恢复端口为16025

改进:
- 页面加载等待时间上限提高到 60秒
- 添加等待 networkidle 状态逻辑
- 前端界面增加等待设置说明
- 提示用户针对验证网站设置 10-30秒等待

使用场景:
- Cloudflare 等验证网站
- 复杂加载过程网站
- 需要用户确认的页面

app.py

@@ -20,8 +20,17 @@ from datetime import datetime
 try:
     from playwright.async_api import async_playwright
     PLAYWRIGHT_AVAILABLE = True
+    # 尝试导入 stealth 反检测
+    try:
+        from playwright_stealth import stealth_async
+        STEALTH_AVAILABLE = True
+    except ImportError:
+        STEALTH_AVAILABLE = False
+        print("⚠️  playwright-stealth 未安装，反爬虫能力降低")
 except ImportError:
     PLAYWRIGHT_AVAILABLE = False
+    STEALTH_AVAILABLE = False
+    print("⚠️  Playwright 未安装，请运行: pip install playwright && playwright install chromium")
 
 # 检查 agent-browser 是否可用
 def check_agent_browser():
@@ -204,16 +213,30 @@ async def capture_with_playwright(
     try:
         async with async_playwright() as p:
             # 启动浏览器，添加反检测配置
-            browser = await p.chromium.launch(
-                headless=True,
-                args=[
-                    '--disable-blink-features=AutomationControlled',  # 反自动化检测
-                    '--disable-dev-shm-usage',
-                    '--no-sandbox',
-                    '--disable-setuid-sandbox',
-                    '--disable-web-security',
-                ]
-            )
+            # 尝试使用真实 Chrome（如果安装了）
+            try:
+                # 首先尝试使用真实 Chrome 浏览器
+                browser = await p.chromium.launch(
+                    headless=True,
+                    channel='chrome',  # 使用真实 Chrome 而不是 Chromium
+                    args=[
+                        '--disable-blink-features=AutomationControlled',
+                        '--disable-dev-shm-usage',
+                        '--no-sandbox',
+                        '--ignore-certificate-errors',
+                    ]
+                )
+            except:
+                # 如果没有 Chrome，使用 Chromium
+                browser = await p.chromium.launch(
+                    headless=True,
+                    args=[
+                        '--disable-blink-features=AutomationControlled',
+                        '--disable-dev-shm-usage',
+                        '--no-sandbox',
+                        '--ignore-certificate-errors',
+                    ]
+                )
             
             viewport_settings = viewport or {"width": 1920, "height": 1080}
             
@@ -223,40 +246,97 @@ async def capture_with_playwright(
                     "width": viewport_settings.get("width", 1920),
                     "height": viewport_settings.get("height", 1080)
                 },
-                user_agent='Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
+                user_agent='Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.6099.109 Safari/537.36',
                 # 添加更多浏览器特征
-                locale='zh-CN',
-                timezone_id='Asia/Shanghai',
-                geolocation={'latitude': 31.2304, 'longitude': 121.4737},  # 上海坐标
-                permissions=['geolocation'],
+                locale='en-US',
+                timezone_id='America/New_York',
+                color_scheme='light',
+                has_touch=False,
+                is_mobile=False,
+                java_script_enabled=True,
+                accept_downloads=True,
             )
             
             # 注入反检测脚本
             await context.add_init_script("""
+                // 移除 webdriver 标记
                 Object.defineProperty(navigator, 'webdriver', {
                     get: () => undefined
                 });
+                
+                // 添加插件
                 Object.defineProperty(navigator, 'plugins', {
-                    get: () => [1, 2, 3, 4, 5]
+                    get: () => {
+                        return [
+                            {name: 'Chrome PDF Plugin', filename: 'internal-pdf-viewer', description: 'Portable Document Format'},
+                            {name: 'Chrome PDF Viewer', filename: 'mhjfbmdgncjokjjpdkfnmlglggk', description: ''},
+                            {name: 'Native Client', filename: 'internal-nacl-plugin', description: ''}
+                        ];
+                    }
                 });
+                
+                // 语言列表
                 Object.defineProperty(navigator, 'languages', {
-                    get: () => ['zh-CN', 'zh', 'en', 'en-US']
+                    get: () => ['en-US', 'en', 'zh-CN']
                 });
+                
+                // 平台
+                Object.defineProperty(navigator, 'platform', {
+                    get: () => 'Win32'
+                });
+                
+                // 硬件并发
+                Object.defineProperty(navigator, 'hardwareConcurrency', {
+                    get: () => 8
+                });
+                
+                // 设备内存
+                Object.defineProperty(navigator, 'deviceMemory', {
+                    get: () => 8
+                });
+                
+                // Chrome 对象
                 window.chrome = {
-                    runtime: {}
+                    runtime: {
+                        PlatformOs: {MAC: 'mac', WIN: 'win', ANDROID: 'android', CROS: 'cros', LINUX: 'linux', OPENBSD: 'openbsd'},
+                        PlatformArch: {ARM: 'arm', X86_32: 'x86-32', X86_64: 'x86-64'},
+                        PlatformNaclArch: {ARM: 'arm', X86_32: 'x86-32', X86_64: 'x86-64'},
+                        Request: function() {},
+                        connect: function() {},
+                        sendMessage: function() {}
+                    },
+                    loadTimes: function() {},
+                    csi: function() {},
+                    app: {}
                 };
             """)
             
             page = await context.new_page()
             
