Re: Ereg() Function

At 23:18 23.05.2003, Daniel J. Rychlik said: --------------------[snip]-------------------- >If I wanted to check for a valid email address, basically and email that has >@ and some domain like .edu, .net or something, why wont this work ? > >if (ereg ("^[\w.-]+@[\w.-]+\.(com|net|org|edu|mil|gov|gob|info|tv|biz) >+(\.[A-Za-z]{2}|\s*)$", $_POST["emailaddy"])) { > > echo "doesnt appear to be a valid mail addy"; >} >Im not understanding why $_POST["emailaddy"] is not being evaluated >correctly. Im putting in non-valid email addresses and I expect this to >return true, but it doesnt. Any Ideas.? >-Dan --------------------[snip]-------------------- I believe it's the '.' character in your character sets above - the dot represents "any character". You need to escape this with a backslash: if (ereg ("^[\w\.-]+@[\w\.-]+\.(com|net|org|edu|mil|gov|gob|info|tv|biz)+(\.[A-Za-z]{ 2}|\s*)$", $_POST["emailaddy"])) { Note that your expression will rule out _all_ national top level domains, like "us", "ca", "br", etc... there are a lot of these, and you might want to check for a valid TLD in an extra step, if at all. I'm using preg_match to validate syntactically correct email addresses using this expression: if (!preg_match('/([a-z0-9\-\.\#\:]+@[a-z0-9\-]+\.[a-z0-9\-\.]+)/i', $value, $aresult)) { I do not know if that allows all RFC compliant addresses (including routing information that _may_ be part of the address), but didn't have any complaint yet, and have _no_ invalid email addresses in our databases so far. BTW, the preg functions are said to be faster than ereg. HTH, -- >O Ernest E. Vogelsinger (\) ICQ #13394035 ^ http://www.vogelsinger.at/