recursive file copy

Question

How can I use the terminal / Linux commands to recursively search a drive for images (jpg) and copy them to my desktop, renaming all the found files using numbers. So the first time it finds should be copied to the desktop as 1.jpg, the second file should be copied as 2.jpg and 50,000th file should be copied as 50000.jpg and so on.

Answer 1

n=1; for file in $(find /media/foo/ -name '*.jpg'); do cp $file ~/Desktop/$n.jpg; n=$((n + 1)); done

Answer 2

There are two problems here.

1. Finding a set of jpeg files.
2. Copying a set of files to a target directory with sequential, numbered names.

For the first problem the obvious and correct solution is find, which fortunately makes this very easy.

find / -type f -iregex '.*\.jpe?g$'

Of course this could be made more sophisticated if you wanted to e.g. do type detection with file instead of by extension.

The second problem is sequential copying. A simple counter is all that is necessary to make this work.

n=1
cp "$source" "$dest/$n.jpeg"
n=$((n + 1 ))

Doing this inside a loop, of course, where $source changes on each iteration.

Putting it all together

#!/usr/bin/env bash

usage() {
    echo "$0: usage: $0 [source directory] [destination directory]"
}

if [ ${#@} -ne 2 ] ; then
    usage
    exit
fi

scan="$1"
dest="$2"

if [ ! -d "$scan" ] ; then
    usage
    printf "\nspecified source does not exist or is not a directory\n"
    exit
fi

if [ ! -d "$dest" ] ; then
    usage
    printf "\nspecified destination does not exist or is not a directory\n"
    exit
fi

n=1
while read -r -d $'\0' source ; do
    cp "$source" "$dest/$n.jpeg"
    n=$((n + 1 ))
done < <(find "$scan" -type f -regextype posix-extended -iregex '.*\.jpe?g$' -print0)

Portability note: -regextype is a GNU find extension; without it -iregex may not match file names containing newlines.