Menu
▾
▴
RE: [Phpwebsite-developers] Refactor of Pagemaster Search
|
From: Roland, R. M <rmr...@in...> - 2003-06-25 18:03:25
|
From Adam's database change, I went through and rewrote the Pagemaster =
search function (which I have included below). Search is a large aspect =
of what we're doing and I tried to 'googleize' the search results a =
little (it's up to you whether this is a good thing or not) to make it =
more useful to our users. My version is also streamlined to just two =
queries. It also sorts by number of matches found in the content of the =
page (ie it strips out all html except META). It also has variables for =
maximum number of results to display per page and number of characters =
to display on either side of a match. I also tried to be as =
non-invasive as possible, so I didn't make any change requirements to =
the database and didn't use any additional functions. The only external =
change is that this function requires the array of search words to be =
passed (ie in the 'mod_search_register' table, the 'search_cols' field =
for the Pagemaster record must be blank). =20
If some people can test it out and see what they think, I welcome any =
feedback.
Thanks,
Ryan Roland
Application Developer
Information Technology Department
Division of Recreational Sports
Indiana University
rmr...@in...
Office: 812.855.9617
Cell: 812.320.0032
/**
* Function used by search module to search pages
*
* @author Ryan Roland <rmr...@in...>
* @param array Array of search words (strings)
* @return array Array of strings of results of search
* @access public
*/
function search($search_array) {
// initialization variables
=20
// characters on either side of a match to display=20
$chars_surrounding_match =3D 35;
=20
// number of results to display per page
$max_results_per_page =3D 5;
=20
// create where clause
$temp_sql =3D "";
$cols_array =3D array("title","text");
for($i=3D0; $i<count($cols_array); $i++) {
for($j=3D0; $j<count($search_array); $j++) {
$temp_sql .=3D $cols_array[$i] ." LIKE '%" . $search_array[$j] . =
"%' ";
if($j<count($search_array))
{ $temp_sql .=3D "OR "; }
}
}
$where_clause =3D "WHERE (" . substr($temp_sql, 0, -3) . ")"; =20
=20
$resultArray =3D array();
// get all sections with matching string occurences
$sql =3D "SELECT * FROM " . $GLOBALS["core"]->tbl_prefix . =
"mod_pagemaster_sections $where_clause";
$result =3D $GLOBALS["core"]->getAll($sql);
if(!DB::isError($result) && is_array($result) && sizeof($result) > =
0) {
$pages =3D array();
$page_count =3D array(); =20
=20
foreach($result as $row) {
$section_title =3D strip_tags($row["title"],"<META>");
$section_text =3D strip_tags($row["text"],"<META>");
=20
$pages[$row["page_id"]][$row["id"]]["title"] =3D =
$section_title;
$pages[$row["page_id"]][$row["id"]]["text"] =3D $section_text;
// init number of matches to 0
$section_match_count =3D 0;
// loop through the words of the search criteria
// counting number of matches found in the section title and =
text=20
// this will elminate sections returned by the DB due to matches =
in HTML
foreach ($search_array as $search_word) {
$section_match_count +=3D =
substr_count(strtolower($section_title),strtolower($search_word));
$section_match_count +=3D =
substr_count(strtolower($section_text),strtolower($search_word));
}
=20
if(isset($page_count[$row["page_id"]]))=20
{ $page_count[$row["page_id"]] =3D =
$page_count[$row["page_id"]] + $section_match_count; }
else { $page_count[$row["page_id"]] =3D $section_match_count; }
}
=20
// sort array ordering by match count highest to lowest
arsort($page_count,SORT_NUMERIC);
$page_ids =3D implode(" OR ", $page_count);
$sql =3D "SELECT id,title,section_order FROM " . =
