初始化 WeChat Agent 项目

MVP Phase 1 核心模块： - src/config/settings.py - 配置管理 - src/vlm/qwen_vl.py - Qwen-VL2 视觉模型接口 - src/wechat/controller.py - 微信客户端控制器（UIAutomation） - src/core/engine.py - 核心引擎（轮询、消息处理、回复） - src/main.py - 主入口文档： - README.md - config.example.yaml - requirements.txt 技术方案：纯视觉AI + UIAutomation - 截图 → Qwen-VL2 识别 → AI 判断 → UIAutomation 操作
2026-04-13 11:12:49 +08:00
commit eb19d8d05f
8 changed files with 1408 additions and 0 deletions
--- a/src/wechat/controller.py
+++ b/src/wechat/controller.py
@@ -0,0 +1,287 @@
+"""
+微信客户端控制器
+WeChat Windows Client Controller using UIAutomation
+"""
+
+import time
+import logging
+from dataclasses import dataclass
+from typing import List, Optional, Tuple
+from pathlib import Path
+
+logger = logging.getLogger(__name__)
+
+# 尝试导入 pywinauto
+try:
+    from pywinauto import Application, WindowSpecification
+    from pywinauto.win32structures import RECT
+    from pywinauto.controls.hwndwrapper import HwndWrapper
+    HAS_PYWINAUTO = True
+except ImportError:
+    HAS_PYWINAUTO = False
+    logger.warning("pywinauto 未安装，请运行: pip install pywinauto")
+
+
+@dataclass
+class ChatMessage:
+    """聊天消息"""
+    sender: str
+    content: str
+    time: str
+    is_self: bool
+
+
+@dataclass
+class WeChatWindow:
+    """微信窗口信息"""
+    hwnd: int
+    title: str
+    rect: RECT
+    isMinimized: bool
+
+
+class WeChatController:
+    """微信客户端控制器"""
+    
+    # 微信窗口类名
+    WEIXIN_WINDOW_CLASS = "WeChatLoginWnd"  # 登录窗口
+    WEIXIN_MAIN_WINDOW_CLASS = "WeChatMainWnd"  # 主窗口
+    
+    def __init__(self, window_title: str = "微信"):
+        if not HAS_PYWINAUTO:
+            raise RuntimeError("pywinauto 未安装，无法控制微信客户端")
+        
+        self.window_title = window_title
+        self.app: Optional[Application] = None
+        self.main_window: Optional[WindowSpecification] = None
+        self._connected = False
+    
+    def connect(self, timeout: float = 10) -> bool:
+        """连接到微信窗口
+        
+        Args:
+            timeout: 超时时间（秒）
+            
+        Returns:
+            是否连接成功
+        """
+        try:
+            # 尝试连接已运行的微信
+            self.app = Application(backend="win32").connect(
+                title=self.window_title,
+                timeout=timeout
+            )
+            self.main_window = self.app.window(title=self.window_title)
+            self._connected = True
+            logger.info("成功连接到微信窗口")
+            return True
+        except Exception as e:
+            logger.error(f"连接微信窗口失败: {e}")
+            return False
+    
+    def find_wechat_window(self) -> Optional[WeChatWindow]:
+        """查找微信窗口"""
+        import ctypes
+        from ctypes import wintypes
+        
+        user32 = ctypes.windll.user32
+        
+        @ctypes.WINFUNCTYPE(wintypes.BOOL, wintypes.HWND, wintypes.LPARAM)
+        def enum_callback(hwnd, lparam):
+            length = user32.GetWindowTextLengthW(hwnd)
+            if length > 0:
+                buff = ctypes.create_unicode_buffer(length + 1)
+                user32.GetWindowTextW(hwnd, buff, length + 1)
+                title = buff.value
+                if self.window_title in title:
+                    rect = RECT()
+                    user32.GetWindowRect(hwnd, ctypes.byref(rect))
+                    is_min = user32.IsIconic(hwnd)
+                    wechat_win = WeChatWindow(
+                        hwnd=hwnd,
+                        title=title,
+                        rect=rect,
+                        isMinimized=bool(is_min)
+                    )
+                    # 存储到列表
+                    windows.append(wechat_win)
+            return True
+        
+        windows = []
+        user32.EnumWindows(enum_callback, 0)
+        
+        if windows:
+            logger.info(f"找到 {len(windows)} 个微信窗口")
+            return windows[0]
+        return None
+    
+    def screenshot(self, output_path: str = None) -> str:
+        """截图
+        
+        Args:
+            output_path: 保存路径，为 None 则保存到临时文件
+            
+        Returns:
+            截图路径
+        """
+        if not self.main_window:
+            raise RuntimeError("未连接微信窗口")
+        
+        if output_path is None:
+            import tempfile
+            output_path = Path(tempfile.gettempdir()) / f"wechat_screenshot_{int(time.time())}.png"
+        else:
+            output_path = Path(output_path)
+        
+        try:
+            # 激活窗口（如果最小化）
+            if self.main_window.is_minimized():
