Re: egrep: question

[Bob Proulx]

Jeremy C. Reed wrote:
> roni wrote:
> > i'd like to search a file, and every time i bump into the word
> > "start", i'd like to print out all lines until i bump into the word "stop".
> > and then proceed.. - to print out all the paragraphs in that file that
> > start with "start" and end with "stop".
> 
> Try sed.
> Have  a look in the "sedfaq". It has examples.

Of course I am sure you can do it in sed.  But the way to have the
process repeat for multiple sets of start and stop blocks escapes me
at this moment.

I think in this case I would use Awk, Perl, or my personal preference
of Ruby.  In this case the Ruby and Perl programs can be identical.
If it was really this simple I would use Awk since it so prevalent.

  awk '/start/,/stop/{print}'

  ruby -lne 'print if /start/ .. /stop/'

  perl -lne 'print if /start/ .. /stop/'

As to your original question about grep/egrep, it does not have
decision making capabilities and switching on and off such as this is
out of scope for it.  You need to use one of the other programs.

Bob