So you're saying you want to match 'hijklmnop', with or without any amount of whitespace appearing between any of the characters, including the possibility of newlines?

use strict;
use warnings;

my $string = <<HERE;
abcdefg
hijk    lm
no  pqr
stu
vwkyz
HERE

my $find = 'hijklmnop';

my $re = join '\s*', split( //, $find );

print "Matched!\n" if $string =~ $re;
[download]

In this situation it's not necessary to use the /s switch, but in some situations, where you're trying to match a pattern across multiple lines, you'll want to use the /s regexp modifier. It causes '.' to match any character including newlines. In the example above, however, \s is used, and that matches whitespace including newlines. This and more is discussed in perlre.

Ultimately, it seems pretty awkward building up a regexp with \s* between each character you're trying to match. It might be wiser to preprocess your text, eliminating newlines, or insignificant whitespace.

As davido and BUU referred to here's an example...
If your input file isn't too big you could do something like this...


#!/usr/bin/perl -W

use strict;

my $string = <<HERE;
abc defg
hijk    lm
no  pqr
stu
vwkyz
HERE
 
my $find = "hijklmnop";

# get a copy to avoid modifying orig
$_ = $string;

# remove all spaces 
s/\s//gs;  # note the /s that davido referred to
 
print "Matched!\n" if $_ =~ $find;

# to see what the subst did.
print $_;
[download]

Not sure which is more efficient though...

Also, it seems like you could do something like this, never modifying the orig string...
But I couldn't get it to work. May the gurus can help...


my $result;
$_ = $string

($result) = /((\w+(?=\s*))+)/s;
# ((1 char) followed by 
# (zero or more whitespace[don't return])
# match one or more)
[download]