支持fastapi服务

reelforge/services/video_generator.py

@@ -56,6 +56,7 @@ class VideoGeneratorService:
         n_scenes: int = 5,  # Only used in generate mode; ignored in fixed mode
         voice_id: str = "zh-CN-YunjianNeural",
         output_path: Optional[str] = None,
+        use_uuid_filename: bool = False,  # Use UUID instead of timestamp for filename
         
         # === LLM Parameters ===
         min_narration_words: int = 5,
@@ -194,10 +195,17 @@ class VideoGeneratorService:
         
         # Auto-generate output path if not provided
         if output_path is None:
-            timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
-            # Use first 10 chars of final_title for filename
-            safe_name = final_title[:10].replace('/', '_').replace(' ', '_')
-            output_path = f"output/{timestamp}_{safe_name}.mp4"
+            if use_uuid_filename:
+                # API mode: use UUID for filename
+                import uuid
+                filename = str(uuid.uuid4()).replace('-', '')
+                output_path = f"output/{filename}.mp4"
+            else:
+                # Default mode: use timestamp + title
+                timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
+                # Use first 10 chars of final_title for filename
+                safe_name = final_title[:10].replace('/', '_').replace(' ', '_')
+                output_path = f"output/{timestamp}_{safe_name}.mp4"
         
         # Ensure output directory exists
         Path(output_path).parent.mkdir(parents=True, exist_ok=True)

reelforge/utils/tts_util.py

@@ -158,6 +158,32 @@ async def edge_tts(
         raise RuntimeError("Edge TTS failed without error (unexpected)")
 
 
+def get_audio_duration(audio_path: str) -> float:
+    """
+    Get audio file duration in seconds
+    
+    Args:
+        audio_path: Path to audio file
+    
+    Returns:
+        Duration in seconds
+    """
+    try:
+        # Try using ffmpeg-python
+        import ffmpeg
+        probe = ffmpeg.probe(audio_path)
+        duration = float(probe['format']['duration'])
+        return duration
+    except Exception as e:
+        logger.warning(f"Failed to get audio duration: {e}, using estimate")
+        # Fallback: estimate based on file size (very rough)
+        import os
+        file_size = os.path.getsize(audio_path)
+        # Assume ~16kbps for MP3, so 2KB per second
+        estimated_duration = file_size / 2000
+        return max(1.0, estimated_duration)  # At least 1 second
+
+
 async def list_voices(locale: str = None, retry_count: int = _RETRY_COUNT, retry_delay: float = _RETRY_DELAY) -> list[str]:
     """
     List all available voices for Edge TTS