Python script to slice large mp3 file by chunks

mp3_slicer.py

from os import path
from pydub import AudioSegment
import sys
# pydub use ffmpeg to file manupulation 
# don't forget to install ffmpeg
sys.path.append('/usr/local/bin/ffmpeg')

def load_songs(fn):
  meta = {}
  songs = []

  file = open_file(fn)
  for line in file:
    if line[0].isdigit():
      song = parse_song(line)
      song.update(meta)
      song['track'] = str(len(songs) + 1)

      if len(songs) > 0:
        songs[-1]['time_end'] = song['time_start']

      songs.append(song)
    else:
      if line.strip():
        key, value = parse_meta(line.strip())
        if value:
          meta[key.lower()] = value

  return songs

def open_file(fn):
  return open(fn)

# returns dict song
def parse_song(line):
  time, title = [string.strip() for string in line.split('-', 1)]
  return { 'time_start': time, 'title': title }

# returns meta key, value
# Example: 'artist', 'Aerosmith'
def parse_meta(line):
  return line.split(':', 1)

def time_to_ms(time):
  if isinstance(time, int):
    return time

  coefficients = [3600000, 60000, 1000]
  time_array = time.split(':')

  if len(time_array) < 3:
    time_array = ['00'] + time_array
  sum = 0
  for idx, element in enumerate(time_array):
    sum += int(element) * coefficients[idx]

  return sum

def extract_song(source, song_data, output_path):
  time_start = song_data['time_start']
  time_end = song_data.get('time_end') or len(source)

  tags = extract_tags(song_data)

  print("extracting %s \n" % song_data['title'])
  song = source[time_to_ms(time_start):time_to_ms(time_end)]
  file_name = compose_filename(song_data)

  song.export(path.join(output_path, file_name), format="mp3", tags=tags)

def extract_tags(song_data):
  tags = ['title', 'album', 'artist', 'year', 'track']

  return {tag: song_data[tag] for tag in tags if tag in song_data}

def compose_filename(song_data):
  file_name = song_data['track'].zfill(2) + " - " + song_data['title'] + ".mp3"
  return file_name.replace("/", '|')

fn = path.normpath('path_to_large_mp3)
output_dir = path.normpath('output_path')
                   
##
# Schema looks like
#
# Artist: Panic! at the Disco
# Album: Pray for the Wicked Tour 2019 - Live at O2 Arena, London 2019
# Year: 2019
# 00:15 - (Fuck A) Silver Lining
# 03:03 - Don't Threaten Me With a Good Time
# 06:34 - Ready to Go (Get Me Out of My Mind)
# ...
##
schema = 'schema_path'

print("reading file \n")
            
load_songs_schema = load_songs(schema)
record = AudioSegment.from_mp3(fn)

for song_data in load_songs_schema:
  extract_song(record, song_data, output_dir)