真实链接测试

douyin_parser/main.py

@@ -1,5 +1,6 @@
 import logging
 import re
+import time
 import tomllib
 import os
 import traceback
@@ -50,27 +51,51 @@ class DouyinParser:
         return cleaned_url
 
     def _get_real_video_url(self, video_url: str) -> str:
-        max_retries = 3
-        retry_delay = 2
+        """获取真实视频链接"""
+        max_retries = 3  # 最大重试次数
+        retry_delay = 2  # 重试延迟秒数
+        proxies = {"http": self.http_proxy, "https": self.http_proxy} if self.http_proxy else None
 
         for retry in range(max_retries):
             try:
                 self.LOG.info("[抖音] 开始获取真实视频链接: %s (第%d次尝试)", video_url, retry + 1)
-                proxy = {"http": self.http_proxy, "https": self.http_proxy} if self.http_proxy else None
-
                 headers = {
-                    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64)...'
+                    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
+                    'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
+                    'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8',
+                    'Range': 'bytes=0-'
                 }
 
-                response = requests.get(video_url, headers=headers, proxies=proxy, allow_redirects=True, timeout=60)
+                response = requests.get(video_url, headers=headers, proxies=proxies, allow_redirects=True, timeout=60)
                 if response.status_code in [200, 206]:
+                    history = [resp.url for resp in response.history]
                     real_url = response.url
-                    self.LOG.info("[抖音] 成功获取真实链接: %s", real_url)
-                    return real_url
+
+                    if history:
+                        self.LOG.debug("[抖音] 重定向历史: %s", history)
+
+                    if real_url != video_url and ('v3-' in real_url.lower() or 'douyinvod.com' in real_url.lower()):
+                        self.LOG.info("[抖音] 成功获取真实链接: %s", real_url)
+                        return real_url
+                    else:
+                        self.LOG.warning("[抖音] 未能获取到真实视频链接，准备重试")
+                        if retry < max_retries - 1:
+                            time.sleep(retry_delay)
+                            continue
+                        return video_url
                 else:
                     self.LOG.error("[抖音] 获取视频真实链接失败, 状态码: %d", response.status_code)
+                    self.LOG.debug("[抖音] 响应头: %s", response.headers)
+                    if retry < max_retries - 1:
+                        time.sleep(retry_delay)
+                        continue
+                    return video_url
             except Exception as e:
-                self.LOG.error("[抖音] 获取真实链接失败: %s", str(e))
+                self.LOG.error("[抖音] 获取真实链接失败: %s (第%d次尝试)", str(e), retry + 1)
+                if retry < max_retries - 1:
+                    time.sleep(retry_delay)
+                    continue
+                return video_url
 
         self.LOG.error("[抖音] 获取真实链接失败，已达到最大重试次数")
         return video_url
@@ -139,70 +164,3 @@ class DouyinParser:
             self.LOG.error("[抖音] 解析过程发生未知错误: %s\n%s", str(e), traceback.format_exc())
             raise DouyinParserError(f"未知错误: {str(e)}")
         return
-
-    def send_xml_video(self, message: WxMsg, title, author, video_url, cover):
-
-        video_title = f"{title[:30]} - {author[:10]}" if author else title[:40],
-        xml_message = f"""
-        <?xml version="1.0"?>
-            <msg>
-                <appmsg appid="" sdkver="1">
-                    <title>{video_title}</title>
-                    <des>点击观看无水印视频</des>
-                    <action>view</action>
-                    <type>5</type>
-                    <showtype>0</showtype>
-                    <content />
-                    <url>{video_url}</url>
-                    <dataurl />
-                    <lowurl />
-                    <lowdataurl />
-                    <recorditem />
-                    <thumburl>{cover}</thumburl>
-                    <messageaction />
-                    <laninfo />
-                    <extinfo />
-                    <sourceusername />
-                    <sourcedisplayname />
-                    <commenturl />
-                    <appattach>
-                        <totallen>0</totallen>
-                        <attachid />
-                        <emoticonmd5 />
-                        <fileext />
-                        <aeskey />
-                    </appattach>
-                    <webviewshared>
-                        <publisherId />
-                        <publisherReqId>0</publisherReqId>
-                    </webviewshared>
-                    <weappinfo>
-                        <pagepath />
-                        <username />
-                        <appid />
-                        <appservicetype>0</appservicetype>
-                    </weappinfo>
-                    <websearch />
-                </appmsg>
-                <fromusername>Jyunere</fromusername>
-                <scene>0</scene>
-                <appinfo>
-                    <version>1</version>
-                    <appname></appname>
-                </appinfo>
-                <commenturl></commenturl>
-            </msg>
-            
-            """
-        # 修改消息数据库里面的消息content 内容
-        text_bytes = xml_message.encode('utf-8')
-        compressed_data = lb.compress(text_bytes, store_size=False).hex()
-
-        data = self.wcf.query_sql('MSG0.db', "SELECT * FROM MSG where type = 49  limit 1")
-        self.wcf.query_sql('MSG0.db',
-                           f"""UPDATE MSG SET CompressContent = x'{compressed_data}', BytesExtra=x'', type=49, SubType=3,
-                  IsSender=0, TalkerId=2 WHERE MsgSvrID={data[0]['MsgSvrID']}"""
-                           )
-
-        result = self.wcf.forward_msg(data[0]["MsgSvrID"], message.roomid)
-        print(f"视频链接:{result}")