Count the number of files in a folder:

$ find ./ -type f | wc -l

List all file extensions found in a folder:

$ find ./ -type f | rev | cut -d "." -f 1 | sort | uniq | rev

List all files sharing the same name within the sub folders:

$ find . -type f -printf "%f\n" | sort | uniq --repeated --all-repeated=separate

List number of files across all subfolders sharing the same name, whatever their extension is:

$ find . -type f -exec basename {} \; | sed 's/\(.*\)\..*/\1/' | sort | uniq -c | grep -v "^[ \t]*1 "

List size in MiB of subfolders and files in the current folder and display them sorted by size:

$ du -cm * | sort -nr

Show the 10 biggest files in MiB found among the current directory and its subfolders:

$ find . -type f -exec du -m "{}" \; | sort -nr | head -n 10

Display the total size used by all PNG files in sub-directories:

$ find ./ -iname "*.png" -exec du -k "{}" \; | awk '{c+=$1} END {printf "%s KB\n", c}'

Case insensitive search from the current folder of all files that have the string dummy in their filename:

$ find ./ -iname "*dummy*"

Recursive and case insensitive content search on non-binary files from the current folder:

$ grep -RiI "string to search" ./*

Same as above but only search string in XML files:

$ find ./* -iname "*.xml" -exec grep -Hi "string to search" "{}" \;

Find all Jpeg images in the system but exclude /home and /var/lib directory:

$ find / -path "/home" -prune -or -path "/var/lib" -prune -or -iname "*.jpg" -print

Get the list of the latest 10 modified files in the current folder tree:

$ find ./ -printf "%TY-%Tm-%Td %TT %p\n" | sort | tail -n10

Same as above but sorted by latest access time:

$ find ./ -printf "%AY-%Am-%Ad %AT %p\n" | sort | tail -n10

Search for string contained in all files named MANIFEST.in, and print their folder path:

$ find . -name "MANIFEST.in" -exec bash -c 'grep --silent "string" "{}" && echo $(dirname "{}")' \;

Search for 4+ characters long upper-cased strings with underscore, in all files but the README.md, LICENSE and Git metadata:

$ grep --only-matching --no-filename --exclude=./{README.md,LICENSE,.git\*} -RIe '[A-Z_]\{4,\}' . | sort | uniq

Search all files starting with a dot, and ending with an extension composed of 6 alphanumeric characters. These are temporary files created by rsync:

$ fd --type file --hidden --ignore-case "^\..+\.[0-9a-z]{6}$"

Create several folder with a similar pattern:

$ mkdir -p ./folder/subfolder{001,002,003}

Create a symbolic link (source):

$ ln -s target link_name

Dump a disk to an image while monitoring the copy progression:

$ pv /dev/da0 > '/mnt/tank/my-data/HDD-part1+part2.img'

Same as above but over SSH:

$ pv /dev/da0 | ssh [email protected] "cat > /mnt/tank/my-data/HDD-part1+part2.img"

Move files containing the date: “2005- string from the ./posts/ subfolder to the other ./2005/ subfolder:

$ grep -R -l 'date: "2005-' ./posts/ | xargs -I '{}' mv '{}' ./2005/

Convert all files in the current folder to lower case:

$ rename 'y/A-Z/a-z/' *

Prefix all files in the current folder:

$ rename 's/(.*)$/prefix-$1/' *

Rename all mp3 files in the current folder by adding a “sub-extension”:

$ rename 's/\.mp3/\.my-sub-extension\.mp3/' *.mp3

Renaming based on regular expression, for files matching another regular expression. The particular example below was used to fix some Dropbox conflicting files:

$ find ./Dropbox -type f -name "* (kev-laptop's conflicted copy 2013-02-01)*" -execdir rename -f -v "s/(.*) \(kev-laptop's conflicted copy 2013-02-01\)(.*)/\1\2/" {} \;

Strip filenames of their leading dot and extension composed of 6 alphanumeric characters. These are temporary files created by rsync:

