+- Jellyfin Forum (https://forum.jellyfin.org)
+-- Forum: Support (https://forum.jellyfin.org/f-support)
+--- Forum: Troubleshooting (https://forum.jellyfin.org/f-troubleshooting)
+---- Forum: Media Scanning & Identification (https://forum.jellyfin.org/f-media-scanning-identification)
+---- Thread: Customizing Scanning/Indexing? (/t-customizing-scanning-indexing)
Customizing Scanning/Indexing? - bitstream - 2025-04-08
I'm a new Jellyfin user, moving away from an old Synology, running JF on Ubuntu in a VM.
As much as I enjoyed the easy setup process and the nice GUI, I was perplex by the results after the first scanning run on my /series/ parent folder: some 3000 entries without any structure and apprx. 10% even with a wrong detection.
Reading the "Shows" guideline I got the impression, JF does expect a rather specific folder structure and naming convention. I got a structure of my own, basically /series/nameofseries/season/episode where season is an integer and the episode's filename contains SxxExx with leading zeros and the title of the episode. However, the structure of the filename isn't the same for all series (it's the same within a series).
Example:
"/series/2 Broke Girls/1/2 Broke Girls - S01E01 - Reboot in Brooklyn.mkv"
"/series/2 Broke Girls/1/2 Broke Girls - S01E02 - Private Grenzen.mkv"
"/series/2 Broke Girls/1/2 Broke Girls - S01E03 - Shopping Krieg.mkv"
...
JF did not manage to match these episodes to a season nor to a series.
Bringing all files to the same strucutre will be quite some work I'd prefer not to do.
Therefore i have two questions:
1) Is there a way to specify an individual RegEx/rule to "guide" JF to understand the directory structure and filename structure and to extract the right information for the right variables? I'd prefer to have a mapping rather then to have to rename all files and folders.
2) Is there a way to make JF to just follow the directory structure for grouping series? So top level = series, sub1 = season, sub2 all the Episodes? Like this the structure will be fine and if the episodes are ordered alphabetically, that would be fine too.
RE: Customizing Scanning/Indexing? - TheDreadPirate - 2025-04-08
1. There is no setting or config file to modify the scanner's regex. You'd have to modify the code directly and recompile.
2. Only if you follow the naming conventions.
Emby and Plex also have some organizational requirements, though I know they are somewhat less strict. The reason for the strictness has to do with special functions like handling "extras" differently than episodes, merge "versions" of a movie as one listing. How does the scanner determine where arbitrary organization ends and where the movies/shows actually begin? How does it differentiate between episodes/movies and the extras? How does it know which files in a multi-version movie setup to "merge"?
By adhering to a set of rules.
There are tools that you can use to mass re-organize your media in a Jellyfin friendly way. Tiny Media Manager is our go-to.
RE: Customizing Scanning/Indexing? - bitstream - 2025-04-09
Thanks a lot @TheDreadPirate for clarifying. I will make a test run with the Tiny Media Manager to get an idea of how much effort this is going to be. My gut feeling is I might be faster with a little script and some RegExes.
Not to complain, just to share a thought with the community: I fully understand that JF needs a specific data structure for metadata as a standardized base for all its functionality. However, the reality out there is diverse, so at the same time, it should also be able to deal with this diversity. Wouldn't it be nice to have a kind of configurable mapper? I.e. it could be a configurable plugin for RegEx based mapping or a scripting interface (i.e. LUA) for more complex operations.