YouTube-Scissors

A simple npm library that allows you to divide a YouTube video into multiple separate videos base on a video's time stamps. Built on top of FFmpeg and JavaScript.

If you are looking for a CLI version of this library look here.

💡 Features

• Can generate multiple videos or extract a single video, based on a YouTube video's time stamps (time stamps from a comment, video description, or chapters)

  • Important: For this library to work, you must either have the YouTube video already downloaded or have a Buffer of the YouTube video

  • Note: For comments, follow this tutorial to figure out how to get a YouTube comment's URL.

• Will automatically download ffmpeg for your current operating system

  • Note: Cannot automatically download ffmpeg for MacOS. You have to download and add it yourself. FFmpeg Downloads

• Can generate time stamps from a YouTube video's chapters, comment, or description.

• You can use this library on top of any YouTube download library / API.

• 100% Open Source (MIT license)

🚀 Install

npm install yt-scissors

Example & Usage

const { getTimeStampList, cutVideo } = require('yt-scissors');
const fs = require('fs');

async function main() {

    // == Different Videos == 
    // Chapters: "https://www.youtube.com/watch?v=iPtPo8Sa3NE"
    // Comment: "https://www.youtube.com/watch?v=89UEYbkHKvg&lc=UgzcUK560Nm4FAxF8-d4AaABAg"
    // Description: "https://www.youtube.com/watch?v=GdzrrWA8e7A"

    let list = await getTimeStampList({ 
        url: "https://www.youtube.com/watch?v=iPtPo8Sa3NE", 
        type: "chapters" });
    
    //Will output a array of time stamps and video titles
    console.log(list);

    //Will generate a video from the 5th video in the array
    let chapter_videos = await cutVideo({
        video: Buffer.from(fs.readFileSync("./Death Grips - Exmilitary [Full Mixtape].mp4")),
        // ffmpegPath: "", (Give it a path to your ffmpeg executable)
        DisableDownloadLogs: false,
        chapters: [list[4]],
        ffmpegOptions: {
            crf: "3"
        },
    });

    fs.writeFileSync("./test.mp4", chapter_videos[0].videoData);
}

main();

📖 API Documentation

getTimeStampList(...)

Description:

Important: Generated time stamps from the description and comments works about 85% of the time. Make sure the video's time stamps are spaced out and have nothing that would make it hard to find them. There is also a bug with any video that is +10 hours long, so video length should be below 10 hours.

• Picks where to get video time stamps and generate array of time stamps from that.

• Can generate time stamps from a video's chapters, comment, or description.

• Returns an array of start and end time for each chapter video.

// getTimeStampList(...) all default values
getTimeStampList({
    url: String,
    type: "chapters" | "comment" | "description" // default is "chapters"  
})

Required	Name	Data Type	Description
Yes	url	String	URL of YouTube video
Yes	type	"chapters" or "comment" or "description"	The data you want to parse to get video time stamps. (default is "chapters")

Returns { Promise<Array<ListVideo_Object>> }

• Returns an array of start and end time for each chapter video.

• Note: Will return a empty array if time stamps couldn't be generated

ListVideo_Object Example:

"ListVideo_Object": {
    title: "{String}", // Title of the chapter
    start_time: "{String or Number}", // Start time stamps of the chapter
    end_time: "{String or Number}" // End time stamps of the chapter
}

cutVideo(...)

Description:

Important: Cannot automatically download ffmpeg for MacOS. You have to download and add it yourself. FFmpeg Downloads

• Using FFmpeg, trims videos into different chapters and encodes theme base on the time stamps given.

• Can automatically download ffmpeg for current operating system, or you can manually install ffmpeg, and give the path to it.

• Returns an array of videos with title and a buffer of the trim down video

// cutVideo(...) all default values
cutVideo ({
    video,
    ffmpegPath = undefined,
    chapters,
    DisableDownloadLogs = false,
    ffmpegOptions = {
        crf: "25",
        preset: "ultrafast",
        ffmpegCmds: undefined,
        ffmpegHide: false
    } })

Required	Name	Data Type	Description
Yes	video	String or Buffer	Video path as a string or a buffer of the video
No	ffmpegPath	String	Path to ffmpeg executable. If none is given then will automatically download ffmpeg. FFmpeg will be downloaded if variable is set to undefined, null, or you don't pass in a string path. (default is undefined)
Yes	chapters	Array<ListVideo_Object>	List of chapters you want to extract from original video get this from getVideoList(...) function
No	DisableDownloadLogs	Boolean	True = disable download logs, false = show download logs. (default is false)
No	ffmpegOptions	Object	FFmpeg commands and options.
No	ffmpegOptions.crf	String	Quality of the video. Lower numbers the better looking the video. (default is 25)
No	ffmpegOptions.preset	String	Speed of encoding video. (default is ultrafast)
No	ffmpegOptions.ffmpegCmds	Array	Add any other ffmpeg commands as a array. Make sure they are String values.
No	ffmpegOptions.ffmpegHide	Boolean	Hide ffmpeg process from being shown in the terminal. (default is false)

Returns { Promise<Array<SaveVideos_Object>> }

• Returns an array object of videos. Videos are store as buffers.

SaveVideos_Object Example:

"SaveVideos_Object": {
    title: "{String}", // Title of the chapter
    videoData: "{Buffer}" // A buffer of the chapter video
}

🗿 Helpful Infomation & Example Code

Example Code

// ( Example 1 ) Using the ffmpegCmds in cutVideo(...) 
await cutVideo({
    video: "Some Buffer of YouTube Video",
    DisableDownloadLogs: false,
    chapters: [list[4]],
    ffmpegOptions: {
        crf: "3",
        //Note: What "-hide_banner -loglevel debug" commands will look like in the array
        ffmpegCmds: ["-hide_banner", "-loglevel", "debug"]
    },
});

Helpful Infomation

• How to find a YouTube comment URL from a video

License

MIT