mic-system/app.py

from __future__ import annotations

from dataclasses import asdict
from pathlib import Path
import os

from flask import Flask, abort, jsonify, render_template, request, send_from_directory
from flask_socketio import SocketIO, emit

from audio_capture import AudioEngine
from recorder import list_recordings, resolve_recording_path
from stt_bridge import SttBridge


BASE_DIR = Path(__file__).resolve().parent
RECORDINGS_DIR = BASE_DIR / "recordings"

app = Flask(__name__, template_folder="templates", static_folder="static")
app.config["SECRET_KEY"] = os.environ.get("MIC_SYSTEM_SECRET", "mic-system-dev")

socketio = SocketIO(app, cors_allowed_origins="*", async_mode="eventlet")
audio_engine = AudioEngine(RECORDINGS_DIR)


def _stt_message_callback(msg):
    """Called when STT worker posts a result via /internal/stt."""
    socketio.emit("stt_message", msg)


stt_bridge = SttBridge(on_message=_stt_message_callback)

# Attach STT bridge to audio engine so it receives processed audio
audio_engine.set_stt_bridge(stt_bridge)

_stream_task_started = False
audio_start_error = ""


def ensure_audio_started() -> None:
    global audio_start_error
    if audio_engine.is_running():
        return

    try:
        audio_engine.start()
        audio_start_error = ""
    except Exception as exc:  # pragma: no cover - runtime environment specific
        audio_start_error = str(exc)


def stream_audio_packets() -> None:
    while True:
        packet = audio_engine.get_latest_packet()
        socketio.emit("audio_data", packet)
        socketio.sleep(0.08)


@app.route("/")
def index() -> str:
    ensure_audio_started()
    return render_template("index.html")


@app.route("/api/status")
def api_status():
    ensure_audio_started()
    status = audio_engine.get_status()
    status["audio_error"] = audio_start_error
    status.update(stt_bridge.get_settings())
    return jsonify(status)


@app.route("/internal/stt", methods=["POST"])
def internal_stt():
    """Receives JSON messages from stt_worker subprocess via HTTP POST."""
    msg = request.get_json(silent=True)
    if msg:
        stt_bridge.handle_worker_message(msg)
    return "", 204


@app.route("/api/recordings")
def api_list_recordings():
    return jsonify(list_recordings(RECORDINGS_DIR))


@app.route("/api/recordings/<path:filename>")
def api_get_recording(filename: str):
    try:
        path = resolve_recording_path(RECORDINGS_DIR, filename)
    except ValueError:
        abort(400, description="Invalid filename")

    if not path.exists():
        abort(404)

    return send_from_directory(RECORDINGS_DIR, path.name, as_attachment=True)


@app.route("/api/recordings/<path:filename>", methods=["DELETE"])
def api_delete_recording(filename: str):
    try:
        path = resolve_recording_path(RECORDINGS_DIR, filename)
    except ValueError:
        abort(400, description="Invalid filename")

    if not path.exists():
        abort(404)

    path.unlink()
    socketio.emit("recordings_updated", {"ok": True})
    return jsonify({"ok": True})


def _handle_command(payload: dict) -> dict:
    msg_type = payload.get("type")

    if msg_type == "settings":
        settings = audio_engine.update_settings(payload)
        stt_keys = {k: v for k, v in payload.items() if k.startswith("stt_")}
        if stt_keys:
            stt_settings = stt_bridge.update_settings(**stt_keys)
            settings.update(stt_settings)
        return {"type": "settings_applied", "settings": settings}

    if msg_type == "stt_settings":
        stt_keys = {k: v for k, v in payload.items() if k.startswith("stt_")}
        stt_settings = stt_bridge.update_settings(**stt_keys)
        return {"type": "stt_settings_applied", "settings": stt_settings}

    if msg_type == "record_start":
        source = str(payload.get("source", "mic1"))
        duration_sec = payload.get("duration_sec")
        record_info = audio_engine.start_recording(source, duration_sec=duration_sec)
        socketio.emit("recordings_updated", {"ok": True})
        return {"type": "record_started", **record_info}

    if msg_type == "record_stop":
        status = audio_engine.stop_recording()
        socketio.emit("recordings_updated", {"ok": True})
        return {"type": "record_stopped", "status": asdict(status)}

    if msg_type == "delete_recording":
        filename = str(payload.get("filename", ""))
        if not filename:
            raise ValueError("Missing filename")
        path = resolve_recording_path(RECORDINGS_DIR, filename)
        if path.exists():
            path.unlink()
            socketio.emit("recordings_updated", {"ok": True})
        return {"type": "recording_deleted", "filename": filename}

    raise ValueError("Unsupported message type: " + str(msg_type))


@socketio.on("connect")
def ws_connect():
    global _stream_task_started

    ensure_audio_started()

    if not _stream_task_started:
        socketio.start_background_task(stream_audio_packets)
        _stream_task_started = True

    status = audio_engine.get_status()
    status.update(stt_bridge.get_settings())
    emit("status", status)
    if audio_start_error:
        emit("server_error", {"message": audio_start_error})


@socketio.on("client_message")
def ws_client_message(payload):
    if not isinstance(payload, dict):
        emit("server_error", {"message": "Invalid payload"})
        return

    try:
        response = _handle_command(payload)
    except Exception as exc:
        emit("server_error", {"message": str(exc)})
        return

    emit("server_ack", response)


@socketio.on("disconnect")
def ws_disconnect():
    return


if __name__ == "__main__":
    ensure_audio_started()
    host = os.environ.get("MIC_SYSTEM_HOST", "0.0.0.0")
    port = int(os.environ.get("MIC_SYSTEM_PORT", "5000"))
    socketio.run(app, host=host, port=port)