Shows/Movies with sample or proof in the name get ignored

[Safihre]

Based on https://forums.sabnzbd.org/viewtopic.php?p=126985

Show name is: The.Moment.of.Proof.S01E01.720p.HEVC.x265-MeGusta
Because of the use of is_sample, the Sorter ignores all files of this job.

I think we can be a bit smarter: if is_sample(NameOfTheDownload) == true, then we allow samples.

Also @jcfp, do we need the to_filepath and is_full_path function really? Aren't the file paths always full-paths?
Probably missing something.

[jcfp]

I think we can be a bit smarter: if is_sample(NameOfTheDownload) == true, then we allow samples.

While an improvement, that would still break sample detection - just for a smaller subset of jobs. We might want to opt for a stricter regex (e.g. one requiring the "sample trigger word" to be at/near the end of the filename) when that condition is true, or fall back to guessit's sample detection for jobs that we know in advance the standard regex-based check will fail to properly detect (it has no trouble with the show name in the example).

As for the functions, not sure: I didn't want that stuff to complicated things further when converting the sorting to guessit with no unittests in place, so I left parts of the old code in place assuming it was there for a good reason.

[thezoggy]

To be proper we should look at media files only. Let alone media file runtime to know if it's a sample.. of course trying to include ffprobe (mediainfo is less reliable on handling stuff as proven by *adarr) can be a mixed bag with supporting arch.

Could use ffprobe-python ?

[Safihre]

While an improvement, that would still break sample detection - just for a smaller subset of jobs.

So this would only be for the Sorting part where we determine what files to rename, a place where we know the job's name.
This is separate from the "Remove Samples" option.
What cases would sample detection break? Job name: Test s01e01, filename: Body of proof.avi.
Is that realistic?

[thezoggy]

I've personally had sab had false positive in the past because it bad naming and ended up just disabling sabs sample removal.