A couple of coding questions

Hi all,

I'm having some difficult with two pieces of code I'm writing and was hoping someone out there could help me out. Any comments or direction would be highly appreciated.

These snippets apply to a database of properties. and my first dilemna is with preg_match_all. I've always had trouble understanding it:

if (isset($_REQUEST["find"])) {
preg_match_all("/[A-Z0-9\']{1,16}/i", $_REQUEST["find"], $words);
$find = implode(" ", $words[0]); //recreate var for FULLTEXT MySQL search later
}

Worked great for a search like breakers west but if you type in breaker's west I get 3 words in my array - breaker, 's, and west. I can't seem to come up with the best way to create an array stripped of all the non alphanumeric chars.

The second one is a little tougher. I'm stuck using MySQL 3.23 and I am doing a FULLTEXT query against name and description fields. I wasn't completely happy with the relevancy of SQL results so I decided to augment the MySQL results with my own relevancy. Problem is that small common words like "a" and "the" are toasting my relevancy results. I was wondering if the best approach would be to test the array against a stopword array loaded with common English words or if you should just limit relevancy check to like words with 4 chars or greater? Or my code might just be on the wrong track completely. Snippet is as follows:

$sql = "SELECT id, name, description, MATCH(name, description) AGAINST ('".addslashes($find)."')
AS score FROM `property`
WHERE MATCH(name, description) AGAINST ('".addslashes($find)."')
AND active='1' ORDER BY score DESC"; //relevancy search against property database

$NumberOfMatches = mysql_num_rows(mysql_query("$sql", $linkID));
if ($NumberOfMatches > 0) //Check for >0 matches else print none found
{
$sqlresult = mysql_query("$sql", $linkID);
for ($x = 0; $x < $NumberOfMatches; $x++) //not using a while loop so I can have a $x for array
{
list ($id, $name, $description) = mysql_fetch_row($sqlresult); //I could have retrieved number of ways but stripslashing the results anyway
$results[$x][id] = $id;
$results[$x][name] = stripslashes($name);
$results[$x][description] = stripslashes($description);
$results[$x][relevancy] = 0;
foreach ($words[0] as $check) {
$results[$x][name] = str_replace(strtoupper($check), "".strtoupper($check)."", $results[$x][name]); //Capitalize and make bold the word match in array
$offset = 0; $position = TRUE;
while ($position!== FALSE) //Step through name string looking for match to $words[0]
{
$position = strpos(strtolower($results[$x][name]), strtolower($check), $offset);
if ($position === FALSE) {break;}
$results[$x][relevancy] += 1.4;
$offset = $position + 1;
}

$results[$x][description] = str_replace(strtoupper($check), "".strtoupper($check)."", $results[$x][description]);
$offset = 0; $position = TRUE;
while ($position!== FALSE) //Step through description string looking for match to $words[0]
{
$position = strpos(strtolower($results[$x][description]), strtolower($check), $offset);
if ($position === FALSE) {break;}
$results[$x][relevancy] += .3;
$offset = $position + 1;
}

}
}
usort($results, compare_relevancy);
}
else {echo "Sorry, I didn't find any properties matching: $find";}

I do wonder if I should even be loading an array this way from MySQL to accomplish this relevancy task. I could foresee memory issues happening if the database got large on down the road. Any help, comments, direction or criticisms on either of my dilemnas would be HIGHLY appreciated,

Thanks,

Burner

if (isset($_REQUEST["find"])) {
print $_REQUEST["find"]; exit;
preg_match_all("/[A-Z0-9\']{1,16}/i", $_REQUEST["find"], $words);
$find = implode(" ", $words[0]); //recreate var for FULLTEXT MySQL search later
}