stt/stt_interface.py

"""
STT (Speech-to-Text) Interface for Flare - Simple Batch Mode
"""
from abc import ABC, abstractmethod
from typing import Optional, List
from dataclasses import dataclass
from enum import Enum


class STTEngineType(Enum):
    NO_STT = "no_stt"
    GOOGLE = "google"
    AZURE = "azure"
    AMAZON = "amazon"
    DEEPGRAM = "deepgram"


@dataclass
class STTConfig:
    """STT configuration parameters"""
    language: str = "tr-TR"
    sample_rate: int = 16000
    encoding: str = "LINEAR16"
    enable_punctuation: bool = True
    enable_word_timestamps: bool = False
    model: str = "latest_long"
    use_enhanced: bool = True
    

@dataclass
class TranscriptionResult:
    """Result from STT engine"""
    text: str
    confidence: float
    timestamp: float
    language: Optional[str] = None
    word_timestamps: Optional[List[dict]] = None


class STTInterface(ABC):
    """Abstract base class for STT providers - Simple batch mode"""
    
    @abstractmethod
    async def transcribe(self, audio_data: bytes, config: STTConfig) -> Optional[TranscriptionResult]:
        """
        Transcribe audio data
        
        Args:
            audio_data: Raw PCM audio data (LINEAR16 format)
            config: STT configuration
            
        Returns:
            TranscriptionResult or None if no speech detected
        """
        pass
    
    @abstractmethod
    def get_supported_languages(self) -> List[str]:
        """Get list of supported language codes"""
        pass
    
    @abstractmethod
    def get_provider_name(self) -> str:
        """Get provider name for logging"""
        pass