FR: --dry-run feature

[Zocker1999NET]

If I did not miss anything, PFERD as of now has no feature to test / try out what files a crawler would download (and how it would save those) without actually downloading them. I would find this very helpful so while creating my configuration file, I can re-assure (to some extend) that PFERD does not fill up my drive with a bunch of files I did not account for.

[Garmelon]

After a successful crawl, you can use --debug-transforms to test changes to your transforms. While this is not a full dry run, it can help in some cases.

[Zocker1999NET]

Thats good to know, but I have already seen that.
I was more interested into using that dry-run feature before downloading a whole new course and especially to avoid producing unnecessary load & using up too much storage when e.g. the course contains unexpectedly big files or (which I have not seen yet) just in case PFERD misclassifies video files and downloads then anyway while having video=no set.

[lstuma]

Would also be very interested in a feature like that :)

[I-Al-Istannen]

As ILIAS doesn't really give you an easy way for estimating the size of files, I am not sure you would gain much in that regard.
If you want to build this feature you would probably need to stub out the handling in the output directory and every crawler (or you return None for every download but log that you found it). It would also not be completely faithful, as some heuristics depend on the local file (and the final sameness check you are sometimes forced to make also depends on the downloaded file).

Would also be very interested in a feature like that :)

What is your reason for that?