scrapetube problems, fix proposal

[night-0909]

Hello,

There are 4 problems with scrapetube for some months :

• Youtube mixes three structures : videoRenderer, lockupViewModel and shortsLockupViewModel.
• Youtube can return some videos when demanded tab don't exists, eg it returns Streams videos if channel has no stream but has regular videos.
  That's the case when Home tab isn't present in channel.
• Youtube tabs don't have filter options when there's not enough videos.
• popular and oldest filter when presents for a tab are not working anymore, as Youtube updated their structure.

My fork : https://github.com/night-0909/scrapetube
videoId, title and is_live are now properties directly available when iterating on scrapetube.get_channel()
Warning with language of title, see #73 (comment)

[coyotle]

@night-0909 many tnx for your fix ❤️

[night-0909]

@night-0909 many tnx for your fix ❤️

You're welcome :)
I took inspiration on yt-dlp and other projects.

[night-0909]

I reverted Accepted-language to "en" in get_session().
I added a warning on title and description returned when scraping Youtube channel :

As Youtube auto-translate some elements (channel title, video title/description, etc...) based on your location, video title/description can be auto-translated.

So when you iterate on scrapetube.get_channel(), title can be auto-translated in session.headers["Accept-Language"] language set below.
To retrieve the title/description from the default language or added language by the channel owner :
get title and description from Youtube Data Api V3 /videos on each videos
or hit watch?v= and get title and description from ytPlayerResponse->videoDetails
or get snippet.defaultLanguage from Youtube Data Api V3 /channels and set it in header Accept-Language or set a cookie name:PREF value:hl=XX // but defaultLanguage isn't always present

In the end, don't use videos title when iterating scrapetube.get_channel().
I personally use Youtube Data Api V3 /videos because its faster and hitting watch?v= too much could trigger anti-bot detection and need loging in to Youtube or solving captcha.