Second, you should be able to test it yourself. In this example, I use Test::More unlike (which will have the test pass if it doesn't match the regex, and fail if it does) and throw a bunch of emails at the regex, to see which ones would be "good" emails and which would be "bad". If you have other outlier emails you wanted to test, you could add more emails into the test list

#!perl

use 5.012; # strict, //
use warnings;
use Test::More;

for (qw/
    test@gmail.com
    test.test@gmail.com
    test.test.test@gmail.com
    test.test.test.test@gmail.com
    test@subdomain.domain.com
    test.test@subdomain.domain.com
    test.test.test@subdomain.domain.com
    test@test@domain.example
    test,test@domain.example
    test@first.example,test@domain.example
/, "contains\nnewline\@fake.address")
{
    unlike $_, qr/\@.*\@|,|\..*\..*\.|\n/i, "test '$_'";
}
done_testing;
__END__
Possible attempt to separate words with commas at C:\Users\peter.jones
+\Downloads\TempData\perl\pm.pl line 18.
ok 1 - test 'test@gmail.com'
ok 2 - test 'test.test@gmail.com'
not ok 3 - test 'test.test.test@gmail.com'
#   Failed test 'test 'test.test.test@gmail.com''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test.test.test@gmail.com'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 4 - test 'test.test.test.test@gmail.com'
#   Failed test 'test 'test.test.test.test@gmail.com''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test.test.test.test@gmail.com'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
ok 5 - test 'test@subdomain.domain.com'
not ok 6 - test 'test.test@subdomain.domain.com'
#   Failed test 'test 'test.test@subdomain.domain.com''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test.test@subdomain.domain.com'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 7 - test 'test.test.test@subdomain.domain.com'
#   Failed test 'test 'test.test.test@subdomain.domain.com''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test.test.test@subdomain.domain.com'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 8 - test 'test@test@domain.example'
#   Failed test 'test 'test@test@domain.example''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test@test@domain.example'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 9 - test 'test,test@domain.example'
#   Failed test 'test 'test,test@domain.example''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test,test@domain.example'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 10 - test 'test@first.example,test@domain.example'
#   Failed test 'test 'test@first.example,test@domain.example''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'test@first.example,test@domain.example'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
not ok 11 - test 'contains
# newline@fake.address'
#   Failed test 'test 'contains
# newline@fake.address''
#   at C:\Users\peter.jones\Downloads\TempData\perl\pm.pl line 20.
#                   'contains
# newline@fake.address'
#           matches '(?^ui:\@.*\@|,|\..*\..*\.|\n)'
1..11
# Looks like you failed 8 tests of 11.
[download]

So in my examples, 3 were "good" emails and the other 8 were "bad".

Since you are the one who is defining what is and isn't a valid email (I disagree with your rules, obviously; but that is irrelevant to the technical perl question of whether you are filtering the emails that you want to filter), only you can decide whether your regex filters enough of them or not.

As others have mentioned, adding a single rule to your checks is unlikely to be an effective way of stopping spam - you may want to consider captchas, established software to prevent spam, and so on, instead of trying to roll your own.

Having said, that, regexes are also not really a good way to parse email addresses, which is why I would use Email::Address; you may also be interested in Email::Valid (which isn't perfect either, but still better than hand-rolled checks).

use warnings;
use strict;

use Email::Address;
sub countdots {
    my @addrs = Email::Address->parse(shift);
    die "expected exactly one address" unless @addrs==1;
    my $x = (my $tmp = $addrs[0]->user) =~ tr/././;
    my $y = ($tmp = $addrs[0]->host) =~ tr/././;
    return wantarray ? ($x,$y) : $x+$y;
}

use Test::More;
my @tests = (
    ['test@gmail.com', 0, 1],
    ['test.test@gmail.com', 1, 1],
    ['test.test.test@gmail.com', 2, 1],
    ['test.test.test.test@gmail.com', 3, 1],
    ['test.test.test.test.test@gmail.com', 4, 1],
    ['test@foo.bar.com', 0, 2],
    ['test.test@foo.bar.com', 1, 2],
);
plan tests => 2*@tests+1;

for my $t (@tests) {
    my ($ad, $exp_x, $exp_y) = @$t;
    my ($got_x, $got_y) = countdots($ad);
    is $got_x, $exp_x;
    is $got_y, $exp_y;
}
is countdots($tests[0][0]), $tests[0][1]+$tests[0][2];
[download]

Have you addressed the issues raised here?

• You'd be better off testing against a blacklist of "rules" (regexes or even better subs) , your approach with or'ed regexes won't be maintainable.
• You'll also need a quality management telling you which rule rejected which email address, at least a log
• If that's still your formmailer, attackers will quickly learn how to beat your rules.
• Who told you I'm Russian?

Russian hackers <...> bogus email addresses <...> more than 2 dots in an email string

Please don't do that, you'd be blocking legitimate addresses too.

Best regards, an owner of an e-mail address that goes like ${firstname}.${lastname}@${division}.${department}.msu.ru

I'm not sure your regex does what you say it does. You say it looks for two at signs and then a comma and then the periods. As I read it, it looks for two at signs OR a comma OR periods OR a return. So any string that has one of these will match. Specifically, the following match:

• ' me @ myself @ I '
• 'Hello, sailor!'
• 'Fee. Fie. Foe. Fum.'
• "\n"

That's correct. That's what I'm after.

Thanks!