function get_search_results($query, $start, $category, $searchtype, $results, $domain)
{
global $sph_messages, $results_per_page, $links_to_next, $show_query_scores, $desc_length;
if ($results != "") {
$results_per_page = $results;
}
if ($searchtype == "phrase") {
$query = str_replace('"', '', $query);
$query = "\"" . $query . "\"";
}
$starttime = getmicrotime();
// catch " if only one time entered
$query = preg_replace("/"/", "\"", $query);
if (substr_count($query, '"') == 1) {
$query = str_replace('"', '', $query);
}
$words = makeboollist($query);
if (isset($words['ignore'])) {
$ignorewords = $words['ignore'];
} else {
$ignorewords = "";
}
$full_result['ignore_words'] = $ignorewords;
if ($start == 0) {
$start = 1;
}
$result = search($words, $category, $start, $results_per_page, $searchtype, $domain);
$query = stripslashes($query);
$entitiesQuery = htmlspecialchars($query);
$full_result['ent_query'] = $entitiesQuery;
$endtime = getmicrotime() - $starttime;
if (isset($result['results'])) {
$rows = $result['results'];
} else {
$rows = "";
}
$time = round($endtime * 100) / 100;
$full_result['time'] = $time;
$did_you_mean = array();
$did_you_mean_b = array();
if (isset($result['did_you_mean']) && is_array($result['did_you_mean'])) {
while (list($key, $alt) = each($result['did_you_mean'])) {
$entities = html_to_latin1(utf8_decode($entitiesQuery));
if ($key != $alt) {
$alt = html_to_latin1(utf8_decode($alt));
$alt = sanitize($alt);
$entities = preg_replace("/"/", "\"", $entities);
if ($alt[0] == "/") {
/* this indicates that the search word is not found and there is no close alternative either */
$alt = substr($alt, 1);
$did_you_mean_b[] = latin1_to_html(str_ireplace($key, "<strike>{$alt}</strike>", $entities));
$did_you_mean[] = str_ireplace($key, "", $entities);
} else {
$did_you_mean_b[] = latin1_to_html(str_ireplace($key, "<b>{$alt}</b>", $entities));
$did_you_mean[] = str_ireplace($key, utf8_encode($alt), $entities);
}
}
}
}
$full_result['did_you_mean'] = $did_you_mean;
$full_result['did_you_mean_b'] = $did_you_mean_b;
$matchword = $sph_messages["matches"];
if ($rows == 1) {
$matchword = $sph_messages["match"];
}
$num_of_results = count($result) - 2;
$full_result['num_of_results'] = $num_of_results;
if ($start < 2) {
saveToLog($query, $time, $rows);
}
$from = ($start - 1) * $results_per_page + 1;
$to = min($start * $results_per_page, $rows);
$full_result['from'] = $from;
$full_result['to'] = $to;
$full_result['total_results'] = $rows;
if ($rows > 0) {
$maxweight = $result['maxweight'];
$i = 0;
while ($i < $num_of_results && $i < $results_per_page) {
if (!isset($result[$i]['url'])) {
$i++;
continue;
}
$url = $result[$i]['url'];
$title = isset($result[$i]['title']) ? $result[$i]['title'] : "";
$summary = $result[$i]['summary'];
$lang = $result[$i]['lang'];
$fulltxt = $result[$i]['fulltxt'];
$page_size = $result[$i]['size'];
$domain = $result[$i]['domain'];
if ($page_size != "") {
$page_size = number_format($page_size, 1) . "kb";
}
$txtlen = strlen($fulltxt);
if ($txtlen > $desc_length) {
$places = array();
foreach ($words['hilight'] as $word) {
$word = latin1_to_html($word);
$tmp = strtolower($fulltxt);
$found_in = strpos($tmp, $word);
$sum = -strlen($word);
while (!($found_in == '')) {
$pos = $found_in + strlen($word);
$sum += $pos;
//FIX!!
