Integrate OpenAI Whisper API Support for Enhanced Transcription Options #137

bot_handlers.py

@@ -1,15 +1,15 @@
 import logging
 import os
 from typing import Dict, Any
-from services import AWSServices, AudioTranscriber, TextSummarizer
+from services import AudioTranscriber, TextSummarizer
 from utils.telegram_utils import send_message, get_telegram_file_url
 from utils.message_utils import format_response, create_tip_button
+from config import Config
 
 logger = logging.getLogger(__name__)
 
 # Initialize services
-aws_services = AWSServices()
-audio_transcriber = AudioTranscriber(aws_services)
+audio_transcriber = AudioTranscriber(Config)
 text_summarizer = TextSummarizer(os.environ.get('MARKETROUTER_API_KEY'))
 
 def handle_update(update: Dict[str, Any]) -> None:

config.py

@@ -6,3 +6,5 @@ class Config:
     MARKETROUTER_API_KEY = os.environ.get('MARKETROUTER_API_KEY')
     AWS_ACCESS_KEY_ID = os.environ.get('AWS_ACCESS_KEY_ID')
     AWS_SECRET_ACCESS_KEY = os.environ.get('AWS_SECRET_ACCESS_KEY')
+    OPENAI_API_KEY = os.environ.get('OPENAI_API_KEY')
+    TRANSCRIPTION_SERVICE = os.environ.get('TRANSCRIPTION_SERVICE', 'aws')  # 'aws' or 'openai'

services.py

@@ -50,7 +50,13 @@ def start_transcription_job(self, job_name, media_uri, media_format='ogg', langu
     def get_transcription_job_status(self, job_name):
         return self.transcribe_client.get_transcription_job(TranscriptionJobName=job_name)
 
-class AudioTranscriber:
+class BaseTranscriber:
+    def _download_audio(self, file_url: str) -> bytes:
+        response = requests.get(file_url)
+        response.raise_for_status()
+        return response.content
+
+class AWSTranscriber(BaseTranscriber):
     def __init__(self, aws_services: AWSServices):
         self.aws_services = aws_services
         self.bucket_name = 'audio-transcribe-temp'
@@ -77,11 +83,6 @@ def transcribe_audio(self, file_url: str) -> str:
             logger.error(f"An error occurred: {e}")
             raise
 
-    def _download_audio(self, file_url: str) -> bytes:
-        response = requests.get(file_url)
-        response.raise_for_status()
-        return response.content
-
     def _wait_for_transcription(self, job_name: str) -> str:
         while True:
             status = self.aws_services.get_transcription_job_status(job_name)
@@ -95,6 +96,61 @@ def _wait_for_transcription(self, job_name: str) -> str:
         else:
             raise Exception("Transcription failed")
 
+class OpenAITranscriber(BaseTranscriber):
+    def __init__(self, api_key: str):
+        self.api_key = api_key
+        self.api_url = "https://api.openai.com/v1/audio/transcriptions"
+
+    def transcribe_audio(self, file_url: str) -> str:
+        try:
+            audio_content = self._download_audio(file_url)
+
+            # Save audio content to a temporary file
+            temp_file = f'/tmp/audio_{uuid.uuid4()}.ogg'
+            with open(temp_file, 'wb') as f:
+                f.write(audio_content)
+
+            # Prepare the request
+            headers = {
+                'Authorization': f'Bearer {self.api_key}'
+            }
+
+            files = {
+                'file': ('audio.ogg', open(temp_file, 'rb'), 'audio/ogg'),
+                'model': (None, 'whisper-1'),
+            }
+
+            response = requests.post(
+                self.api_url,
+                headers=headers,
+                files=files
+            )
+            response.raise_for_status()
+
+            # Clean up temporary file
+            os.remove(temp_file)
+
+            return response.json()['text']
+        except Exception as e:
+            logger.error(f"An error occurred with OpenAI transcription: {e}")
+            raise
+
+class AudioTranscriber:
+    def __init__(self, config):
+        self.config = config
+        if config.TRANSCRIPTION_SERVICE == 'aws':
+            aws_services = AWSServices(region_name=config.AWS_REGION)
+            self.transcriber = AWSTranscriber(aws_services)
+        elif config.TRANSCRIPTION_SERVICE == 'openai':
+            if not config.OPENAI_API_KEY:
+                raise ValueError("OpenAI API key is required for OpenAI transcription service")
+            self.transcriber = OpenAITranscriber(config.OPENAI_API_KEY)
+        else:
+            raise ValueError(f"Unsupported transcription service: {config.TRANSCRIPTION_SERVICE}")
+
+    def transcribe_audio(self, file_url: str) -> str:
+        return self.transcriber.transcribe_audio(file_url)
+
 class TextSummarizer:
     def __init__(self, api_key: str):
         self.api_key = api_key