[Adoby]

If you know a little about shell scripting then it is possible to write a script that renames all the files and add the names of the folders to the file name, with suitable separators. And copies the renamed files to a separate folder.

One problem might be that Windows may not like long file names, so avoid making the file names longer than the old file name with the path. Or this problem has perhaps been fixed in Windows? Or just do this in Linux...

When all the files have been renamed you can import them to a separate Calibre "repository" library and make sure to write file name to some custom tag.

After all this you can use the search and replace functions in Calibre to add tags and fix titles and author names to the books in the repository library. It will still be a lot of manual work. There are tools in Calibre that might help by extracting ISBN from inside books and download meta data and cover based on ISBN.

You might keep the repository and now and then search it and pull out a few books from there to add to your actual library. One author at a time, for instance. Or one specialized subject at a time. Or when you are researching something.

Fix tags and other meta data over time and move books to your main calibre library and make sure that it is perfect. Normalized!

You can create virtual libraries for each genre or subject. Or Year. Or first letter of the first author's middle name. Or based on publisher. Or all this at the same time. You can adopt some real library coding system. You just have to decide how you want to access the books. I use type of publication (periodical, fiction, nonfiction) as base for which library to use. And a simple genre tag after that. You can even have a sub genre tag if you want to.

At any time you can export your fine normalized Calibre library to some wonderful folder structure based on tags and other meta data. But over time I'm sure you will prefer to access the books from Calibre itself.