From 0c05b260eae0670223f4df2b7c6e1561d9d7167d Mon Sep 17 00:00:00 2001
From: SayItDitto <123747090+SayItDitto@users.noreply.github.com>
Date: Sun, 23 Aug 2026 09:11:53 +0800
Subject: [PATCH] =?UTF-8?q?fix:=20CloakBrowser=20=E8=8E=B7=E5=8F=96?=
 =?UTF-8?q?=E9=A1=B5=E9=9D=A2=E6=BA=90=E7=A0=81=E5=9C=A8=20CF=20=E8=B4=A8?=
 =?UTF-8?q?=E8=AF=A2=E9=A1=B5=E8=B6=85=E6=97=B6=20(#6402)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 app/adapters/network/browser.py | 30 +++++++++++++++++++++++++++---
 1 file changed, 27 insertions(+), 3 deletions(-)

diff --git a/app/adapters/network/browser.py b/app/adapters/network/browser.py
index 458424407..ae6a4c1a2 100644
--- a/app/adapters/network/browser.py
+++ b/app/adapters/network/browser.py
@@ -1118,6 +1118,7 @@ class PlaywrightHelper:
         :param headless: 是否无头模式
         :param timeout: 超时时间
         """
+        timeout = timeout or 60
         source = None
         # 如果配置为 FlareSolverr，则直接调用获取页面源码
         if self.__browser_emulation() == "flaresolverr":
@@ -1140,10 +1141,33 @@ class PlaywrightHelper:
                 if cookies:
                     page.set_extra_http_headers({"cookie": cookies})
 
-                page.goto(url)
-                page.wait_for_load_state("networkidle", timeout=timeout * 1000)
+                page.goto(url, wait_until="load", timeout=timeout * 1000)
 
-                source = page.content()
+                # 修复: 部分站点(如 Cloudflare 质询页)会持续轮询请求,
+                # 导致 networkidle 永不触发而超时。改为等待页面加载完成后
+                # 轮询检查标题, 直到不再停留在质询/加载页。
+                challenge_titles = ("just a moment", "请稍候", "loading")
+                deadline = time.time() + timeout
+                while time.time() < deadline:
+                    try:
+                        current_title = (page.title() or "").strip().lower()
+                    except Exception:
+                        current_title = ""
+                    if current_title and not any(
+                            t in current_title for t in challenge_titles):
+                        break
+                    time.sleep(2)
+
+                # 页面跳转中 content() 可能失败, 重试几次
+                source = None
+                for _attempt in range(5):
+                    try:
+                        source = page.content()
+                        if source:
+                            break
+                    except Exception:
+                        source = None
+                        time.sleep(2)
 
             except Exception as e:
                 logger.error(f"获取网页源码失败: {str(e)}")