src/tts_service.py

"""Text-to-Speech Service using ElevenLabs"""
import os
from typing import Optional, List, Dict
from elevenlabs import ElevenLabs, VoiceSettings

# Voice options mapping (name -> voice_id)
VOICE_OPTIONS = {
    "Rachel (Female)": "21m00Tcm4TlvDq8ikWAM",  # Default - Rachel
    "Lera (Female)": "EXAVITQu4vr4xnSDxMaL",  # Lera - female voice
    "Bella (Female)": "EXAVITQu4vr4xnSDxMaL",  # Alternative female
    "Antoni (Male)": "ErXwobaYiN019PkySvjV",  # Antoni - male voice
    "Arnold (Male)": "VR6AewLTigWG4xSOukaG",  # Arnold - male voice
    "Adam (Male)": "pNInz6obpgDQGcFmaJgB",  # Adam - male voice
    "Domi (Female)": "AZnzlk1XvdvUeBnXmlld",  # Domi - female voice
    "Elli (Female)": "MF3mGyEYCl7XYWbV9V6O",  # Elli - female voice
    "Josh (Male)": "TxGEqnHWrfWFTfGW9XjX",  # Josh - male voice
    "Sam (Male)": "yoZ06aMxZJJ28mfd3POQ",  # Sam - male voice
}


class TTSService:
    """Text-to-Speech service using ElevenLabs API"""
    
    def __init__(self, api_key: str, voice_id: str = "21m00Tcm4TlvDq8ikWAM"):
        """
        Initialize TTS service
        
        Args:
            api_key: ElevenLabs API key
            voice_id: Voice ID to use (default: Rachel)
        """
        self.api_key = api_key
        self.voice_id = voice_id
        self.client = None
        self.available = False
        
        if api_key:
            try:
                self.client = ElevenLabs(api_key=api_key)
                self.available = True
            except Exception as e:
                print(f"Error initializing ElevenLabs client: {e}")
    
    def text_to_speech(self, text: str, voice_id: Optional[str] = None) -> Optional[bytes]:
        """
        Convert text to speech
        
        Args:
            text: Text to convert
            voice_id: Optional voice ID override
            
        Returns:
            Audio bytes or None if error
        """
        if not self.client:
            print("ElevenLabs client not initialized. Please check API key.")
            return None
        
        try:
            voice_to_use = voice_id or self.voice_id
            
            # Generate audio using newer model (free tier compatible)
            audio = self.client.generate(
                text=text,
                voice=voice_to_use,
                model="eleven_turbo_v2_5"  # Free tier compatible model
            )
            
            # Convert generator to bytes
            audio_bytes = b""
            for chunk in audio:
                audio_bytes += chunk
            
            return audio_bytes
            
        except Exception as e:
            print(f"Error generating speech: {e}")
            return None
    
    def text_to_speech_stream(self, text: str, voice_id: Optional[str] = None):
        """
        Convert text to speech with streaming
        
        Args:
            text: Text to convert
            voice_id: Optional voice ID override
            
        Yields:
            Audio chunks
        """
        if not self.client:
            print("ElevenLabs client not initialized. Please check API key.")
            return
        
        try:
            voice_to_use = voice_id or self.voice_id
            
            # Stream audio using newer model (free tier compatible)
            audio_stream = self.client.generate(
                text=text,
                voice=voice_to_use,
                model="eleven_turbo_v2_5",
                stream=True
            )
            
            for chunk in audio_stream:
                yield chunk
                
        except Exception as e:
            print(f"Error streaming speech: {e}")
            return
    
    def save_audio(self, audio_bytes: bytes, filename: str) -> bool:
        """
        Save audio bytes to file
        
        Args:
            audio_bytes: Audio data
            filename: Output filename
            
        Returns:
            Success status
        """
        try:
            with open(filename, 'wb') as f:
                f.write(audio_bytes)
            return True
        except Exception as e:
            print(f"Error saving audio: {e}")
            return False
    
    def get_available_voices(self) -> List[Dict[str, str]]:
        """
        Get list of available voices
        
        Returns:
            List of voice information
        """
        if not self.client:
            return []
        
        try:
            voices = self.client.voices.get_all()
            return [{"voice_id": v.voice_id, "name": v.name} for v in voices.voices]
        except Exception as e:
            print(f"Error getting voices: {e}")
            return [
                {"voice_id": "21m00Tcm4TlvDq8ikWAM", "name": "Rachel (Default)"},
                {"voice_id": "ErXwobaYiN019PkySvjV", "name": "Antoni"},
                {"voice_id": "MF3mGyEYCl7XYWbV9V6O", "name": "Elli"},
            ]