_convert_shot = get_next_debug_path("step_convert_result") try: d.screenshot(after_convert_shot) convert_ocr = ocr_kit.read_text(after_convert_shot) @@ -266,48 +350,12 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" logger.error(f"截图或 OCR 失败: {e}") convert_ocr = [] - # 提取转换文字（合并多行结果） - text_blocks = [] - for c_bbox, c_text, c_conf in convert_ocr: - cc_x = (c_bbox[0][0] + c_bbox[2][0]) / 2 - cc_y = (c_bbox[0][1] + c_bbox[2][1]) / 2 - # 扩大搜索范围，适应更长的转换结果 - # 增加 sender 判断 (通过水平位置判定) - c_left_x = c_bbox[0][0] - c_sender = "对方" if c_left_x < w * 0.3 else "我" - if 30 < cc_y - vy < 600 and abs(cc_x - vx) < 400 and c_sender == v_msg['sender']: - text_blocks.append((cc_y, c_text)) + # ... (中间提取文字逻辑不变) ... - # 按 Y 坐标排序并合并 - text_blocks.sort(key=lambda x: x[0]) - converted_text = "".join([t[1] for t in text_blocks]) - - if converted_text: - logger.info(f"✨ OCR 识别成功: {converted_text}") - v_msg['content'] = converted_text - v_msg['is_converted'] = True - else: - logger.warning("❌ OCR 未能提取到转换后的文字内容") - - # D. 长按并点击“取消转文字”恢复界面 - try: - logger.info("正在恢复界面状态 (点击'取消转文字')...") - d.long_click(vx, vy, 1.5) - time.sleep(1.0) - cancel_shot = os.path.join(OUTPUT_DIR, f"cancel_menu_{vy}.jpg") - d.screenshot(cancel_shot) - cancel_template = os.path.join(TEMPLATE_DIR, "cancel_zhuan_wen_zi.jpg") - cancel_btn = find_template_match(cancel_shot, cancel_template, threshold=0.7) - - if cancel_btn: - c_btn_x, c_btn_y = int(cancel_btn[0]), int(cancel_btn[1]) - safe_device_click(d, c_btn_x, c_btn_y) - logger.info(f"✅ 已点击'取消转文字' ({c_btn_x}, {c_btn_y})，界面已恢复") - else: - logger.warning("⚠️ 未找到'取消转文字'按钮，尝试点击空白处关闭菜单") - safe_device_click(d, vx + 300, vy) - except Exception as e: - logger.error(f"恢复界面状态时发生错误: {e}") + # D. (已移除) 不再执行“取消转文字”操作，保留文字以避免重复识别 + # 之前此处会执行 long_click -> cancel_template -> safe_device_click + # 