Open Beelden data loader

ob.php

<?php // written by AJ Slaghuis (c) 2013

$input = 'FIN-180912-3681a_hd'; // the video name we are looking for
// let's not bang too much on the openbeelden server
if (!file_exists('download.xml')) {
    $file = file_get_contents('http://www.openbeelden.nl/feeds/oai/?verb=ListRecords&metadataPrefix=oai_dc&from=2012-01-01&until=2012-12-31');
    file_put_contents('download.xml', $file);
}
$xml = simplexml_load_file('download.xml'); // print_r(scandir('.'));

$namespaces = $xml->getNamespaces(true); //print_r($namespaces);
$records = $xml->ListRecords->record;
foreach ($records as $record) {
    // $record->header->identifier, datestamp and setSpec
    //   print_r($record->header);

    $data = $record->metadata->children($namespaces['oai_dc']);
    $rows = $data->children($namespaces['dc']);

    /* print_r($rows), $rows->title, creator, subject, description, publisher, contributor,
     * date, type, format[i], identifier, source, relation[?], language, coverage[i], rights
     */

    /* use this for debugging your XML marker and checking namespaces
      $rows->yourpointer->asXML('php://output');
     */
    foreach ($rows->title as $title) {
        if ($title == $input) {
            print($title . "\n");
            //print_r($rows->format);
            $uris = (array) $rows->format;
            //print_r($uris);
            foreach ($uris as $uri) {
                $arr = explode('.', $uri); //print(end($arr) . "\n");
                if (end($arr) == 'png') {// print($uri);                   
                    $farr = explode('/', $arr[sizeof($arr) - 2]);
                    $filename = end($farr) . '.' . end($arr); // print($filename);
                    if (!file_exists($filename)) {
                        $content_file = file_get_contents($uri);
                        file_put_contents($filename, $content_file);
                    }
                }
            }
        }
    }
}
?>

Nice example, but uncommon input. Why not something more dynamic, with an hint of preventing information overflow?

I thought to illustrate a bridge from searching the web interface (seeing a picture or categorie one likes ) to actually getting it.

Making it more dynamic. We could do some random stuf, in the other Gist (NieuwsQuiz) I actually do that.