Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!sundc!rlgvax!bdmrrr!shprentz
From: shprentz@bdmrrr.bdm.com (Joel Shprentz)
Newsgroups: news.software.b
Subject: Re: Another way to count news articles by date
Message-ID: <819@bdmrrr.bdm.com>
Date: Fri, 16-Jan-87 09:27:08 EST
Article-I.D.: bdmrrr.819
Posted: Fri Jan 16 09:27:08 1987
Date-Received: Thu, 22-Jan-87 06:26:37 EST
References: <286@desint.UUCP>
Organization: BDM Corporation, McLean, Virginia
Lines: 23
Summary: Fast script for 2.11 news history files

In article <286@desint.UUCP>, geoff@desint.UUCP (Geoff Kuenning) writes:
> > From: stephen@comp.lancs.ac.uk (Stephen J. Muir)
> > ...  So here is a wee awk script to print out the number of news articles
> > received on each date in the history file:
> 
> I liked the idea, but found the script slow.  Here's what I came up with:
> 
>     sed 's/[^\t]*\t\([A-Za-z]* *[0-9]*\).*/\1/' /usr/lib/news/history \
>       | uniq -c

I liked Geoff's improvement, but found it didn't work here.  I suspect it
is for an older style of history file.  Here is a similar script that
works for 2.11 history files.  As with Geoff's script, replace \t with
a tab (in two places) before running it.

sed  -e 's/[^\t]*\t *//' -e '/[a-z]/!d' -e '/cancelled/d' \
     -e 's/ .*//' /usr/lib/news/history  | uniq -c

-- 
Joel Shprentz                   Phone:  (703) 848-7305
BDM Corporation                 Uucp:  seismo!bdmrrr!shprentz
7915 Jones Branch Drive                shprentz@bdmrrr.bdm.com
McLean, Virginia  22102