Quickly find duplicate files in directories (CLI utility)
Quickly find duplicate files in directories.
CapiDup recursively crawls through all the files in a list of directories and identifies duplicate files. Duplicate files are files with the exact same content, regardless of their name, location or timestamp.
This program is designed to be quite fast. It uses a smart algorithm to detect and group duplicate files using a single pass on each file (that is, CapiDup doesn’t need to compare each file to every other).
CapiDup fully supports both Python 2 and Python 3.
The capidup-cli package is the command-line utility. It depends on the capidup library package, which actually implements the functionality.
Using CapiDup is quite simple:
$ capidup /media/sdcard/DCIM ~/photos /media/sdcard/DCIM/DSC_1137.JPG /home/user/photos/Lake001.jpg ------------------------------ /media/sdcard/DCIM/DSC_1138.JPG /home/user/photos/Lake002.jpg ------------------------------ /home/user/photos/Woman.jpg /home/user/photos/portraits/Janet.jpg ------------------------------
Here we find out that /media/sdcard/DCIM/DSC_1137.JPG is a duplicate of ~/photos/Lake001.jpg, DSC_1138.JPG is a duplicate of Lake002.jpg, and ~/photos/Woman.jpg is a duplicate of photos/portraits/Janet.jpg.
CapiDup crawls the directories and gathers the list of files. Then, it takes a 3-step approach:
There is a very small possibility of false positives. For any given file, there is a 1 in 264 (1:18,446,744,073,709,551,616) chance of some other random file being detected as its duplicate by mistake.
The reason for this is that two different files may have the same hash: this is called a collision. CapiDup uses MD5 (which generates 128 bit hashes) for detecting whether the files are equal. It cannot distinguish between a case where both files are equal and a case where they just happen to generate the same MD5 hash.
The odds of this happening by accident for two files of the same size, are, then, extremely low. For normal home use, dealing with movies, music, source code or other documents, this concern can be disregarded.
There is one case when care should be taken: when comparing files which might have been intentionally manipulated by a malicious attacker.
While the chance of two random files having the same MD5 hash are really very low (as stated above), it is possible for a malicious attacker to purposely manipulate a file to have the same MD5 as another. The MD5 algorithm is not secure against intentional disception.
This may be of concern for example when comparing things such as program installers. A malicious attacker could infect an installer with malware, and manipulate the rest of the file in such a way that it still has the same MD5 as the original. Comparing the two files, CapiDup would show them as duplicates when they are not.
Future plans for CapiDup include having a configurable option to use a different hashing algorithm, such as SHA1 which has a larger hash size of 160 bits, or SHA2 which allows hashes up to 512 bits and has no publicly known collision attacks. SHA2 is currently used for most cryptographic purposes, where security is essential. False positives, random or maliciously provoked, would be practically impossible. Duplicate detection will of course be slower, depending on the chosen algorithm.
For the extremely paranoid case, there could be an additional setting which would check files with two different hashing algorithms. The tradeoff in speed would not be worthwhile for any normal use case, but the possibility could be there.