Script que obtiene las veces que Juan Villalvazo dice la frase <vida de programador>

contar-frase.py

import re
import youtube_dl
import audioread
from google.cloud import storage
from google.cloud import speech
from google.cloud.speech import enums
from google.cloud.speech import types

BUCKET_NAME = "audio-devlife"
BLOB_PATH = "audio.flac"
LANG_CODE = 'es-MX'
VIDEO_LIST = ["https://youtu.be/QVMDXf_mfH0"]
AUDIO_RATE = 16000
AUDIO_CHANNELS = 2

ORIGINAL_PHRASE = r'(?P<phrase>(vi|bi){1}da de programador)'
POSSIBLE_PHRASE = r'(?P<phrase>(vi|bi){1}.{0,3}(de)? programador)'
DEV_PHRASE = r'programador'


def download_audio():
    ydl_opts = {
        "format":
        "bestaudio",
        "outtmpl":
        "audio.%(ext)s",
        "postprocessors": [{
            "key": "FFmpegExtractAudio",
            "preferredcodec": "flac",
            "preferredquality": "192",
        }],
    }
    with youtube_dl.YoutubeDL(ydl_opts) as ydl:
        ydl.download(VIDEO_LIST)
        save_in_bucket()


def save_in_bucket():
    storage_client = storage.Client()
    bucket = storage_client.get_bucket(BUCKET_NAME)
    blob = bucket.blob(BLOB_PATH)
    with open(BLOB_PATH, "rb") as my_file:
        blob.upload_from_file(my_file)
        get_audio_info()


def get_audio_info():
    with audioread.audio_open(BLOB_PATH) as f:
        AUDIO_CHANNELS = int(f.channels)
        AUDIO_RATE = int(f.samplerate)
        get_text()


def get_text():
    gcs_uri = "gs://{}/{}".format(BUCKET_NAME, BLOB_PATH)
    client = speech.SpeechClient()

    audio = types.RecognitionAudio(uri=gcs_uri)
    config = types.RecognitionConfig(
        encoding=enums.RecognitionConfig.AudioEncoding.FLAC,
        language_code=LANG_CODE,
        audio_channel_count=AUDIO_CHANNELS,
        enable_separate_recognition_per_channel=True)

    operation = client.long_running_recognize(config, audio)
    response = operation.result()
    contador_phx = 0
    contador_php = 0
    contador_dev = 0

    for result in response.results:
        for alternative in result.alternatives:
            print('=' * 20)
            text_block = alternative.transcript
            print(text_block)
            phx = re.findall(ORIGINAL_PHRASE, text_block,
                             re.MULTILINE & re.IGNORECASE)
            php = re.findall(POSSIBLE_PHRASE, text_block,
                             re.MULTILINE & re.IGNORECASE)
            devp = re.findall(DEV_PHRASE, text_block,
                              re.MULTILINE & re.IGNORECASE)

            if phx:
                contador_phx += len(phx)

            if php:
                contador_php += len(php)

            if devp:
                contador_dev += len(devp)

    print("\nFRASE ORIGINAL: {}\tFRASE POSIBLE: {}\t PALABRA PROGRAMADOR:{}".
          format(contador_phx, contador_php, contador_dev))


if __name__ == "__main__":
    download_audio()