y/readme.md

52 lines
1.8 KiB
Markdown
Raw Normal View History

2023-06-23 20:07:30 +02:00
# y
> algorithmically-enhanced [youtube-dl](https://yt-dl.org) wrapper for audio-downloads
## install
1. look at [`install.sh`](./install.sh) file for instructions
1. adjust `PREFIX` at wish
1. make sure the final directory is on `$PATH`
## use
`y https://orax.bandcamp.com/track/ectoplasmic`
(adjust [config.sh](./config.sh) file to set your prefered download paths)
## requirments
- [`youtube_dl`](https://github.com/ytdl-org/youtube-dl) the main thing
- [`ffmpeg`](https://ffmpeg.org/download.html) to convert files of diffrent formats
- [`sd`](https://github.com/chmln/sd) a better `sed`
- [`lynx`](https://github.com/lynx/lynx) to easily scrap web-pages
## site-specific improvements
### bandcamp
1. prefers mp3-320 (checks every track)
2. downloads entire discography if given root-domain or `/music` url
### twitter
file-name is now twitter-username and tweet-id
(instead of full text of a tweet, which can result in strings over 255 characters long, aborting the entire process)
### reddit
file-name is now subreddit, post-id, and url-slug
(instead of full text of a title, which can result in strings over 255 characters long, aborting the entire process)
### youtube
allows to set several predetermined filesystem-paths to keep file-categories separately:
- `y m <youtube_URL>` music (automatically chosen for `music.youtube.com`)
- `y a <youtube_URL>` audiobooks
### soundcloud
1. occasional wav-files are converted to mp3 320kb/s
1. artist is set based on url (instead of having none by default)
### instagram
- file-name is now username, video id, and nubmer of the video if there are multiple
(instead of "Video by username", which refuses to download multiple videos from different id's: "Video by username.mp4 has already been downloaded")
- TODO: explose playlist's `%(uploader_id)s`, `nr_current`, `nr_total` to the naming api