$tmp = substr($tmp, $pos);
$places[] = $sum;
$found_in = strpos($tmp, $word);
}
}
sort($places);
$x = 0;
$begin = 0;
$end = 0;
while (list($id, $place) = each($places)) {
while (isset($places[$id + $x]) && $places[$id + $x] - $place < $desc_length && $x + $id < count($places) && $place < strlen($fulltxt) - $desc_length) {
$x++;
$begin = $id;
$end = $id + $x;
}
}
if (!isset($places[$begin])) {
$places[$begin] = 0;
}
$begin_pos = max(0, $places[$begin] - 30);
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
if ($places[$begin] > 0) {
$begin_pos = strpos($fulltxt, " ");
}
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
$fulltxt = substr($fulltxt, 0, strrpos($fulltxt, " "));
$fulltxt = $fulltxt;
}
$weight = number_format($result[$i]['weight'] / $maxweight * 100, 2);
if ($title == '') {
/* for an untitled document, use the filename without the path */
$pos = strrpos($url, "/");
if ($pos >= 0) {
$pos++;
} else {
$pos = 0;
}
$title = substr($url, $pos);
}
$regs = array();
if (strlen($title) > 80) {
$title = substr($title, 0, 76) . "...";
}
foreach ($words['hilight'] as $change) {
$change = latin1_to_html($change);
$count = 0;
while (preg_match("/[ .,;\\(\\)\\'\"](" . $change . ")[ .,;\\(\\)\\'\"]/i", " " . $title . " ", $regs) && ++$count < 20) {
$title = preg_replace("/([ .,;\\(\\)\\'\"])" . $regs[1] . "([ .,;\\(\\)\\'\"])/i", "\$1<b>" . $regs[1] . "</b>\$2", $title);
}
$count = 0;
while (preg_match("/[ .,;\\(\\)\\'\"](" . $change . ")[ .,\\(\\)\\'\"]/i", " " . $fulltxt . " ", $regs) && ++$count < 20) {
$fulltxt = preg_replace("/([ .,;\\(\\)\\'\"])" . $regs[1] . "([ .,;\\(\\)\\'\"])/i", "\$1<b>" . $regs[1] . "</b>\$2", $fulltxt);
}
}
$num = $from + $i;
$full_result['qry_results'][$i]['num'] = $num;
$full_result['qry_results'][$i]['weight'] = $weight;
$full_result['qry_results'][$i]['url'] = $url;
$full_result['qry_results'][$i]['title'] = $title;
$full_result['qry_results'][$i]['summary'] = $summary;
$full_result['qry_results'][$i]['lang'] = $lang;
$full_result['qry_results'][$i]['fulltxt'] = $fulltxt;
$full_result['qry_results'][$i]['page_size'] = $page_size;
$full_result['qry_results'][$i]['domain_name'] = $domain;
$i++;
}
}
$pages = ceil($rows / $results_per_page);
$full_result['pages'] = $pages;
$prev = $start - 1;
$full_result['prev'] = $prev;
$next = $start + 1;
$full_result['next'] = $next;
$full_result['start'] = $start;
$full_result['query'] = $entitiesQuery;
if ($from <= $to) {
$firstpage = $start - $links_to_next;
if ($firstpage < 1) {
$firstpage = 1;
}
$lastpage = $start + $links_to_next;
if ($lastpage > $pages) {
$lastpage = $pages;
}
for ($x = $firstpage; $x <= $lastpage; $x++) {
$full_result['other_pages'][] = $x;
}
}
return $full_result;
}
function get_search_results($query, $start, $category, $searchtype, $results, $domain)
{
global $sph_messages, $results_per_page, $links_to_next, $show_query_scores, $mysql_table_prefix, $desc_length;
if ($results != "") {
$results_per_page = $results;
}
