无头浏览器

schoolNewsCrawler/crawler/xhw/XhwCrawler.py

@@ -110,26 +110,41 @@ class XhwCrawler(BaseCrawler):
         self.driver = self._init_driver()
     
     def _init_driver(self):
-        """初始化并返回Chrome WebDriver实例"""
+        """初始化并返回Chrome WebDriver实例（无头模式）"""
         chrome_options = Options()
-        # 确保浏览器可见，不使用无头模式
-        # 或者完全删除这行，因为默认就是有界面模式
+        
+        # 无头模式配置
+        chrome_options.add_argument('--headless=new')  # 使用新的headless模式
         chrome_options.add_argument('--no-sandbox')
         chrome_options.add_argument('--disable-dev-shm-usage')
+        chrome_options.add_argument('--disable-gpu')
+        
+        # 设置窗口大小（headless模式必需）
+        chrome_options.add_argument('--window-size=1920,1080')
+        
+        # 反检测配置
         chrome_options.add_argument('--disable-blink-features=AutomationControlled')
         chrome_options.add_experimental_option("excludeSwitches", ["enable-automation"])
         chrome_options.add_experimental_option('useAutomationExtension', False)
         chrome_options.add_argument('--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')
-        # 确保浏览器可见
-        chrome_options.add_argument('--start-maximized')
-        chrome_options.add_argument('--disable-gpu')
+        
+        # 其他安全配置
         chrome_options.add_argument('--disable-web-security')
         chrome_options.add_argument('--allow-running-insecure-content')
         chrome_options.add_argument('--disable-features=VizDisplayCompositor')
+        
+        # 根据系统选择chromedriver路径和chrome二进制文件路径
         chrome_driver_path = 'win/chromedriver.exe'
-
+        chrome_binary_path = 'win/chrome-headless/chrome-headless-shell-win64/chrome-headless-shell.exe'
+        
         if platform.system() == 'Linux':
             chrome_driver_path = 'linux/chromedriver'
+            chrome_binary_path = 'linux/chrome-headless/chrome-headless/chrome-headless-shell'  # Linux使用系统安装的Chrome
+        
+        # 指定Chrome二进制文件路径（用于chrome-headless-shell）
+        if chrome_binary_path and os.path.exists(chrome_binary_path):
+            chrome_options.binary_location = chrome_binary_path
+            logger.info(f"使用Chrome二进制: {chrome_binary_path}")
 
         service = Service(executable_path=chrome_driver_path)
         

schoolNewsCrawler/crawler/xxqg/XxqgCrawler.py

@@ -160,27 +160,41 @@ class XxqgCrawler(BaseCrawler):
         self.driver = self._init_driver()
     
     def _init_driver(self):
-        """初始化并返回Chrome WebDriver实例"""
+        """初始化并返回Chrome WebDriver实例（无头模式）"""
         chrome_options = Options()
-        # 确保浏览器可见，不使用无头模式
-        # 或者完全删除这行，因为默认就是有界面模式
+        
+        # 无头模式配置
+        chrome_options.add_argument('--headless=new')  # 使用新的headless模式
         chrome_options.add_argument('--no-sandbox')
         chrome_options.add_argument('--disable-dev-shm-usage')
+        chrome_options.add_argument('--disable-gpu')
+        
+        # 设置窗口大小（headless模式必需）
+        chrome_options.add_argument('--window-size=1920,1080')
+        
+        # 反检测配置
         chrome_options.add_argument('--disable-blink-features=AutomationControlled')
         chrome_options.add_experimental_option("excludeSwitches", ["enable-automation"])
         chrome_options.add_experimental_option('useAutomationExtension', False)
         chrome_options.add_argument('--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')
-        # 确保浏览器可见
-        chrome_options.add_argument('--start-maximized')
-        chrome_options.add_argument('--disable-gpu')
+        
+        # 其他安全配置
         chrome_options.add_argument('--disable-web-security')
         chrome_options.add_argument('--allow-running-insecure-content')
         chrome_options.add_argument('--disable-features=VizDisplayCompositor')
-        # 判断系统类型获取对应的chromedriver路径
+        
+        # 根据系统选择chromedriver路径和chrome二进制文件路径
         chrome_driver_path = 'win/chromedriver.exe'
-
+        chrome_binary_path = 'win/chrome-headless/chrome-headless-shell-win64/chrome-headless-shell.exe'
+        
         if platform.system() == 'Linux':
             chrome_driver_path = 'linux/chromedriver'
+            chrome_binary_path = 'linux/chrome-headless/chrome-headless/chrome-headless-shell'  # Linux使用系统安装的Chrome
+        
+        # 指定Chrome二进制文件路径（用于chrome-headless-shell）
+        if chrome_binary_path and os.path.exists(chrome_binary_path):
+            chrome_options.binary_location = chrome_binary_path
+            logger.info(f"使用Chrome二进制: {chrome_binary_path}")
 
         service = Service(executable_path=chrome_driver_path)
         

schoolNewsCrawler/linux/chrome-headless/ABOUT

@@ -0,0 +1,9 @@
+Google Chrome
+
+Copyright 2025 Google LLC. All rights reserved.
+
+Chrome is made possible by the Chromium open source project
+(https://www.chromium.org/) and other open source software
+(chrome://credits).
+
+See the Terms of Service at chrome://terms.

schoolNewsCrawler/linux/chrome-headless/deb.deps

@@ -0,0 +1,30 @@
+ca-certificates
+fonts-liberation
+libasound2 (>= 1.0.17)
+libatk-bridge2.0-0 (>= 2.5.3)
+libatk1.0-0 (>= 2.11.90)
+libatspi2.0-0 (>= 2.9.90)
+libc6 (>= 2.17)
+libcairo2 (>= 1.6.0)
+libcups2 (>= 1.6.0)
+libcurl3-gnutls | libcurl3-nss | libcurl4 | libcurl3
+libdbus-1-3 (>= 1.9.14)
+libexpat1 (>= 2.1~beta3)
+libgbm1 (>= 17.1.0~rc2)
+libglib2.0-0 (>= 2.39.4)
+libgtk-3-0 (>= 3.9.10) | libgtk-4-1
+libnspr4 (>= 2:4.9-2~)
+libnss3 (>= 2:3.35)
+libpango-1.0-0 (>= 1.14.0)
+libudev1 (>= 183)
+libvulkan1
+libx11-6 (>= 2:1.4.99.1)
+libxcb1 (>= 1.9.2)
+libxcomposite1 (>= 1:0.4.4-1)
+libxdamage1 (>= 1:1.1)
+libxext6
+libxfixes3
+libxkbcommon0 (>= 0.5.0)
+libxrandr2
+wget
+xdg-utils (>= 1.0.2)