+                self.main_window.restore()
+            
+            # 截图
+            self.main_window.capture_as_image().save(str(output_path))
+            logger.debug(f"截图已保存: {output_path}")
+            return str(output_path)
+        except Exception as e:
+            logger.error(f"截图失败: {e}")
+            raise
+    
+    def get_chat_list(self) -> List[str]:
+        """获取聊天列表"""
+        # 聊天列表在左侧，需要根据具体窗口结构调整
+        # 这里需要根据实际 UI 结构调整
+        try:
+            # 查找子窗口
+            chat_list = self.main_window.window(
+                class_name="ChatListBox"  # 假设的类名
+            )
+            items = chat_list.items()
+            return [item.text() for item in items]
+        except Exception as e:
+            logger.warning(f"获取聊天列表失败: {e}")
+            return []
+    
+    def click_on_chat(self, chat_name: str) -> bool:
+        """点击聊天"""
+        try:
+            # 查找聊天项
+            chat_list = self.main_window.window(class_name="ChatListBox")
+            chat_item = chat_list.window(title=chat_name)
+            chat_item.click()
+            logger.info(f"点击聊天: {chat_name}")
+            return True
+        except Exception as e:
+            logger.error(f"点击聊天失败: {e}")
+            return False
+    
+    def get_message_list(self, count: int = 10) -> List[ChatMessage]:
+        """获取消息列表
+        
+        Args:
+            count: 最多获取的消息数
+            
+        Returns:
+            消息列表
+        """
+        messages = []
+        try:
+            # 查找消息列表窗口
+            msg_list = self.main_window.window(class_name="ChatMessageList")
+            
+            # 获取消息项
+            items = msg_list.items()[-count:]  # 取最后 N 条
+            for item in items:
+                text = item.text()
+                # 解析消息格式
+                # 格式: [发送者] 时间\n内容
+                msg = self._parse_message(text)
+                if msg:
+                    messages.append(msg)
+                    
+        except Exception as e:
+            logger.warning(f"获取消息列表失败: {e}")
+        
+        return messages
+    
+    def _parse_message(self, text: str) -> Optional[ChatMessage]:
+        """解析消息文本"""
+        import re
+        
+        # 简单解析
+        # 格式: 发送者 时间\n内容
+        pattern = r"(.+?)\s+(\d{2}:\d{2})\n([\s\S]+)"
+        match = re.match(pattern, text)
+        if match:
+            sender = match.group(1).strip()
+            time_str = match.group(2).strip()
+            content = match.group(3).strip()
+            is_self = sender == "我"
+            return ChatMessage(sender=sender, content=content, time=time_str, is_self=is_self)
+        return None
+    
+    def send_text(self, text: str) -> bool:
+        """发送文本消息
+        
+        Args:
+            text: 要发送的文本
+            
+        Returns:
+            是否发送成功
+        """
+        try:
+            # 查找输入框
+            input_box = self.main_window.window(class_name="Edit")
+            
+            # 清空并输入
+            input_box.set_edit_text("")
+            input_box.type_keys("^a")  # 全选
+            input_box.type_keys("{DELETE}")
+            input_box.set_edit_text(text)
+            
+            # 按回车发送
+            input_box.type_keys("{ENTER}")
+            
+            logger.info(f"发送消息: {text[:20]}...")
+            return True
+        except Exception as e:
+            logger.error(f"发送消息失败: {e}")
+            return False
+    
+    def find_button(self, name: str) -> Optional[HwndWrapper]:
+        """查找按钮"""
+        try:
+            btn = self.main_window.window(title=name, class_name="Button")
+            return btn
+        except Exception:
+            return None
+    
+    def click_button(self, name: str) -> bool:
+        """点击按钮"""
+        try:
+            btn = self.find_button(name)
+            if btn:
+                btn.click()
+                return True
+            return False
+        except Exception as e:
+            logger.error(f"点击按钮失败: {e}")
+            return False
+    
+    def is_connected(self) -> bool:
+        """检查是否已连接"""
+        return self._connected and self.app is not None
+    
+    def disconnect(self):
+        """断开连接"""
+        if self.app:
+            self.app = None
+            self.main_window = None
+            self._connected = False
+            logger.info("已断开微信连接")
+    
+    def __enter__(self):
+        self.connect()
+        return self
+    
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        self.disconnect()