if ($searchtype == "phrase") {
$query = str_replace('"', '', $query);
$query = "\"" . $query . "\"";
}
$starttime = getmicrotime();
// catch " if only one time entered
if (substr_count($query, '"') == 1) {
$query = str_replace('"', '', $query);
}
$words = makeboollist($query);
$ignorewords = $words['ignore'];
$full_result['ignore_words'] = $words['ignore'];
if ($start == 0) {
$start = 1;
}
$result = search($words, $category, $start, $results_per_page, $searchtype, $domain);
$query = stripslashes($query);
$entitiesQuery = htmlspecialchars($query);
$full_result['ent_query'] = $entitiesQuery;
$endtime = getmicrotime() - $starttime;
$rows = $result['results'];
$time = round($endtime * 100) / 100;
$full_result['time'] = $time;
$did_you_mean = "";
if (isset($result['did_you_mean'])) {
$did_you_mean_b = $entitiesQuery;
$did_you_mean = $entitiesQuery;
while (list($key, $val) = each($result['did_you_mean'])) {
if ($key != $val) {
$did_you_mean_b = str_replace($key, "<b>{$val}</b>", $did_you_mean_b);
$did_you_mean = str_replace($key, "{$val}", $did_you_mean);
}
}
}
$full_result['did_you_mean'] = $did_you_mean;
$full_result['did_you_mean_b'] = $did_you_mean_b;
$matchword = $sph_messages["matches"];
if ($rows == 1) {
$matchword = $sph_messages["match"];
}
$num_of_results = count($result) - 2;
$full_result['num_of_results'] = $num_of_results;
if ($start < 2) {
saveToLog(addslashes($query), $time, $rows);
}
$from = ($start - 1) * $results_per_page + 1;
$to = min($start * $results_per_page, $rows);
$full_result['from'] = $from;
$full_result['to'] = $to;
$full_result['total_results'] = $rows;
if ($rows > 0) {
$maxweight = $result['maxweight'];
$i = 0;
while ($i < $num_of_results && $i < $results_per_page) {
$title = $result[$i]['title'];
$url = $result[$i]['url'];
$fulltxt = $result[$i]['fulltxt'];
$page_size = $result[$i]['size'];
$domain = $result[$i]['domain'];
if ($page_size != "") {
$page_size = number_format($page_size, 1) . "kb";
}
$txtlen = strlen($fulltxt);
if ($txtlen > $desc_length) {
$places = array();
foreach ($words['hilight'] as $word) {
$tmp = strtolower($fulltxt);
$found_in = strpos($tmp, $word);
$sum = -strlen($word);
while (!($found_in == '')) {
$pos = $found_in + strlen($word);
$sum += $pos;
//FIX!!
$tmp = substr($tmp, $pos);
$places[] = $sum;
$found_in = strpos($tmp, $word);
}
}
sort($places);
$x = 0;
$begin = 0;
$end = 0;
while (list($id, $place) = each($places)) {
while ($places[$id + $x] - $place < $desc_length && $x + $id < count($places) && $place < strlen($fulltxt) - $desc_length) {
$x++;
$begin = $id;
$end = $id + $x;
}
}
$begin_pos = max(0, $places[$begin] - 30);
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
if ($places[$begin] > 0) {
$begin_pos = strpos($fulltxt, " ");
}
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
$fulltxt = substr($fulltxt, 0, strrpos($fulltxt, " "));
$fulltxt = $fulltxt;
}
$weight = number_format($result[$i]['weight'] / $maxweight * 100, 2);
if ($title == '') {
