dedupe_copy /

Filename Size Date modified Message
47 B
1.5 KB
2.9 KB
1.0 KB

Find duplicates / copy and restructure file layout command-line tool.

This is a simple multi-threaded file copy tool designed for consolidating and restructuring sprawling file systems.

The most common use case is for backing up data into a new layout, ignoring duplicated files.

Other uses include:
  1. Getting a .csv file describing all duplicated files
  2. Comparing different file systems
  3. Restructuring existing sets of files into different layouts (such as
sorted by extension or last modification time)

This tool is NOT a Robocopy or rsync replacement and does not try to fill the role those play.

As with all code that walks a file tree, please use with caution and expect absolutely no warranty! :)

Command examples:

Generate a duplicate file report for a path:
dedupecopy -p /Users/ -r dupes.csv -m manifest
Copy all .jpg files from multiple paths to a /YYYY_MM/.jpg structure
dedupecopy -p C:pics -p D:pics -e jpg -R jpg:mtime -c X:pics

Copy all files from two drives to a single target, preserving the path for all extensions:

dedupecopy -p C:-p D:-c X:-m X:manifest -R *:no_change
Resume an interrupted run (assuming "-m manifest" used in prior run):
dedupecopy -p /Users/ -r dupes_2.csv -i manifest -m manifest

Sequentially copy different sources into the same target, not copying duplicate files (2 sources and 1 target):

1.) First record manifests for all devices
dedupecopy -p \targetshare -m target_manifest dedupecopy -p \source1share -m source1_manifest dedupecopy -p \source2share -m source2_manifest
2.) Copy each source to the target (specifying --compare so manifests from

other sources are loaded but not used as part of the set to copy and --no-walk to skip re-scan of the source): dedupecopy -p \source1share -c \targetshare -i source1_manifest

--compare source2_manifest --compare target_manifest --no-walk
dedupecopy -p \source2share -c \targetshare -i source2_manifest
--compare source1_manifest --compare target_manifest --no-walk

Complete example:

Assuming you start with a set of files laid out as follows:

photo1.jpg photo2.jpg




And you run the command: > dedupecopy -p C:pics -p D:pics -e jpg -R jpg:mtime -c X:photos

You would end up with the following (photo3.jpg was a true duplicate):

photo1.jpg photo2.jpg photo3.jpg

This project is on bitbucket: This project is on pypi: