#!/usr/bin/env python3 import argparse import datetime import glob import json import music_tag import os from pathlib import Path import sys import youtube_dl def cleanup_metadata_files(music_directory): files = glob.glob(os.path.join(music_directory, '*.json')) for file in files: os.remove(file) def download_song(song_url, ytdl): """ Download a song using youtube url and song title """ return ytdl.extract_info(song_url, download=True) def format_youtube_date(date): default = "Unknown Year" try: fmt = "%Y%m%d" d = datetime.datetime.strptime(date, fmt) return d.year except Exception as ex: print(ex) return default def get_all_files(directory): things = glob.glob(os.path.join(directory, '*.mp3')) files = [] for thing in things: if os.path.isfile(thing): files.append(thing) return files def get_command_line_options(): parser = argparse.ArgumentParser(description="Download songs from YouTube Music") parser.add_argument("url", metavar="string", type=str, help="Playlist or Song URL to download") parser.add_argument("-c", "--cookies", metavar="string", type=str, help="Cookie file to use.") parser.add_argument("-o", "--output", metavar="string", type=str, help="Output directory to use") return parser.parse_args() # TODO: switch command line args to argparse def get_playlist_url(): return sys.argv[1] def get_video_urls_in_playlist(playlist_url, ytdl): videos = ytdl.extract_info(playlist_url, download=False) urls = [] for vid in videos['entries']: if 'webpage_url' in vid.keys() and vid['webpage_url'] is not None: urls.append(vid['webpage_url']) return urls def move_file(file, metadata, output_directory): # TODO: Pass a jamos specific metadata object/dict # Then we can do this validation all at once artist = 'unknownartist' album = 'unknownalbum' title = 'unknownsong' if ('artist' in metadata.keys()) and (metadata['artist'] is not None): if len(metadata['artist'].split(',')) > 1: # If there are multiple artists, pick the first one # NOTE: This may break if the artist has a comma in their name artist = metadata['artist'].split(',')[0].replace(' ', '_').lower() else: artist = metadata['artist'].replace(' ', '_').lower() if ('album' in metadata.keys()) and (metadata['album'] is not None): album = metadata['album'].replace(' ', '_').replace(',', '').replace("'", "").replace('"', "").lower() if ('title' in metadata.keys()) and (metadata['title'] is not None): title = metadata['title'].replace(' ', '_').replace('/', '').lower() final_directory = os.path.join( output_directory, artist, album) Path(final_directory).mkdir(parents=True, exist_ok=True) # TODO: Include album title in filename # TODO: Research converting to mp3 instead of just naming it such. # TODO: Research better file formats over mp3? os.rename( file, os.path.join(final_directory, artist + '_' + title + '.mp3')) def write_metadata_to_song_file(file, metadata): f = music_tag.load_file(file) artist = 'unknownartist' album = 'unknownalbum' title = 'unknownsong' year = 9999 if ('artist' in metadata.keys()) and (metadata['artist'] is not None): if len(metadata['artist'].split(',')) > 1: # If there are multiple artists, pick the first one # NOTE: This may break if the artist has a comma in their name artist = metadata['artist'].split(',')[0].replace(' ', '_').lower() else: artist = metadata['artist'].replace(' ', '_').lower() artist = artist.replace('&', 'and') if ('album' in metadata.keys()) and (metadata['album'] is not None): album = metadata['album'].replace(' ', '_').lower() if ('title' in metadata.keys()) and (metadata['title'] is not None): title = metadata['title'].replace(' ', '_').lower() if ('release_date' in metadata.keys()): year = format_youtube_date(metadata['release_date']) f['name'] = title f['artist'] = artist f['album'] = album f['year'] = year f.save() def create_downloader(music_directory, cookies): audio_options = { 'format': 'mp3/bestaudio/best', 'cookiefile': cookies, 'outtmpl': music_directory + '%(title)s.%(ext)s', 'postprocessors': [ { 'key': 'FFmpegExtractAudio', 'preferredcodec': 'mp3', 'preferredquality': '192', }, {'key': 'FFmpegMetadata'}, ], 'writeinfojson': True } return youtube_dl.YoutubeDL(audio_options) if __name__ == "__main__": args = get_command_line_options() # Get the playlist url from the command line playlist_url = args.url music_directory = args.output or "~/Music" cookies = args.cookies or "~/cookies.txt" ytdl = create_downloader(music_directory, cookies) # TODO: Save urls to file so we can start in the # middle of the playlist if needed urls = get_video_urls_in_playlist(playlist_url, ytdl) for url in urls: try: download_song(url, ytdl) except Exception as ex: # TODO: Handle this better print(ex) files = get_all_files(music_directory) for f in files: json_data = None with open(f.replace('.mp3', '.info.json')) as json_file: json_data = json.load(json_file) write_metadata_to_song_file(f, json_data) move_file(f, json_data, music_directory) cleanup_metadata_files(music_directory)