His problem has nothing to do with commas at all. my (@array1, @array2) = ...; will always assign all of the right side to the first array. This is just the w

His problem has nothing to do with commas at all. my (@array1, @array2) = ...; will always assign all of the right side to the first array. This is just the way perl works with assigning lists to arrays. There's a few other ways you can do this instead. # with array references my ($arr_ref1, $arr_ref2) = (index($str2, '-') != -1) ? ([split /-/,$str2], [split /;/, $str4]) : ([split /:/,$str2], [split /;/, $str4]); The above will assign two array references to each scalar variable. The other option is similar but only works in newer perls with a feature called "refaliasing" and "declared_refs", which are still considered experimental, but which are expected to be marked as stable in 5.30 (If my memory serves) use feature qw/refaliasing declared_refs/; my (\@arr1, \@arr2) = (index($str2, '-') != -1) ? ([split /-/,$str2], [split /;/, $str4]) : ([split /:/,$str2], [split /;/, $str4]); This takes the array references and assigns binds them to the arrays similar to what the author originally intended. This is a result of the same problem that perl's default argument passing has with list flattening. sub foo { my (@a, @b) = @_; # load the arguments that were passed into the subroutine ... } foo(@data, @names); That will result in both sets of values to be assigned to @a and nothing in @b, because @a ends up consuming everything. Ref-aliasing with perl's prototypes will allow you to make this work like you might expect if you aren't experienced with perl's quirks but generally when you see this pattern you should be passing a reference to begin with and not trying to use the "magic" to hide that from the callers. You can play around with these examples in an pastebin that allows running live perl code (server side), here https://perl.bot/p/gblj94 https://perl.bot/p/gblj94 EDIT: fix the code slightly, and provide interactive pastebin.