From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jens Knoell <jens@surefoot.com>
Subject: Re: Need Help with making a bash script
Date: Fri, 28 May 2004 16:40:32 -0600
Sender: linux-admin-owner@vger.kernel.org
Message-ID: <40B7BFE0.7070503@surefoot.com>
References: <c98bdh$7s8$1@sea.gmane.org>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Return-path: <linux-admin-owner@vger.kernel.org>
In-Reply-To: <c98bdh$7s8$1@sea.gmane.org>
List-Id: <linux-admin.vger.kernel.org>
Content-Type: text/plain; charset="us-ascii"; format="flowed"
To: Fernando <ferosspublic@comcast.net>
Cc: linux-admin@vger.kernel.org

Hi Fernando

answers inline.

Fernando wrote:
> I have two problems.
> 
> 
> 1. I'm making a script to add IP's to my blacklist...
> 
> I need to add three <tabs>, the word 'REJECT' and a #DDMMYYYY that it's 
> being done to a text file that has the IP's in it... ie this is the 
> original list;
> 
> 80.25.148.216
> 80.43.197.197
> 
> needs to end up like this...
> 
> 80.25.148.216            REJECT    #28052004
> 80.43.197.197            REJECT    #28052004

I just happen to have almost exactly such a script running here. The 
relevant part is this one here, modified to your needs. It requires that 
you set SOURCEFILE to whatever file you use. Same for DESTFILE. I didn't 
put any secure checks for either source or dest in there, just FYI.

rm $DESTFILE
cat $SOURCEFILE | while read TCPREMOTEIP
do
   echo "$TCPREMOTEIP<insert TABs here>REJECT #`date +%Y%m%d`" >> $DESTFILE
done

The same thing can probably also be done faster with sed.

> 2. The second is a little bit more complicated.
> 
> My server is having trouble deleting virus emails, I want to help it 
> along with a little cron job that will clear out virus emails every 10 
> minutes or so. I get the list of infected emails from the maillog but in 
> the maillog only the email ID is displayed. In the actual queue there 
> are two files per 1 email ID.
> 
> ie... this is what the log spits out as emails that are infected that I 
> want to delete...
> 
> foo1
> foo2
> foo3
> 
> but in the queue there are two files per email, a qffoo and a dffoo so 
> in the queue the previous emails would actually be...
> 
> dffoo1
> qffoo1
> dffoo2
> qffoo2
> dffoo3
> qffoo3
> 
> So, is it possible to run rm with a wildcard to compensate for the df qf 
> letters and still feed it the list of email ID's which make up the last 
> part of the df qf files?

Yes, simply run it with wildcards like this: rm ??foo1

Hope this helps

Jen