Question about awk/if/else (not too difficult, I think!)

[SuaSwe]

Hi all,

I'm having some problems with one section of a script I'm writing. The code I have currently looks as follows:



cat test | uniq -c | awk '{ if ($1 >1) print $1,$2 }' > test.output


... where 'test' is a file containing the following info:

item1
item2
item3
item4
item5
item5
item6
item7

As it stands, the script outputs the count and name(s) of any repeated items in 'test' regardless of whether there is a repeat count or not - so if there isn't, I just get a blank file. What I want is for the script to output the count and name into 'test.output' only if an item is repeated in 'test'. Could anyone tip me off on how to do this?

Cheers!

[ghostdog74]

use


uniq -D file

[SuaSwe]

Guessing that's uniq -d (cap D gives illegal option) and that does look an interesting one - problem is I need the instance count in there as well as the item name. Any ideas? uniq -c obviously means that the rows are all unique, so adding uniq -d doesn't give me any output. Any thoughts on this, such as a way to make uniq look to a certain column? Is that possible?

[Blackbug]

If I understood you correctly..what you want needs to preserve both {print$1} thats the count and {print$2} the value from file.

I wrote a script for it


# storing the output (count)
arr1=(`cat test | uniq -c | awk '{print$1}'`)

# string the output items
arr2=(`cat test | uniq -c | awk '{print$2}'`)

e=0

while [ $e -lt ${#arr1 } ]
do
echo "Count : ${arr1[e]}"

# checks for multiple occurences
if [ ${arr1[e]} -gt 1 ]
then
echo "Multiple Count: ${arr1[e]}"
fi

echo ${arr2[e]}>>file.Output
e=`expr $e + 1`
done


I am not sure what exactly you meant, but if above script is useful, let me know.

[ghostdog74]

$ awk 'a[$0]++ { print a[$0],$0 }' file
2 item5

[SuaSwe]

That sounds interesting, will try that when I'm next able, cheers!