Update app.py

app.py

@@ -134,14 +134,20 @@ async def get_transcript(youtube_url: str):
             'skip_download': True,
             'writesubtitles': True,
             'writeautomaticsub': True,
-            # 同时支持中英文及变体
-            'subtitleslangs': ['zh*', 'en*'],  # 包含所有中文和英语变体
+            # 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,
             'subtitlesformat': 'best',
             'outtmpl': '%(id)s.%(ext)s',
             'noplaylist': True,
             'cookiefile': "firefox-cookies.txt",
-            'retries': 3,
-            'sleep_interval': 10,  # 增加请求间隔
+            # 优化网络请求参数
+            'retries': 5,
+            'sleep_interval': 15,
             'ignoreerrors': True,
         }
 
@@ -151,56 +157,72 @@ async def get_transcript(youtube_url: str):
             info = ydl.extract_info(youtube_url, download=True)
             video_id = info['id']
             
-            # 语言优先级列表（中文 > 英文 > 其他）
-            priority_order = [
-                'zh', 'zh-Hans', 'zh-Hant', 'zh-CN', 'zh-TW',  # 中文
-                'en', 'en-US', 'en-GB',  # 英文
-            ]
+            # 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())
             
-            # 获取所有字幕文件并按优先级排序
-            subtitle_files = []
+            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
+            }
+            
+            # 查找最佳匹配字幕文件
+            best_sub = None
             for f in os.listdir('.'):
                 if f.startswith(video_id):
-                    lang_part = f.split('.')[-2]
-                    # 检查是否包含优先语言代码
-                    for lang in priority_order:
-                        if lang in lang_part:
-                            subtitle_files.append(f)
-                            break
-            
-            # 按优先级排序文件
-            subtitle_files = sorted(
-                subtitle_files,
-                key=lambda x: min(
-                    priority_order.index(lang) 
-                    for lang in priority_order 
-                    if lang in x
-                )
-            )
-
-            if subtitle_files:
-                selected_file = subtitle_files[0]
-                with open(selected_file, 'r', encoding='utf-8') as f:
+                    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]
+                with open(subtitle_file, 'r', encoding='utf-8') as f:
                     content = f.read()
 
-                # 统一清理逻辑
+                # 增强型文本清理
                 clean_text = re.sub(
-                    r'\d{2}:\d{2}:\d{2}[\.\,]\d{3}.*?(\n|$)|'
-                    r'<.*?>|{.*?}|^WEBVTT$',
+                    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": selected_file.split('.')[-2]
+                    "detected_lang": subtitle_file.split('.')[-2],
+                    "available_languages": available_langs
                 }
 
-            return {"error": "No subtitles found"}
+            return {
+                "error": "No subtitles available",
+                "available_languages": available_langs
+            }
 
     except Exception as e:
-        return {"error": str(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"
+            ]
+        }