A pydub based Python script used to merged fractions of audio segments.

presets_variations_merger.py

import argparse
import os
import mimetypes

# You need to:
# python -m pip install pydub tqdm
# And to install ffmpeg
# https://github.com/jiaaro/pydub
# https://github.com/jiaaro/pydub/blob/master/API.markdown
from pydub import AudioSegment
from tqdm import tqdm

PYDUB_DEBUG = False
if PYDUB_DEBUG:
    import logging
    l = logging.getLogger("pydub.converter")
    l.setLevel(logging.DEBUG)
    l.addHandler(logging.StreamHandler())


def ensure_path(path: str) -> None:
    """Ensure that the specified path exist."""
    if not os.path.isdir(path):
        os.makedirs(path)


def safe_mime_guess(path: str) -> list:
    """Always return a iterable when trying to guess the MIME type."""
    guess = mimetypes.guess_type(path)[0]
    return guess if guess is not None else []


if __name__ == '__main__':
    # TODO:
    #  Add argparse to select input directories or files
    parser = argparse.ArgumentParser(
        prog='Presets variations merger',
        description='This script cut fragments of specified files and merge them in a unique file',
        epilog='')
    parser.add_argument(
        'input', help='A directory of audio files or a list of audio files to merge together', nargs='+')
    parser.add_argument(
        '-o', '--output', help='The output file', required=True, type=str)
    parser.add_argument(
        '-f', '--format', help='The format of the output file (by default it is a mp3 file)', default='mp3', )

    args = parser.parse_args()

    # Create outputs folder tree if it does not exist
    output_file = os.path.abspath(args.output)
    ensure_path(os.path.dirname(output_file))

    # Create the list of input files
    input_files = args.input

    if len(input_files) == 1:
        path = os.path.abspath(input_files[0])
        input_files = tuple(os.path.join(path, file) for file in os.listdir(
            path) if 'audio' in safe_mime_guess(file))

    else:
        old_files = input_files
        input_files = tuple(
            file for file in input_files if 'audio' in safe_mime_guess(file))

        if len(old_files) > len(input_files):
            print('One or more files have been dropped since their MIME type is not "audio/[...]"')

    if len(input_files) < 2:
        print(f'Not enough valid input files have been provided:\n{input_files[0]}')
    # Import audio files
    audio_segments = {file: AudioSegment.from_file(
        file, mimetypes.guess_type(file)[1]) for file in input_files}

    file_format = args.format

    # It is not assumed that each steps is the same
    # size thus we will get a fraction of each steps
    number_of_steps = len(input_files)
    merged_audio = AudioSegment.empty()

    index = 0
    print('Concatenate audio segments')
    for filename, segment in tqdm(audio_segments.items()):
        subsegment_length = len(segment)//number_of_steps
        indexes = (subsegment_length * index, subsegment_length * (index + 1))
        index += 1

        if index == number_of_steps:
            indexes = (indexes[0], -1)

        merged_audio += segment[indexes[0]:indexes[1]]

    print('Save result to file')
    with open(output_file, 'wb') as file:
        merged_audio.export(file, format=file_format)