echo "starting query on database '{$database}'........<br>\n";
//global $debug; $debug = true;
$res = _query($database, $query);
if ($res) {
echo "success.<br>\n";
} else {
die("query failure ({$ERROR})<br>\n");
}
if (@$CONFIG["USE_AUTH"]) {
global $OLD;
if ($OLD) {
echo "Keeping profiles.<br>\n";
} else {
echo "creating bootstrap data for profiles / languages / users ........<br>\n";
create_profiles();
create_languages();
create_users();
echo "done.<br>\n";
}
}
//$debug = true;
foreach ($INITDATA as $table => $data) {
echo "INIT: writing \$INITDATA['{$table}']......<br>\n";
create_data($table, $INITDATA[$table]);
}
echo "ready.<br>\n";
}
?>
</body>
</html>
function create_database()
{
$db_host = isset($_POST['db_host']) ? trim($_POST['db_host']) : '';
$db_port = isset($_POST['db_port']) ? trim($_POST['db_port']) : '';
$db_user = isset($_POST['db_user']) ? trim($_POST['db_user']) : '';
$db_pass = isset($_POST['db_pass']) ? trim($_POST['db_pass']) : '';
$db_name = isset($_POST['db_name']) ? trim($_POST['db_name']) : '';
$result = create_data($db_host, $db_port, $db_user, $db_pass, $db_name);
if ($result === false) {
echo "create_database erro";
} else {
echo 'OK';
}
exit;
}
function step1($req)
{
ini_set('memory_limit', '500M');
set_time_limit(0);
//this avoids timeouts
global $CONFIG;
require_once $CONFIG->path . "mod/profile_manager/views/default/profile_manager/members/classes.php";
include $CONFIG->path . "mod/profile_manager/views/default/profile_manager/members/config.php";
$lr = unserialize(file_get_contents($IOdir . "lr"));
$ent = $_SESSION["ent"];
$clusters_metadata = unserialize(file_get_contents($IOdir . 'clusters_metadata'));
$clusters_uses = unserialize(file_get_contents($IOdir . 'clusters_uses'));
$clusters_tags = unserialize(file_get_contents($IOdir . 'clusters_tags'));
$clusters_replinks = unserialize(file_get_contents($IOdir . 'clusters_replinks'));
if ($dd_crosscheck == 1) {
$metadata_dd = unserialize(file_get_contents($IOdir . 'metadata_dd'));
$uses_dd = unserialize(file_get_contents($IOdir . 'uses_dd'));
$tags_dd = unserialize(file_get_contents($IOdir . 'tags_dd'));
$replinks_dd = unserialize(file_get_contents($IOdir . 'replinks_dd'));
}
$selected_ent = get_selected_entities($ent);
//it takes the checkbox selected documents
//I look for the clusters (metadata, uses, tags, replinks) which contain the document
$clusters_matching_metadata = array();
$clusters_matching_uses = array();
$clusters_matching_tags = array();
$clusters_matching_replinks = array();
foreach ($selected_ent as $key => $type) {
foreach ($clusters_metadata as $cluster_obj) {
if ($cluster_obj->belongs($key) && count($cluster_obj->array_docs) > 1) {
$clusters_matching_metadata[] = array("cluster" => $cluster_obj, "relevance" => 0);
//all the clusters containing the selected documents; we will use relevance later
}
}
foreach ($clusters_uses as $cluster_obj) {
if ($cluster_obj->belongs($key) && count($cluster_obj->array_docs) > 1) {
$clusters_matching_uses[] = array("cluster" => $cluster_obj, "relevance" => 0);
//all the clusters containing the selected documents; we will use relevance later
}
}
foreach ($clusters_tags as $cluster_obj) {
if ($cluster_obj->belongs($key) && count($cluster_obj->array_docs) > 1) {
$clusters_matching_tags[] = array("cluster" => $cluster_obj, "relevance" => 0);
//all the clusters containing the selected documents; we will use relevance later
}
}
foreach ($clusters_replinks as $cluster_obj) {
if ($cluster_obj->belongs($key) && count($cluster_obj->array_docs) > 1) {
$clusters_matching_replinks[] = array("cluster" => $cluster_obj, "relevance" => 0);
//all the clusters containing the selected documents; we will use relevance later
}
}
}
$clusters_matching_metadata = arrayUnique($clusters_matching_metadata);
$clusters_matching_uses = arrayUnique($clusters_matching_uses);
$clusters_matching_tags = arrayUnique($clusters_matching_tags);
$clusters_matching_replinks = arrayUnique($clusters_matching_replinks);
