Email Extractor (newbie)

[Jellz]

I'm brand new to php, while I've read books on it, I've only recently began generating my own code.

My idea was to take each row of a column, break it up into individual words, and scan each of those words for the @ symbol. If it is found, store it in a seperate array until the process is complete, if not, proceed to next row.

I have commented out my thought process so it should be fairly easy for someone to narrow down my error, but I am fairly certain it is the way I am using the "foreach" or the "if $value == $findme" (probably both). ANY ideas or suggestions would be very greatful. (I connect using the include from a diff file, don't worry about that it's done correctly as ive used it plenty of times in other scripts).

<?php 
//convert email column to string
$result = mysql_query("SELECT email FROM users");
//create please array to store valid email addresses
$please=array();
$findme = "@";
//for each row of 'email' in table 'users'
    while($row = mysql_fetch_array($result)) {
      $address = (string)$row['email'];
//take the entire string and explode each word into array $words using the ' ' as a delimiter 
   $words=array(explode(' ', $address));
//loop over $words applying its value to $value
   foreach($words as $value) { 
//if the $value has $findme (the @)
   if($value == "$findme"){
//echo $value and push it to $please array
   echo "I have a @ in my string! <br />";
        array_push($please, "$value);

   }else{

   echo "Doesnt have @ in string! <br />"
        unset($value);
}
}
?>

[bpat1434]

The == operator will actually check for equality, not containment. If you want to see if a string has a character in it, use [man]strpos/man.

Of course the easier way would be to use [man]preg_match_all/man with a pattern like:

/ [^\s]+@[^\s]+ /Us

If you were to use preg_match, your code would turn into

<?php

$result = array(
    array('email' => 'Some random text wih an [e-mail] email address in it.'),
    array('email' => 'Hello my name is [e-mail] and I am selling you this product.  To receive said product send your details to [e-mail] who will forward your information to me and I will email you at [e-mail] and you will receive our product.')
);

$please = array();
// This is in place of while($row = mysql_fetch_array($result))
foreach($result as $row)
{
    $address = (string)$row['email'];

preg_match_all('/\s?([^\s]+@[^\s]+)\s/Us', $address, $matches);
array_shift($matches);

$please = array_merge($please, $matches[0]);
}

print_r($please);

To see it in action, visit: here

[Jellz]

Finally got it. Thanks for suggestions.

 <?php
$result = mysql_query("SELECT email FROM users");
$valid_Emails=array();
$findme = '@';
	 while($row = mysql_fetch_array($result)) {
		$address = (string)$row['email'];
		$words=array(explode(' ', $address));


	foreach ($words as $value) 
		{
		$numofrows = count($value);
		for ($i = 0; $i < $numofrows; $i++) {
			if(strpos($value[$i],$findme) === false) {
			} else {
				echo "@ was found in " . $value[$i] . "<br />";
				array_push($valid_Emails, $value[$i]);	
			}
		}	
		}
 }

 $csv_file = implode(",", $valid_Emails);
?> 

[nrg_alpha]

bpat1434;10925753 wrote:

Of course the easier way would be to use [man]preg_match_all/man with a pattern like:

/ [^\s]+@[^\s]+ /Us

Granted, both the U and s modifiers are not necessary in this case (especially the s modifier, as this only applies to dot_match_all wildcards for the sake of including new lines. But character classes can match these by default).