Open Vrihub opened 6 months ago
--simulate
is a really old option that was, more or less, only added because youtube-dl had a similar option (2df1a15fb801bc5e1e050df673b86b66bc77ecce). It writes to archives and potentially caches login sessions, but nothing else.
--no-download
was added a bit later (62097284fe7ce70c517af198ccd5de1395ad871e) and, as you assume, does everything except download (extractor.download
). It was added primarily to allow running post processors without having to download files.
-g
and -G
only print download URLs and do nothing else.
So:
--no-download
: doesn't download files (shows file local paths), does everything else
--simulate
: doesn't download files (shows file names), updates download-archive if used
-g|-G
: shows file URLs, does nothing else
-j
: shows json data, does nothing else
-E
: shows some extractor settings, does nothing else
-K
: shows available keywords, does nothing else
Ideally, --simulate
should not write anything to disk (as in youtube-dl and other programs that have this option), and the use case you mentioned in the commit message (just update the download-archive) can be fulfilled by --no-download
; so I'd suggest to fix job.py -> SimulationJob
to not write anything to disk (i.e. only show file names).
Or, let's just leave it as is and clarify the documentation: see the attached patch for option.py
The docs could be improved in explaining the subtle differences between the
--simulate
and--no-download
options, and maybe the related-g
and-G
.AFAIK
--simulate
seems to avoid most writing operations, while--no-download
does everything except for downloading files, right?If you can provide more details about how different features are affected (metadata? archive?....) I will try to update the docs.