[PLUG] edit lines in a text file

[Paul Heinlein]

On Tue, 29 Mar 2005, Steve Bonds wrote:

> On Tue, 29 Mar 2005 11:36:33 -0800, Maodhog wrote:
>
>> I want to take a text file, made up of email addresses and chop the 
>> users name off the domain name so I can uniq -u the domains.
>>
>> I can't remember what the command is to remove all before the 
>> character "X"(which in this case is the @ sign).
>
> awk -F@ '{print $2}' < your_file
> perl -p -e 's/^[^\@]+\@//' < your_file
> perl -n -e 'chomp; if (/\@(.*)$/) { print "$1\n" }' < your_file

sed 's!^.*@!!' your_file

cut -d@ -f2 < your_file

I find that GNU sed is typically faster than any other solution, but 
the real time difference is probably insignificant in a job like 
yours.

Oh, to sort the domain names in a way that takes subdomains into 
account, try

sed 's!^.*@!!' your_file | sort -t. -k 3,3 -k 2,2 -k 1,1 | uniq

This recipe assumes no more than 3 levels of subdomains in any given 
address. I assume you can figure out how to adjust the recipe if you 
know that you have more or fewer...

--Paul Heinlein <heinlein at madboa.com>