split full email addresses into name and email? - php

Question

There seems to be many acceptable email address formats in the To: and From: raw email headers ...
person#place.com
person <person#place.com>
person
Another Person <person#place.com>
'Another Person' <person#place.com>
"Another Person" <person#place.com>
After not finding any effective PHP functions for splitting out names and addresses, I've written the following code.
You can DEMO IT ON CODEPAD to see the output...
// validate email address
function validate_email( $email ){
return (filter_var($email, FILTER_VALIDATE_EMAIL)) ? true : false;
}
// split email into name / address
function email_split( $str ){
$name = $email = '';
if (substr($str,0,1)=='<') {
// first character = <
$email = str_replace( array('<','>'), '', $str );
} else if (strpos($str,' <') !== false) {
// possibly = name <email>
list($name,$email) = explode(' <',$str);
$email = str_replace('>','',$email);
if (!validate_email($email)) $email = '';
$name = str_replace(array('"',"'"),'',$name);
} else if (validate_email($str)) {
// just the email
$email = $str;
} else {
// unknown
$name = $str;
}
return array( 'name'=>trim($name), 'email'=>trim($email) );
}
// test it
$tests = array(
'person#place.com',
'monarch <themonarch#tgoci.com>',
'blahblah',
"'doc venture' <doc#venture.com>"
);
foreach ($tests as $test){
echo print_r( email_split($test), true );
}
Am I missing anything here? Can anyone recommend a better way?

Answer 1

I have managed to make one regex to your test cases:
person#place.com
person <person#place.com>
person
Another Person <person#place.com>
'Another Person' <person#place.com>
"Another Person" <person#place.com>
using preg_match with this regex will surely help you bit.
function email_split( $str ){
$sPattern = "/([\w\s\'\"]+[\s]+)?(<)?(([\w-\.]+)#((?:[\w]+\.)+)([a-zA-Z]{2,4}))?(>)?/g";
preg_match($sPattern,$str,$aMatch);
if(isset($aMatch[1]))
{
echo $aMatch[1] //this is name;
}
if(isset($aMatch[3]))
{
echo $aMatch[3] //this is EmailAddress;
}
}
Note: I just noticed that single "person" i.e. your third test case could be discarded with this regex (just that because of space constraint in regex) so,at first line of your email_split function, append space at last place of your string.
Then it would be bang on target.
Thanks, Hope this helps.
Code I tried:
<?php
// validate email address
function validate_email($email) {
return (filter_var($email, FILTER_VALIDATE_EMAIL)) ? true : false;
}
// split email into name / address
function email_split($str) {
$str .=" ";
$sPattern = '/([\w\s\'\"]+[\s]+)?(<)?(([\w-\.]+)#((?:[\w]+\.)+)([a-zA-Z]{2,4}))?(>)?/';
preg_match($sPattern, $str, $aMatch);
//echo "string";
//print_r($aMatch);
$name = (isset($aMatch[1])) ? $aMatch[1] : '';
$email = (isset($aMatch[3])) ? $aMatch[3] : '';
return array('name' => trim($name), 'email' => trim($email));
}
// test it
$tests = array(
'person#place.com',
'monarch <themonarch#tgoci.com>',
'blahblah',
"'doc venture' <doc#venture.com>"
);
foreach ($tests as $test) {
echo "<pre>";
echo print_r(email_split($test), true);
echo "</pre>";
}
Output I got:
Array
(
[name] =>
[email] => person#place.com
)
Array
(
[name] => monarch
[email] => themonarch#tgoci.com
)
Array
(
[name] => blahblah
[email] =>
)
Array
(
[name] => 'doc venture'
[email] => doc#venture.com
)

Answer 2

How about this:
function email_split($str) {
$parts = explode(' ', trim($str));
$email = trim(array_pop($parts), "<> \t\n\r\0\x0B");
$name = trim(implode(' ', $parts), "\"\' \t\n\r\0\x0B");
if ($name == "" && strpos($email, "#") === false) { // only single string - did not contain '#'
$name = $email;
$email = "";
}
return array('name' => $name, 'email' => $email);
}
Looks like this is about twice as fast as the regex solution.
Note: the OPs third test case (for my purposes) is not needed. But in the interest of answering the OP I added the if stmt to produce the OPs expected results. This could have been done other ways (check the last element of $parts for '#').

Answer 3

use preg_match in php, http://php.net/manual/en/function.preg-match.php
or in my opinion, you can make your own function (let say get_email_address), it catch # character and then get the 'rest-left-string' from # until '<' character and 'rest-right-string' from # until '>' character.
for example, string monarch <themonarch#tgoci.com> will return 'rest-left-string' = themonarch and 'rest-right-string' = tgoci.com . finally, your function get_email_address will return themonarch#tgoci.com
hopefully it help.. :)

Answer 4

unfortunately the regex fails in a couple of conditions of the fullname:
non alphanumeric chars (eg. "Amazon.it")
non printable chars
emojs
i adjusted the expression this way
$sPattern = '/([^<]*)?(<)?(([\w-\.]+)#((?:[\w]+\.)+)([a-zA-Z]{2,4}))?(>)?/';
and now all chars are correctly recognized and splitted.
tested with
$address = "Test User # `` . !! 🔥 <test#email.com";
after 7 years, hope this helps :)