#!/bin/sh

BASE=data

find $BASE -type f -exec md5sum {} \; > liste.txt
sort liste.txt > liste.txt.sort
rm -f mygleiche.*
N=1
F=`head -1 liste.txt.sort |  cut -d' ' -f1`
while read LINE; do
   A1=`echo $LINE | cut -d' ' -f1`
   A2=`echo $LINE | cut -d' ' -f2`
   if [ "${A1}" == "${F}" ]; then
      echo "$A2 " >> mygleiche.${N}
   else
      let N=N+1
      echo "$A2 " >> mygleiche.${N}
   fi
   F="$A1"
done < liste.txt.sort

for i in mygleiche.*; do
  LLL=`cat $i | wc -l `
  if [ $LLL -gt 1 ]; then
     sort -n $i > $i.s
     while read LINE2; do
        echo -n "$LINE2 " >> mygleiche.alle
     done < $i.s
     echo >> mygleiche.alle
  fi
done 
sort mygleiche.alle | sed -e "s/ $//" > gleiche.out