Update app.py

app.py

@@ -132,23 +132,16 @@ async def get_transcript(youtube_url: str):
     try:
         ydl_opts = {
             'skip_download': True,
-            'writesubtitles': True,
-            'writeautomaticsub': True,
-            # 精确指定中英文语言代码
-            'subtitleslangs': [
-                'zh', 'zh-Hans', 'zh-Hant', 'zh-CN', 'zh-TW',  # 中文
-                'en', 'en-US', 'en-GB', 'en-AU'  # 英文
-            ],
-            # 允许自动生成字幕作为后备
-            'subtitleslangs_automatic': True,
+            'writesubtitles': True,       # 仅下载手动上传的字幕
+            'writeautomaticsub': False,   # 禁用自动生成字幕
+            'subtitleslangs': ['original'],  # 关键参数：只获取原始语言
             'subtitlesformat': 'best',
             'outtmpl': '%(id)s.%(ext)s',
             'noplaylist': True,
             'cookiefile': "firefox-cookies.txt",
-            # 优化网络请求参数
-            'retries': 5,
-            'sleep_interval': 15,
-            'ignoreerrors': True,
+            # 禁用所有翻译功能
+            'postprocessors': [],
+            'compat_opts': {'no-sub-translate'},
         }
 
         env_to_cookies_from_env("firefox-cookies.txt")
@@ -157,72 +150,35 @@ async def get_transcript(youtube_url: str):
             info = ydl.extract_info(youtube_url, download=True)
             video_id = info['id']
             
-            # 获取实际可用的字幕语言
-            available_langs = []
-            for sub_type in ['subtitles', 'automatic_captions']:
-                subs = info.get(sub_type, {})
-                available_langs.extend(subs.keys())
+            # 获取原始语言代码（如视频是英语则返回en）
+            original_lang = info.get('original_lang') or 'en'
             
-            logger.info(f"Available subtitle languages: {list(set(available_langs))}")
-
-            # 构建优先级列表（中文 > 英文 > 其他）
-            priority_map = {
-                'zh': 0, 'zh-Hans': 1, 'zh-Hant': 2, 'zh-CN': 3, 'zh-TW': 4,
-                'en': 5, 'en-US': 6, 'en-GB': 7, 'en-AU': 8
-            }
+            # 查找匹配原始语言的字幕文件
+            subtitle_pattern = f"{video_id}.{original_lang}.*"
+            subtitle_files = [f for f in os.listdir('.') 
+                             if re.match(subtitle_pattern, f)]
             
-            # 查找最佳匹配字幕文件
-            best_sub = None
-            for f in os.listdir('.'):
-                if f.startswith(video_id):
-                    parts = f.split('.')
-                    if len(parts) >= 3:
-                        lang_code = parts[-2]
-                        # 处理复合语言代码（如 en-US）
-                        base_lang = lang_code.split('-')[0]
-                        priority = priority_map.get(base_lang, 99)
-                        if not best_sub or priority < best_sub[1]:
-                            best_sub = (f, priority)
-
-            if best_sub:
-                subtitle_file = best_sub[0]
+            logger.info(f"原始语言字幕文件: {subtitle_files}")
+
+            if subtitle_files:
+                # 选择第一个匹配的文件
+                subtitle_file = subtitle_files[0]
                 with open(subtitle_file, 'r', encoding='utf-8') as f:
                     content = f.read()
-
-                # 增强型文本清理
-                clean_text = re.sub(
-                    r'\d{2}:\d{2}:\d{2}[,.]\d{3}.*?(\n|$)|'  # 时间戳
-                    r'<[^>]+>|{[^}]+}|^\s*WEBVTT\s*$|'        # HTML/样式标签
-                    r'^\d+\s*$',                               # 序号行
-                    '', 
-                    content,
-                    flags=re.MULTILINE
-                ).strip()
-
-                # 合并重复的空行
-                clean_text = re.sub(r'\n{3,}', '\n\n', clean_text)
-
+                
                 return {
-                    "transcript": clean_text,
-                    "detected_lang": subtitle_file.split('.')[-2],
-                    "available_languages": available_langs
+                    "transcript": content,
+                    "language": original_lang,
+                    "is_translated": False
                 }
-
+            
             return {
-                "error": "No subtitles available",
-                "available_languages": available_langs
+                "error": "无原始语言字幕",
+                "available_langs": list(info.get('subtitles', {}).keys())
             }
 
     except Exception as e:
-        logger.error(f"Error: {str(e)}")
-        return {
-            "error": f"Processing failed: {str(e)}",
-            "advice": [
-                "Try adding '&tlang=en' to URL for translated subs",
-                "Check if cookies are still valid",
-                "Reduce request frequency if seeing 429 errors"
-            ]
-        }
+        return {"error": str(e)}