mirror of
https://github.com/jellyfin/jellyfin.git
synced 2026-01-15 15:48:03 +00:00
6dc61a430b
Because the Nfo files emit the collections as they are in-memory, the files are not stable in format, genres, tags, albums, people, etc. are emitted in random orders. Add ordering of the collections when emitting the Nfo files so the file remains stable (unchanged) when underlying media information doesn't change. In the process of this, it became clear that most of the providers and probes don't trim the strings like people's names, genre names, etc. so did a pass of Trim cleanup too. Specific ordering: (alphabetical/numeric ascending after trimming blanks and defaulting to zero for missing numbers) BaseItem: Directors, Writers, Trailers (by Url), Production Locations, Genres, Studios, Tags, Custom Provider Data (by key), Linked Children (by Path>LibraryItemId), Backdrop Images (by path), Actors (by SortOrder>Name) AlbumNfo: Artists, Album Artists, Tracks (by ParentIndexNumber>IndexNumber>Name) ArtistNfo: Albums (by Production Year>SortName>Name) MovieNfo: Artists Fix Debug build lint Fix CI debug build lint issue. Fix review issues Fixed debug-build lint issues. Emits the `disc` number to NFO for tracks with a non-zero ParentIndexNumber and only emit `position` if non-zero. Removed the exception filtering I put in for testing. Don't emit actors for MusicAlbums or MusicArtists Swap from String.Trimmed() to ?.Trim() Addressing PR feedback Can't use ReadOnlySpan in an async method Removed now-unused namespace