Search for duplicate file names within folder hierarchy?

Question

I have a folder called img, this folder has many levels of sub-folders, all of which containing images. I am going to import them into an image server.

Normally images (or any files) can have the same name as long as they are in a different directory path or have a different extension. However, the image server I am importing them into requires all the image names to be unique (even if the extensions are different).

For example the images background.png and background.gif would not be allowed because even though they have different extensions they still have the same file name. Even if they are in separate sub-folders, they still need to be unique.

So I am wondering if I can do a recursive search in the img folder to find a list of files that have the same name (excluding extension).

Is there a command that can do this?

Solution 1

FSlint is a versatile duplicate finder that includes a function for finding duplicate names:

The FSlint package for Ubuntu emphasizes the graphical interface, but as is explained in the FSlint FAQ a command-line interface is available via the programs in /usr/share/fslint/fslint/. Use the --help option for documentation, e.g.:

$ /usr/share/fslint/fslint/fslint --help
File system lint.
A collection of utilities to find lint on a filesystem.
To get more info on each utility run 'util --help'.

findup -- find DUPlicate files
findnl -- find Name Lint (problems with filenames)
findu8 -- find filenames with invalid utf8 encoding
findbl -- find Bad Links (various problems with symlinks)
findsn -- find Same Name (problems with clashing names)
finded -- find Empty Directories
findid -- find files with dead user IDs
findns -- find Non Stripped executables
findrs -- find Redundant Whitespace in files
findtf -- find Temporary Files
findul -- find possibly Unused Libraries
zipdir -- Reclaim wasted space in ext2 directory entries
$ /usr/share/fslint/fslint/findsn --help
find (files) with duplicate or conflicting names.
Usage: findsn [-A -c -C] [[-r] [-f] paths(s) ...]

If no arguments are supplied the $PATH is searched for any redundant
or conflicting files.

-A reports all aliases (soft and hard links) to files.
If no path(s) specified then the $PATH is searched.

If only path(s) specified then they are checked for duplicate named
files. You can qualify this with -C to ignore case in this search.
Qualifying with -c is more restictive as only files (or directories)
in the same directory whose names differ only in case are reported.
I.E. -c will flag files & directories that will conflict if transfered
to a case insensitive file system. Note if -c or -C specified and
no path(s) specifed the current directory is assumed.

Example usage:

$ /usr/share/fslint/fslint/findsn /usr/share/icons/ > icons-with-duplicate-names.txt
$ head icons-with-duplicate-names.txt 
-rw-r--r-- 1 root root    683 2011-04-15 10:31 Humanity-Dark/AUTHORS
-rw-r--r-- 1 root root    683 2011-04-15 10:31 Humanity/AUTHORS
-rw-r--r-- 1 root root  17992 2011-04-15 10:31 Humanity-Dark/COPYING
-rw-r--r-- 1 root root  17992 2011-04-15 10:31 Humanity/COPYING
-rw-r--r-- 1 root root   4776 2011-03-29 08:57 Faenza/apps/16/DC++.xpm
-rw-r--r-- 1 root root   3816 2011-03-29 08:57 Faenza/apps/22/DC++.xpm
-rw-r--r-- 1 root root   4008 2011-03-29 08:57 Faenza/apps/24/DC++.xpm
-rw-r--r-- 1 root root   4456 2011-03-29 08:57 Faenza/apps/32/DC++.xpm
-rw-r--r-- 1 root root   7336 2011-03-29 08:57 Faenza/apps/48/DC++.xpm
-rw-r--r-- 1 root root    918 2011-03-29 09:03 Faenza/apps/16/Thunar.png

Solution 2

find . -mindepth 1 -printf '%h %f\n' | sort -t ' ' -k 2,2 | uniq -f 1 --all-repeated=separate | tr ' ' '/'

As the comment states, this will find folders as well. Here is the command to restrict it to files:

find . -mindepth 1 -type f -printf '%p %f\n' | sort -t ' ' -k 2,2 | uniq -f 1 --all-repeated=separate | cut -d' ' -f1

Solution 3

Save this to a file named duplicates.py

#!/usr/bin/env python

# Syntax: duplicates.py DIRECTORY

import os, sys

top = sys.argv[1]
d = {}

for root, dirs, files in os.walk(top, topdown=False):
    for name in files:
        fn = os.path.join(root, name)
        basename, extension = os.path.splitext(name)

        basename = basename.lower() # ignore case

        if basename in d:
            print(d[basename])
            print(fn)
        else:
            d[basename] = fn

Then make the file executable:

chmod +x duplicates.py

Run in e.g. like this:

./duplicates.py ~/images

It should output pairs of files that have the same basename(1). Written in python, you should be able to modify it.

