Add Telegram bot endpoint

Dockerfile

@@ -1,6 +1,8 @@
 FROM python:3.7
 COPY . /app
 WORKDIR /app
+RUN apt-get update
+RUN apt-get install -y ffmpeg
 RUN wget https://github.com/robinhad/voice-recognition-ua/releases/download/v0.3/uk.tflite
 RUN wget https://github.com/mozilla/DeepSpeech/releases/download/v0.9.1/deepspeech-0.9.1-models.tflite
 RUN pip install -r requirements.txt

main.py

@@ -1,7 +1,29 @@
 from flask import Flask, render_template, request
 from io import BytesIO
 from client import client
+import telebot
+import logging
+import os
+import warnings
+from client import client
+from io import BytesIO
+import pydub
+
+warnings.simplefilter('ignore')
+TOKEN = os.environ['TOKEN']
+
+if not TOKEN:
+    print('You must set the TOKEN environment variable')
+    exit(1)
+
+START_MSG = '''Вітання!
+Цей бот створений для тестування перекладу українських аудіозаписів в текст.
+Група для обговорення: https://t.me/speech_recognition_uk'''
 
+FIRST_STEP = '''Використовувати бота просто: надішліть аудіоповідомлення і чекайте відповіді'''
+
+
+bot = telebot.TeleBot(TOKEN, parse_mode=None)
 app = Flask(__name__,)
 app.config['MAX_CONTENT_LENGTH'] = 120 * 1024
 
@@ -22,5 +44,55 @@ def recognize():
     return result
 
 
+@app.route('/bot/' + TOKEN, methods=['POST'])
+def getMessage():
+    bot.process_new_updates(
+        [telebot.types.Update.de_json(request.stream.read().decode("utf-8"))])
+    return "!", 200
+
+
+@app.route("/bot")
+def webhook():
+    bot.remove_webhook()
+    bot.set_webhook(
+        url='https://voice-recognition-ua.herokuapp.com/bot/' + TOKEN)
+    return "!", 200
+
+
+@bot.message_handler(commands=['start', 'help'])
+def send_welcome(message):
+    bot.reply_to(message, START_MSG)
+    bot.reply_to(message, FIRST_STEP)
+
+
+@bot.message_handler(content_types=['voice'])
+def process_voice_message(message):
+    # download the recording
+    file_info = bot.get_file(message.voice.file_id)
+    downloaded_file = bot.download_file(file_info.file_path)
+    # create in-memory representation of files
+    source_audio = BytesIO()
+    source_audio.write(downloaded_file)
+    source_audio.seek(0)
+    output_audio = BytesIO()
+    ogg_file = pydub.AudioSegment.from_ogg(
+        source_audio)
+
+    # convert ogg to wav
+    ogg_file.set_frame_rate(16000).set_channels(
+        1).export(output_audio, "wav", codec="pcm_s16le")
+    output_audio.seek(0)
+
+    # do the recognition
+    # get the recognized text
+    text = client(source_audio)
+    # no results
+    if not text:
+        bot.reply_to(message, 'Я не зміг розпізнати 😢')
+    else:
+        # send the recognized text
+        bot.reply_to(message, text)
+
+
 if __name__ == '__main__':
     app.run(host='0.0.0.0')

requirements.txt

@@ -1,4 +1,6 @@
 Flask==1.1.2
 deepspeech-tflite==0.9.1
 numpy==1.17.0
-uwsgi==2.0.19.1
+uwsgi==2.0.19.1
+pytelegrambotapi==3.7.6
+pydub==0.24.1