Re: blank lines as "^\s*$"

--- In [email protected], "Ruud H.G. van Tol" <rvtol@...> wrote:
>
> On 07/07/2013 09:38, RAKESH wrote:
> 
> > We could also do it making use of the -0 777 option of perl where
> > the whole file is slurped in, & then we proceed to
> > chop off the leading & trailing blank lines, like as if chopping the front/tails of a carrot.
> >
> > perl -0777wple 's/\A\s+//ms;s/\s+\z//ms' yourfile
> 
> That has issues:
> 
> - the 'ms' modifiers are superfluous
> - the starting whitespace of the first non-whitespace line is removed
> - the newline after the last non-whitespace line is removed
> 

That was a very good catch, (especially point #2),
I was not checking the code b4 posting.

perl -0777wple 's/\A\s*\n(?:\s*$)*//m;s/(?:^\s*$)*\z//m' yourfile

Or this:

perl -0777wple '1 while s/\A\s*\n//m;1 while s/^\s*\z//m' yourfile

As for the 'ms' modifier, I actually have a habit of putting them at the end of every regex that I write just to make me remember less their functionality. Actually, the "-l" option is redundant here as it's being set to the 0777 char in this case.

-Rakesh