Limit to one working cache file when using classic plugins in flow

[davidfdezalcoba]

Is your feature request related to a problem? Please describe.
When using flows to preprocess a file before transcoding (i.e. cleaning unwanted subtitle tracks, cleaning unwanted language tracks, etc) with classic plugins in a sequence, a new working cache file is created per step. This means that, when having a flow with, let's say, 8 classic plugins that work on an original file of, let's say, 28Gi, we'll end up with 8*28=224Gi of transcode cache for a single original file, in case every step of the flow has processed it.

As an example, I have a flow that works like this:

and the transcode cache for this flow looks like this:

.
├── 1729593261553
├── 1729593262216
├── 1729593262830
├── 1729593263436
├── 1729593264027
├── 1729593264635
├── 1729593266487
│   └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv
├── 1729593880588
│   └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv
├── 1729593920464
│   └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv
├── 1729594030455
├── 1729594031208
│   └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv
├── 1729594077278
│   └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv
└── 1729594152261
    └── The Aristocats (1970) [imdbid-tt0065421] - [Bluray-1080p][EAC3 5.1][x264]-playHD.mkv

with a total directory size of 42Gi for a file of 8Gi:

4.0K    ./1729593261553
8.0G    ./1729594031208
8.0G    ./1729593880588
8.0G    ./1729594077278
4.0K    ./1729593262216
8.0G    ./1729593266487
8.0G    ./1729593920464
4.0K    ./1729594030455
4.0K    ./1729593264027
4.0K    ./1729593264635
4.0K    ./1729593262830
4.0K    ./1729593263436
1.9G    ./1729594152261
42G     .

This makes it really difficult to have more than one concurrent transcode, since original file size will be multiplied by quite a factor, filling up the transcode cache.

Describe the solution you'd like
I think that tdarr should automatically delete the previous working file, once a step on the flow has processed it and produced a new output file.

Describe alternatives you've considered
I tried using the "Delete file" plugin, but then the flow will not work as intended.

Additional context
This has been tried in latest v2.27.02.

[kagoromo]

I can confirm the same behavior on my Tdarr 2.27.02. It's especially problematic when you are trying to process a huge BD like 80GB in size. The classic plugins I have in my flow were spawning so many cache files that they overflowed my Unraid cache pool. I believe the constant copying was making the flow way slower than it should as well.