Data Hoarder

I want to create a searchable catalog of various media (physical documents, physical books, digital movies, scanned recipe cards, anything) where I can create categories/types (photo, historical document, recipe card, movie, etc) with unique metadata information, link relevant files (for scanned documents link the scan, for movies link a poster image, etc) and then be able to search through it.

The purpose of the system (which might help you understand what I want) is to (1) be able to create a searchable index of all the physical documents I'm scanning/archiving and (2) be able to also include information about digital items of importance (be it films, books, artwork, etc).

The system doesn't need to manage large file sizes or hard drives (anything large like a movie file will simply be linked and information about where it's stored).

My main requirements are:

(1) Ideally Self-Hosted Web Server (but I'm open)

(2) The ability to define my own metadata structure for each item type I create.

(3) Flexible/advanced search function (to clarify, it does not need to search within a file's contents, just the metadata about the file)

(4) Item relationships (optional) - The ability to assign relationships between items, for example, if I have an item that's a person (Steven Spielberg) , the ability to link items about them (biography of Spielberg) or published by them (Jaws).

Any advice or suggestions would be super appreciated. I did research about various CMS software, but most of them were either far too complicated for my needs or were headless (no frontend) and I'd rather not design my own search frontend.