Extract links to music websites from Mastodon's outbox.json and download them with youtube-dl in parallel with covers and metadata

extract_music_urls.py

#! /usr/bin/python
"""Extract links to music websites from Mastodon's outbox.json
outbox.json contains all of your toots
"""
import json
from bs4 import BeautifulSoup as Soup


def extract_music_urls():
    with open("outbox.json") as f:
        j = json.loads(f.read())

    prefixes = ("https://youtu.be", "https://youtube.com",
                "https://www.youtube.com", "https://soundcloud.com", 
                "https://m.soundcloud.com", "https://vimeo.com")
    links = []
    for m in j["orderedItems"]:
        if "content" in m["object"]:
            html = Soup(m["object"]["content"], 'html.parser')
            hrefs = [a['href'] for a in html.find_all('a') if a['href'].startswith(prefixes)]
            if hrefs:
                links.append(hrefs[0])
    return sorted(links)



if __name__ == "__main__":
    urls = extract_music_urls()
    print("\n".join(urls))

To download audio tracks from all of the extracted music URLs, run:
./extract_music_urls.py | parallel 'LC_ALL=en_US.UTF-8 youtube-dl -f bestaudio --extract-audio --audio-format mp3 --audio-quality 0 --embed-thumbnail --add-metadata {}'

if you want to download the files without transcoding to mp3 then:
./extract_music_urls.py | parallel 'LC_ALL=en_US.UTF-8 youtube-dl -f bestaudio --extract-audio --embed-thumbnail --add-metadata {}'

to get rid of some special chars from filenames you can use iconv:
find . -type f -exec bash -c 'mv "$1" "${1%/*}/$(iconv -f UTF8 -t ASCII//TRANSLIT <<< ${1##*/})"' -- {} \;