How to compare list of files with different extensions and delete extra files?

Question

I have 2 folders one contains images files and the second contains text files each text file have the same name as the image file and contains information about the image.eg:

   -Labels:
     -1.txt
     -2.txt
     -3.txt
     -6.txt
   -Images:
     -1.jpg
     -2.jpg
     -3.jpg
     -4.jpg
     -5.jpg
     -6.jpg

I want to delete images who has not a text file(in this example:4.jpg,5.jpg), I found a method how to determinate the different files but I can't delete them.

diff <(ls -1 ./Images | sed s/.jpg//g) <( ls -1 ./Labels | sed s/.txt//g)

Answer 1

Here is a small bash script that can help you to solve this task:

#!/bin/bash
for file in Images/*.jpg
do
    if [[ ! -f "Labels/$(basename ${file%.*}).txt" ]]
    then
        echo rm "$file"
    fi
done

• Remove echo to do the actual changes.

The script must be executed in the parent directory, here it is formatted as inline command:

for f in Images/*.jpg; do if [[ ! -f "Labels/$(basename ${f%.*}).txt" ]]; then echo rm "$f"; fi; done

Answer 2

I think I would do it like this:

for i in Images/*; do file=`echo $i | sed -e 's/jpg/txt/' -e 's/Images/Labels/'`; if [ ! -f "$file" ] ; then rm $i ; fi; done

If you want to make sure it works , before actually using it try this first:

for i in Images/*; do file=`echo $i | sed -e 's/jpg/txt/' -e 's/Images/Labels/'`; if [ ! -f "$file" ] ; then echo rm $i ; fi; done

It will show which commands will be executed.

Answer 3

cd /path/to/Images
LIST=$(find /path/to/Labels -iname *.txt -printf "%f|" | sed 's/.txt/.jpg/g')
rm -i !($LIST)

1. Generates a list of txt files.
2. Changes their extension to jpg.

3. Removes anything that is not on the list.

   • -i is just for safety.