Spaces:
Running
Running
File size: 8,771 Bytes
52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c 52fad77 913182c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
"""
STT Lifecycle Manager for Flare - Batch Mode
===============================
Manages STT instances and audio collection
"""
import asyncio
from typing import Dict, Optional, Any
from datetime import datetime
import traceback
import base64
from chat_session.event_bus import EventBus, Event, EventType, publish_error
from chat_session.resource_manager import ResourceManager, ResourceType
from stt.stt_factory import STTFactory
from stt.stt_interface import STTInterface, STTConfig, TranscriptionResult
from stt.voice_activity_detector import VoiceActivityDetector
from utils.logger import log_info, log_error, log_debug, log_warning
class STTSession:
"""STT session with audio collection"""
def __init__(self, session_id: str, stt_instance: STTInterface):
self.session_id = session_id
self.stt_instance = stt_instance
self.is_active = False
self.config: Optional[STTConfig] = None
self.created_at = datetime.utcnow()
# Audio collection
self.audio_buffer = []
self.vad = VoiceActivityDetector()
# Stats
self.total_chunks = 0
self.total_bytes = 0
def reset(self):
"""Reset session for new utterance"""
self.audio_buffer = []
self.vad.reset()
self.total_chunks = 0
self.total_bytes = 0
class STTLifecycleManager:
"""Manages STT instances lifecycle"""
def __init__(self, event_bus: EventBus, resource_manager: ResourceManager):
self.event_bus = event_bus
self.resource_manager = resource_manager
self.stt_sessions: Dict[str, STTSession] = {}
self._setup_event_handlers()
self._setup_resource_pool()
def _setup_event_handlers(self):
"""Subscribe to STT-related events"""
self.event_bus.subscribe(EventType.STT_STARTED, self._handle_stt_start)
self.event_bus.subscribe(EventType.STT_STOPPED, self._handle_stt_stop)
self.event_bus.subscribe(EventType.AUDIO_CHUNK_RECEIVED, self._handle_audio_chunk)
self.event_bus.subscribe(EventType.SESSION_ENDED, self._handle_session_ended)
async def _handle_stt_start(self, event: Event):
"""Handle STT start request"""
session_id = event.session_id
config_data = event.data
try:
log_info(f"🎤 Starting STT", session_id=session_id)
# Get or create session
if session_id not in self.stt_sessions:
# Acquire STT instance from pool
resource_id = f"stt_{session_id}"
stt_instance = await self.resource_manager.acquire(
resource_id=resource_id,
session_id=session_id,
resource_type=ResourceType.STT_INSTANCE,
cleanup_callback=self._cleanup_stt_instance
)
# Create session
stt_session = STTSession(session_id, stt_instance)
self.stt_sessions[session_id] = stt_session
else:
stt_session = self.stt_sessions[session_id]
stt_session.reset()
# Build STT config
locale = config_data.get("locale", "tr")
stt_config = STTConfig(
language=self._get_language_code(locale),
sample_rate=config_data.get("sample_rate", 16000),
encoding=config_data.get("encoding", "LINEAR16"),
enable_punctuation=config_data.get("enable_punctuation", True),
model=config_data.get("model", "latest_long"),
use_enhanced=config_data.get("use_enhanced", True),
)
stt_session.config = stt_config
stt_session.is_active = True
log_info(f"✅ STT started in batch mode", session_id=session_id, language=stt_config.language)
# Notify STT is ready
await self.event_bus.publish(Event(
type=EventType.STT_READY,
session_id=session_id,
data={"language": stt_config.language}
))
except Exception as e:
log_error(
f"❌ Failed to start STT",
session_id=session_id,
error=str(e),
traceback=traceback.format_exc()
)
# Clean up on error
if session_id in self.stt_sessions:
await self._cleanup_session(session_id)
# Publish error event
await publish_error(
session_id=session_id,
error_type="stt_error",
error_message=f"Failed to start STT: {str(e)}"
)
async def _handle_audio_chunk(self, event: Event):
"""Process audio chunk through VAD and collect"""
session_id = event.session_id
stt_session = self.stt_sessions.get(session_id)
if not stt_session or not stt_session.is_active:
return
try:
# Decode audio data
audio_data = base64.b64decode(event.data.get("audio_data", ""))
# Add to buffer
stt_session.audio_buffer.append(audio_data)
stt_session.total_chunks += 1
stt_session.total_bytes += len(audio_data)
# Process through VAD
is_speech, silence_duration_ms = stt_session.vad.process_chunk(audio_data)
# Check if utterance ended (silence threshold reached)
if not is_speech and silence_duration_ms >= 2000: # 2 seconds of silence
log_info(f"💬 Utterance ended after {silence_duration_ms}ms silence", session_id=session_id)
# Stop STT to trigger transcription
await self.event_bus.publish(Event(
type=EventType.STT_STOPPED,
session_id=session_id,
data={"reason": "silence_detected"}
))
# Log progress periodically
if stt_session.total_chunks % 100 == 0:
log_debug(
f"📊 STT progress",
session_id=session_id,
chunks=stt_session.total_chunks,
bytes=stt_session.total_bytes,
vad_stats=stt_session.vad.get_stats()
)
except Exception as e:
log_error(
f"❌ Error processing audio chunk",
session_id=session_id,
error=str(e)
)
async def _handle_stt_stop(self, event: Event):
"""Handle STT stop request and perform transcription"""
session_id = event.session_id
reason = event.data.get("reason", "unknown")
log_info(f"🛑 Stopping STT", session_id=session_id, reason=reason)
stt_session = self.stt_sessions.get(session_id)
if not stt_session:
log_warning(f"⚠️ No STT session found", session_id=session_id)
return
try:
if stt_session.is_active and stt_session.audio_buffer:
# Combine audio chunks
combined_audio = b''.join(stt_session.audio_buffer)
# Transcribe using batch mode
log_info(f"📝 Transcribing {len(combined_audio)} bytes of audio", session_id=session_id)
result = await stt_session.stt_instance.transcribe(
audio_data=combined_audio,
config=stt_session.config
)
# Publish result if we got transcription
if result and result.text:
await self.event_bus.publish(Event(
type=EventType.STT_RESULT,
session_id=session_id,
data={
"text": result.text,
"is_final": True,
"confidence": result.confidence
}
))
else:
log_warning(f"⚠️ No transcription result", session_id=session_id)
# Mark as inactive and reset
stt_session.is_active = False
stt_session.reset()
# Send STT_STOPPED event
await self.event_bus.publish(Event(
type=EventType.STT_STOPPED,
session_id=session_id,
data={"reason": reason}
))
log_info(f"✅ STT stopped", session_id=session_id)
except Exception as e:
log_error(
f"❌ Error stopping STT",
session_id=session_id,
error=str(e)
) |