[Animal]

Hello,

What determines a valid (or invalid) email address? I'm writing a script which can determine whether the formatting is correct or not. For example, you need an "@" sign, a ".", etc. You can't have spaces or backslashes in an email, etc.

Does someone have a complete list of rules?

Thanks,
Steve

[cyrus]

Most java scripts that are written to check for a valid e-mail only check for the "@" and that it contains a "." that's it ;)

[zylstra]

From the CGI/Perl CookBook:

sub email_check {
    local($email) = $_[0];

    # Check that the email address doesn't have 2 @ signs, a .., a @., a 
    # .@ or begin or end with a .

    if ($email =~ /(@.*@)|(\.\.)|(@\.)|(\.@)|(^\.)|(\.$)/ || 

        # Allow anything before the @, but only letters numbers, dashes and 
        # periods after it.  Also check to make sure the address ends in 2 or 
        # three letters after a period and allow for it to be enclosed in [] 
        # such as [164.104.50.1]
    
        ($email !~ /^.+\@localhost$/ && 
         $email !~ /^.+\@\[?(\w|[-.])+\.[a-zA-Z]{2,3}|[0-9]{1,3}\]?$/)) {
        return(0);
    }

    # If it passed the above test, it is valid.
    
    else {
        return(1);
    }
}

Here's some more Perl code:

$a = "john\@example.com"; 
print &CheckEmailAddress($a); 
sub CheckEmailAddress() { 
# This regexp validates the format of an email address. It returns the cleaned 
# version of the email address, or blank if the address was not valid. 
# 
# An email address must be of the form: 
# 1) (trim any spaces or tabs or linefeeds or carriage returns) 
# 2) (possibly one quotation mark) 
# 3) (one or more characters for username, excepting a quotation mark) 
# 4) (possibly one quotation mark) 
# 5) @ 
# 6) (one or more characters for hostname(s), excepting [ <>\t]) 
# 7) . 
# 8) (two or more characters for top-level-domain, excepting [ <>\t]) 
# 9) (trim any spaces or tabs or linefeeds or carriage returns) 
# 
# 1 2 3 4 56 7 8 9 
# .............'''.......'''.'''''''''..''''''''''''''''''............. 
    $_[0] =~ /[ |\t|\r|\n]*\"?([^\"]+\"?@[^ <>\t]+\.[^ <>\t][^ <>\t]+)[ |\t|\r|\n]*/; 
    return $1; 
}

Or, if you're using Perl you can use the module Mail::RFC822::Address.

Or, if you like reading specs, it's supposed to be in RFC822, http://www.faqs.org/rfcs/rfc822.html, but I couldn't find it.

-Z