Basic summary information for duplicates

[howardjones]

Added a one-line summary of each album (in lib, and import target) so
you can easily tell if (for example) you are about to overwrite your
FLAC copy with a low-bitrate mp3 from somewhere else.

[howardjones]

Leaving a new output of:

This album is already in the library!
OLD: 13 items. 13 FLAC. 847141 average bitrate. 2988s total length.
NEW: 13 items. 13 MP3. 205687 average bitrate. 2989s total length.
[S]kip new, Keep both, Remove old?

[sampsyo]

Looks like this doesn't need to be a method, right (i.e., it doesn't use self)? Can we make it a top-level utility function?

[howardjones]

True enough. I will change it.

[sampsyo]

This is awesome! ✨ 🚀 👍 Can't wait to ship it.

I've got a few suggestions for finishing this off. Happy to help with any of them.

[howardjones]

I think that's a bit nicer now.

The next logical thing is to make an automated function to do scoring, a la #116 I guess. This is what my old perl mp3-auto-filing script did. Score 1000 for each FLAC, mp3_bitrate for each mp3.

Not sure how that (or beets generally) will deal with replacing the "deluxe edition" of an album with the regular version of the same album - i.e. the first 12 tracks are identical, but the existing lib version has 6 bonus tracks. It looks like it will see them as identical currently (it only checks the import's items against the lib, not the other way around).

[sampsyo]

Awesome; this is looking great. Thanks for taking those comments into consideration.

Yep, scoring would be useful but certainly complicated!

One other possible next step would be making this work for "in-flight" albums. You may have noticed that resolve_duplicate gets called even when find_duplicates returns nothing—specifically when the album is in seen_idents. The reason is that albums that were just imported won't be in the database yet and therefore won't be found by find_duplicates.

[sampsyo]

All merged up! Thanks again.