microsoft_speech_api_stt.py

import requests
import http.client
import urllib.parse
import uuid
import json
import io
import os
import sys
import time
from datetime import timedelta
import argparse


#We need to get our API credentials in the code for authentication that we have stored as Environment Variables locally
Ocp_Apim_Subscription_Key = os.environ.get("KEY_SPEECH")


#Following line is used to save all the console output into a text file
sys.stdout = open('speech_api_test_text_stt_output.txt', 'w')

start_time = time.monotonic()


def input_file(speech_file_path):
    global content
    if os.path.isfile(speech_file_path):
      with io.open(speech_file_path, 'rb') as audio_file:
        content = audio_file.read()
    else:
      print("File doesn't exist in the directory!")


def speech_stt():
    params = ""
    headers = {
        # NOTE: Replace the "Ocp-Apim-Subscription-Key" value with a valid subscription key.
        'Ocp-Apim-Subscription-Key': Ocp_Apim_Subscription_Key,
    }

    AccessTokenHost = "api.cognitive.microsoft.com"
    path = "/sts/v1.0/issueToken"

    #Connect to server to get the Access Token
    print("Connect to server to get the Access Token")
    conn = http.client.HTTPSConnection(AccessTokenHost)
    conn.request("POST", path, params, headers)
    response = conn.getresponse()
    print(response.status, response.reason)

    data = response.read()
    conn.close()

    accesstoken = data.decode("UTF-8")

    endpoint = 'https://speech.platform.bing.com/recognize'
    request_id = uuid.uuid4()
    # Params from Microsoft Example
    params = {'scenarios': 'ulm',
              'appid': 'D4D52672-91D7-4C74-8AD8-42B1D98141A5',
              'locale': 'en-US',
              'version': '3.0',
              'format': 'json',
              'instanceid': '565D69FF-E928-4B7E-87DA-9A750B96D9E3',
              'requestid': uuid.uuid4(),
              'device.os': 'windows'}
    content_type = "audio/wav; codec = ""audio/pcm""; samplerate = 16000"

    headers = {'Authorization': 'Bearer ' + accesstoken,
                       'Content-Type': content_type}
    response = requests.post(endpoint,
                         params = params,
                         data = content,
                         headers = headers)
    val = json.loads(response.text)
    print('Text: ', val["results"][0]["name"])
    print('Confidence: ', val["results"][0]["confidence"])


if __name__ == '__main__':
    parser = argparse.ArgumentParser(
        description = __doc__,
        formatter_class = argparse.RawDescriptionHelpFormatter)
    parser.add_argument(
        'speech_file_path',
        help = 'The complete file path of the speech file you want to convert from speech to text.')
    args = parser.parse_args()

    input_file(args.speech_file_path)
    speech_stt()


end_time = time.monotonic()
print("Execution_Time:", timedelta(seconds = end_time - start_time))
print('\n')