Update app.py

app.py

@@ -26,6 +26,13 @@ import logging
 from textblob import TextBlob
 import whisper
 import time
+import os
+import openai 
+from openai import OpenAI 
+
+client = OpenAI(
+    api_key= os.environ.get("openAI_api_key"),  # This is the default and can be omitted
+)
 
 def silence(duration, fps=44100):
     """
@@ -240,7 +247,7 @@ def process_entry(entry, i, video_width, video_height, add_voiceover, target_lan
     audio_segment = None
     if add_voiceover:
         segment_audio_path = f"segment_{i}_voiceover.wav"
-        generate_voiceover([entry], target_language, segment_audio_path)
+        generate_voiceover_OpenAI([entry], target_language, segment_audio_path)
         audio_clip = AudioFileClip(segment_audio_path)
         # Get and log all methods in AudioFileClip        
         logger.info("Methods in AudioFileClip:")
@@ -324,7 +331,46 @@ def generate_voiceover(translated_json, language, output_audio_path):
         tts.save(output_audio_path)
     except Exception as e:
         raise ValueError(f"Error generating voiceover: {e}")
-    
+
+def generate_voiceover_OpenAI(translated_json, language, output_audio_path):
+    """
+    Generate voiceover from translated text for a given language using OpenAI TTS API.
+    """
+    # Concatenate translated text into a single string
+    full_text = " ".join(entry["translated"] for entry in translated_json)
+
+    # Define the voice based on the language (for now, use 'alloy' as default)
+    voice = "alloy"  # Adjust based on language if needed
+
+    # Define the model (use tts-1 for real-time applications)
+    model = "tts-1"
+
+    max_retries = 3
+    retry_count = 0
+
+    while retry_count < max_retries:
+        try:
+            # Create the speech using OpenAI TTS API
+            response = client.audio.speech.create(
+                model=model,
+                voice=voice,
+                input=full_text
+            )
+            # Save the audio to the specified path
+            with open(output_audio_path, 'wb') as f:
+                for chunk in response.iter_bytes():
+                    f.write(chunk)
+            logging.info(f"Voiceover generated successfully for {output_audio_path}")
+            break
+
+        except Exception as e:
+            retry_count += 1
+            logging.error(f"Error generating voiceover (retry {retry_count}/{max_retries}): {e}")
+            time.sleep(5)  # Wait 5 seconds before retrying
+
+    if retry_count == max_retries:
+        raise ValueError(f"Failed to generate voiceover after {max_retries} retries.")
+
 def upload_and_manage(file, target_language, mode="transcription"):
     if file is None:
         logger.info("No file uploaded. Please upload a video/audio file.")