Only download the video file without extracting info

[Interopnl]

Is it prossible to instruct Youtube-dl to only download the video file by providing the video download url and not extracting the video information?

[phihag]

No, not yet - many features (for example, output template or video format selection) are only useful with information about the video. May I ask what your overarching goal is? How did you get the video URL in the first place?

[jaimeMF]

If you know only the url and want to save it to a specific file, using python is not difficult, for example:

from youtube_dl import *
ydl = YoutubeDL({'outtmpl':''})
ydl.fd._do_download(u'example.jpeg', {'url':"https://upload.wikimedia.org/wikipedia/commons/3/31/Faro%2C_Puerto_de_la_Cruz%2C_Tenerife%2C_Espa%C3%B1a%2C_2012-12-13%2C_DD_01.jpg"})

But as phihag has said, you can't use the ouput template, for that you need an info_dict with all the information

[Interopnl]

Okay, thanks for your quick reply.

The goal is that once the you have extracted the video information from youtube for example, in a json file, there is no need to roundtrip to youtube again just to download the video file. I would be a nice to have to instruct youtube-dl to use a json file with the extracted video information.

[phihag]

Oh, that sound's reasonable simple to implement (@jaimeMF already provided an outline we can use, although I'm not sure that it includes post-processing). However, we need to move the cookies to the .info.json first. Just keep following this issue, we'll update it when it gets implemented.

[jaimeMF]

@phihag it just downloads from the url, it's not useful for this case.
It's not hard to implement, once the cookies are read, the only thing that needs to be done is pass the info_dict to ydl.process_info, which will execute post-processors, save subtitles...

[oliverthered]

the bugs I raised #1566 is kid of the reverse process of this since the part file contains the watch url info not the actual video file info but the solution is similar (or could be)

[jaimeMF]

I would save the original url (or a simplified one) in the json info, so that we could reobtain the video url if the link has expired, but we would first try to download the already extracted info (to save time).

[jaimeMF]

With the last version once you have written the info to a .info.json file with the --write-info-json, you can run youtube-dl --load-info somefilename.info.json and it will use the saved info to download the video.

[Interopnl]

Thanks for implementing this!