Edge Repair

src/gui.py

@@ -48,7 +48,7 @@ class ScraperThread(QThread):
                 else:
                     self.log.emit(f"<font color='green'><b>[√] {msg}</b></font>")
 
-            scraper = Scraper1688(headless=self.headless, status_callback=status_cb)
+            scraper = Scraper1688(headless=self.headless, status_callback=status_cb, log_callback=self.log.emit)
             
             # 使用流式生成器抓取
             collected_count = 0

src/scraper.py

@@ -24,19 +24,19 @@ from selenium.webdriver.common.action_chains import ActionChains
 from selenium_stealth import stealth
 
 class Scraper1688:
-    def __init__(self, headless=True, status_callback=None):
+    def __init__(self, headless=True, status_callback=None, log_callback=None):
         self.headless = headless
         self.status_callback = status_callback
+        self.log_callback = log_callback # 新增：用于向 GUI 发送普通日志
         self.user_data_path = os.path.abspath(os.path.join(os.getcwd(), "1688_user_data"))
         self.driver = None
         
-        # 1. 强力锁定长效 Session
         edge_path = self._find_edge()
         if edge_path:
-            print(f"[*] 【长效稳定模式】正在启动 Edge 独立环境...")
+            print(f"[*] 【极致稳定模式】正在启动 Edge 深度伪装环境...")
             self._cleanup_processes()
-            # 使用固定且持久的 Session 目录，确保 12 小时以上免登录
-            edge_user_data = os.path.join(os.getcwd(), "1688_edge_stable_session")
+            # 使用固定且持久的 Session 目录，确保长效免登录
+            edge_user_data = os.path.join(os.getcwd(), "1688_edge_ultimate_session")
             cmd = [
                 edge_path, 
                 "--remote-debugging-port=9222", 
@@ -48,37 +48,32 @@ class Scraper1688:
             if headless: cmd.append("--headless")
             try:
                 subprocess.Popen(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
-                time.sleep(5) # 预留更充足的启动时间
+                time.sleep(6) 
                 opts = EdgeOptions()
                 opts.add_experimental_option("debuggerAddress", "127.0.0.1:9222")
-                try:
-                    self.driver = webdriver.Edge(options=opts)
-                    print("[+] Edge 长效环境连接成功！")
-                except:
-                    from webdriver_manager.microsoft import EdgeChromiumDriverManager
-                    service = EdgeService(EdgeChromiumDriverManager().install())
-                    self.driver = webdriver.Edge(service=service, options=opts)
+                self.driver = webdriver.Edge(options=opts)
+                print("[+] Edge 极致稳定环境接管成功！")
             except Exception as e:
                 print(f"[!] Edge 启动失败: {e}")
         
         if not self.driver:
-            print("[*] 正在启动备用 Chrome 模式...")
             self._init_chrome(headless)
 
         if self.driver:
-            # 基础特征隐藏补丁
+            # 深度擦除自动化指纹
             try:
                 self.driver.execute_cdp_cmd("Page.addScriptToEvaluateOnNewDocument", {
-                    "source": "Object.defineProperty(navigator, 'webdriver', { get: () => undefined });"
+                    "source": """
+                        Object.defineProperty(navigator, 'webdriver', { get: () => undefined });
+                        Object.defineProperty(navigator, 'languages', { get: () => ['zh-CN', 'zh'] });
+                        Object.defineProperty(navigator, 'plugins', { get: () => [1, 2, 3, 4, 5] });
+                    """
                 })
             except: pass
 
