associates 2 columns of data using the first one's elements as key

column_correlate.sh

#!/bin/bash

# parses an unsorted document with 2 columns :
#
# ip date
#
# where same ip's might appear multiple times and even the exact same ip date pair
# and outputs
#
# ip  | # of occurences | list of dates of occurences
# 
# each ip appears only once in the output (key) and multiple exactly identical ip-date tuples from the
# initial list are counted in only once
#

awk '{
	if( 0 == match( date[ $1 ], $2 ) ) {
		occurences[ $1 ] = occurences[ $1 ] + 1
		date[ $1 ] = $2" "date[ $1 ]
	}
}
END {
	for( ip in date )
		printf( "%s\t|\t%d\t|\t%s\n", ip, occurences[ ip ], date[ ip ] );
}'