From: Dossy <dossy@...>
Date: 2002-06-14T09:27:47+09:00
Subject: Re: UNIX pipes

On 2002.06.14, David Douthitt <DDouthitt@cuna.coop> wrote:
> I thought I could do:
> 
> @snaptime = `/usr/contrib/bin/gunzip -lv #{file} | /usr/bin/sed 's/^[^ ]* [^ ]* \(... .. ..:..\).*/\1/'`.chomp
> p @snaptime
> 
> ...and it should (to me, anyway) return:
> 
> "Jun 12 13:54"
> 
> ...instead of the odd result of:
> 
> "\001"

The \001 is probably happening because the \1 in your sed pattern is
getting turned into a \001 by Ruby, instead of getting passed to
sed as the string "\1" ...

Try:

@snaptime = `/usr/contrib/bin/gunzip -lv #{file} | /usr/bin/sed 's/^[^ ]* [^ ]* \(... .. ..:..\).*/\\1/'`.chomp

Escape the \ of \1 as \\1.

> What's the accepted way to use a UNIX pipe?  I wound up using Ruby
> commands instead of sed...

Why use sed when Ruby already has regex?  I'd certainly just take the
output of gunzip and process it in Ruby rather than with another
process (sed).

-- Dossy

-- 
Dossy Shiobara                       mail: dossy@panoptic.com 
Panoptic Computer Network             web: http://www.panoptic.com/ 
  "He realized the fastest way to change is to laugh at your own
    folly -- then you can let go and quickly move on." (p. 70)