-            try:
-                await page.goto(url, wait_until="networkidle", timeout=30000)
-            except:
-                await page.goto(url, wait_until="domcontentloaded", timeout=30000)
+            # 应用 stealth 反检测（如果可用）
+            if STEALTH_AVAILABLE:
+                await stealth_async(page)
             
+            try:
+                await page.goto(url, wait_until="domcontentloaded", timeout=30000)
+            except Exception as e:
+                # 如果 domcontentloaded 超时，尝试 commit
+                try:
+                    await page.goto(url, wait_until="commit", timeout=30000)
+                except:
+                    pass
+            
+            # 固定等待时间（用于验证、加载等过程）
             if wait_time > 0:
                 await page.wait_for_timeout(wait_time)
             
+            # 等待页面基本稳定
+            try:
+                await page.wait_for_load_state("networkidle", timeout=5000)
+            except:
+                # networkidle 超时没关系，继续执行
+                pass
+            
             if scroll_times > 0:
                 for i in range(scroll_times):
                     await page.evaluate("window.scrollBy(0, 800)")
@@ -284,6 +364,59 @@ async def capture_with_playwright(
         return {"success": False, "error": str(e)}
 
 
+async def capture_with_cdp(
+    url_hint: str = "",
+    action: str = "screenshot",
+    cdp_port: int = 9222
+):
+    """
+    使用 CDP 连接到已打开的 Chrome 浏览器
+    用于方案三：手动验证后自动截图
+    """
+    if not PLAYWRIGHT_AVAILABLE:
+        return {"success": False, "error": "Playwright not installed"}
+    
+    try:
+        async with async_playwright() as p:
+            # 连接到已运行的 Chrome
+            browser = await p.chromium.connect_over_cdp(f"http://localhost:{cdp_port}")
+            
+            # 获取所有页面
+            contexts = browser.contexts
+            result_data = []
+            
+            for context in contexts:
+                for page in context.pages:
+                    current_url = page.url
+                    title = await page.title()
+                    
+                    # 如果 URL 包含提示词或者是唯一页面
+                    if url_hint in current_url or not url_hint:
+                        if action == "screenshot":
+                            temp_file = CAPTURE_DIR / f"cdp_capture_{uuid.uuid4()[:8]}.png"
+                            await page.screenshot(path=str(temp_file), full_page=True)
+                            result_data.append({
+                                "url": current_url,
+                                "title": title,
+                                "file_path": str(temp_file)
+                            })
+                        elif action == "html":
+                            html = await page.content()
+                            result_data.append({
+                                "url": current_url,
+                                "title": title,
+                                "html": html
+                            })
+            
+            if result_data:
+                return {"success": True, "pages": result_data}
+            else:
+                return {"success": False, "error": "No matching page found"}
+    
+    except Exception as e:
+        return {"success": False, "error": f"CDP connection failed: {str(e)}"}
+
+
 def capture_webpage(
     url: str,
     action: str = "screenshot",
@@ -387,7 +520,8 @@ def capture():
         "full_page": false,
         "viewport": {"width": 1920, "height": 1080},
         "wait_time": 2000,
-        "backend": "auto" | "agent-browser" | "playwright"
+        "backend": "auto" | "agent-browser" | "playwright" | "chrome-cdp",
+        "cdp_port": 9222  // 用于 chrome-cdp 后端，连接到已打开的 Chrome
     }
     """
     data = request.get_json()
@@ -410,6 +544,8 @@ def capture():
     viewport = data.get("viewport")
     wait_time = int(data.get("wait_time", 2000))
     backend = data.get("backend", "auto")
+    cdp_port = int(data.get("cdp_port", 9222))
+    url_hint = data.get("url_hint", "")
     
     result = capture_webpage(
         url=url,
@@ -419,7 +555,9 @@ def capture():
         full_page=full_page,
         viewport=viewport,
         wait_time=wait_time,
-        backend=backend
+        backend=backend,
+        
+        
     )
     
     if not result["success"]:

