grep by paragraph

[Suvayu Ali]

On Fri, Aug 30, 2013 at 11:54:20AM -0500, Michael Hennebry wrote:
> On Fri, 30 Aug 2013, Timothy Murphy wrote:
> 
> >Dave Mitchell wrote:
> >
> >>>that will print out a whole paragraph -
> >>>defined as the section between two blank lines -
> >>>containing a given word or phrase?
> >>>
> >>>Such as the above 4 lines.
> >>>If not, can anyone suggest a simple script that will do this?
> >>
> >>The perl one-liner below demonstrates this. Setting the $/ (input record
> >>separator) var to the empty string causes perl to read in "lines" a
> >>paragraph at a time.
> >
> >>    $ perl -e'$/=""; while (<>) { print if /green/ }' /tmp/text
> >
> >Thanks for the solutions.
> >
> >I gather there is no method using grep or one of its variants?
> 
> sed could do it.
> I don't have a recipe on me at the moment.

Try this:

$ sed -e '/[^[:space:]]\+/{H;d}' -e 'x;/<search_string>/!d' path_to_file.txt

This has the side-effect of adding a newline in the front.  You can
partially deal with it by using `1h;1!H' instead of `H' in the first
expression; but it only works if the match is in the 1st paragraph.  I
do not think it is possible to deal with it properly in sed.

Cheers,

PS: Modified version of an example from this article:
    <http://www.thegeekstuff.com/2009/12/unix-sed-tutorial-7-examples-for-sed-hold-and-pattern-buffer-operations/>

-- 
Suvayu

Open source is the future. It sets us free.