speech-scoring/speech_tools.py

import os
import math
import string
import threading
import multiprocessing
import pandas as pd
import numpy as np
import pyaudio
from pysndfile import sndio as snd
# from matplotlib import pyplot as plt
from speech_spectrum import plot_stft, generate_spec_frec

SAMPLE_RATE = 22050
N_CHANNELS = 2

def step_count(n_records,batch_size):
    return int(math.ceil(n_records*1.0/batch_size))

def file_player():
    p_oup = pyaudio.PyAudio()
    def play_file(audiopath,plot=False):
        print('playing',audiopath)
        samples, samplerate, form = snd.read(audiopath)
        stream = p_oup.open(
            format=pyaudio.paFloat32,
            channels=2,
            rate=samplerate,
            output=True)
        one_channel = np.asarray([samples, samples]).T.reshape(-1)
        audio_data = one_channel.astype(np.float32).tobytes()
        stream.write(audio_data)
        stream.close()
        if plot:
            plot_stft(samples, SAMPLE_RATE)
    def close_player():
        p_oup.terminate()
    return play_file,close_player

def record_spectrogram(n_sec, plot=False, playback=False):
    # show_record_prompt()
    N_SEC = n_sec
    CHUNKSIZE = int(SAMPLE_RATE * N_SEC / N_CHANNELS)  # fixed chunk size
    input('Press [Enter] to start recording sample... ')
    p_inp = pyaudio.PyAudio()
    stream = p_inp.open(
        format=pyaudio.paFloat32,
        channels=N_CHANNELS,
        rate=SAMPLE_RATE,
        input=True,
        frames_per_buffer=CHUNKSIZE)
    data = stream.read(CHUNKSIZE)
    numpydata = np.frombuffer(data, dtype=np.float32)
    multi_channel = np.abs(np.reshape(numpydata, (-1, 2))).mean(axis=1)
    one_channel = np.asarray([multi_channel, -1 * multi_channel]).T.reshape(-1)
    mean_channel_data = one_channel.tobytes()
    stream.stop_stream()
    stream.close()
    p_inp.terminate()
    if plot:
        plot_stft(one_channel, SAMPLE_RATE)
    if playback:
        p_oup = pyaudio.PyAudio()
        stream = p_oup.open(
            format=pyaudio.paFloat32,
            channels=2,
            rate=SAMPLE_RATE,
            output=True)
        stream.write(mean_channel_data)
        stream.close()
        p_oup.terminate()
    ims, _ = generate_spec_frec(one_channel, SAMPLE_RATE)
    return ims


def _apply_df(args):
    df, func, num, kwargs = args
    return num, df.apply(func, **kwargs)

def apply_by_multiprocessing(df,func,**kwargs):
  cores = multiprocessing.cpu_count()
  workers=kwargs.pop('workers') if 'workers' in kwargs else cores
  pool = multiprocessing.Pool(processes=workers)
  result = pool.map(_apply_df, [(d, func, i, kwargs) for i,d in enumerate(np.array_split(df, workers))])
  pool.close()
  result=sorted(result,key=lambda x:x[0])
  return pd.concat([i[1] for i in result])

def square(x):
    return x**x

# if __name__ == '__main__':
#     df = pd.DataFrame({'a':range(10), 'b':range(10)})
#     apply_by_multiprocessing(df, square, axis=1, workers=4)

def hms_string(sec_elapsed):
    h = int(sec_elapsed / (60 * 60))
    m = int((sec_elapsed % (60 * 60)) / 60)
    s = sec_elapsed % 60.
    return "{}:{:>02}:{:>05.2f}".format(h, m, s)

def rm_rf(d):
    for path in (os.path.join(d,f) for f in os.listdir(d)):
        if os.path.isdir(path):
            rm_rf(path)
        else:
            os.unlink(path)
    os.rmdir(d)

def create_dir(direc):
    if not os.path.exists(direc):
        os.makedirs(direc)
    else:
        rm_rf(direc)
        create_dir(direc)


def format_filename(s):
    """
    Take a string and return a valid filename constructed from the string.
    Uses a whitelist approach: any characters not present in valid_chars are
    removed. Also spaces are replaced with underscores.

    Note: this method may produce invalid filenames such as ``, `.` or `..`
    When I use this method I prepend a date string like '2009_01_15_19_46_32_'
    and append a file extension like '.txt', so I avoid the potential of using
    an invalid filename.
    """
    valid_chars = "-_.() %s%s" % (string.ascii_letters, string.digits)
    filename = ''.join(c for c in s if c in valid_chars)
    filename = filename.replace(' ','_') # I don't like spaces in filenames.
    return filename

#################### Now make the data generator threadsafe ####################

class threadsafe_iter:
    """Takes an iterator/generator and makes it thread-safe by
    serializing call to the `next` method of given iterator/generator.
    """
    def __init__(self, it):
        self.it = it
        self.lock = threading.Lock()

    def __iter__(self):
        return self

    def __next__(self): # Py3
        with self.lock:
            return next(self.it)

    def next(self):     # Py2
        with self.lock:
            return self.it.next()


def threadsafe_generator(f):
    """A decorator that takes a generator function and makes it thread-safe.
    """
    def g(*a, **kw):
        return threadsafe_iter(f(*a, **kw))
    return g