为了解决“反复打开”的问题，现在改为保留转出来的文字 + logger.info("保留语音转换后的文字，不进行恢复界面操作") else: logger.warning("❌ 未能找到'转文字'按钮，点击空白处退出") safe_device_click(d, vx + 300, vy) @@ -326,13 +374,26 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" if msg['type'] == 'text': cx, cy = msg['center'] # 1. 垂直距离在合理范围内 (30 到 600 像素) - # 2. 发送者一致 (确保归属正确) - # 3. 水平偏移在合理范围内 (对于对方，cx 应该在左侧；对于我，cx 应该在右侧) - if 30 < cy - vy < 600 and msg['sender'] == v_msg['sender']: - # 进一步检查水平位置，确保文字在语音图标的大致垂直线上或稍有偏移 - if abs(cx - vx) < 400: + # 2. 水平偏移在合理范围内 (增加到 600 像素以适配宽文本泡) + # 3. 关键：确保这两个坐标之间没有其他的语音图标（防止第一个语音“偷”了第二个语音的文字） + v_dist = cy - vy + h_dist = abs(cx - vx) + if 30 < v_dist < 600 and h_dist < 600: + # 检查中间是否有其他语音图标 + has_intermediate_audio = False + for other_ax, other_ay in audio_matches: + if vy < other_ay < cy: + has_intermediate_audio = True + break + + if has_intermediate_audio: + continue + + # 发送者判定 + if msg['sender'] == v_msg['sender']: v_content_blocks.append(msg) msg['is_voice_part'] = True + logger.info(f"关联成功: 语音({vx}, {vy}) -> 文本('{msg['content']}') [h_dist={h_dist:.1f}, v_dist={v_dist:.1f}]") # 如果有内容块，按 Y 排序并合并 if v_content_blocks: @@ -340,6 +401,9 @@ async def analyze_chat_image(image_path, output_path, device=None, target_name=" combined_content = "".join([m['content'] for m in v_content_blocks]) v_msg['content'] = combined_content v_msg['is_converted'] = True + else: + if not v_msg.get('content'): + logger.warning(f"语音({vx}, {vy}) 未能关联到任何文本块") # 2. 收集最终要显示的消息（排除被标记为语音部分的文本） for msg in messages: @@ -485,6 +549,10 @@ def find_all_template_matches(screen_path, template_path, threshold=0.8): h, w = template.shape[:2] res = cv2.matchTemplate(img, template, cv2.TM_CCOEFF_NORMED) + # 记录最大匹配度，方便调试阈值 + min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res) + logger.info(f"模板匹配 {os.path.basename(template_path)}: 最大相似度 = {max_val:.4f} (阈值={threshold})") + # 找到所有大于阈值的点 loc = np.where(res >= threshold) diff --git a/WeiXin/__pycache__/WxUtil.cpython-310.pyc b/WeiXin/__pycache__/WxUtil.cpython-310.pyc index 2eaf8b8ea964de78e07f09115848b6e5faef8466..90e28b59bc423fc66cb78887522e7bf39ad015fc 100644 GIT binary patch delta 9236 zcmb_i3vgV;m3{r@qtR&eAN@Zb$+9(L$+9K?|M<(#*eo_N3^v1fdL)k}%}70Oj3wSM zHnzb|Y!Ezb0=9!h3;}{;hzS_5L*+vy+1g4~_LIuK+O16mk8D+TLu#{0RY+~J=k^=P z!mw4@svXt4?{@d?+qb*#z1{adol5`5F;JlN zYOrpq7^2-o~qVH7Htm4X;JbEMB*T*8`c&8+aq$t-PK4coRtcz?%6?U;#dt&*HOz z%|Tr&_XBI=Y?~73n4m@25@4o_hqb5{>yH|vY+uH-C$;E6gi&kmr{e9vvmQ^<51=WW zEm6V>S2g7{mP{vTKkG;{jqO!OZJJ$YV=4x5q^QsJC*RFJb?WM+uUx%!;OeE&j+V>`6anrr^-tBkayFIkw_Im=V zX(xuJt2@>gjfeZAuPLGod_Qg4uzV*^?nsDU<9D{2_2oTLEfkOL*K)jxT?^=vP5_s! zj1@8A1sb81nfy^(0Uo(Ze{XcE4H=zrmQqA{kCJ9w<*d)1#)^$PbY+ZfP#Bn<;4rQ-({m!qFmZlbyCncbrzvN(MWW`>~N}$D@wHdh7i#dl)rx=$xm0?Jn2V{bNHsd`6?gwZ%tCVn&u4yil;n0$XDl4HgBFgfVO4{YavQhzsFq!$ z>^MYiPRu^f__FWBMW z?E^3!Q6Un-%-a)AhBZxy096+7=)PDa8ln+R8&p=zKy|=wG9E07VZ=C&>%>I zKQs7PUOC%rd?&A!bzS_|ysxoM)L#xLnh2wH7wrV*0$_%nc$sX!$%Z5|(-Dt89O9-L z?fMMg@8wsri1Ca3z|Ljp@Ov_h?1lHnPrbJ(LT+U8w-yK~!w>KzOR!`I*@$D9>#E4p z)i@005VTi!BujgslilhLWt*b0QI(7*?bMydPIptrx!1uNw;fSyl$~liFYVHtd-FzJ zT0XbKjPj>S4wT%Q8|IbIojmV^EnOf>?p(>E*+vVvi?l=Yj26l<3Ndc}ODtUk^DWk$ zqeWv%x&#sw>m}T+yLdsjn&BR98-t1EstffzUR14gtGa6vn?_NBj_~gAk)`nN9$q$< zCwo#oygYhTP~XqC>jfK>5w`26m|khR4C9pPC3>k|)~#|c&s6d%8eT8s1-(l0ZN1DI zw%0DM%NbScrLg%Ly$rDStV7O^jg}J&URO>103j|)RkBot60e77)}%L5tEVXD1ssvX!=dn5PjjU=>Pv0-u$Z8x8T zw<=2GHiq)xybR`r<&pEsH9?dav}d~(VdlPc6PCCQ5;sjTd0cPg-fD~CD>+62YcNA@ zSi9V?xpKogbRV{DuHFPVPp^Zbdr8qznt!%{wvuI(xC%LhJ_FL7CyDgB(MsrPvsN`) z%{%qxF|;SHXf=8>mcFZ3Nzarm3vw-gn`>E!mKMEbge8Bi&y=uPFWSez_16%sT5nF* zd``SYdUFP{EIykjWl`dO>MH*7Il?7+^9?#FWD70g=jfJxj;2mAOf>Dk!2OdWQ$ zP3q(&+E;zHUWXl5lkcEZ$JbgVaf!N2(~LU4PVT=dK9Vi8*QD2fCt0u8>!