fix: update douyin profile link parse

src/room.py

@@ -18,6 +18,10 @@ no_proxy_handler = urllib.request.ProxyHandler({})
 opener = urllib.request.build_opener(no_proxy_handler)
 
 
+class UnsupportedUrlError(Exception):
+    pass
+
+
 HEADERS = {
     'User-Agent': 'Mozilla/5.0 (Linux; Android 11; SAMSUNG SM-G973U) AppleWebKit/537.36 (KHTML, like Gecko) '
                   'SamsungBrowser/14.2 Chrome/87.0.4280.141 Mobile Safari/537.36',
@@ -28,9 +32,7 @@ HEADERS = {
 HEADERS_PC = {
     'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.5845.97 '
                   'Safari/537.36 Core/1.116.438.400 QQBrowser/13.0.6070.400',
-    'Cookie': 'sessionid=7494ae59ae06784454373ce25761e864; __ac_nonce=0670497840077ee4c9eb2; '
-              '__ac_signature=_02B4Z6wo00f012DZczQAAIDCJJBb3EjnINdg-XeAAL8-db;  '
-              's_v_web_id=verify_m1ztgtjj_vuHnMLZD_iwZ9_4YO4_BdN1_7wLP3pyqXsf2; ',
+    'Cookie': '',
     }
 
 
@@ -61,38 +63,44 @@ async def get_sec_user_id(url: str, proxy_addr: str | None = None, headers: dict
                     room_id = str(redirect_url).split('?')[0].rsplit('/', maxsplit=1)[1]
                     return room_id, sec_user_id
                 else:
-                    print("Could not find sec_user_id in the URL.")
+                    raise RuntimeError("Could not find sec_user_id in the URL.")
             else:
-                print("The redirect URL does not contain 'reflow/'.")
+                raise UnsupportedUrlError("The redirect URL does not contain 'reflow/'.")
+    except UnsupportedUrlError as e:
+        raise e
     except Exception as e:
-        print(f"An error occurred: {e}")
-    return None
+        raise RuntimeError(f"An error occurred: {e}")
 
 
 # 获取抖音号
 async def get_unique_id(url: str, proxy_addr: str | None = None, headers: dict | None = None) -> str | None:
     if not headers or all(k.lower() not in ['user-agent', 'cookie'] for k in headers):
-        headers = HEADERS_PC
+        headers = HEADERS
 
     try:
         proxy_addr = utils.handle_proxy_addr(proxy_addr)
         async with httpx.AsyncClient(proxy=proxy_addr, timeout=15) as client:
             response = await client.get(url, headers=headers, follow_redirects=True)
             redirect_url = str(response.url)
+            if 'reflow/' in str(redirect_url):
+                raise UnsupportedUrlError("Unsupported URL")
             sec_user_id = redirect_url.split('?')[0].rsplit('/', maxsplit=1)[1]
-
-            user_page_response = await client.get(f'https://www.douyin.com/user/{sec_user_id}', headers=headers)
-            matches = re.findall(r'undefined\\"},\\"uniqueId\\":\\"(.*?)\\",\\"customVerify',
-                                 user_page_response.text)
+            headers['Cookie'] = ('ttwid=1%7C4ejCkU2bKY76IySQENJwvGhg1IQZrgGEupSyTKKfuyk%7C1740470403%7Cbc9a'
+                                 'd2ee341f1a162f9e27f4641778030d1ae91e31f9df6553a8f2efa3bdb7b4; __ac_nonce=06'
+                                 '83e59f3009cc48fbab0; __ac_signature=_02B4Z6wo00f01mG6waQAAIDB9JUCzFb6.TZhmsU'
+                                 'AAPBf34; __ac_referer=__ac_blank')
+            user_page_response = await client.get(f'https://www.iesdouyin.com/share/user/{sec_user_id}',
+                                                headers=headers, follow_redirects=True)
+            matches = re.findall(r'unique_id":"(.*?)","verification_type', user_page_response.text)
             if matches:
                 unique_id = matches[-1]
                 return unique_id
             else:
-                print("Could not find unique_id in the response.")
-                return None
+                raise RuntimeError("Could not find unique_id in the response.")
+    except UnsupportedUrlError as e:
+        raise e
     except Exception as e:
-        print(f"An error occurred: {e}")
-        return None
+        raise RuntimeError(f"An error occurred: {e}")
 
 
 # 获取直播间webID

src/spider.py

@@ -25,7 +25,7 @@ import urllib.request
 from . import JS_SCRIPT_PATH, utils
 from .utils import trace_error_decorator
 from .logger import script_path
-from .room import get_sec_user_id, get_unique_id
+from .room import get_sec_user_id, get_unique_id, UnsupportedUrlError
 from .http_clients.async_http import async_req
 
 
@@ -113,12 +113,11 @@ async def get_douyin_app_stream_data(url: str, proxy_addr: OptionalStr = None, c
             room_data = json_data['data'][0]
             room_data['anchor_name'] = json_data['user']['nickname']
         else:
-            data = await get_sec_user_id(url, proxy_addr=proxy_addr)
-
-            if data:
+            try:
+                data = await get_sec_user_id(url, proxy_addr=proxy_addr)
                 _room_id, _sec_uid = data
                 room_data = await get_app_data(_room_id, _sec_uid)
-            else:
+            except UnsupportedUrlError:
                 unique_id = await get_unique_id(url, proxy_addr=proxy_addr)
                 return await get_douyin_stream_data(f'https://live.douyin.com/{unique_id}')
 
@@ -3003,4 +3002,4 @@ async def get_faceit_stream_data(url: str, proxy_addr: OptionalStr = None, cooki
         result['anchor_name'] = anchor_name
     else:
         result = {'anchor_name': anchor_name, 'is_live': False}
-    return result
+    return result