Valid email?

[Anon]

Hi guys,

I'm makin' a regex statement for validating e-mail addresses and URL's. The regex itself is no problem, but does anybody know the EXACT rules an e-mail address and an URL has to have? I mean, for example, which characters are valid before or after the "at"-sign (@) in an e-mail and MUST there be a period (.) after the "at"-sign (@)?? Stuff like that. TIA.

Greetz,

Vincent Driessen
Venlo, The Netherlands

[Anon]

I have found that before the @ sign the following
characters are allowed:

[_.'0-9a-z-]

After the @ sign, domain name rules apply, for which
the following expression is handy:

([0-9a-z][0-9a-z-]+.)+[a-z]{2,3}

Of course this does not cover the case of illegal domain
names like domain-.com, which has recently judged as
not kosher.

This the PHP expression of my liking is:

@eregi("^{[_.\'0-9a-z-]+@([0-9a-z][0-9a-z-]+.)+[a-z]{2,3}$",} $email);

Regards
Jaime

Vincent Driessen wrote:

Hi guys,

I'm makin' a regex statement for validating e-mail addresses and URL's. The regex itself is no problem, but does anybody know the EXACT rules an e-mail address and an URL has to have? I mean, for example, which characters are valid before or after the "at"-sign (@) in an e-mail and MUST there be a period (.) after the "at"-sign (@)?? Stuff like that. TIA.

Greetz,

Vincent Driessen
Venlo, The Netherlands

[spyrral]

I just posted a question on how to modify the very nice regex included in the code library. Basically, I wanted to allow the + character in the first half of the email address, such as [e-mail]. This is allowed, and is used by MS Exchange Server as a way of creating aliases. I'm using it for testing purposes. So the info you got from the previous poster may not be 100% accurate.

Spyrral

[Anon]

Try

@eregi("^{[+_.\'0-9a-z-]+@([0-9a-z][0-9a-z-]+.)+[a-z]{2,3}$",} $email);Jaime

[nocrom99]

To allow a + in the username portion of the email address use this:

eregi("^{0-9a-z@0-9a-z\.[a-z]{2,3}$",} $email);

[Anon]

Also note that addresses containting IP addresses are valid (user@164.34.34.13) as well as having any number of subdomains ([e-mail]).

You can also have the following in a local network: user@host (where there are no periods). But again, that would only be good for intra-net setups.

Don't forget some of the other characters that are allowed before the '@':
_-.+%
(i can't remember, but I thing I had also seen some of the following used as well:
!#$^&*()=~`)

I ended up making a regex that was used for a web-site verifying that what the user entered was at least in a valid form with the following:
"^{[a-z0-9-]+(%[a-z0-9-]){0,1}[(.[_a-z0-9-]+)@((a-z0-9[a-z0-9]*.)+[a-z]{2,6}|[1-2]{0,1}[0-9]{1,2}.[1-2]{0,1}[0-9]{1,2}.[1-2]{0,1}[0-9]{1,2}.[1-2]{0,1}[0-9]{1,2})$"}

It's really long, but I have not seen any addresses that were valid not go through (I was logging to verify the regex).

The pattern will let invalid IP addresses through, but that's not part of the form, but more of a "address doesn't exist" problem.