     def _find_edge(self):
         import winreg
-        reg_paths = [
-            (winreg.HKEY_LOCAL_MACHINE, r"SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\msedge.exe"),
-            (winreg.HKEY_CURRENT_USER, r"SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\msedge.exe")
-        ]
+        reg_paths = [(winreg.HKEY_LOCAL_MACHINE, r"SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\msedge.exe"), (winreg.HKEY_CURRENT_USER, r"SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\msedge.exe")]
         for hkey, subkey in reg_paths:
             try:
                 with winreg.OpenKey(hkey, subkey) as key:
@@ -97,10 +92,7 @@ class Scraper1688:
             opts = uc.ChromeOptions()
             opts.add_argument(f"--user-data-dir={self.user_data_path}")
             return opts
-        try:
-            self.driver = uc.Chrome(options=create_options(), headless=headless)
-        except:
-            self.driver = uc.Chrome(options=create_options(), headless=headless)
+        self.driver = uc.Chrome(options=create_options(), headless=headless)
 
     def check_for_captcha(self):
         def is_blocked():
@@ -115,69 +107,76 @@ class Scraper1688:
             if self.status_callback: self.status_callback(True, msg)
             while is_blocked(): time.sleep(3)
             if self.status_callback: self.status_callback(False, "验证通过")
-            print("[*] 解封成功，强制进入 60 秒安全静默期...")
-            time.sleep(60) 
+            
+            cool_msg = "[*] 监测到干预完成，进入 120 秒深度冷却期以重置风控权重..."
+            print(cool_msg)
+            if self.log_callback: self.log_callback(f"<font color='orange'>{cool_msg}</font>")
+            time.sleep(120) 
         return True
 
-    def _human_interact(self):
-        """ 深度模拟人类交互，降低被识别概率 """
-        try:
-            actions = ActionChains(self.driver)
-            # 1. 随机小幅度鼠标移动
-            for _ in range(random.randint(3, 6)):
-                actions.move_by_offset(random.randint(-10, 10), random.randint(-10, 10)).perform()
-                time.sleep(random.uniform(0.3, 0.8))
-            # 2. 随机滚动一小段
-            self.driver.execute_script(f"window.scrollBy(0, {random.randint(100, 400)});")
-            time.sleep(random.uniform(1.0, 2.0))
-        except: pass
+    def _human_behavior(self, duration=10):
+        """ 高级拟人化行为模拟 """
+        start_time = time.time()
+        while time.time() - start_time < duration:
+            try:
+                # 1. 随机滚动
+                scroll_y = random.randint(200, 600)
+                self.driver.execute_script(f"window.scrollBy(0, {scroll_y});")
+                # 2. 随机鼠标晃动
+                actions = ActionChains(self.driver)
+                actions.move_by_offset(random.randint(-5, 5), random.randint(-5, 5)).perform()
+                time.sleep(random.uniform(1.5, 4.0))
+                # 3. 概率性往回滚
+                if random.random() > 0.7:
+                    self.driver.execute_script(f"window.scrollBy(0, -{random.randint(100, 300)});")
+            except: break
 
     def search_products_yield(self, keyword, total_count=200, existing_links=None):
         gbk_keyword = urllib.parse.quote(keyword, encoding='gbk')
         base_url = f"https://s.1688.com/selloffer/offer_search.htm?keywords={gbk_keyword}&n=y&netType=1%2C11%2C16"
         
         self.driver.get("https://www.1688.com")
+        time.sleep(random.randint(3, 6))
         self.check_for_captcha()
 
         all_links = existing_links if existing_links is not None else set()
         page, initial_count = 1, len(all_links)
+        # 随机设定下一次深度冷却的阈值 (5-12条之间)
+        next_cool_threshold = random.randint(5, 12)
         
         while len(all_links) < total_count + initial_count:
-            print(f"[*] 正在处理列表页: 第 {page} 页...")
+            print(f"[*] 正在模拟搜索: 第 {page} 页...")
             self.driver.get(f"{base_url}&beginPage={page}&page={page}")
             self.check_for_captcha()
             
