I tried Text::Balanced the other night, however, the output wasn't useful to me. Here is the code I used...

#!/usr/bin/perl
use strict;
use warnings FATAL => qw( all );

use Text::Balanced qw(extract_bracketed);
use Data::Dumper;

my $text = 'A line with B<bold>, I<italic>, and B<I<bold and italic>> 
+text.';

my @line = extract_bracketed( $text, '<>');

print Dumper(\@line);
[download]

Here is the returned results...

$VAR1 = [
          undef,
          'A line with B<bold>, I<italic>, and B<I<bold and italic>> t
+ext.',
          undef
        ];
[download]

Either I did something horribly wrong, or it doesn't extract anything just returns the original string with undefs in an array.

Hi Aleena,

The extract_* functions are meant to operate on the start of a string, not from an arbitrary point. As mentioned in the Text::Balanced description, you may skip a prefix before the start of the balanced text, but by default this will only skip whitespace.

So if you were to change text to:

my $text = '  <bold>, I<italic>, and B<I<bold and italic>> text.';
[download]

Your output would be:

$VAR1 = [
          '<bold>',
          ', I<italic>, and B<I<bold and italic>> text.',
          '  '
        ];
[download]

Where the return is a triple of the bracketed text, the remaining string, and the prefix that was bypassed before the bracketed text was found.

If you leave your $text input as it was in your example but change the function call to consider everything preceding a < as a prefix:

my @line = extract_bracketed($text, '<>', qr(.*?(?=<)));
[download]

You'll get:

$VAR1 = [
          '<bold>',
          ', I<italic>, and B<I<bold and italic>> text.',
          'A line with B'
        ];
[download]

Where the prefix is again everything before the <. but includes the bold code at the end, which you'd have to deal with appropriately.

HTH