How it works, under the hood?

[VFansss]

Hi. First of all, thanks for your efforts.

I've found Sharedown while I was searching on how to download video from Sharepoint. Incredibly, this seems the ONLY solution working I've found online, in either involving ffmpeg and yt-dlp.

I'm asking myself: what actually is Sharedown under the hood?

• You paste the sharepoint video URL on Sharedown, and you press "Start downloading"
• "You log in"
• You land on the video page through url
• "something happen in the background"
  • I guess Pupeteer (or something like that) ispect the page/traverse the DOM and find the video URL
  • Incredibly, I still didn't understood WHAT is the video URL
• It pass the URL and I guess some authentication token to yt-dlp or ffmpeg
  • What paremeters it actually pass?

Can I have some detail on what's performed on the bold parts of the process?

Because I would truly like to massive download video/content using yt-dlp (or ffmpeg) but for the love of me I can't find any way to find the right video url to pass to yt-dlp

[MaximAndreevich]

there is a guide on reddit how to use ffmpeg without any wrappers. I guess this is one of the scenarios.
https://www.reddit.com/r/sharepoint/comments/nuk8q0/is_there_any_way_to_download_view_only_videos/
didn't work for me anyway :/

Also you can enable logging and check messages, there are steps described.

[kylon]

the app is open source so all steps are in https://github.com/kylon/Sharedown/blob/main/preload.js
there is no url in the page, it tries to catch a known response with the video data and build video manifest url, the manifest url is what ffmpeg and yt-dlp will get

[soufhehe]

yt-dlp -f- --allow-unplayable-format -o "output" "manifest url" and type the audioID+videoID

without --allow-u it says that it is DRM protected.