-            # 高级模拟滚动：滑下去，停顿，回滑一点
-            steps = random.randint(6, 10)
-            for i in range(1, steps + 1):
-                self.driver.execute_script(f"window.scrollTo(0, document.body.scrollHeight * {i/steps});")
-                time.sleep(random.uniform(1.2, 3.0))
-                if i == random.randint(3, 5):
-                    self.driver.execute_script("window.scrollBy(0, -250);")
-                    time.sleep(1.0)
+            # 列表页模拟“翻找”行为
+            for _ in range(random.randint(5, 8)):
+                self.driver.execute_script(f"window.scrollBy(0, {random.randint(400, 800)});")
+                time.sleep(random.uniform(1.5, 3.5))
+                if random.random() > 0.8:
+                    self.driver.execute_script("window.scrollBy(0, -300);")
 
             page_results = self._extract_all_methods()
-            if not page_results:
-                print(f"[!] 第 {page} 页无结果，尝试刷新...")
-                self.driver.refresh()
-                time.sleep(5)
-                page_results = self._extract_all_methods()
-
             page_batch = []
             for it in page_results:
                 clean_url = self.clean_url(it["link"])
                 if clean_url and clean_url not in all_links:
                     all_links.add(clean_url)
                     
-                    # --- 核心频率控制：深度保护机制 ---
-                    current_new = len(all_links) - initial_count
-                    if current_new > 0 and current_new % 8 == 0:
-                        rest = random.randint(90, 240)
-                        print(f"[*] 已连续作业8个详情，触发深度冷却 {rest} 秒以维持 12 小时 Session...")
+                    # --- 核心订正：随机深度冷却 ---
+                    new_processed = len(all_links) - initial_count
+                    if new_processed >= next_cool_threshold:
+                        rest = random.randint(120, 300)
+                        cool_msg = f"[*] 随机触发深度保护 (已处理{new_processed}条)，睡眠 {rest} 秒模拟休息..."
+                        print(cool_msg)
+                        if self.log_callback: self.log_callback(f"<font color='orange'><b>{cool_msg}</b></font>")
                         time.sleep(rest)
+                        next_cool_threshold += random.randint(5, 12) # 设定下一个随机检查点
 
-                    print(f"  [>] 详情解析: {clean_url}")
-                    time.sleep(random.uniform(4.0, 7.0)) # 访问前随机静默
+                    print(f"  [>] 详情仿真采集: {clean_url}")
+                    
+                    # 访问前大幅随机停顿
+                    time.sleep(random.uniform(5, 12)) 
                     
                     detail_results = self.scrape_detail(clean_url)
                     if detail_results: page_batch.extend(detail_results)
@@ -187,22 +186,24 @@ class Scraper1688:
                         yield page_batch
                         page_batch = []
                     
-                    # --- 极低频率抓取：详情页间超长等待 ---
-                    time.sleep(random.uniform(20, 45)) 
+                    # 详情页之间的大跨度等待
+                    time.sleep(random.uniform(30, 60)) 
                     
                     if len(all_links) >= total_count + initial_count: break
             
             if page_batch: yield page_batch
             page += 1
-            if page > 100: break
+            # 每翻 3 页随机回一次 1688 首页，消除路径单一性
+            if page % 3 == 0:
+                self.driver.get("https://www.1688.com")
+                time.sleep(random.randint(10, 20))
         return list(all_links)
 
     def scrape_detail(self, url):
-        """ 拟人化的详情数据提取 """
         try:
             self.driver.get(url)
-            time.sleep(random.uniform(5.0, 10.0)) # 充分加载
-            self._human_interact() # 执行随机交互
+            # --- 核心改进：详情页留存仿真 ---
+            self._human_behavior(duration=random.randint(12, 25))
             self.check_for_captcha()
             
             model = self.driver.execute_script(
@@ -264,7 +265,6 @@ class Scraper1688:
         return url
 
     def _extract_all_methods(self):
-        """ 全力提取列表页链接 """
         results = []
         try:
             res = self.driver.execute_script("return JSON.stringify(window.data || window.__INITIAL_DATA__)")