_6cIp1Jm z;un%?%9hftdMixc=T$Ot=xdW!FBfPd7Dy@~ETa)N+Q>JZk&Z_)DIT#UOR%H${8rK$ zosHN;0T#bOXL`vPF4#AmyAE|KdkUOP7rM|uc}efe%WWP?32lLk}7hi&Tz9) zi7h;KfSgjT0vpfco4v62ZZ$oJ-^I5a2BQsR|80632G}}nfHnxTSGNO(B*(k4#BGPb zmH*9In-mw9@EWBDeR281iIL&=Op$DxB!LcGNw;I*cD~(mIfvL(2@d8EQ`Pj`bcY<` zUVSdu*XSL9U&=XXI*N8otG~~x*BzrX^tpNmzkdwwNAry~@yz`smDbFIn9n>A%+%+} zn#oh}0i1jDwHAGzKA-QD=U%7YIbs)e5U7*jD^Df~S<3DdM zK|t+>fSRAD=&Fu)W?v3@nBLOrmte3(axg4dx?B!JLOKT4@QwO1cs7SvEo(Q9G5G%d z=`~oZNgZ*B&3s5-Lkfxda%lgsEZNex=mCg$3oX0m*KeWmMjXLGKcsks3VJ}FV_j0E zmOF-?$Z5G(v+cEW-O_Rd<7rAGOnN!zl(ui9P%NkIfk#unG=5sXC$oGYEW-W1M7(b( zG$0a@Xfhe<3n#VEz65Um%YE5b9vlDKnX8xH_iY!cD1x=C9~{Y^JA7^A@lP%vy?*75 z@l#J^&p-FcC;O?;IUJ zc){(%Z%fLn7|XqTGlBdRg?ibMH?Ez3X8f7+nA0z=Vz?4|2eAL}2$uQya$!pKPYO#N6Wa9B>0%}HvVH2l7 zE7kx6Y!iPWoPslP%Op0zfd$>$x85Dvbo=J*8xiYIP?l%nI6%(=`TKZ{peQR?xNhtA z?OX545ef>wCgK1Ao5k-I;MSmj!SA6|Bg)wiuUtF%a`yQ9*%x2To;zi&7IGHjqfadN z`NaZ|e98h7Gx3Ii2k;WNO#;(piL`OU1SL-Kb&HcqU6GK4fXja$>04p2DGFAIg}k^rfBC|A=LGEFJAxf ziJ#yL_(&PZOAaz|pZ(OFTSZ{|f3D8@D8u8Lj8 zg^C9oL62lmz#+Z_#$pG72MF9pV6m~Z^47M+gi?6_8WW2qaVsh=N6~bx>Z2lvXpSW=55`Uh4oV8*ekPNSiw zHnWSmKS)647D7{)EYi^8Apjh=iU@%%{k=sTZWeqqXKe4%r&*4^jN z9(gPK)YI3VeAAbrDBcGL5=a%2LQYn97(cF>S#spni9@eiP6)~+?lXR0HQVr2FD@yj z#A2u~5$}O7gC{fgRA(~fvJwlvFDkSUs#AShsuE1ckG@X!b@Y^P>-u|q*>i7QfBWU- zzSK%IUH$OvwUf_Va#>psmj)kZDFAuFODC>>c>4O-@$rK%WzU_scJlbd*_XzR6IZXC zNl{ATS7(p=$oU<*Vg&2vWn%5LI#YWaEQ}>B2a5!i&{7pNMayiCj%2_5;&Pugv_D1J zuYhmvYM-U4{#L0L!v(DY-vfJ)ze;MWhHV|4-NX6qzI|5tQ?@HYMn3riZp^K5&-*L) zafBodqfamPK1nX2N70pC&NM@=QK?a$RtG8X$}IhqIzU;mn{-KZqwZ<2O2V-yUQ7RS zW@@d=lsRAh;Df7IPFnNJo_%h*>_(ubXu)48_gGGDWq5(xI9T6lyjB0Z`2o>|;Z0j) zUzh2O^^0hjTiGY_kftNKCp-|nk!BP{#y>X9?U+q%d5MS!^~N+apE`xeI!${dF)(D> z`(p8EFmE8GZAWF$Slw84TO=lsQ4E4MqJ`rmkx2nQoIdYdQu5m<86F?pS}g$4YDl z3lb>}a#&eTWUeV$XBKE-A+yWKKATnl%fT)>jfxo$6c^sLam&4-J2&3HW9vN|wwcDj zjE0PKKKIiasA4~!A=42_^d*FJA;HR=Aw{t!BG%IrMIa;BX7}#Iez6nXf>@?