Normalising filenames

I use Linux. I don’t like spaces (or any other special characters) in filenames, so have a shellscript which works well to normalise my filenames, so they match the regex “^[a-z_-0-9]*$”
If I import into beets, despite using %lower{$pathspec}, I get what to me is the infuriating set of filenames ‘A Nother Album/…’ - including single quotes.

Can beets be told to leave my carefully selected pathnames alone?

Thanks
Oh, and please, if you want to say “those are perfectly good filenames, why would you want to change them?” , keep it to yourself.

Sorry to comment to my own question, but what I’d really like to do is to have beets import my music (which I ripped from CD’s - and so which is sometimes in a format I dislike) and transform the filenames to the pathspec which I prefer - ie the regex (plus forward slashes, of course ) which I gave.
As an aside, in case you think I’m being finicky, look at what AWS suggest as sensible filename characters for S3 - they make a lot of sense to me.

Hmm; %lower certainly shouldn’t be doing that. I agree that using that and similar path-format functions should be the way to get the filenames you want.

Can you include the paths: configuration you’re currently using?