wallyweezle has asked for the wisdom of the Perl Monks concerning the following question:

Hi, I've been tinkering with this for about an hour now trying to figure out why it won't catch intentional mistakes that I throw in. What I'm doing is trying to validate an email address, with something like this: 
if ($MailTo !~ /\@*\./) {
#set a kill flag
#print a message
}

[download]
So, to clear up the seemingly random characters up there, that's a / to open it, and escaped @ sign, a *, an escaped ., then another / to close it. Essentially I'm looking for an @ sign then a dot. On the console, this line works great, catches mistakes and everything. But when I run it through a web page as a CGI script it doesn't catch it unless I don't put anything in the field. Any ideas? Here's all of the code, just to see if anybody catches a minor mistake in my code: 
use Net::SMTP;
use CGI ':standard';

#Get parameters from http string#
my $ServerName = param('server');
my $MailFrom   = param('from');
my $MailTo     = param('to');
my $Subject    = param('subject');
my $Content    = param('content');

print "Content-type: text/html\n\n";
print "<html>\n
       <head>\n
       <title>Mail Results</title>\n
       </head>\n
       <body>\n";

my $kill = 0;

if (!($MailFrom =~/\@*\./)) {

    $kill = 1;
    print "Invalid return address in From box.
";

}

if (!($MailTo !~/\@*\./)) {
    $kill = 1;
    print "Invalid destination address.  Please notify the webmaster o
+f this problem by sending an email to 

wvhs-web\@charter.net.
";

}

if ($kill == 1) {

    die("One or more errors have occurred which prevent the script fro
+m continuing further.
");

} else {

    $smtp = Net::SMTP->new($ServerName, Debug => 0);
    die "Couldn't connet to server: $!" unless $smtp;

    $smtp->mail($MailFrom);
    $smtp->to($MailTo);
    $smtp->data();
    $smtp->datasend("To:  $MailTo\n");
    $smtp->datasend("From:  $MailFrom\n");
    $smtp->datasend("Subject:  $Subject\n");
    $smtp->datasend("\n");
    $smtp->datasend("$Content\n\n");
    $smtp->dataend();
    $smtp->quit();

    print "Mail Sent Successfully\n
\n";

}

print  "</body>\n
        </html>";

[download]
Thanks, Eric

edited by ybiC: Balanced <code> and <readmore> tags, retitle from "Regex problem"

Replies are listed 'Best First'.
Re: Regex problem
by edoc (Chaplain) on Jul 31, 2003 at 02:03 UTC

    check out Email::Valid

    if (!($MailTo !~/\@*\./)){ .. }

    you've got a double negative there, so as I read it, it says:

    if $Mailto contains zero or more @'s (\@*) followed by a '.' (\.) then $kill = 1

    or, more simply.. if $Mailto contains a '.' then $kill = 1

    cheers,

    J

[reply]
[d/l]
Re: Regex problem
by antirice (Priest) on Jul 31, 2003 at 02:04 UTC

    Your regex isn't doing what you want. If you set $MailTo to "bob@@@tom.@blah@." then it will think that you have a fully qualified e-mail address. I would recommend that you use Email::Valid and its rfc822 method.

    To figure out where your regex went wrong, check out perlre for more info.

    antirice    
    The first rule of Perl club is - use Perl
    The     ith rule of Perl club is - follow rule i - 1 for i > 1

[reply]
Re: Regex problem
by graff (Chancellor) on Jul 31, 2003 at 02:21 UTC
    Double negatives in code always confuse me a bit... You have these two conditions: 
    if (!($MailFrom =~/\@*\./)) {
    $kill = 1;
    print "Invalid return address in From box.
";
}

if (!($MailTo !~/\@*\./)) {
    $kill = 1;
    print "Invalid destination address...
";
}

    [download]
    The following pseudo-code properly "translates" this logic, so the question is: does it also express your true intent?
    • if the "from" parameter string contains at least a single period (dot) character, which optionally follows "@", then it's okay, otherwise, complain about it.
    • if the "to" parameter string does not contain a period, which optionally follows "@", then it's okay, otherwise complain about it.

    Now if the criterion is supposed to be "@" followed by period, you should not have the asterisk after "@" -- this means "zero or more occurrences" of "@" preceding a period, so both "." and "@." match your /\@*\./ regex (but neither of them looks good to me as an email address).

[reply]
[d/l]
[select]
Re: email address validation regexp
by nega (Scribe) on Jul 31, 2003 at 13:10 UTC
[reply]

Back to Seekers of Perl Wisdom