rmdoop

A CLI tool for recursively removing duplicate files (by content) from different directories.

Work in progress

This project is still in development. While a minimum viable product is available, only the autonomous mode is implemented so far. I plan to add a prompted, user-friendly mode for more curated workflows.

Installation

curl --proto '=https' --tlsv1.2 -LsSf https://github.com/Zitronenjoghurt/rmdoop/releases/download/v0.1.0/rmdoop-v0.1.0-installer.sh | sh

Usage

Basic concept

rmdoop distinguishes between source and target files:

• Source files are never deleted, but are used for comparison.
• Target files are candidates for deletion.

A group of identical files needs at least one source file for deduplication to happen.

# All files in the current directory become targets.
# Without any source files, nothing will be deleted.
rmdoop ./

Provide one or more source directories using --sources (or -s):

rmdoop ./ -s ~/home/original ~/home/other-originals

Promotion

If you just want to deduplicate files in a directory and don't care about which files are the actual source files, you can use --promote (or -p):

# Promotes the first file of each identical group to a source file,
# deleting the rest.
rmdoop ./ -p

Safe mode

Use --list (or -l) to preview which files would be removed without actually deleting anything:

rmdoop ./ -l

Combine it with --promote to preview deduplication of a single directory:

rmdoop ./ -lp

Autonomous mode

Use --autonomous (or -a) to skip confirmation prompts. rmdoop will still only delete duplicates that have at least one source file.

# Automatically deduplicate all files in the current directory.
rmdoop ./ -ap