dirkf

The fake cookie prevents the redirect. Sadly the extractor can't find the video in the page despite the several different tactics used for extraction. This is because the page returned...

There is a login routine that tries to POST data including username and password to https://www.facebook.com/login.php?next=http%3A%2F%2Ffacebook.com%2Fhome.php&login_attempt=1. The problem with these routines is that sites are prone to change them randomly...

The GET is to extract some nonce data sent by FB that gets replayed in the subsequent POST, but, as you say, this isn't working and my original assumption is...

m.facebook.com might be less tricky than the login for the full JS app (but as long as the login cookies are valid for facebook.com as a whole that's no problem)....

Thanks. From that, we need to modify the extractor's `_login()` method thusly: * use m.facebook.com/login.php * use the correct URL to POST the form (retrievable from the form action, or...

In the extractor the tahoe endpoint is used when other tactics fail. Is it possible that it may still work in some cases? Ofc, it's more likely to be a...

>the Android app uses HTTP/3 for the JSON requests, But [not the iPhone app](https://caniuse.com/http3)?

I can reproduce this. The URL works with yt-dlp: we need to back-port the extractor, like this: ```shellsession $ python -m youtube_dl -v -F 'https://www.la7.it/piazzapulita/rivedila7/piazzapulita-puntata-del-13102022-14-10-2022-455682' [debug] System config: [u'--prefer-ffmpeg'] [debug]...

Ah. Maybe the single executable file installation (is that what you used?) appears to have installed itself in `/usr/local/bin` while it's running. When a PR appears to fix this I'll...

See https://github.com/yt-dlp/yt-dlp/issues/5360#issuecomment-1292387041 for another necessary change to be rolled in.