ytcs

ytcs.sh

This script is a simple tool to fetch YouTube subscriptions via RSS feed and watch them using mpv and youtube-dl, thus enabling using your login cookies from Firefox, and allowing the viewing of videos.

The script presents the subscriptions in a rofi or fzf menu, making it easy to navigate and select the videos you want to watch.

Updated to include https://github.com/Brainicism/bgutil-ytdlp-pot-provider

Prerequisites

The script requires the following programs to be installed:

youtube-dl or yt-dlp
mpv
rofi or fzf
wget

Installation

This tool can be installed by cloning the repository to your local machine and then running the script.

Usage

Playing a single video

It is possible to simply use this as a way to pipe a YouTube url to mpv. It will also include that in your “seen” videos. Simply

ytcs.sh [url]

where the url is either the full url to the video, or the shortened https://youtu.be/ format.

Browsing Subscriptions.

There are several commands available with this script when used to browse videos:

ytcs.sh [--loud] [--help] [--import FILENAME] [--refresh] [--subscription|--grouped|--time]

You may use any combination of these:

–loud : This command provides more verbose output for debugging and notify-send. Should be FIRST.
–refresh : Refresh your subscriptions.
–help or -h : Display help context (this, basically).
–import or -i : Import subscriptions from a csv file.

Choose ONE of the three of these:

–grouped or -g : Display the videos grouped by subscriptions.
–time –chronological -t or -c : Arrange and display the videos in chronological order.
–subscription or -s : Select and view videos by browsing subscriptions.

For example, you can refresh your subscriptions using the –refresh command like this:

./ytcs.sh --refresh

Ideally, you’d do that with a cron job; it does take a moment, and caches the results for quick user input.

You can use the –import command to import your subscriptions from a csv file like this:

./ytcs.sh --import /path/to/csv

See the notes below.

You can also browse your videos by group, by time or by subscription using the –grouped , –time , and –subscription commands respectively. In grouped and time based view, it will compare to a watched file, and update it if you’ve watched and videos.

Keep in mind that since it’s powered by rofi or fzf, you can search among the titles in any of these views.

./ytcs.sh --grouped 

./ytcs.sh --time 

./ytcs.sh --subscription 

Configuration

Rename the file ytcs.env.example to ytcs.env and edit it. The example file has the defaults:

export ROFI_THEME="arthur"
export MAX_CHANNEL_AGE=182
export MAX_GROUPED_VIDS=10
#export watchtop=4
export LOUD=0
export YTDLP_COOKIES="firefox"
export MARK_AGE="TRUE"
export GEOMETRY1="1366x768+50%+50%"
export GEOMETRY2="1366x768"

export YTPOT_BASEURL="youtubepot-bgutilhttp:base_url=http://127.0.0.1:8080" is used ONLY if you have the BGUtil POTS provider on a non-default host or port as described in its documentation.

The value watchtop is used for process control; while parsing data, ytcs will attempt to use a number of subshells equal to your processor cores unless watchtop is set to a positive value to limit the number of cores.

I personally use the “arthur_modified” rofi theme included here which has some small tweaks.

YTDLP_COOKIES is used to control which browser yt-dlp gets cookies from.

MARK_AGE determines whether the visual marking of video age is displayed, change to anything but “TRUE” to turn off.

GEOMETRY1 and GEOMETRY2 are for mpv; they correspond to --geometry=${GEOMETRY1} --autofit=${GEOMETRY2}.

Notes

Nearly all channel data and views are cached to optimize speed when you’re actually watching videos. So you will not see “watched” changes until it rebuilds that cached view after you quit the program. Likewise, when you change a variable such as MARK_AGE, you must either use the refresh command, or delete time_data.txt and grouped_data.txt in the cache directory in order to see the change.
When the tool plays a video, it attempts to automatically use xclip and/or copyq to put the video URL in the clipboard, should you want to share it, etc.
Ensure that the CSV file used during the import operation is formatted properly, each line consisting of channel id, url, channel name and without a comma at the end. A sample is enclosed. This is the same export format that FreeTube, for example, uses for CSV export. You can also manually find channelID using a tool like the one at https://www.streamweasels.com/%20tools/youtube-channel-id-and-%20user-id-convertor/
This script is intended for personal use and not to be utilized for streaming without permission from the respective YouTube creators. Ensure you comply with the terms and conditions from YouTube.

TODO

TODO - Demo TODO - I mean, it’s spaghetti code. It works, it is nowhere near optimized, and is probably got issues in some way
TODO - sort grouped by channel age (this is REALLY kicking my ass.) TODO - filter EXCLUDE terms (new feature)

This site is open source. Improve this page.