Re: Splitting a string ...

On 3/14/2010 9:54 PM, Ashley M. Kirchner wrote:
> I'm not a regexp person (wish I was though), and I'm hoping someone can give
> me a hand here.  Consider the following strings:
>
>
>
> -          domain\[email protected]
>
> -          domain\username
>
> -          the same as above but with / instead of \  (hey, it happens)
>
> -          [email protected]
>
> -          username
>
>
>
> Essentially I have a sign-up form where folks will be typing in their
> username.  The problem is, in our organization, when you tell someone to
> enter their username, it could end up being any of the above examples
> because they're used to a domain log in procedure where in some cases they
> type the whole thing, in other cases just the e-mail, or sometimes just the
> username.
>
>
>
> So what I'd like is a way to capture just the 'username' part, regardless of
> what other pieces they put in.  In the past I would write a rather
> inefficient split() routine and eventually get what I need.  With split()
> getting deprecated, I figured I may as well start looking into how to do it
> properly.  There's preg_split(), str_split(), explode() . possibly others.
>
>
>
> So, what's the proper way to do this?  How can I capture just the part I
> need, regardless of how they typed it in?
>
>
>
> Thanks!
>
>
>
> A
>
>

The basic problem is that the slashes are legitimate characters for usernames 
per the RFC 5322; http://en.wikipedia.org/wiki/E-mail_address

However, per the "Notwithstanding the addresses permitted by these 
standards...." paragraph, I disallow the "! # $ % * / ? ^ ` { | } ~" and have 
not found a problem.

You didn't mention whether you had control over the whole submission process. If 
so, I'd suggest checking the submitted address and sending back to the client a 
message asking them to correct their submission and resending it.

You can check for any of the above characters and bounce the submission back to 
the client.

Check out filter_var($emailAddr, FILTER_VALIDATE_EMAIL) It will catch a lot of 
errors.

Also, you'll find the helpful preg_match("%[[:alnum:][:punct:]]%", $addr); These 
are legit characters.  First remove everything following the @ and the @ itself. 
Obviously, you can use if(!preg_match....) to catch anything not valid.

There are several str functions that'll do it or simply preg_replace("%@.*%", 
'', $addr)