UNIX for Beginners Questions & Answers

file:
Data has to be grouped on 1st field; And in each group, print 2nd field IF 1st field doesn't match with any of the values in 2nd field..
o/p to be:
I ended up writing as below, any help ??

Code:

awk -F, '!($1 in a){a[$2]++;}END{for (i in a)print i}' file

I'm not quite getting your requirement. but your code will not exclude the way you seem to want. You have to read the entire file into an array first, then loop back through it, set your exclusions in the second loop, i.e., set the array element "count" from 1 to zero. Now your array will work for a simple read through it.

Why? because encountering later records in the file invalidates your first choice of "yes, print this" based on an incomplete data set. ...assuming I understood your request.

Code:

awk -F, '{a[$0]=$0} $1==$2 {b[$1]=$2} END {for (i in a) {$0=a[i]; if (!($1 in b)) print $2}}' file

Try also

Code:

awk -F, '{T[$1] = T[$1] ORS $2} END {for (t in T) if (!(T[t] ~ t)) print substr(T[t],2)}' file
B2
C2