skiptirengu / media-split

Split media files

Media-split

Split audio (and video) files

Install

Make sure you have ffmpeg installed and added to your path then run:

npm install [-g] media-split

If you don't have ffmpeg installed, you can install this or this package which comes with a bundled ffmpeg, and media-split will automatically detect and use it.

npm install [-g] ffmpeg-static

or

npm install [-g] @ffmpeg-installer/ffmpeg

CLI

media-split comes with a built in CLI tool. Type media-split -h to list all options and see an usage example. For input you can use either a YouTube link or a local file.

The template file should be in the following format. For more info, checkout ffmpeg's duration syntax page.

[([hh:]mm:ss[.ms...]) [ - ([hh:]mm:ss[.ms...])]] My Music Name

A template file usually looks like this:

[00:00] eli filosov [ p h i l o ] - oneofone_rwrk
[01:30] Swishers [SwuM x bsd.u]
[03:28] sweetbn _ i held you so close i forgot the world
[05:52 - 07:49] emune - Gretchen
[07:52 - 09:50] jhfly - sheets
[10:00] arbour - elusive
[11:30] tomppabeats - will you stay here with me
[12:40] tomppabeats - lonely but not when you hold me
[13:31 - 15:30] Bassti - To All The Ladys In The Place
[15:37] wish you still felt this way [ sophie meiers x 90sFlav ]

Known issues

Sometimes, when downloading videos from YouTube, media-split can fail with the error message "Too many redirects". This is caused by a third party library and is already being tracked here.

Library

You can also use media-split as a library.

let MediaSplit = require('media-split');
let split = new MediaSplit({ input: 'myaudio.mp3', sections: ['[01:30] My audio'] });
split.parse().then((sections) => {
  for (let section of sections) {
    console.log(section.name);      // filename
    console.log(section.start);     // section start
    console.log(section.end);       // section end
    console.log(section.trackName); // track name
  }
});

new MediaSplit(options)

Returns: MediaSplit

Param	Type	Description
options	object
options.downloadCover	boolean	Whether to download the cover from the YouTube video
options.input	string	The input file. Can be either a file path or a YouTube url
options.concurrency	number	Number of parallel workers MediaSplit will spawn at once
options.sections	Array.<string>	Sections to extract from the input source. Supported formats are [01:30 - 03:50] File or [01:30] File
options.output	string	Output path
options.format	string	Output format (mp3, m4a, flac, etc)
options.audioonly	boolean	Force download only audio files when using a url as input
options.quality	string	The download quality when downloading from YouTube (highest/lowest/highestaudio/lowestaudio/highestvideo/lowestvideo)
options.inputParams	Array.<string>	Additional input parameters for FFMpeg
options.outputParams	Array.<string>	Additional output parameters for FFMpeg

MediaSplit.parse() ⇒ Promise.<Array.<object>>

Split the media

"url" (input, info, cached)

URL event. This event is emitted only once.

Param	Type	Description
input	string	The input file
info	object	The video info
cached	boolean	Whether the file was cached or not

"data" (sections)

Data event. This event is emitted only once.

Param	Type	Description
sections	Array.<object>	Array with the parsed sections

"beforeSplit" (info, index)

Emitted before a section is split.

Param	Type	Description
info	object	Section info
index	number	Index

"afterSplit" (info, index)

Emitted after a section is split.

Param	Type	Description
info	object	Section info
index	number	Section index

"downloadProgress" (chunk, downloaded, total)

Download progress.

Param	Type	Description
chunk	number	Chunk length in bytes
downloaded	number	Total downloaded in bytes
total	number	Total download length in bytes

"downloadLength" (length)

Total download length. This event is emitted only once.

Param	Type	Description
length	number	Length in bytes

License

Licensed under the incredibly permissive MIT license

media-split's Issues

Possibility to define output format

Hello @skiptirengu and thank you for this script.

FFMPEG is powerfull tool. Maybe a way for us to define output file format?
I can convert m4a into mp3... but would be fun if it could stay m4a :) 👍

Regex does not handle local folders with a dot in the name

If I have a folder name with a dot, like "/Users/tommie/Git/Github.com/my-project/my-file.mp3", then it is recognized as a url and tries to find a youtube video by id instead, which of course will fail.

I played around a bit with the regex and tried multiple urls + linux/unix and windows folders with dots and it seem to work as expected with my proposed fix below (do not allow any slash before the url, by this in the start: ^(?!\/)):

Today:
const urlRe = /(http(s)?:\/\/.)?(www\.)?[-a-zA-Z0-9@:%._+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9@:%_+.~#?&/=]*)/g

Replace with:
const urlRe = /^(?!\/)(http(s)?:\/\/.)?(www\.)?[-a-zA-Z0-9@:%._+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9@:%_+.~#?&/=]*)/g

Update dependency ytdl-core to more recent version

Running media-split from a fresh install errors with:

Error: Cannot find module 'm3u8stream/lib/parse-time'
Require stack:
- <path>/ytdl-core/lib/info-extras.js
- <path>/node_modules/ytdl-core/lib/info.js
- <path>/node_modules/ytdl-core/lib/index.js
- <path>/node_modules/media-split/lib/MediaSplit.js

media-split currently relies on ytdl-core version 0.29.7, which imports parseTime from m3u8stream/lib/parse-time, however m3u8stream has since renamed /lib/ to /dist/.

Updating the ytdl-core dependency to the version containing the change (^1.0.0) will fix.

See:

• fent/node-ytdl-core#519

Commit in ytdl-core updating path:

• fent/node-ytdl-core@0d730d2

Readme.MD error

In your readme,you still write npm install [-g] mp3-split, what should be npm install [-g] media-split

video resize

im using this great package and was wandering: is there an ability to resize the splitted videos?