Merge pull request #765 from KlzXS/dups

Modified dups to allow selecting files for removal
This commit is contained in:
Mischievous Meerkat 2020-11-08 17:51:11 +05:30 committed by GitHub
commit 7b3b41093e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 58 additions and 3 deletions

View File

@ -4,12 +4,67 @@
#
# Source: https://www.commandlinefu.com/commands/view/3555/find-duplicate-files-based-on-size-first-then-md5-hash
#
# Dependencies: find md5sum sort uniq xargs
# Dependencies: find md5sum sort uniq xargs gsed
#
# Shell: POSIX compliant
# Note: bash compatible required for mktemp
#
# Shell: bash
# Authors: syssyphus, KlzXS
find . -size +0 -type f -printf "%s %p\n" | sort -rn | sed -n 'N; /^\([0-9]*\) .*\n\1.*$/p;$d;D' | awk '{printf("%s\0", substr($0, index($0, $2)))}' | xargs -0 md5sum | sort | uniq -w32 --all-repeated=separate
# If the size of a file has more that $size_digits digits the file will be misplaced
# 12 digits fit files up to 931GiB
EDITOR="${EDITOR:-vi}"
TMPDIR="${TMPDIR:-/tmp}"
size_digits=12
tmpfile=$(mktemp "$TMPDIR/.nnnXXXXXX")
printf "\
## This is an overview of all duplicate files found.
## Comment out the files you wish to remove. You will be given an option to cancel.
## Lines with double comments (##) are ignored.
## If you choose to remove, you will be given a choice between removing files with force or interactively.\n
" > "$tmpfile"
# shellcheck disable=SC2016
find . -size +0 -type f -printf "%${size_digits}s %p\n" | sort -rn | uniq -w"${size_digits}" -D | sed -E '
s/^ {,12}([0-9]{,12}) (.*)$/printf "%s %s\\n" "$(md5sum "\2")" "d\1"/
' | tr '\n' '\0' | xargs -0 -n1 sh -c | sort | { uniq -w32 --all-repeated=separate; echo; } | sed -nE '
h
s/^(.{32}).* d([0-9]*)$/## md5sum: \1 size: \2 bytes/p
g
:loop
N
/.*\n$/!b loop
p' | sed -E 's/^.{32} (.*) d[0-9]*$/\1/' >> "$tmpfile"
"$EDITOR" "$tmpfile"
printf "Remove commented files? (yes/no) [default=n]: "
read -r commented
if [ "$commented" = "y" ]; then
sedcmd="/^(##|[^#]).*/d; /^$/d; s/^# *(.*)$/\1/"
else
printf "Press any key to exit"
read -r _
exit
fi
printf "Remove with force or interactive? (f/i) [default=i]: "
read -r force
if [ "$force" = "f" ]; then
#shellcheck disable=SC2016
sed -E "$sedcmd" "$tmpfile" | tr '\n' '\0' | xargs -0 sh -c 'rm -f "$0" "$@" </dev/tty'
else
#shellcheck disable=SC2016
sed -E "$sedcmd" "$tmpfile" | tr '\n' '\0' | xargs -0 sh -c 'rm -i "$0" "$@" </dev/tty'
fi
rm "$tmpfile"
printf "Press any key to exit"
read -r _