$title = $sph_messages["Untitled"];
}
$regs = array();
if (strlen($title) > 80) {
$title = substr($title, 0, 76) . "...";
}
foreach ($words['hilight'] as $change) {
while (preg_match("/[^\\>](" . $change . ")[^\\<]/i", " " . $title . " ", $regs)) {
$title = preg_replace("/" . $regs[1] . "/i", "<b>" . $regs[1] . "</b>", $title);
}
while (preg_match("/[^\\>](" . $change . ")[^\\<]/i", " " . $fulltxt . " ", $regs)) {
$fulltxt = preg_replace("/" . $regs[1] . "/i", "<b>" . $regs[1] . "</b>", $fulltxt);
}
$url2 = $url;
while (preg_match("/[^\\>](" . $change . ")[^\\<]/i", $url2, $regs)) {
$url2 = preg_replace("/" . $regs[1] . "/i", "<b>" . $regs[1] . "</b>", $url2);
}
}
$num = $from + $i;
$full_result['qry_results'][$i]['num'] = $num;
$full_result['qry_results'][$i]['weight'] = $weight;
$full_result['qry_results'][$i]['url'] = $url;
$full_result['qry_results'][$i]['title'] = $title;
$full_result['qry_results'][$i]['fulltxt'] = $fulltxt;
$full_result['qry_results'][$i]['url2'] = $url2;
$full_result['qry_results'][$i]['page_size'] = $page_size;
$full_result['qry_results'][$i]['domain_name'] = $domain;
$i++;
}
}
$pages = ceil($rows / $results_per_page);
$full_result['pages'] = $pages;
$prev = $start - 1;
$full_result['prev'] = $prev;
$next = $start + 1;
$full_result['next'] = $next;
$full_result['start'] = $start;
$full_result['query'] = $entitiesQuery;
if ($from <= $to) {
$firstpage = $start - $links_to_next;
if ($firstpage < 1) {
$firstpage = 1;
}
$lastpage = $start + $links_to_next;
if ($lastpage > $pages) {
$lastpage = $pages;
}
for ($x = $firstpage; $x <= $lastpage; $x++) {
$full_result['other_pages'][] = $x;
}
}
return $full_result;
}
function get_search_results($query, $start, $category, $searchtype, $results, $domain)
{
global $sph_messages, $results_per_page, $all_wild, $show_meta_description, $title_length, $links_to_next, $wildsearch, $show_warning, $mark, $type, $show_query_scores, $index_host, $url_length, $query_hits, $mysql_table_prefix, $desc_length, $utf8, $case_sensitive;
if ($results != "") {
$results_per_page = $results;
}
if ($searchtype == "phrase") {
$query = str_replace('"', '', $query);
$query = "\"" . $query . "\"";
}
if ($utf8 == 1 && $case_sensitive == 0 && $searchtype != "phrase") {
$query = lower_case($query);
}
$starttime = getmicrotime();
// catch " if only one time entered
if (substr_count($query, '"') == 1) {
$query = str_replace('"', '', $query);
}
$words = makeboollist($query);
$ignorewords = $words['ignore'];
$full_result['ignore_words'] = $words['ignore'];
if ($start == 0) {
$start = 1;
}
$result = search($words, $category, $start, $results_per_page, $searchtype, $domain);
$query = stripslashes($query);
$entitiesQuery = htmlspecialchars(str_replace("\"", "", $query));
$full_result['ent_query'] = $entitiesQuery;
$endtime = getmicrotime() - $starttime;
$rows = $result['results'];
$time = round($endtime * 100) / 100;
$full_result['time'] = $time;
$did_you_mean = "";
if (isset($result['did_you_mean'])) {
