Update helper.py

helper.py

@@ -12,16 +12,17 @@ from moviepy.editor import VideoFileClip
 import wave
 
 # -------------------------------
-# Logging
+# Logging setup
 logger = logging.getLogger(__name__)
-logging.basicConfig(level=logging.INFO)
+logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
 
 # Hugging Face credentials
 HF_TOKEN = os.getenv("HF_TOKEN")
 DATASET_REPO = "EllenBeta/Voxai-data"
 
 # -------------------------------
-
+# Upload to Hugging Face Dataset
+# -------------------------------
 def save_to_dataset_repo(file_path, hf_path, file_name):
     """Upload audio file to Hugging Face dataset."""
     try:
@@ -43,7 +44,8 @@ def save_to_dataset_repo(file_path, hf_path, file_name):
         raise Exception(f"Upload failed: {str(e)}")
 
 # -------------------------------
-
+# Save metadata to PostgreSQL
+# -------------------------------
 def save_audio(user_id, audio_url, title, text, duration):
     """Save audio metadata to PostgreSQL."""
     conn = create_connection()
@@ -66,82 +68,122 @@ def save_audio(user_id, audio_url, title, text, duration):
             conn.close()
 
 # -------------------------------
-
+# Utility: Generate random filename
+# -------------------------------
 def generate_random_filename(extension="mp3", length=12):
     letters = string.ascii_letters + string.digits
     random_str = ''.join(random.choice(letters) for _ in range(length))
     return f"{random_str}.{extension}"
 
 # -------------------------------
-
+# Cut video safely
+# -------------------------------
 def cut_video(video_path, max_duration=2):
     """Cut video to a maximum duration using MoviePy."""
-    clip = VideoFileClip(video_path)
-    if clip.duration > max_duration:
-        clip = clip.subclip(0, max_duration)
-    return clip
+    try:
+        clip = VideoFileClip(video_path)
+        if clip.duration > max_duration:
+            clip = clip.subclip(0, max_duration)
+        return clip
+    except Exception as e:
+        raise Exception(f"Failed to open or trim video: {e}")
 
 # -------------------------------
-
+# Convert video to audio safely
+# -------------------------------
 def video_to_audio(video_input, output_path="/tmp/temp_reference.wav", max_duration=2):
-    """Convert video (or base64 video) to WAV audio using MoviePy."""
+    """
+    Convert video (file path, data URL, or base64) to WAV audio using MoviePy.
+    Handles all input formats safely and avoids oversized subprocess arguments.
+    """
     temp_video_path = None
+    clip = None
+
     try:
-        # Decode base64 input if needed
-        if video_input.startswith("data:video/"):
-            header, encoded = video_input.split(",", 1)
-            temp_video = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4")
-            temp_video_path = temp_video.name
-            with open(temp_video_path, "wb") as f:
-                f.write(base64.b64decode(encoded))
-            video_path = temp_video_path
+        # Detect and decode base64 or data URLs
+        if isinstance(video_input, str):
+            if video_input.startswith("data:video/"):  # full data URL
+                logger.info("📹 Detected data URL video input")
+                header, encoded = video_input.split(",", 1)
+                decoded_bytes = base64.b64decode(encoded)
+            elif len(video_input) > 500 and all(c.isalnum() or c in "+/=\n" for c in video_input.strip()):
+                logger.info("📹 Detected raw base64 video input")
+                decoded_bytes = base64.b64decode(video_input)
+            elif os.path.exists(video_input):  # already a valid path
+                logger.info(f"📁 Detected video file path: {video_input}")
+                video_path = video_input
+                clip = cut_video(video_path, max_duration)
+            else:
+                raise Exception("Invalid video input format (not file path or base64)")
         else:
-            video_path = video_input
+            raise Exception("Invalid video input type")
 
-        # Cut video if needed
-        clip = cut_video(video_path, max_duration)
-
-        # Extract audio
-        clip.audio.write_audiofile(output_path, fps=22050, nbytes=2, codec="pcm_s16le")
+        # If we got decoded bytes, write to a temp file
+        if clip is None:
+            with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as tmp:
+                tmp.write(decoded_bytes)
+                temp_video_path = tmp.name
+            video_path = temp_video_path
+            clip = cut_video(video_path, max_duration)
+
+        # Extract audio safely
+        clip.audio.write_audiofile(
+            output_path,
+            fps=22050,
+            nbytes=2,
+            codec="pcm_s16le",
+            verbose=False,
+            logger=None
+        )
         clip.close()
 
+        # Validate output
         if not os.path.exists(output_path) or os.path.getsize(output_path) == 0:
             raise Exception("Output audio file not created or empty")
 
-        logger.info(f"✅ Video converted to audio: {output_path}")
+        logger.info(f"✅ Video converted successfully → {output_path}")
         return output_path
 
     except Exception as e:
-        logger.error(f"Video→Audio conversion failed: {e}")
+        logger.error(f"❌ Video→Audio conversion failed: {e}")
         raise
+
     finally:
-        if temp_video_path and os.path.exists(temp_video_path):
+        if clip:
             try:
-                os.remove(temp_video_path)
+                clip.close()
             except:
                 pass
+        if temp_video_path and os.path.exists(temp_video_path):
+            try:
+                os.remove(temp_video_path)
+                logger.debug(f"🧹 Deleted temp file {temp_video_path}")
+            except Exception as e:
+                logger.warning(f"⚠️ Failed to delete temp file: {e}")
 
 # -------------------------------
-
+# Ensure WAV format
+# -------------------------------
 def ensure_wav_format(input_path):
-    """Convert to WAV if not already valid WAV."""
+    """Convert audio to WAV if not already valid WAV."""
     try:
         with wave.open(input_path, 'rb'):
-            return input_path
+            return input_path  # valid WAV
     except:
         try:
             audio = AudioSegment.from_file(input_path)
             wav_path = os.path.splitext(input_path)[0] + "_converted.wav"
             audio.export(wav_path, format="wav")
-            logger.info(f"Converted {input_path} → WAV ({wav_path})")
+            logger.info(f"🎵 Converted {input_path} → WAV ({wav_path})")
             return wav_path
         except Exception as e:
             raise Exception(f"Failed to convert to WAV: {e}")
 
 # -------------------------------
-
+# Validate audio file
+# -------------------------------
 def validate_audio_file(file_path, max_size_mb=10):
-    """Validate audio file: size, format, duration."""
+    """Validate audio file: size, format, and duration."""
     try:
         if not os.path.exists(file_path):
             return False, "File not found"