my $nr_of_lines = $str =~ tr/\n//;
[download]

> Add 1 if you want to count an incomplete last line, and the string doesn't end with a newline.

better use scalar split, counting the delimiter is unnecessary complexity.

  DB<1> $a="1\n2\n3";$b=$a."\n"
 
  DB<2> print scalar split /\n/ for $a,$b
33
  DB<3> print scalar tr/\n// for $a,$b
23
[download]

Cheers Rolf

Two problems with scalar split: 1) pre-5.12, it will globber @_. 2) it's incorrect:

print scalar split /\n/ for "1\n2\n3\n\n";  # Prints 3 instead of 4
[download]

Using a negative third argument doesn't solve the problem:

print scalar split /\n/, $_, -1 for "1\n2\n3\n\n";  # Prints 5
[download]

you'd have to subtract 1, but only if the string ends in a new line - so you end up looking at the end as well.

As for the complexity:

print tr/\n// + !/\n\z/;
[download]

even gives the correct results in list context.

Thanks, that's a nice neat bit of working code with the brackets added. Thanks a lot.

Just operate on the string directly.

(At least the technique described in perlfaq5 can be easily used that way).

Here's is the another way.Treating the scalar as a file handle.

use strict;
use warnings;
use IO::Scalar; 

my $string = "one\ntwo\nthree\nfour\nfive\nsix";
my $scalar_fh = IO::Scalar->new( \$string );    # creating a scalar fi
+le handle.

my $count = 0;
$count++ while (<$scalar_fh> );
print $count;
[download]

This is not efficient than using tr. But, you can process a string as file by this way.

Sorry misread the thread: my replies gives the total number of lines in a FILE not in a STRING