UNIX for Dummies Questions & Answers

yeah - the short version would be:

Code:

grep 'N 0.0' *.txt | nawk -F'[()]' 'NR {print $4}' | sed -e 's/:/./g'

but legibilty/maintainabilty is never a bad thing :-)

one thing - you might want to reduce this down to the unique set, rather than having duplicates i.e. use sort -u as in e.g.:

orig o/p:

Code:

 00.11.50.d7.c7.3c
 00.03.93.ec.21.93
 00.19.e8.d8.c2.e0
 00.11.50.d7.c7.3c
 00.03.93.ec.21.93
 00.19.e8.d8.c2.e0

new code:

Code:

grep 'N 0.0' inf* | nawk -F'[()]' 'NR {print $4}' | sed -e 's/:/./g' | sort -u

which produces:

Code:

 00.03.93.ec.21.93
 00.11.50.d7.c7.3c
 00.19.e8.d8.c2.e0

ymmv

Very cool!

Thanks a lot for those refinements.

In this instance, having duplicates is not an issue, as the results are simply compared against another list, looking for matches. If we get a double match rather than a single match, it doesn't really matter. (we shouldn't get any matches actually, so in the final output there would presumably be no difference)

However, its neater and quicker if there aren't duplicates, so I think I am going to use your sort -u addition.

Thanks a lot. I really appreciate the feedback.

Cheers