Printing arbitrary line range from files

[lisa-asket]

Yes, it seemed to run once total.



But I want to return lines $na trough $nb from each file matched.






From: Seth David Schoen <schoen@loyalty.org>
To: lisa-asket@perso.be
Subject: Re: Printing arbitrary line range from files
Date: 28/06/2021 09:35:07 Europe/Paris
Cc: Greg Wooledge <greg@wooledge.org>;
   help-bash@gnu.org

lisa-asket@perso.be writes:

> Does not function very well when I do
> 
> print-lines 5 8 .
> 
> print-lines ()
>   {
>     na=$1
>     nb=$2
>     dir=$3
> 
>     find "$dir" \( -name \*.org -o -name \*.texi \)  \
>       | xargs sed -n "$na,${nb}p"
>   }

I see two things that could easily go wrong:

* You probably want -print0 in find, and -0 in xargs (so that it will
work properly for files whose names contain spaces).

* You probably want -n 1 in xargs so that it will run the sed command
once per file, instead of (likely) once total. xargs constructs
command lines using multiple file arguments per command, unless you
use -n to limit this. When you want it to run the command separately
for each individual file, you then need -n 1.

The reason for the second behavior is that many programs will take
multiple file arguments and do the expected thing with all of them. If
you wanted to compress a lot of files with gzip, you could use

find /somedir -print0 | xargs -0 gzip

and xargs would run gzip "as few times as possible" in some sense, but
gzip would still compress each file.

-- 
Seth David Schoen <schoen@loyalty.org> | Qué empresa fácil no pensar
http://www.loyalty.org/~schoen/ | en un tigre, reflexioné.
| -- Borges, "El Zahir"