manual_capture_guide.md

@@ -0,0 +1,69 @@
+# 手动 Chrome 捕获指南
+
+## 方案三完整流程
+
+### 1️⃣ 启动真实 Chrome（带调试端口）
+
+在您的本地机器或服务器上运行：
+
+```bash
+# Linux
+google-chrome --remote-debugging-port=9222 https://www.techpowerup.com/gpu-specs/rtx-pro-6000-blackwell.c4272
+
+# 或者使用已有的 Chrome 进程（需要先关闭所有 Chrome）
+google-chrome-stable --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-debug
+```
+
+### 2️⃣ 手动验证（如 Cloudflare）
+
+- 在浏览器中手动完成验证过程
+- 等待页面完全加载
+- 确认可以看到正常内容
+
+### 3️⃣ 连接到 Chrome 并截图
+
+验证完成后，运行脚本：
+
+```bash
+# 连接到 Chrome 并截图
+python3.12 /tmp/connect_chrome.py techpowerup
+
+# 或者捕获当前所有页面
+python3.12 /tmp/connect_chrome.py
+```
+
+### 4️⃣ 查看结果
+
+```bash
+# 查看截图
+ls -lh /tmp/chrome_capture_*.png
+
+# 查看HTML
+head -100 /tmp/chrome_html_0.html
+```
+
+## API 集成方案
+
+如果需要通过 API 获取内容，可以：
+
+1. 在服务代码中添加 CDP 连接选项
+2. 用户先手动验证
+3. API 连接到已验证的浏览器实例
+
+## 注意事项
+
+⚠️ **重要**：
+- Chrome 必须带 `--remote-debugging-port=9222` 启动
+- 所有 Chrome 进程必须先关闭（或使用新的 user-data-dir）
+- 验证完成后才能运行脚本
+- 截图后浏览器保持打开，可以继续使用
+
+## 一键脚本
+
+```bash
+# 启动 Chrome 并等待用户验证
+google-chrome --remote-debugging-port=9222 "https://www.techpowerup.com/gpu-specs/rtx-pro-6000-blackwell.c4272" &
+echo "请在浏览器中完成验证，然后按 Enter 继续截图..."
+read
+python3.12 /tmp/connect_chrome.py techpowerup
+```

templates/index.html

@@ -295,20 +295,21 @@
                 </div>
                 
                 <div class="form-group">
-                    <label>滚动设置（用于加载动态内容）</label>
+                    <label>等待设置（重要！）</label>
                     <div class="row">
                         <div>
-                            <label for="scrollTimes">滚动次数</label>
+                            <label for="waitTime">页面加载等待(ms) ⏱️</label>
+                            <input type="number" id="waitTime" value="2000" min="0" max="60000" step="500">
+                            <small style="color: #666; display: block; margin-top: 4px;">某些网站需要较长验证过程，可设置为 10000-30000</small>
+                        </div>
+                        <div>
+                            <label for="scrollTimes">滚动次数（可选）</label>
                             <input type="number" id="scrollTimes" value="0" min="0" max="20">
                         </div>
                         <div>
                             <label for="scrollDelay">滚动间隔(ms)</label>
                             <input type="number" id="scrollDelay" value="1000" min="500" max="5000" step="100">
                         </div>
-                        <div>
-                            <label for="waitTime">页面加载等待(ms)</label>
-                            <input type="number" id="waitTime" value="2000" min="0" max="10000" step="500">
-                        </div>
                     </div>
                 </div>
                 
@@ -350,16 +351,23 @@
         </div>
         
         <div class="card api-docs">
-            <h3>📚 API 文档</h3>
-            <p>POST /api/capture</p>
-            <pre><code>{
+            <h3>📚 使用说明</h3>
+            <p><strong>重要提示：</strong></p>
+            <ul style="margin: 10px 0; line-height: 1.8;">
+                <li>⏱️ <strong>页面加载等待</strong>：某些网站需要验证过程（如 Cloudflare），请设置较长时间（10000-30000ms）</li>
+                <li>🔄 <strong>滚动次数</strong>：用于加载动态内容（如微博、推特等），建议 3-5 次</li>
+                <li>🎯 <strong>全页截图</strong>：滚动加载后建议开启此选项</li>
+            </ul>
+            <p style="margin-top: 15px;"><strong>API 调用：</strong></p>
+            <pre><code>POST /api/capture
+{
   "url": "https://example.com",
   "action": "screenshot",  // 或 "html"
-  "scroll_times": 3,
-  "scroll_delay": 1000,
-  "full_page": true,
-  "viewport": {"width": 1920, "height": 1080},
-  "wait_time": 2000
+  "wait_time": 15000,     // 重要！验证网站需设置较长等待
+  "scroll_times": 3,      // 可选：加载动态内容
+  "scroll_delay": 1000,  // 可选：滚动间隔
+  "full_page": true,     // 可选：全页截图
+  "backend": "playwright" // 可选：playwright 或 agent-browser
 }</code></pre>
         </div>
     </div>