A tool for efficiently scanning over large numbers of files, to find and remove duplicates. Mac, Linux, and Windows binaries are here.
I wrote this for a bit of practice using Go, and to help me finally get around to cleaning up that mess of photo backups.
Here's an example use case for this tool:
I have spent a while organising most of my photos, by year etc. These are stored in /photos.
I have a copy in /backup from when I was part of the way through the re-organization.
I have a directory /unsorted with photos that were downloaded from iCloud (in /unsorted/icloud), Amazon Photos (in /unsorted/amazon), Google Photos and others.
Some photos are only backed up in one place, some are duplicated, it's a big mess.
What I'd like to do is remove all duplicates from /backup and /unsorted so that I can organise the ones that aren't in /photos yet. I can do that with just one command:
dedupe --trash=/trash /photos /backup /unsorted
I created a /trash directory to move the duplicates that are 'trashed'.
They won't be deleted, and I can undo the move of some/all of them by manually copying them back to where they were originally.
Duplicates will be moved to /trash/backup, /trash/unsorted/amazon, etc. based on where they started.
If I'm nervous about what could get moved I could leave off the --trash option, and it would just find duplicates and output what would be moved.
dedupe /photos /backup /unsorted
It decides what to move based on the order of the directories- higher priority first.
If I have some directories within /backup that are better organised even than my main /photos directory I can make those higher priority:
dedupe --trash=/trash /backup/2015move /backup/2020aprilfun /photos /backup /unsorted
If you don't want to build it yourself you can pull down a precompiled binary from here:
Here's the help text which the tool outputs, if it's unclear let me know or file an issue:
Usage: dedupe [OPTION] DIRECTORY...
dedupe --trash=<trash> [OPTION]... DIRECTORY...
Search DIRECTORY(ies)... for duplicate files and optionally moves them to <trash> without user interaction.
DIRECTORY order is used for priority, highest first. Higher priority files are left untouched and lower priority files are moved.
Mandatory parameters:
Options:
--trash root directory for moved duplicates, (default: files not moved)
--compare-time compare file modification time (default: false)
--compare-name compare file name (default: false)
--compare-size compare file size (default: true)
--compare-hash compare file hash (default: true)
--compare-contents compare whole file contents (default: false)
--min-size minimum file size, bytes or human readable e.g. 4M, 5G (default 1)
--follow-symlinks follow symbolic links, false ignores them (default false)
--verbose emit verbose information (default: false)
--version output version and license information and exit
Advanced options:
--scan-buffer size of the scan buffer (default: 100)
--scanners number of scanner coroutines (default: 2)
--match-buffer size of the match buffer (default: 100)
--matchers number of matcher coroutines (default: 4)
--move-buffer size of the move buffer (default: 100)
--movers number of mover coroutines (default: 2)
--max-cpus maximum CPUs to use (default: system setting)
Copyright 2020 Alan Davies <alan@galax.xyz>
Distributed under the MIT license <https://opensource.org/licenses/MIT>.
See <https://github.com/glxxyz/dedupe> for documentation and help.