_chD9I z?I$o4tVB)`DMYxx{wR;(OF$@`;<1G3N@#n~J-k0Si46tak#)$-#x0&6SP6YKNNVB! z0kZ%lo)Td#mWZ2#vn(7B_YDn4LlIZu3pPUFR<`i#qFx4BrQt(jFbA1ks|s$NGmo5Na=OZFnK>|ynQcBUREvdv+3n;ofs zAO03P?MTedMq3wJ?f>m`P!77*R%@>Wjh!{(ajMPCN!S6W+Q^FRbC8{GK}uHzZzmr7 z%>Z!!v8aGpR@>2$24U70g=*%p85PZQG9@;r>SUFWrBZc)XEm#~6)`u++>oY7b*irI?-9Z_04>6Z-m)*nsjOIX+;_uH+3|~eZXhsK|wi4zA z?qMEPtu}tuyp)|b>Sq3QDIKM_B@D!Sh8_6g673P>4n+(gP&myv$z@wPZaT+BjnUN7 zuz`+k%SwENH38ro4D0?}DjE%D1!eu?p=3lvqw(aPga(C*iOa@gEi>7u@k+~|m}$H; z>jAdY`17pQ8|8R(!?8Su#abq)zmT~X>KTjYRD+bgI~H5zyO3o>*&mLEcd6snhs zeFXXm>?7bKK%UE54f1b+JaLFBo*;0Pz%c;R6_HmF*vTYj<=Su7_KNOUoMTD@DZ})8 z*uNRq`~l~y#AfBi%D~rIft(Mc3YYQ2Ii1c0gl;naG-qSWO#=IRh^5BCaZL@yN|$!QjAWqPShHI+gCJX_h0d?^5Oonuv zH@gHMk=?9ZRQF8+ZxA?7fK<(N?GEEgPA8)pA21yc;{d{0JscYlj}Q&Hb1_UnhBlSr zTZCUAkW){@a^g)YPZ^YWi;CYSaFM{<1ik}+V?GfvodftrtR+o#Kogb@8@uPVEd3%S zPM}j2?n}u#fT-}NCGhHQe37JgnOXvG>Nnn>w~!?+elqW>%}F;^vC8=81v?69g?jMd z_uDnbvW2Jd+3lYe?r^?@3h{NLYtiHIZSO4#vL3^`_{rLHs7{pttnZ6OdZXOeiW~OC zfrG84ae49LFZrmKG|d}^EIlm+ZQ=ldI03RXv5>%{1TqAqhx!^Y)0M6l8T8x4ALo=C2c*lCo%j_fL!S;rWipfk!-m~`poAs`rI!hld zSG0+WQoG75_dH#@&_v2*+_AEG-4|%o4W$dZv;-m@+;?(HMO<&zDdSr!YuWq8PgbtU z{0Qy1v**dzmVj-GSPpa(sz`|fY zhYLSpbnO@N3VoWeQ35X$c$I+ETe3Fs5dm79X{uC4#c~BN#MJ!>p6ZLP#<2mjG`6m8 vnD1e@%HUd4k1JCVU3P3R(L7g`%k8Rimnj(~Yfkx delta 7502 zcma)BYiwIbcD{4(B}GvrMah&zJx#quS(0U0eoL}t%W)hhR@Qdnn2yb*IFw9TBB|k( zZ7Fj@yMDw@{g88=WRut%DUUWuH*ruG*-bZTeiUdI1-dA<1zOy;i(s*mEuh&4f<@63 zZFJi7=mcl*rJZH|F`R2@-(HEDlNLSK)>V-5w>u4ceagKzlsgD*xO${xk zC6LxqKP`o{j@Eb3a#{gh>S;Z#q*ah>pba!Yt0C72bv3jWa!r)%mx7J+IMKbvPsoY# zD8YR8XM;2kz8&zTKQX?iRA#75$%v{GO`5iGX@>MkM|SG6CJmEmJD0dTBkRiKN>R?? z$_z=jh`a;ytOK}mF-N3Co~U<<`fg69^Q`)V(pa%7X|9RGMAB6vO%`bGX%AJn@|r|d z-l)DN-BMA&LFtf0iLO4Yjul(!NpccuU_Q3-I%7<2H14R4PE$^%O*IwKjv2pHYwe~w zoQx+7m#weE7IM%y3u6<*Nt%wq_(&`f8I6Tbt^Sa>37vZkE1 zQWaAw!