Received: by 10.66.83.35 with SMTP id n3mr3472787pay.23.1350352593209;
        Mon, 15 Oct 2012 18:56:33 -0700 (PDT)
Path: s9ni9014pbb.0!nntp.google.com!npeer01.iad.highwinds-media.com!news.highwinds-media.com!feed-me.highwinds-media.com!border3.nntp.dca.giganews.com!border1.nntp.dca.giganews.com!border4.nntp.dca.giganews.com!border2.nntp.dca.giganews.com!border2.nntp.ams.giganews.com!border3.nntp.ams.giganews.com!border1.nntp.ams.giganews.com!nntp.giganews.com!news.panservice.it!feeds.phibee-telecom.net!zen.net.uk!dedekind.zen.co.uk!reader02.nrc01.news.zen.net.uk.POSTED!not-for-mail
Newsgroups: comp.unix.shell
From: Geoff Clare <ge...@clare.See-My-Signature.invalid>
Subject: Re: A strange issue when using awk to substitute a field in specific line.
References: <k53tif$n9g$1@aspen.stu.neva.ru>
User-Agent: XPN/1.2.6 (Street Spirit ; Linux)
MIME-Version: 1.0
Date: Thu, 11 Oct 2012 13:33:17 +0100
Message-ID: <dtqik9-cne.ln1@leafnode-msgid.gclare.org.uk>
Lines: 47
Organization: Zen Internet
NNTP-Posting-Host: 7d946b44.news.zen.co.uk
X-Trace: DXC=iBJOm30>2D_oS>6CUf2S_]]G;bfYi23hT=dR0\ckLKGPWeZ<[7LZNRVJE]RN?L4]9U4CJ\2i>aFoT>P83MgUBGaZo=Jo3M9UM9T
X-Complaints-To: abuse@zen.co.uk
Bytes: 2564
X-Received-Bytes: 2762
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Hongyi Zhao wrote:

> $ cat sub_test.awk 
> baseDirForScriptSelf="$(cd "$(dirname "$0")"; pwd)"
> awk  -v a=$baseDirForScriptSelf  '{if($1=="set" && $2=="base_path") sub(/
> ^.*$/,a"/apt-mirror",$3);print}'  ./for_test 1<>./for_test
>
> But, after I run the sub_test.awk, I found that the for_test file become 
> the following one:
>
> $ cat for_test 
> set base_path /home/werner/Desktop/test_awk_sub/apt-mirror
> clean http://mirror.bjtu.edu.cn/debian
> clean http://mirror.bjtu.edu.cn/debian-multimedia
> ltimedia
>
> As you can see, the final for_test becomes four lines in it.  The last 
> line "ltimedia" is appeared when I issue the command:
>
> $ ./sub_test.awk
>
> I cann't figure out why this should happen.  Could you please give me 
> some hints on this strange thing?

Because you are using 1<>file to redirect standard output, the shell
does not truncate the file before executing awk.  When awk writes to
the file it is overwriting existing data.  Your transformation
shortens the data, and therefore there is some of the old data
left at the end when awk has finished writing the new data.

The 1<>file trick is neat, but it should only really be used when
you are doing a one-to-one transformation.  It's also only safe if
the transformation is idempotent (i.e. repeating the change does
not alter the result).  This is so that if awk is killed part way
through, you can just restart it from the beginning.

An example of a safe usage is:

$ cat for_test
abcabc
$ awk '{ gsub(/b/, "B"); print }' for_test 1<>for_test
$ cat for_test
aBcaBc

-- 
Geoff Clare <netn...@gclare.org.uk>