md5sumを用いて重複ファイルを削除するスクリプトを書いた / Using md5sum, this script removes duplicated files

rm_duplication.sh

#!/bin/bash
# this script removes duplicated files 
# using md5sum.
#
# $1 : target directory
#
tmptable=$(mktemp temp.XXXXX)

\ls -t ${1%/}/* | while read line
do
 md5 "$line" | sed "s,^MD5 (,," | sed "s/) = / /" 
done > $tmptable

# duplicate check
cat $tmptable | awk '{print $NF}' | sort | uniq -c \
| awk '{if($1 > 1) print $NF}' \
| while read ff
do 
 cat $tmptable | egrep " $ff$" | sed "s/ $ff$//" | awk 'NR>1{print}'
done | while read rmfile
do
  echo "removed "$rmfile
  rm $rmfile
done

rm $tmptable