alternate caption scraper

[tomByrer]

I saw that you added the ability to download transcripts via GUI. That can work, but incase the transcripts GUI doesn't work, there will likely still be the XML file hidden somewhere that you can grab. Might be more efficient:
(My special fork; I wanted to keep some HTML markup. WebVTT output would be easy to add.

https://github.com/tomByrer/youtube-captions-scraper/blob/master/src/index.js

[absidue]

YouTube.js already extracts the captions information, it has supported captions for much longer than transcripts.

They also have two different use-cases, which is why YouTube has them separately and YouTube.js supports both. Showing captions on a video where you might want styling or a transcript where you want pure text.

Also there is a reason YouTube calls it a base URL, because you can add a bunch more query parameters to it, YouTube mainly uses it to add tracking parameters but there are two useful ones:

• fmt: can be set to vtt, ttml, srv1, srv2, srv3 or json3
• tlang: can be set to any of the languages in the translation languages array, to make YouTube spit out a machine translation of that caption track, into your desired language.

So if you want a WebVTT file just add &fmt=vtt to the captions base URL, no need to do the conversion yourself.