$ rename --force --dry-run 's/^\.(.+)\.[0-9a-zA-Z]{6}$/$1/' *

Delete all empty files and folders (run this command several times to remove nested empty directories):

$ find ./ -empty -print -delete

Remove empty directories found in all subfolders starting with prefix:

$ find ./ -type d -empty -ipath "./prefix*" -print -delete

Delete files ending with .thumbnail.jpg or .thumbnail.png files (case insensitive):

$ find ./ -iregex ".*\.thumbnail\.\(jpg\|png\)$" -print -delete

Same as above but instead for files ending with their dimensions, like image-640x480.jpg or photo-2400x3200.png:

$ find ./ -iregex ".*-[0-9]+x[0-9]+\.\(jpg\|png\)$" -print -delete

Here is how I clean-up copies of external drives from accumulated cruft over the past decades:

# Remove metadata at volume's root.
$ find . -name "System Volume Information"  -type d -depth 1 -mount -print -delete
$ find . -name ".DocumentRevisions-V*"      -type d -depth 1 -mount -print -delete
$ find . -name ".TemporaryItems"            -type d -depth 1 -mount -print -delete
$ find . -name "\$AVG8.VAULT\$"             -type d -depth 1 -mount -print -delete
$ find . -name ".Spotlight-V*"              -type d -depth 1 -mount -print -delete
$ find . -name "\$RECYCLE.BIN"              -type d -depth 1 -mount -print -delete
$ find . -name ".VolumeIcon.*"              -type f -depth 1 -mount -print -delete
$ find . -name "autorun.inf"                -type f -depth 1 -mount -print -delete
$ find . -name ".fseventsd"                 -type d -depth 1 -mount -print -delete
$ find . -name ".Trash-*"                   -type d -depth 1 -mount -print -delete
$ find . -name "RECYCLER"                   -type d -depth 1 -mount -print -delete
$ find . -name "Recycled"                   -type d -depth 1 -mount -print -delete
$ find . -name "found.*"                    -type d -depth 1 -mount -print -delete
$ find . -name "\$AVG"                      -type d -depth 1 -mount -print -delete

# Remove metadata file and folders artifacts.
$ find . -name "desktop.ini"    -type f -mount -print -delete
$ find . -name "__MACOSX"       -type d -mount -print -delete
$ find . -name "Thumbs.db"      -type f -mount -print -delete
$ find . -name ".DS_Store"      -type f -mount -print -delete
$ find . -name "._*"            -type f -mount -print -delete

# Remove empty directories (repeat until none left).
$ find -type d -empty -mount -print -delete
$ find -type d -empty -mount -print -delete
$ find -type d -empty -mount -print -delete

Delete all files and folders in the current directory except the README.txt file:

$ ls ./ -I "README.txt" | xargs rm -rf

Remove all duplicates within the whole pool of files (including —hidden ones) build up from folder-1, folder-2 and folder-3 directories. In a set of duplicates, the first file in alphabeticcaly sorted named path is kept (-S p option).

$ rmlint --progress --hidden -S p ./folder-1 ./folder-2 ./folder-3
$ ./rmlint.sh

Remove all duplicates in backup-set1 and backup-set2 if and only if they’re already present in backup-set3 (i.e. the reference folder tagged after the // separator), but do not alter the latter in anyway (thanks to the —keep-all-tagged option). To make things extra-safe we use —no-crossdev to not jump to other physical file systems:

$ rmlint --progress --hidden --no-crossdev --keep-all-tagged ./backup-set1/ ./backup-set2/ // ./backup-set3/
$ ./rmlint.sh

Download and refresh local ClamAV virus definition database:

$ freshclam

Check all files, only display infected files and ring a bell when found (really slow scan):

$ clamscan -r --bell -i /

Initialize and start a backup with restic:

$ restic init
$ restic backup --one-file-system ~/

Remove old backups:

$ restic forget --keep-hourly 24 --keep-daily 15 --keep-weekly 13 --keep-monthly 12 --keep-yearly 3 --prune