Removes duplicate files in a directory, replacing them with hard links.

lndupes.sh

#!/bin/sh

dir=$1
if [ -z $dir ]; then
	dir="."
fi;

# create a temporary file name
TMPFILE="/tmp/$(echo -n "$dir" | md5sum | awk '{print $1}' | tr -d '\n')_$(date +%Y.%m.%d.%H.%M)_dupes.txt"

#find the duplicates and output the list to the temporary file
echo "Looking for duplicates..."
nice fdupes -qr1 $dir | egrep -i "\.(mzxml|mzdata|mzml|csv|cdf|dlt|sig|txt)\s*\$" | tee $TMPFILE 

# overwrite duplicate files with hard links.
cat $TMPFILE | while read dupe; do
	firstfile=""
	nextdupe=""
	for i in $dupe; do
		if [ -z "$firstfile" ]; then
			firstfile="$i";
		elif [ ! -f "$firstfile" ]; then
			firstfile="$firstfile $i"

		elif [ -z "$nextdupe" ]; then
			nextdupe="$i";
		elif [ ! -f "$nextdupe" ]; then
			nextdupe="$nextdupe $i";
		fi;
		if [ -f "$firstfile" -a -f "$nextdupe" ]; then
			echo -e "\033[01;32mLinking $nextdupe to $firstfile\033[0m"
			ln -f "$firstfile" "$nextdupe"
			nextdupe=""
		fi;
	done;
		
	
done;

#rm $TMPFILE