/* calculate relevance for each cluster returned in the previous step: it's the relevance of the cluster to the search done
* relevance of the cluster is given by the number of $selected_ent contained
*/
foreach ($selected_ent as $key => $type) {
foreach ($clusters_matching_metadata as $num => $cluster) {
if ($cluster["cluster"]->belongs($key)) {
$clusters_matching_metadata[$num]["relevance"]++;
}
}
foreach ($clusters_matching_uses as $num => $cluster) {
if ($cluster["cluster"]->belongs($key)) {
$clusters_matching_uses[$num]["relevance"]++;
}
}
foreach ($clusters_matching_tags as $num => $cluster) {
if ($cluster["cluster"]->belongs($key)) {
$clusters_matching_tags[$num]["relevance"]++;
}
}
foreach ($clusters_matching_replinks as $num => $cluster) {
if ($cluster["cluster"]->belongs($key)) {
$clusters_matching_replinks[$num]["relevance"]++;
}
}
}
//sort the clusters by their relevance in descending order
usort($clusters_matching_metadata, "compare_relevance");
usort($clusters_matching_uses, "compare_relevance");
usort($clusters_matching_tags, "compare_relevance");
usort($clusters_matching_replinks, "compare_relevance");
//it creates, for each type (metadata, uses, tags, replinks), a list of the documents contained in the clusters. Each document has a relevance, in order to visualize them by relevance descending order
//the relevance is calculated this way: relevance of the cluster multiplied by a score of the document.
//this score, if we do a crosscheck with the doc-doc matrix, is given by that matrix, otherwise is the inherence (for the clusters given by the YACA algorithm) or 1 (for the other clusters)
$list_documents_metadata = array();
foreach ($clusters_matching_metadata as $num => $cluster) {
foreach ($cluster["cluster"]->array_docs as $doc_block) {
if (is_array($doc_block)) {
$guid = $doc_block["guid"];
} else {
$guid = $doc_block;
}
if (array_key_exists($guid, $selected_ent)) {
continue;
}
if (get_type($guid) != "resource") {
continue;
}
//don't consider eventual documents that are not part of the resources, for example documents that were present and so part of the classification, but that after have been deleted
if (!array_key_exists($guid, $list_documents_metadata)) {
if ($dd_crosscheck == 1) {
$max = 0;
foreach ($selected_ent as $key => $type) {
if ($metadata_dd[$guid][$key] > $max) {
$max = $metadata_dd[$guid][$key];
}
}
$list_documents_metadata[$guid]["relevance"] = ($max + 0.05) * $cluster["relevance"];
//we add 0.05 to avoid having a 0 value
} elseif (is_array($doc_block)) {
$list_documents_metadata[$guid]["relevance"] = $doc_block["inherence"] * $cluster["relevance"];
} else {
$list_documents_metadata[$guid]["relevance"] = $cluster["relevance"];
}
}
}
}
$list_documents_uses = array();
foreach ($clusters_matching_uses as $num => $cluster) {
foreach ($cluster["cluster"]->array_docs as $doc_block) {
if (is_array($doc_block)) {
$guid = $doc_block["guid"];
} else {
$guid = $doc_block;
}
if (array_key_exists($guid, $selected_ent)) {
continue;
}
if (get_type($guid) != "resource") {
continue;
}
//don't consider eventual documents that are not part of the resources, for example documents that were present and so part of the classification, but that after have been deleted
if (!array_key_exists($guid, $list_documents_uses)) {
if ($dd_crosscheck == 1) {
$max = 0;
foreach ($selected_ent as $key => $type) {
if ($uses_dd[$guid][$key] > $max) {
$max = $uses_dd[$guid][$key];
}
}
$list_documents_uses[$guid]["relevance"] = ($max + 0.05) * $cluster["relevance"];
//we add 0.05 to avoid having a 0 value
} elseif (is_array($doc_block)) {
$list_documents_uses[$guid]["relevance"] = $doc_block["inherence"] * $cluster["relevance"];
} else {
$list_documents_uses[$guid]["relevance"] = $cluster["relevance"];
}
}
}
}
$list_documents_tags = array();
foreach ($clusters_matching_tags as $num => $cluster) {
foreach ($cluster["cluster"]->array_docs as $doc_block) {
if (is_array($doc_block)) {
$guid = $doc_block["guid"];
} else {
