#How to use #Vosk -- the Offline Speech Recognition Library for Python
#https://www.youtube.com/watch?v=3Mga7_8bYpw

#how to load wav file
#https://towardsdatascience.com/transcribe-large-audio-files-offline-with-vosk-a77ee8f7aa28
import os
import sys
import json
#from vosk import Model, KaldiRecognizer
from vosk import Model, KaldiRecognizer, SetLogLevel
import pyaudio

import wave
import json

import ssl
ssl._create_default_https_context = ssl._create_unverified_context



#wav_file = 'test.wav'

wav_file = "00_00_00_Nameless_Neutral__Yes,[breath] I'm planning to get out of here, [breath]and if you want to join me, come down to the harbor. We'll meet at the ship.wav"



#wf = wave.open(wav_file, "rb")
wf = wave.open(wav_file, "rb")

theModelDir = (r"E:\Amo_Files\ZDIECIA\Reaction_FACES\ponyface\ai\_whl\vosk-model-en-us-0.22")


audioFrameHz = 48000
model = Model(theModelDir, model_name=None, lang=None)


#recognizer = KaldiRecognizer(model, audioFrameHz)

rec = KaldiRecognizer(model, wf.getframerate())


transcription = []



while True:
    data = wf.readframes(4000)
    if len(data) == 0:
        break
    if rec.AcceptWaveform(data):
        # Convert json output to dict
        result_dict = json.loads(rec.Result())
        # Extract text values and append them to transcription list
        transcription.append(result_dict.get("text", ""))



final_result = json.loads(rec.FinalResult())
transcription.append(final_result.get("text", ""))

# merge or join all list elements to one big string
transcription_text = ' '.join(transcription)
print(transcription_text)







x="""
while True:
    data = stream.read(4096)
    #if len(data) ==0:
        #break
    if recognizer.AcceptWaveform(data):
        text = recognizer.Result()
        print(text)
        print(text[14:-3])

"""