http://stackoverflow.com/a/10755010/418413

bench

#!/bin/bash

generate_lines() {
	local length=$1 # Generate this many lines.
	local input_file=$2 # Leave $2 and $3 empty to generate an independant file.
	local -i input_length # The number of lines in the input file.
	local freq_dupes=$(( ${3:-length+1} )) # Copy a line from $input every $freq_dupes iterations
	if [[ -r $input_file ]]; then
		input_length=$(( $(wc -l < "$input_file") ))
	fi
	for (( i=0; i<length; i++ )); do
		if (( i && (i % freq_dupes == 0) )); then
			# If i > input_length, input_line = i % input_length, otherwise i
			awk "$((i > input_length ? i % input_length : i))"'==NR { print $0; exit }' "$input_file"
		else
			echo $RANDOM
		fi
	done
}
time_comm() {
	echo 'Testing comm method.' >&2
	time bash comm_method a b
}
time_grep() {
	echo 'Testing grep method.' >&2
	time bash grep_method a b
}

setup_tests() {
	echo "Setting up tests"
	local comm_path="$PWD/comm_method"
	local grep_path="$PWD/grep_method"
	if ! pushd "$(mktemp -dt "${0##*/}")"; then
		echo "fail."
		exit 1
	fi
	ln -s "$comm_path"
	ln -s "$grep_path"
}

generic_test() {
	local -i length=$1
	local -i rep=$2
	(( rep=rep?rep:10 )) # rep can't be zero
	echo "Creating ${length}-line files with $rep repetitions in the second file."
	generate_lines $length > a
	generate_lines $length a $rep > b
	echo '----------'
	time_comm a b > /dev/null
	echo
	echo '----------'
	time_grep a b > /dev/null
}

test_10() {
	generic_test 10 $(( RANDOM % 10 ))
}

test_100() {
	generic_test 100 $(( RANDOM % 100 ))
}

test_1000() {
	generic_test 1000 $(( RANDOM % 1000 ))
}

test_10000() {
	generic_test 10000 $(( RANDOM % 10000 ))
}

test_100000() {
	generic_test 100000 $(( RANDOM % 100000 ))
}

run() {
	setup_tests
	echo
	echo '=========='
	echo
	test_10
	echo
	echo '=========='
	echo
	test_100
	echo
	echo '=========='
	echo
	test_1000
	echo
	echo '=========='
	echo
	test_10000
	echo
	echo '=========='
	echo
	test_100000
}

run

comm_method

#!/bin/bash

while read; do
	if [[ $REPLY = $'\t'* ]] ; then
		printf "%s\t0\n" "${REPLY#?}"
	else
		printf "%s\t1\n" "${REPLY}"
	fi
done < <(comm -2 <(sort "$1") <(sort "$2"))

grep_method

#!/bin/bash

grep -xF -f "$@" | sed $'s/$/\t1/'
grep -vxF -f "$@" | sed $'s/$/\t0/'