From mboxrd@z Thu Jan  1 00:00:00 1970
From: erik quanstrom <quanstro@quanstro.net>
Date: Fri,  7 Jun 2013 09:18:26 -0400
To: 9fans@9fans.net
Message-ID: <bdd43bfc464622878b43260e4c0f4134@kw.quanstro.net>
In-Reply-To: <1UkwIw-2xe-00@marmaro.de>
References: <1UkwIw-2xe-00@marmaro.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Subject: Re: [9fans] Duff's rc paper: Why awk?
Topicbox-Message-UUID: 64742152-ead8-11e9-9d60-3106f5b1d025

> I've read [0], which is enlightening btw, but there is one thing,
> in Section 27, which I don't understand: Why is awk(1) used there?
>
> 	fn read{
> 		$1=`{awk '{print;exit}'}
> 	}
>
> [0] http://static.tobold.org/rc/rc-duff.html
>
> I rather would have used sed(1), which is less distracting in this
> case:
>
> 	fn read{
> 		$1=`{sed q}
> 	}
>
> This use of awk is unexpected to me, it draws my attention on it,
> thus I'm searching for the strange hidden detail that might be
> emphasized. (Such as the use of `if not' instead of `else'.) But I
> can't find it.
>
> Maybe there is no such hidden detail. Maybe there is no real
> reason behind the use of awk here. I'm not really sure ...

plan 9 sed reads a second line before quitting (note the "def" in the example
below); sed does not work.

- erik
------
; fn read{
	$1=`{sed q}
	echo read `{whatis $1}
}
; read x
abc
def
read x=abc
; fn read{
	$1=`{awk '{print;exit}'}
	echo read `{whatis $1}
}
; read x
abc
read x=abc