On Tue, 25 Nov 2003 14:22:08 -0600, Steve Cowles wrote > On Tue, 2003-11-25 at 13:01, Mike Vanecek wrote: > > I have a bash command that is essentially grep | sed | sort. It creates a file > > of sorted records. What I would like to do next is create a report that lists > > each unique item and its count. > > > > 1 > > 1 > > 1 > > 2 > > 3 > > 3 > > > > would be > > 1 - 3 > > 2 - 1 > > 3 - 2 > > > > Can this be done via a bash command and if yes, how? Otherwise, I would assume > > it needs to be fed to a perl script? > > Type: man uniq > > In particular, the -c option will give you a good start a generating > your final report. i.e. given the results of your script are stored > in report.txt > your_script | uniq -c > 1 3 > 2 1 > 3 2 Perfect, thank you. I modified an idea by Ulrich Scholler to (zcat /var/log/maillog.1.gz; cat /var/log/maillog) | grep "reject" | sed -re 's/.*\[([0-9]+\.[0-9]+\.[0-9]+\.[0-9]+)\].*/\1/' | sort | uniq -c quickly shows troublesome ips. -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