$GLOBALS["core"]->tbl_prefix . "mod_pagemaster_pages WHERE $page_ids";
$result =3D $GLOBALS["core"]->getAll($sql);
$ordered_page_results =3D $page_count;
if(!DB::isError($result) && is_array($result) && sizeof($result) > =
0) {
=20
// loop through all pages
foreach($result as $row) {
$page_results =3D array();
=20
$page_id =3D $row["id"];
$page_title =3D $row["title"];
$section_order =3D =
unserialize($row["section_order"]);
$needed_results =3D $max_results_per_page;
$pages[$page_id]["page_title"] =3D $page_title;
=20
// loop through all sections finding occurrences and parsing =
text
for ($i=3D0;$i<=3Dsizeof($section_order);$i++) {
if ($needed_results > 0) { =20
$section_id =3D $section_order[$i]; =20
$title =3D $pages[$page_id][$section_id]["title"];
$text =3D $pages[$page_id][$section_id]["text"];
=20
foreach ($search_array as $search_word) {
// Find occurrences of the search_word(s) in the =
sections' title
$prev_pos =3D 0;
// while there exists another result and we haven't =
found all needed results
while (($next_pos =3D =
strpos(strtolower($title),strtolower($search_word),$prev_pos)) && =
$needed_results > 0) {
// # of chars from the occurence to the end of the =
string
$chars_to_end =3D strlen($title) - $next_pos;
// add bold tags around the occurence in the text
$bolded =3D =
substr_replace(substr_replace($title,"<b>",$next_pos,0),"</b>",($next_pos=
+ 3) + strlen($search_word),0);
=20
// set prev_pos to the next char after the current =
position
$prev_pos =3D $next_pos + 1;
// get up to 'chars_surrounding_match' on either side =
of the occurrence,=20
// or to the begin/end of the string, whichever comes =
first
$match_result =3D "..." . substr($bolded,$next_pos - =
min($next_pos,$chars_surrounding_match), $chars_surrounding_match * 2) . =
"...";
=20
array_push($page_results,$match_result);
$needed_results--;
}// end while
// Find occurrences of the search_word(s) in the =
sections' text =20
$prev_pos =3D 0;
// while there exists another result and we haven't =
found all needed results
while (($next_pos =3D =
strpos(strtolower($text),strtolower($search_word),$prev_pos)) && =
$needed_results > 0) {
// # of chars from the occurence to the end of the =
string
$chars_to_end =3D strlen($title) - $next_pos;
// add bold tags around the occurence in the text
$bolded =3D =
substr_replace(substr_replace($text,"<b>",$next_pos,0),"</b>",($next_pos =
+ 3) + strlen($search_word),0);
=20
// set prev_pos to the next char after the current =
position
$prev_pos =3D $next_pos + 1;
// get up to 'chars_surrounding_match' on either side =
of the occurrence,=20
// or to the begin/end of the string, whichever comes =
first
$match_result =3D "..." . substr($bolded,$next_pos - =
min($next_pos,$chars_surrounding_match), $chars_surrounding_match * 2) . =
"...";
=20
array_push($page_results,$match_result);
$needed_results--;
}// end while
}// end foreach
}// end if
}// end for
$ordered_page_results[$page_id] =3D $page_results;
}// end foreach
=20
// loop through results and consolidating them to be returned
while($next_page =3D each($page_count)) {
$page_id =3D $next_page["key"];
$count =3D $next_page["value"];
=20
$count =3D=3D 1 ? $count_string =3D "match" : $count_string =
=3D "matches";
=20
$matches_string =3D "";
foreach($ordered_page_results[$page_id] as $next_match)
{ $matches_string .=3D $next_match . "<br />"; }
=20
$resultArray[$page_id] =3D "<u><b>" . =
$pages[$page_id]["page_title"] . "</b> - <b> ". $count ." =
</b> ". $count_string ." found : first <b> " . =
sizeof($ordered_page_results[$page_id]) . " </b> displayed.</u><br />" . =
$matches_string;
}// end while
}// end if
}// end if
return $resultArray;
}// END FUNC search
|
