define('FILENAME', '/tmp/DLRM-Dan.csv');
define('CORRECTED_FILENAME', '/tmp/corectat.csv');
// First build a hashmap of 4.1 lexems.
$lexem41Map = array();
$dbResult = db_execute("select id, form, isLoc, concat(modelType, modelNumber, restriction) from LOC_4_1.Lexem");
while (!$dbResult->EOF) {
$lexem41Map[$dbResult->fields[0]] = array($dbResult->fields[1], intval($dbResult->fields[2]), $dbResult->fields[3]);
$dbResult->MoveNext();
}
// Next, load the corrected file (Matei provided this as an errata to Dan's file)
$lexemDlrmMap = array();
$f = fopen(CORRECTED_FILENAME, 'r');
while (($fields = fgetcsv($f)) !== false) {
$formNoAccent = locNotationToDexNotation($fields[0]);
$model = $fields[1] ? $fields[1] : 'I1';
$lexem = Lexem::get("formNoAccent = '{$formNoAccent}' and concat(modelType, modelNumber, restriction) = '{$model}'");
if ($lexem) {
$lexemDlrmMap[$lexem->id] = true;
}
}
fclose($f);
// Next, load the file and build a hashmap of DLRM lexems to keep in LOC
$f = fopen(FILENAME, 'r');
while (($fields = fgetcsv($f)) !== false) {
$formNoAccent = locNotationToDexNotation($fields[1]);
$models = preg_split('/\\s+/', trim($fields[2]));
$modelStrings = '';
foreach ($models as $m) {
if ($m) {
if ($modelStrings) {
$modelStrings .= ',';
