doit#

Opinionated wrapper around yt-dlp.

Why and what?#

This is not a music downloader for everyone. It works for specific use cases only (hence, opinionated). Let me explain.

I have a decently sized local library of music, and the library grows incrementally. I organize the music in 2 ways:

• author/album/songs
  • this means that the structure looks something like this:

    author1/
      album1/
        song1.1
        sont1.2
        song1.3
      album2/
        song2.1
        song2.2
        song2.3
      singles/
        song1
        song2
    

• year/month/songs (this is the important one)
  • I organize the music by the time when I heard the song for the first time, or just when I listened to the song a bunch.
  • so the structure looks something like this:

    2023/
      January/
        song1
        song2
        song3
      February/
        song1
        song2
        song3
    

This means that for each month, I have a folder and inside that folder a file named links in which I collect the music over a period of one full month, and I occasionally run doit on that file.

The links file looks something like this:

Author - Song https://example.com
Author - Song 2 https://example2.com
Author 2 - Song A https://hi.com
-- and so on

You might think that it's just too much work writing every author and title by hand, but that is required since there is no other way to guaranty the authors and titles are correct and/or properly formatted. Plus, since English input method is the only common input method on all my devices I listen to my music, all names/titles need to be transliterated.

If your workflow is similar, doit is the program for you. Enjoy :)

Note: you can now use generate-doit-links to speed up the process of generating links of new albums/playlists (for non-incremental builds of links).

Is it any good?#

Yes.

Options#

Available flags/options (with examples). All flags can be used with both single and double dash ( - and -- ).

index <number>             - Enables indexing and sets the starting index, or 
                             disables it if the index is negative. Default: 1
file <string>              - Set file to be used as input file. Default: "./links"
dir <string>               - Set/create directory to store downloads. Default: "./"
batch <number>             - Set number of parallel downloads. Default: 6

Example usage#

./doit -index 1 -file the_prodigy.txt -dir "The Prodigy/" -batch 15

or, if you put links for songs in file 'links' and those songs have their own authors in that file (format: author-song name https://example.com) you can also just us ./doit and it will parse the links file and download the songs in the current directory.

Dependencies#

Given the nature of the script (wrapper), doit depends on yt-dlp (which, I believe, depends on glibc, so as for now, doit won't work on muslc systems).

TODO#

• add options and levels of verbosity
  • better logging
• add tests for files
• add options to retry failed downloads
• figure out how to retain the metadata
  • take a look at something like %(title)s 1.%(ext)s', '-x', '--audio-format', 'mp3', '--ignore-config', '--hls-prefer-native', '--no-mtime', '--embed-thumbnail', '--add-metadata', '--verbose', '--no-post-overwrites', '<url>']
  • https://www.reddit.com/r/youtubedl/comments/138bhzp/ytdlp_mp3_tagging/

Licence#

This code is licensed under MIT licence (see LICENCE for details).