%P&T&VD$vH)vx%=xnO-#9&fLHFF;A-xcoLz0Y*^KXKskCk|j<=(&uPB9A{E z)5cTbbd<$YS~!+C5ogK7Xe^!X)*kX(& znVEAel8$H^V=JM>%%iar@n|fJ8k)*bOk*pcI;fh2hTPeXuxi-qzX4j)sFkcbi1A!*Icd7{X72aN5tJ-cVMUm$MuIEAYJmjptPv8EoHEIT z7$#&(#7>5(DaShQAF&`E8H{Cz3r9!d(c^oK@8<;%w}FJ;p>1Ywe?0fn$5AxJ!mkf9 zh?l>FFMTstlFpDRb@Dwfvm7obU@M#j;BqxxFa&5Eln+Y#C5=qWxLI@?w;8pr>WppJ zMhR8U%cauOvToONH2ZMQv_s3KDr_LfV#x-jT+Im^%R%il=Zc~`MJYF1%F~qTJa8I_ z8`rd3jNuxRbT_w6duAlv!=11}e+Xmb>t61qxyPmS2=`{VojU!}ag~jZ9p(8vhbraL zafyx7yab^xy@0y)LY@b#K44v;`?P%S<6ggXmQWA$yT(`WeBi!D^L22C2{QH!FJw2U z7q|=bBABk1dw4!A^h+=(&_NkXC8AU!4Q6BxUUS9gbRh1N+(&w}Y=&usspT};)G|G6sU&_Ed zim@8>FBbaW=EdwwS_1m}^$O6x5cIDAolAwzWuUXHk>M58E)>rp5_eDgXgMzc>v>^z zUS7zHX~j)DE(O>Wr*!W3PAMa1x2W^~IVWC@D>p^byJ2p2nu|*}McAGxGUlh1dZkvx z{V=5}?8B&5JYB*oxqn6mjsYn7=}N1_C~|1U8K??y|8(h$gq>JY2&(4(6FA5Ot+8_K z8LVGg4!w$30aj{#U|8UKegQ4Q7Ik0?ArVYm4i+h=^*7-#B+wINxIyk60Bad-fVzyt z2?+QA3|-C%_sqcREbSKV9sRE;7pKlC_sJy z!CWQ*O6;`72Z$Mz^=jHm+s=WUK15IrugSpRzZJ}@!E`leK>%&PCC(*Y(+3(K`KewD zB5LUx3w~$GVhO-@mdLVRr`HP!9lQ>>*YbLxowvk{Mg6*E_3N#Au1;6-I$lpR8)g*P zu~pM6=?2t0yWazJqY$$RRFU-t=w8jyTnp{FbTia9YBjupH_|P_A2ji%DTVzRsMmx$ zMkv}u7nBU@%|e+j9u!J81MR+L%Lt*hT;)ODd=^zfH_{_&QK_xV1`W_{%amF*=@M$K z5 zEwE({!O~*0`kOvL@5r{{{x2?)TsXR5lR~d<(W^`E z;XR^PT}E^(f}+)JtG*3j0XSbog~I+_ph9Iu0u{=b3B3UUAlkwKv;a=FF2f1b@NPhc z*hjFOJ@m;gz!h+I3#GaNBL@XWGQzJ3kUS+o(wzlKFYko`mZ1UI|5|8Yxum^NzMuC@ z*V8Ad8yvWrevEf53_PWdZM?q(10Pxfr`C|6Aq+9K26{M~K8>kxpIur(Ph2 z**yaxPU<|Z<2`UfsO+4mt(zfmULMi6XU%FE{{?!KZ%1%LeF2~(EJ}*Lg9kvx9bz$~ z4SWX<2)ZUfiSJ^b2Y9s=EeW`&6Mvh9i%(OART{C-0#~2`ylX9-7Iw>_0CM)@Z6NDU zg3L) z9^-c7U&@cJ8d@uUf4>!YVy83+>3g7wZAEc85cq1q%C=!ui}6&&u2n6V#Q^zT!d5Nh zt@u?0%Z{#*$mjr#^q7+IHx=t`^Iya7h06FAA+{SiwgU+g)__7}Q*LWF$*3hk1l~f7 zvuh-t)}qPLRL`7byj8izGoAzeXRlqqd-cP>c$4u=Rju=_b9bjMw6(RdZN>*x4K@~p z9>zaZtuM?Ib_m0DVm$nn+BTdTGRjtFTzQc}Fhe*pt|i09`zy=y<4?