$did_you_mean_b = $entitiesQuery;
$did_you_mean = $entitiesQuery;
while (list($key, $val) = each($result['did_you_mean'])) {
if ($key != $val) {
$did_you_mean_b = str_replace($key, "<b>{$val}</b>", $did_you_mean_b);
$did_you_mean = str_replace($key, "{$val}", $did_you_mean);
}
}
}
$full_result['did_you_mean'] = $did_you_mean;
$full_result['did_you_mean_b'] = $did_you_mean_b;
$matchword = $sph_messages["matches"];
if ($rows == 1) {
$matchword = $sph_messages["match"];
}
$num_of_results = count($result) - 2;
$full_result['num_of_results'] = $num_of_results;
if ($start < 2) {
saveToLog(addslashes($query), $time, $rows);
}
$from = ($start - 1) * $results_per_page + 1;
$to = min($start * $results_per_page, $rows);
$full_result['from'] = $from;
$full_result['to'] = $to;
$full_result['total_results'] = $rows;
if ($rows > 0) {
$maxweight = $result['maxweight'];
$i = 0;
while ($i < $num_of_results && $i < $results_per_page) {
$title = $result[$i]['title'];
$url = $result[$i]['url'];
$fulltxt = $result[$i]['fulltxt'];
$page_size = $result[$i]['size'];
$domain = $result[$i]['domain'];
if ($page_size != "") {
$page_size = number_format($page_size, 1) . " kb";
}
// If available, enable part of a word highlighting in result report
if ($all_wild) {
$words = makeboollist($all_wild);
}
$txtlen = strlen($fulltxt);
//$refreshed = ereg_replace("[*!]", '',trim($query)); // works also for *wildcard search
if ($show_meta_description === 1 || $txtlen > $desc_length) {
$places = array();
$strictpos = strpos($query, '!');
if ($strictpos === 0) {
// if !strict search enter here
if ($case_sensitive == '1') {
$recovered = str_replace('!', '', trim($query));
$tmp = $fulltxt;
} else {
$recovered = str_replace('!', '', trim(lower_case($query)));
$tmp = lower_case($fulltxt);
}
$words['hilight'][0] = "{$recovered}";
// replace without ' ! '
$strict_length = strlen($recovered);
$found_in = '1';
// pointer position start
$pos_absolut = '0';
foreach ($words['hilight'] as $word) {
while (!($found_in == '')) {
$found_in = strpos($tmp, $word);
$tmp_front = substr($tmp, $found_in - 1);
// one character before found match position
$pos = $found_in + strlen($word);
$pos_absolut = $pos_absolut + $found_in;
$tmp = substr($tmp, $pos);
// get rest of fulltxt
// check weather found match is realy strict
$found_before = preg_match("/[(a-z)-_*.\\/\\:&@\\w]/", substr($tmp_front, 0, 1));
$found_behind = preg_match("/[(a-z)-_*.,\\/\\:&@\\w]/", substr($tmp, 0, 1));
if ($found_before === 0 && $found_behind === 0) {
$places[] = $pos_absolut;
// remind absolut position of match
$found_in = '';
}
}
}
} else {
// if not !strict search enter here (standard search)
foreach ($words['hilight'] as $word) {
if ($case_sensitive == '0') {
$tmp = lower_case($fulltxt);
$word = lower_case($word);
} else {
$tmp = $fulltxt;
}
$found_in = strpos($tmp, $word);
$sum = -strlen($word);
while (!($found_in == '')) {
$pos = $found_in + strlen($word);
$sum += $pos;
//FIX!!