Solution 4

I'm assuming you only need to see these "duplicates", then handle them manually. If so, this bash4 code should do what you want I think.

declare -A array=() dupes=()
while IFS= read -r -d '' file; do 
    base=${file##*/} base=${base%.*}
    if [[ ${array[$base]} ]]; then 
        dupes[$base]+=" $file"
    else
        array[$base]=$file
    fi
done < <(find /the/dir -type f -print0)

for key in "${!dupes[@]}"; do 
    echo "$key: ${array[$key]}${dupes[$key]}"
done

See http://mywiki.wooledge.org/BashGuide/Arrays#Associative_Arrays and/or the bash manual for help on the associative array syntax.

Solution 5

This is bname:

#!/bin/bash
#
#  find for jpg/png/gif more files of same basename 
#
# echo "processing ($1) $2"
bname=$(basename "$1" .$2)
find -name "$bname.jpg" -or -name "$bname.png"

Make it executable:

chmod a+x bname 

Invoke it:

for ext in jpg png jpeg gif tiff; do find -name "*.$ext" -exec ./bname "{}" $ext ";"  ; done

Pro:

• It's straightforward and simple, therefore extensible.
• Handles blanks, tabs, linebreaks and pagefeeds in filenames, afaik. (Assuming no such thing in the extension-name).

Con:

• It finds always the file itself, and if it finds a.gif for a.jpg, it will find a.jpg for a.gif too. So for 10 files of same basename, it finds 100 matches in the end.

View more solutions

Comments

[Eliah Kagan]

@DavidFoerster You're right! I have no idea why I had thought this might be a duplicate of How to find (and delete) duplicate files, but clearly it is not.

[JD Isaacks]

How do I execute a command like that in a terminal? Is this something I need to save to a file first and execute the file?

[geirha]

@John Isaacks You can copy/paste it into the terminal or you can put it in a file and run it as a script. Either case will achieve the same.

[JD Isaacks]

Thanks this worked. Some of the results are in purple and some are in green. Do you know off hand what the different colors mean?

[RusGraf]

@John It looks like FSlint is using ls -l to format its output. This question should explain what the colors mean.

[Navin]

FSlint has a lot of dependencies.

[David Foerster]

I changed the solution so that it returns the full (relative) path of all duplicates. Unfortunately it assumes that path names don’t contain white-space because uniq doesn’t provide a feature to select a different field delimiter.

[cp.engr]

@DavidFoerster, your rev 6 was an improvement, but regarding your comment there, since when is sed obsolete? Arcane? Sure. Obsolete? Not that I'm aware of. (And I just searched to check.)

[David Foerster]

@cp.engr: sed isn't obsolete. It's invocation became obsolete after another change of mine.

[cp.engr]

@DavidFoerster, obsolete doesn't seem like the right word to me, then. I think "obviated" would be a better fit. Regardless, thanks for clarifying.

[David Foerster]

@cp.engr: Thanks for the suggestion! I didn't know that word but it appears to fit the situation better.

[Rolf]

It doesn't seem to work properly. It detects P001.ORF and P001 (1).ORF as duplicates and also it seems to think that 60% of my files are duplicates which is wrong I'm pretty sure. fslint found a realstic number of duplicate filenames which is close to 3%.

[Dennis Golomazov]

Doesn't work on mac: find: -printf: unknown primary or operator

[T'n'E]

While this is on the AskUbuntu exchange, this will work on mac (from previous edits): find src -exec basename {} \; | sed 's/\(.*\)\..*/\1/' | sort | uniq -c | grep -v "^[ \t]*1 "

[ojblass]

I do not like how complicated my simple answer has morphed to over the years.