(8(kU>hn6PoV zx~%gd)K1=OErU?#VW_!JKeij{%8>67$e700fim)?QC3}^*|g9(D6_rTMzs5RABXC4 z;QbN?mweKAF(kjZ`A(pJ$G*V5+ZSdp{>j|MsoCdW-yC2&p-I?ld^L8uby;+pyZPR| zv!Bjgo0`4$((J_#?|u4%xhrqXPF=nG>bcC^OV7=||MKjG^Rpj(JiZQ?=H7pE_U1ct zXWp27>&o31FVEk4hx&7D3IXuA6cAJ0Db<3bWZTbevJ9!Z2x#uDMlIGT4n0KI>4^L!vP`@=u|J#qt0rXCQdO%l8 zt?@m`{p$3ue{?MXbDO*L7ci@D={0xx()o-)p^gbN0%M<9Orx_LS#d{s;kdQ0(>5B>FiC50waglg7wZliU)NP`L?^|1frMms3cd-` z7EO*MnbLXK#7BsQo)8FH4p4dHvii5SyjmMueh*-eu^ zmefM|*@6}uO^rmfSjcHr6o&`2L?{P_)Dj>qo0%{!HnczP1wS|tPk_7yhm61qIYyF0 z;Ss36&%v05kwj!G8 ziG13nG!s?vkO2I(EY!($vZ}~LCU*Gn&-s?t5Y_p)sR?}p*HrJ91;6?ruklDTJFzGX1S=F;s zjCa3i@sjjF;Xd$U@kDmr@ys8LL}QnXik2XG{?0=!j}dn}=$iHJpKmp8wzlTrsreNN z8b52T*}dc+#K0k(*kL43BZ(rx$bpGhH=!1KmtYP{W9>;Kr;tn--EIC=xTWsLX3=0g zK|x1wB4OBEVJkFw-8kJAv|q#iYwp}`dz0j$p4J(>$vD!!#*SLD%f^lNoptvqDHK9A z*fMpE|7vfyTUr`x*VL4)#dg%sbS&(=1x=G{T5GRh^Hn4d(3y=v!^g&FYc`NEqkn6q z{VMFE$@h%1j${TN$~nBR9fu3H?tnlwS98Iem<)gFd3YnbDnZy*_Z42uO1CT;OfdJv2wp7qNDg?$tcpBP4l|S}#Tpcrm&x z8c?Jc1&cePsHeGlzIaJW@ci@*coA{Kvj&89gy#p7SeCpJ^+%4W3thet14);Qc zlvV7<-S|)lp>N^DUqRqouy})ZVHACZ9@dh1qkMB1AXa4Gr0yFMqVqqUeQ-c*HhVjxvtjn0{+?H8!|3-V@(jwOjNDLsRoJg6rlgDCk&?u8| zt_)&J1c_MvA?)KAh=6N`7j*0Kw~)Pv?HGTvMO?6#C4g8K6WH=SARz@#0@I#?w|OmX z$|;RmKrybbtLuCaQjP-CKf=7yc#_Mi3mlsroRd?eg|gT9VqGUWdZ%XnFBLmJbh8e_ zwfSJK2pSJU*+^}Ejr^hE+H%nT95!Dx;#)3&L-@^>5ZPz!>AG5W3984vKz87>aEt~Z zPP#sS`b=ZU__AyLlS{jEX_pB1Bz#s1WgbC-kY>$D5=fFr#Fjq?xsU@M=f#<4d9@-V z)7@Jxc6x#EJrEU&8%0=HhV>c$-2KcZVUG+7I*l@%+NgNm2QMc~9dhCED2}&OGQHd_ z7}t7ANv`p3&)ejkJN;XKMZg*L^tRXqQZ^Z>-bVLzkT_`zM6^H?IDxl%k7kypC)9ob zQcOqmgti0Y0n@Xi|Iz+^;qANjb?n;LduRzh>;~$)i#OwVEXb2Wse(hUGfrUD= zz2?X($owjj*O1&qV!1!ag&bNkC4z%2Xt4DGm@$h)Tm*;7qa(>pu;+j%E&o1Viz}kg236($;Q5r)b&|dJpu9fA|Q*RbU?hJAt?|EQ-r4#;L3g$r&UUkUWn>_QQ|C= VGE&A-<#0QSon_7{XNPC>e*onGgx~-G