Duplicate plugin behavior when moving files

[nsteins]

I'm struggling with the behavior of the duplicates plugin. I would like to make a copy of all the duplicate files in another directory and then delete all of the duplicates (alternatively, I could just move them). However, when I run the beets duplicate -k artist -k title -k album --copy dupes, beets correctly copies the file from music to dupes, but then updates the file path of the duplicate to the file in the dupes folder. So now I have two files in the music folder, and one in the dupes folder, and two entries in the db, one pointing to the music folder and one pointing to dupes. So if I run beets duplicate -k artist -k title -k album --delete, the file in the dupes folder is deleted, leaving me with the same duplicate files in the music folder, but only one entry in the db.

Any suggestions on the way to actually accomplish my goal here? (delete duplicates but make a backup first) Also, how can I best import the copied files in music which are no longer in the db?

[benpage2]

If you run beet duplicates -o PATH [other options] it will copy the duplicates to PATH. Although, I'm not sure if it removes them from your library database.

You can check this by running beet ls PATH. If stuff shoes up, they're still in your database and you need to run 'beet rm PATH`. For the sake of clarity this won't delete your backups but just remove them from the database.

Then run beet duplicates -d [other options] to delete the duplicates form your library and disk, but keep your backups at PATH

[nsteins]

Thank you for your help, but the issue is that there are now three copies of the file (2 in the library folder, 1 in the dupes folder), but only 2 entries in the db. So after I remove the copied duplicates using beet rm /dupes, running beet duplicates no longer turns up duplicates because the original duplicates are not in the db.

[benpage2]

Apologies, I got a bit confused. Just to clarify, you want to use duplicates to delete duplicate tracks, but keep a copy of the ones that were deleted just in case there was a mistake?

[nsteins]

yes, that is ultimately my goal. I would like a back up copy of the files, but the duplicates removed from my library folder and from the beets database.

[benpage2]

If you haven't you should add the old tracks back to your library. Then the following should work for you.

If you want to move of all copies of the duplicate tracks (i.e. if you have two files of the same song, move them both), run the following:

beet duplicates -F -m PATH -k artist -k title -k album

• -F tells it to act on all copies of the duplicated track
• -m will move the tracks from your library to PATH
• I've included your keys as is

Then beet remove PATH to remove the entries for all the items in PATH from your database.

Alternatively, If you want duplicates to leave one copy of the duplicated item in your library, just remove the -F flag:

beet duplicates -m PATH -k artist -k title -k album

• You can control which version it chooses to leave in your library by adding to the tiebreak: option in to config (see the docs)

Then same as above, run beet remove PATH

Let me know if something doesn't work right, I'll keep watching my inbox.

[nado]

I also got confused with this unintuitive meaning. I get the logic of it mirroring beet move/copy/delete, but only after having actually tested the different results on a test beet db.
Could we add a beet duplicates --export PATH which would mimic beet move -e and add a mention of the library update in case of a --copy or --move?