jarvis-server-v2/jarvis/api.py

import json
import logging
import sys
import tempfile
from threading import Lock

import openai
from flask import Flask, request
from flask_socketio import SocketIO, emit, join_room, leave_room, \
    rooms

from jarvis.utils import chat_utils, whisper_utils, chatgpt_utils, faster_whisper_utils

# Set this variable to "threading", "eventlet" or "gevent" to test the
# different async modes, or leave it set to None for the application to choose
# the best option based on installed packages.
async_mode = None

app = Flask(__name__)
app.config['SECRET_KEY'] = 'secret!'
socketio = SocketIO(app, async_mode=async_mode)
thread = None
thread_lock = Lock()

openai.api_key = sys.argv[1]


@app.route('/')
def index():
    return "Welcome to Jarvis Server API !"


@socketio.event
def process_message(message):
    message = json.loads(message)
    logging.info("New PROCESS request from room " + message['uuid'])
    logging.info("Message : " + message['data'])

    if message['uuid'] not in rooms():
        logging.warning("Room not found, creating it")
        join_room(message['uuid'])

    # TODO: maybe implement grammar check and correction ?

    # intent_manager.recognise(message['data'], message['uuid'])
    if message['data'] != "":
        response = chatgpt_utils.chatgpt_recognise(message['data'], message['uuid'])
        text_response = chatgpt_utils.get_answer_from_response(response)
        # response = "Tokens are expensive ya know?"

        chat_utils.send_jarvis_message_to_room(text_response, message['uuid'])


@socketio.event
def join(message):
    message = json.loads(message)

    logging.info("New client joined room " + message['uuid'])
    join_room(message['uuid'])


@socketio.event
def leave(message):
    leave_room(message['uuid'])


@socketio.event
def connect():
    global thread
    emit('my_response', {'data': 'Connected', 'count': 0})


@socketio.event
def clear_chat(message):
    message = json.loads(message)

    emit('clear_chat', {}, to=message['uuid'])
    chatgpt_utils.clear_chat(message['uuid'])


# .WAV (i.e.) FILE REQUEST
@app.route("/get_text_from_audio", methods=['POST'])
def get_text_from_audio():
    logging.info("New STT request from " + request.remote_addr)

    audio_temp_file = tempfile.NamedTemporaryFile(prefix='jarvis-audio_', suffix='_client')
    audio_temp_file.write(request.data)

    # text = whisper_utils.whisper_cpp_stt(audio_temp_file.name)
    text = faster_whisper_utils.faster_whisper_stt(audio_temp_file.name)

    logging.info("STT result for " + request.remote_addr + " : " + text)

    return {"data": text}


"""
@app.route("/process_text", methods=['POST'])
def process_text():
    print("[" + request.remote_addr + "] - New TXT request")

    text = request.values['text']

    answer = intent_manager.recognise(text, request.headers.get('Client-Ip'), request.headers.get('Client-Port'))

    return {"transcription": text, "answer": answer}"""


def start_api():
    logging.info("Starting Jarvis Server API...")
    socketio.run(app, host='0.0.0.0', port=6000)