Update app.py

app.py

@@ -393,24 +393,24 @@ def root():
 class TTSResponse(BaseModel):
     fileName: str
     duration: float  # seconds
-    audioBase64: str   
+    audioPath: str   
 
-######## Convert your audio to Base64
-import base64
-import torchaudio
-import io
+# ######## Convert your audio to Base64
+# import base64
+# import torchaudio
+# import io
 
-def audio_to_base64(audio_path: str) -> (str, float):
-    # load audio to get duration
-    waveform, sr = torchaudio.load(audio_path)  # waveform shape: [channels, samples]
-    duration = waveform.shape[1] / sr  # seconds
+# def audio_to_base64(audio_path: str) -> (str, float):
+#     # load audio to get duration
+#     waveform, sr = torchaudio.load(audio_path)  # waveform shape: [channels, samples]
+#     duration = waveform.shape[1] / sr  # seconds
 
-    # read file bytes
-    with open(audio_path, "rb") as f:
-        audio_bytes = f.read()
+#     # read file bytes
+#     with open(audio_path, "rb") as f:
+#         audio_bytes = f.read()
 
-    audio_b64 = base64.b64encode(audio_bytes).decode("utf-8")
-    return audio_b64, duration
+#     audio_b64 = base64.b64encode(audio_bytes).decode("utf-8")
+#     return audio_b64, duration
 
 #---------------------------concatenate text with tags ---------------------------
 
@@ -474,10 +474,57 @@ def generate_tagged_text(text: str, emotion_enum: str, intensity_enum: str) -> s
 
 #     return response
 
+
+# async def run_tts_pipeline(task_id: str, story: StoryCreationDTO):
+#     try:
+#         await generate_story_audios(story, base_output=OUTPUT_DIR)
+
+#         final_story_path = os.path.join(
+#             OUTPUT_DIR,
+#             story.storyId,
+#             f"{story.storyId}_full.wav"
+#         )
+
+#         final_generated_story_path = await concat_story_audio(
+#             story,
+#             base_output=OUTPUT_DIR,
+#             final_path=final_story_path
+#         )
+
+#         audio_b64, duration = audio_to_base64(final_generated_story_path)
+
+#         tasks[task_id] = {
+#             "status": "completed",
+#             "result": {
+#                 "fileName": os.path.basename(final_generated_story_path),
+#                 "duration": duration,
+#                 "audioPath": audio_b64
+#             }
+#         }
+
+#     except Exception as e:
+#         print(f"Exception caught at run tts pipeline {str(e)} and status is now failed")
+#         tasks[task_id] = {
+#             "status": "failed",
+#             "error": str(e)
+#         }
+
+import os
+import uuid
+from supabase import create_client, Client
+from pydub import AudioSegment  # For duration in seconds
+
+# Initialize Supabase client
+SUPABASE_URL = "https://kvlxvhdgacktsgykyckm.supabase.co/"
+SUPABASE_KEY = "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6Imt2bHh2aGRnYWNrdHNneWt5Y2ttIiwicm9sZSI6InNlcnZpY2Vfcm9sZSIsImlhdCI6MTc3MTk2MTQ5MSwiZXhwIjoyMDg3NTM3NDkxfQ.tzfHcbzwzctHDDDp3vk4JGz30ajN2szncAV-1wK7_pM"
+supabase: Client = create_client(SUPABASE_URL, SUPABASE_KEY)
+
 async def run_tts_pipeline(task_id: str, story: StoryCreationDTO):
     try:
+        # 1️⃣ Generate story audios
         await generate_story_audios(story, base_output=OUTPUT_DIR)
 
+        # 2️⃣ Concatenate final story audio
         final_story_path = os.path.join(
             OUTPUT_DIR,
             story.storyId,
@@ -490,19 +537,36 @@ async def run_tts_pipeline(task_id: str, story: StoryCreationDTO):
             final_path=final_story_path
         )
 
-        audio_b64, duration = audio_to_base64(final_generated_story_path)
+        # 3️⃣ Calculate duration
+        audio_segment = AudioSegment.from_file(final_generated_story_path)
+        duration_seconds = len(audio_segment) / 1000  # pydub gives length in milliseconds
+
+        # 4️⃣ Prepare the file for upload
+        file_name = f"{uuid.uuid4()}_{os.path.basename(final_generated_story_path)}"
+        storage_path = f"{story.storyId}/final/{file_name}"
+
+        # 5️⃣ Upload to Supabase
+        with open(final_generated_story_path, "rb") as f:
+            supabase.storage.from_("story-audio-files").upload(
+                storage_path,
+                f,
+                content_type="audio/wav"
+            )
+
+        # 6️⃣ Get public URL
+        audio_url = supabase.storage.from_("story-audio-files").get_public_url(storage_path)
 
+        # 7️⃣ Update task status with audio URL and duration
         tasks[task_id] = {
             "status": "completed",
             "result": {
                 "fileName": os.path.basename(final_generated_story_path),
-                "duration": duration,
-                "audioBase64": audio_b64
+                "duration": duration_seconds,
+                "audioPath": audio_url
             }
         }
 
     except Exception as e:
-        print(f"Exception caught at run tts pipeline {str(e)} and status is now failed")
         tasks[task_id] = {
             "status": "failed",
             "error": str(e)
@@ -564,7 +628,7 @@ async def get_results(task_id: str):
 
     # Ensure result exists and has all required fields
     result = task.get("result")
-    if result and all(k in result for k in ("fileName", "duration", "audioBase64")):
+    if result and all(k in result for k in ("fileName", "duration", "audioPath")):
         #clearing cache
         for file_path in download_cache.values():
             if os.path.exists(file_path):