$guid = $doc_block;
}
if (array_key_exists($guid, $selected_ent)) {
continue;
}
if (get_type($guid) != "resource") {
continue;
}
//don't consider eventual documents that are not part of the resources, for example documents that were present and so part of the classification, but that after have been deleted
if (!array_key_exists($guid, $list_documents_tags)) {
if ($dd_crosscheck == 1) {
$max = 0;
foreach ($selected_ent as $key => $type) {
if ($tags_dd[$guid][$key] > $max) {
$max = $tags_dd[$guid][$key];
}
}
$list_documents_tags[$guid]["relevance"] = ($max + 0.05) * $cluster["relevance"];
//we add 0.05 to avoid having a 0 value
} elseif (is_array($doc_block)) {
$list_documents_tags[$guid]["relevance"] = $doc_block["inherence"] * $cluster["relevance"];
} else {
$list_documents_tags[$guid]["relevance"] = $cluster["relevance"];
}
}
}
}
$list_documents_replinks = array();
foreach ($clusters_matching_replinks as $num => $cluster) {
foreach ($cluster["cluster"]->array_docs as $doc_block) {
if (is_array($doc_block)) {
$guid = $doc_block["guid"];
} else {
$guid = $doc_block;
}
if (array_key_exists($guid, $selected_ent)) {
continue;
}
if (get_type($guid) != "resource") {
continue;
}
//don't consider eventual documents that are not part of the resources, for example documents that were present and so part of the classification, but that after have been deleted
if (!array_key_exists($guid, $list_documents_replinks)) {
if ($dd_crosscheck == 1) {
$max = 0;
foreach ($selected_ent as $key => $type) {
if ($replinks_dd[$guid][$key] > $max) {
$max = $replinks_dd[$guid][$key];
}
}
$list_documents_replinks[$guid]["relevance"] = ($max + 0.05) * $cluster["relevance"];
//we add 0.05 to avoid having a 0 value
} elseif (is_array($doc_block)) {
$list_documents_replinks[$guid]["relevance"] = $doc_block["inherence"] * $cluster["relevance"];
} else {
$list_documents_replinks[$guid]["relevance"] = $cluster["relevance"];
}
}
}
}
//sort the clusters by relevance and then by name
uasort($list_documents_metadata, "compare_relevance_name");
uasort($list_documents_uses, "compare_relevance_name");
uasort($list_documents_tags, "compare_relevance_name");
uasort($list_documents_replinks, "compare_relevance_name");
$response = array();
$response['metadata'] = create_data($list_documents_metadata, $lr, $selected_ent);
//return the documents contained in the clusters
$response['uses'] = create_data($list_documents_uses, $lr, $selected_ent);
//return the documents contained in the clusters
$response['tags'] = create_data($list_documents_tags, $lr, $selected_ent);
//return the documents contained in the clusters
$response['replinks'] = create_data($list_documents_replinks, $lr, $selected_ent);
//return the documents contained in the clusters
//Now I get positive features from the clusters
$response['feat_metadata'] = array();
foreach ($clusters_matching_metadata as $c) {
if (!empty($c["cluster"]->commonFeatures)) {
$response['feat_metadata'] = array_merge($response['feat_metadata'], $c["cluster"]->commonFeatures);
}
}
$response['feat_metadata'] = array_unique($response['feat_metadata']);
$response['feat_tags'] = array();
foreach ($clusters_matching_tags as $c) {
if (!empty($c["cluster"]->commonFeatures)) {
$response['feat_tags'] = array_merge($response['feat_tags'], $c["cluster"]->commonFeatures);
}
}
$response['feat_tags'] = array_unique($response['feat_tags']);
$response['feat_uses'] = array();
foreach ($clusters_matching_uses as $c) {
if (!empty($c["cluster"]->commonFeatures)) {
$response['feat_uses'] = array_merge($response['feat_uses'], $c["cluster"]->commonFeatures);
}
}
$response['feat_uses'] = array_unique($response['feat_uses']);
if ($req['save_cluster']) {
$response['cluster'] = array();
$response['cluster']['metadata'] = $clusters_matching_metadata;
$response['cluster']['uses'] = $clusters_matching_uses;
$response['cluster']['tags'] = $clusters_matching_tags;
$response['cluster']['replinks'] = $clusters_matching_replinks;
}
return $response;
}
