#! /bin/bash

CLEANUP=1 ## Delete temp files on exit, unless 'k' switch if give on command line.
BASEDIR=$PWD

#Parse command line for options
for a in 1 $*; do if [ $a = -v ] || [ $a = -V ]; then  VERBOSE=1; else VERBOSE=0; fi; done
for b in 1 $*; do if [ $b = -h ] || [ $b = -H ]; then  HELP=1; else HELP=0; fi; done
for c in 1 $*; do if [ $c = -O ]; then  OUT2FILE="1"; else OUT2FILE="0"; fi; done
for d in 1 $*; do if [ $d = -o ]; then  OUT2FILE="2"; fi; done
for e in 1 $*; do if [ $e = -k ] || [ $b = -K ]; then  CLEANUP=0; fi; done

help ()
{
  cat <<EOF
                  dupelist - a duplicate file lister - 
  Dupelist calculates, and collects, MD5SUMS of all regular files in the current directory
  then recurses all subdirectories calculating and logging MD5SUMs for all regular files found.
  After collection, MD5SUMS are sorted and checked for duplicates.
  List of duplicates is output in sets seperated by blank lines.
  
  OPTIONS:
   -h Help: display this usage message and exit.
   -o Output to file: Output list of duplicates found to file 'DuplicateFiles.txt' only.
   -O Output to file: Output list of duplicates found to file 'DuplicateFiles.txt' and stdout.
   -k Keep: Don't delete temp files on exit.
   -v (Very) Verbose: optput description of each action as it's done.
  exit 0
EOF
}

seek ()
{
  if [ -e ./dupelist.tmp ]; then rm dupelist.tmp; fi
  if [ -e ./sums.tmp ]; then rm sums.tmp; fi
  # Referencing files by inode numbers rather than file name prevents breakage by
  # filenames that contain blank spaces
  for a in $(ls -i ) 
  do if [ ! -e "$a" ]
         then 
           b=$(find -inum $a 2>/dev/null | sed s#./##)
               if [ -L "$b" ]; then continue
             elif [ -f "$b" ]; then
                echo -n "$PWD/$b:" >> $BASEDIR/dupelist.tmp
                md5sum "$b" >> $BASEDIR/dupelist.tmp
             elif [ -d "$b" ]; then
               cd "$b"
               seek
               cd ..
             fi
       fi
  done
}

parse()
{
  cut -d ":" -f 2 dupelist.tmp | cut -b -33 | sort > sums.tmp
  if [ $OUT2FILE = 1 ]
  then rm DuplicateFiles.txt 2>/dev/null
       for c in $(uniq -d sums.tmp)
       do echo | tee -a DuplicateFiles.txt
          grep $c dupelist.tmp | cut -d ":" -f 1 | tee -a DuplicateFiles.txt
          echo | tee -a DuplicateFiles.txt
       done
  elif [ $OUT2FILE = 2 ]
  then rm DuplicateFiles.txt 2>/dev/null
      for c in $(uniq -d sums.tmp)
      do echo  >>DuplicateFiles.txt
         grep $c dupelist.tmp | cut -d ":" -f 1 >>DuplicateFiles.txt
         echo >>DuplicateFiles.txt
      done
  elif [ $OUT2FILE = 0 ]
  then for c in $(uniq -d sums.tmp)
       do echo 
          grep $c dupelist.tmp | cut -d ":" -f 1
          echo
       done
  fi
}

if [[ $HELP = 1 ]]; then help; fi

seek 
parse

if [[ $CLEANUP = 1 ]]; then rm dupelist.tmp sums.tmp; fi