tts_interface.py

"""
TTS Interface and Implementations
"""

from abc import ABC, abstractmethod
from typing import Optional, Dict, Any
import httpx
import os
from datetime import datetime
import sys

def log(message: str):
    timestamp = datetime.now().strftime("%H:%M:%S.%f")[:-3]
    print(f"[{timestamp}] {message}")
    sys.stdout.flush()

class TTSInterface(ABC):
    """Abstract base class for TTS providers"""
    
    @abstractmethod
    async def synthesize(self, text: str, voice_id: Optional[str] = None, **kwargs) -> bytes:
        """
        Convert text to speech and return audio bytes
        
        Args:
            text: Text to convert to speech
            voice_id: Optional voice ID specific to the provider
            **kwargs: Additional provider-specific parameters
            
        Returns:
            Audio data as bytes (MP3 or WAV format)
        """
        pass
    
    @abstractmethod
    def get_supported_voices(self) -> Dict[str, str]:
        """Get list of supported voices"""
        pass


class ElevenLabsTTS(TTSInterface):
    """ElevenLabs TTS implementation"""
    
    def __init__(self, api_key: str):
        self.api_key = api_key
        self.base_url = "https://api.elevenlabs.io/v1"
        self.default_voice_id = "2thYbn2sOGtiTwd9QwWH" # Avencia

                # Debug log
        masked_key = f"{api_key[:4]}...{api_key[-4:]}" if len(api_key) > 8 else "***"
        log(f"🔑 ElevenLabsTTS initialized with key: {masked_key}")
        
    async def synthesize(self, text: str, voice_id: Optional[str] = None, **kwargs) -> bytes:
        """Convert text to speech using ElevenLabs API"""
        try:
            voice = voice_id or self.default_voice_id
            url = f"{self.base_url}/text-to-speech/{voice}"
            
            headers = {
                "xi-api-key": self.api_key,
                "Content-Type": "application/json"
            }
            
            # Default parameters
            data = {
                "text": text,
                "model_id": kwargs.get("model_id", "eleven_multilingual_v2"),
                "voice_settings": kwargs.get("voice_settings", {
                "stability": 1,
                "similarity_boost": 0.85, 
                "style": 0.7,             
                "speed": 1.14,
                "use_speaker_boost": True
                })
            }
            
            # Add optional parameters
            if "output_format" in kwargs:
                params = {"output_format": kwargs["output_format"]}
            else:
                params = {"output_format": "mp3_44100_128"}
            
            log(f"🎤 Calling ElevenLabs TTS for {len(text)} characters")
            
            async with httpx.AsyncClient(timeout=30) as client:
                response = await client.post(
                    url,
                    headers=headers,
                    json=data,
                    params=params
                )
                
                response.raise_for_status()
                audio_data = response.content
                
                log(f"✅ ElevenLabs TTS returned {len(audio_data)} bytes")
                return audio_data
                
        except httpx.HTTPStatusError as e:
            log(f"❌ ElevenLabs API error: {e.response.status_code} - {e.response.text}")
            raise
        except Exception as e:
            log(f"❌ TTS synthesis error: {e}")
            raise
    
    def get_supported_voices(self) -> Dict[str, str]:
        """Get default voices - full list can be fetched from API"""
        return {
            "21m00Tcm4TlvDq8ikWAM": "Rachel (Female)",
            "EXAVITQu4vr4xnSDxMaL": "Bella (Female)",
            "ErXwobaYiN019PkySvjV": "Antoni (Male)",
            "VR6AewLTigWG4xSOukaG": "Arnold (Male)",
            "pNInz6obpgDQGcFmaJgB": "Adam (Male)",
            "yoZ06aMxZJJ28mfd3POQ": "Sam (Male)",
        }


class BlazeTTS(TTSInterface):
    """Placeholder for future Blaze TTS implementation"""
    
    def __init__(self, api_key: str):
        self.api_key = api_key
        
    async def synthesize(self, text: str, voice_id: Optional[str] = None, **kwargs) -> bytes:
        raise NotImplementedError("Blaze TTS not implemented yet")
    
    def get_supported_voices(self) -> Dict[str, str]:
        return {}


def create_tts_provider(engine: str, api_key: Optional[str] = None) -> Optional[TTSInterface]:
    """Factory function to create TTS provider instances"""
    if engine == "elevenlabs" and api_key:
        return ElevenLabsTTS(api_key)
    elif engine == "blaze" and api_key:
        return BlazeTTS(api_key)
    elif engine == "no_tts":
        return None
    else:
        log(f"⚠️ Unknown or unconfigured TTS engine: {engine}")
        return None