my $str ="word1,word2, word3,word4,word5";
my @arr = split /,(?=\S)/, $str;
[download]

The code above is a look-ahead; see perlre for more info (look at the chapter on "Look-Around Assertions" there).

Or if you want to split in the case of a trailing comma at the end of the input, you can also use a negative look-ahead:

my @ar = split /,(?!\s), $str;
[download]

Perl 6 - second systems done right

You don't need lookarounds. A word boundary (\b) will do the trick:

use strict;
use warnings;
use 5.010;

my $str ="word1,word2, word3,word4,word5";

my @pieces = split /\b,\b/, $str;

for my $piece (@pieces) {
    say "-->$piece<---";
}


output:
-->word1<---
-->word2, word3<---
-->word4<---
-->word5<---
[download]

A space isn't a word character, therefore it cannot be a word boundary when next to a comma.

Literally true -- but only applicable if you think no further than splitting on a comma followed by a non-whitespace character; nonsense, otherwise. You need to think through the alternatives.

• What would happen if you merely split on comma?
• What if you state the regex in split as a negative?
• Didn't you read the docs (and thereby find the methods recommended by previous responders)?