[UPHPU] keyword substring scoring

[Brandon Stout]

I'd like to search text strings for instances of a keyword, give the
string a score based on them, and group by the scores.  The obvious
problem is the while loop within a while loop.  This splits up the
scoring when an item has more than one keyword (I think...), but I can't
think of any other way. In addition, I'll want to group by scores
(descending).  Right now, I'm doing this:

[code]
// iterate over each item row (haystack)

while ($item = mysql_fetch_array($items, MYSQL_ASSOC)) {
 global $keywdqry;
 $keywdqry = mysql_query($keywdsql);

// iterate over each keyword (needle) within each row (haystack)

 while ($keywd = mysql_fetch_array($keywdqry, MYSQL_ASSOC)) {

// search the link field for any of the keywords, multiply by pointvalue

  $linksrch  = (
   substr_count($item['link'], $keywd['keyword'])
   * $keywd['pointvalue']
  );

// search the title field for any of the keywords, multiply by pointvalue

  $titlesrch = (
   substr_count($item['title'], $keywd['keyword'])
   * $keywd['pointvalue']
  );

// search the desc field for any of the keywords, multiply by pointvalue

  $descsrch  = (
   substr_count($item['desc'], $keywd['keyword'])
   * $keywd['pointvalue']
  );

// total the score

  $srchtotal = $linksrch + $titlesrch + $descsrch;

// print to screen any items that at least have a score (herodic)

  if ($srchtotal > 0) {
   echo  <<<str
  <li><a href="{$item['link']}" rel="external"
   title="{$item['title']}">{$item['title']}</a>
  </li>\r\n
str;
  }
 }
}
[/code]

Any ideas?

Thank you,

Brandon Stout
http://mscis.org