email addressing sucks

I know that most people know how much email addresses suck, but I found this message in Mutt while looking for something else, and I thought it was a nice illustration.

A user of Email::Valid writes:

I was doing some testing with Email::Valid, and I came across this:

$ perl -MEmail::Valid -e \
    "print Email::Valid->address('dave@#%^&*$.com') \
     ? 'yes' : 'no'"
yes
$

This prints “yes”, and I don’t understand why that address would be seen as valid. It doesn’t seem to matter what characters are in the domain part of the address, it always gets marked as valid.

I replied:

Let’s have a look!

dave@#%^&*$.com

You don’t think this should be valid. Let’s consult the RFC.

Section 6.1 says that an address must be . That is:

local-part "@" domain

I’m guessing you don’t dispute that “dave” is a valid local-part. A

must be: sub-domain *("." sub-domain) I'm also guessing you don't dispite that "com" is a valid sub- domain. So, is "#%^&*$"? Well, let's see... sub-domain = domain-ref / domain-literal domain-literal = "[" *(dtext / quoted-pair) "]" domain-ref = atom There are no square brackets, so there is no domain-literal involved. That means that if "#%^&*$" is a valid atom, it is also a valid domain-ref, and the address matches the addr-spec pattern. The atom pattern is: atom = 1*<any CHAR except specials, SPACE and CTLs> None of those characters are space or control characters. The specials are: specials = "(" / ")" / "<" / ">" / "@" ; Must be in quoted- / "," / ";" / ":" / "\" / <"> ; string, to use / "." / "[" / "]" None of them are specials, either! Are you shocked? Don't be! RFC822 sucks! I hope this helped. Good night! :)

Written on April 23, 2007