$tmp = substr($tmp, $pos);
$places[] = $sum;
$found_in = strpos($tmp, $word);
}
}
}
sort($places);
$x = 0;
$begin = 0;
$end = 0;
while (list($id, $place) = each($places)) {
while ($places[$id + $x] - $place < $desc_length && $x + $id < count($places) && $place < strlen($fulltxt) - $desc_length) {
$x++;
$begin = $id;
$end = $id + $x;
}
}
$begin_pos = max(0, $places[$begin] - 80);
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
if ($places[$begin] > 0) {
$begin_pos = strpos($fulltxt, " ");
}
$fulltxt = substr($fulltxt, $begin_pos, $desc_length);
$fulltxt = substr($fulltxt, 0, strrpos($fulltxt, " "));
$fulltxt = $fulltxt;
}
if ($query_hits == '0') {
// calculate percentage of weight
$weight = number_format($result[$i]['weight'] / $maxweight * 100, 1);
} else {
$weight = number_format($result[$i]['weight']);
}
if ($title == '') {
$title = $sph_messages["Untitled"];
}
$regs = array();
if (strlen($title) > $title_length) {
// if necessary shorten length of title in result page
$length_tot = strpos($title, " ", $title_length);
// find end of last word for shortened title
if ($length_tot) {
$title = substr($title, 0, $length_tot) . " ...";
}
}
$url2 = $url;
if (strlen($url) > $url_length) {
// if necessary shorten length of URL in result page
$url2 = substr($url, 0, $url_length) . "...";
}
if ($places[0] == '' && $query_hits == 1 && $type != 'tol') {
// if nothing found in HTML text and query hits as result output
$weight = '0';
}
if ($places[0] == '' && $show_warning == '1' && $type != 'tol' || $show_warning == '1' && $weight == '0') {
// if no HTML text to highlight
$warnmessage = $sph_messages['showWarning'];
$fulltxt = "<span class='warn'>{$warnmessage}</span>";
}
if ($mark == 'markbold') {
$highlight = "span class='mak_1'";
}
if ($mark == 'markblue') {
$highlight = "span class='mak_2'";
}
if ($mark == 'markyellow') {
$highlight = "span class='mak_3'";
}
if ($mark == 'markgreen') {
$highlight = "span class='mak_4'";
}
foreach ($words['hilight'] as $change) {
if (!($strictpos === 0)) {
// no marking in title and url if strict search
if ($case_sensitive == '1') {
// if we have to search case sensetive, enter here
while (@ereg("[^\\>](" . $change . ")[^\\<]", " " . $title . " ", $regs)) {
$title = ereg_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $title);
}
if ($index_host == '1') {
while (@ereg("[^\\>](" . $change . ")[^\\<]", $url2, $regs)) {
$url2 = ereg_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $url2);
}
}
} else {
// mark upper and lower case match
while (@eregi("[^\\>](" . $change . ")[^\\<]", " " . $title . " ", $regs)) {
$title = eregi_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $title);
}
if ($index_host == '1') {
while (@eregi("[^\\>](" . $change . ")[^\\<]", $url2, $regs)) {
$url2 = eregi_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $url2);
}
}
}
}
if ($strictpos === 0) {
// if strict search mark only the real result with blanks before and behind
$change = " {$change} ";
}
if ($case_sensitive == '1') {
// mark fulltext case sensitive
while (@ereg("[^\\>](" . $change . ")[^\\<]", " " . $fulltxt . " ", $regs)) {
$fulltxt = ereg_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $fulltxt);
}
} else {
// mark all in fulltext
while (@eregi("[^\\>](" . $change . ")[^\\<]", " " . $fulltxt . " ", $regs)) {
$fulltxt = eregi_replace($regs[1], "<{$highlight}>" . $regs[1] . "</span>", $fulltxt);
}
}
}
$places = array();
$num = $from + $i;
$full_result['qry_results'][$i]['num'] = $num;
$full_result['qry_results'][$i]['weight'] = $weight;
$full_result['qry_results'][$i]['url'] = $url;
$full_result['qry_results'][$i]['title'] = $title;
$full_result['qry_results'][$i]['fulltxt'] = $fulltxt;
$full_result['qry_results'][$i]['url2'] = $url2;
$full_result['qry_results'][$i]['page_size'] = $page_size;
$full_result['qry_results'][$i]['domain_name'] = $domain;
$i++;
}
}
$pages = ceil($rows / $results_per_page);
$full_result['pages'] = $pages;
$prev = $start - 1;
$full_result['prev'] = $prev;
$next = $start + 1;
$full_result['next'] = $next;
$full_result['start'] = $start;
$full_result['query'] = $entitiesQuery;
if ($from <= $to) {
$firstpage = $start - $links_to_next;
if ($firstpage < 1) {
$firstpage = 1;
}
$lastpage = $start + $links_to_next;
if ($lastpage > $pages) {
$lastpage = $pages;
}
for ($x = $firstpage; $x <= $lastpage; $x++) {
$full_result['other_pages'][] = $x;
}
}
return $full_result;
}
