function _scrapeLexPage($html, $url)
{
global $vocab;
global $rdfData;
$dom = new simple_html_dom();
$dom->load($html);
$h2a = $dom->find("h2[@class='headline'] a");
$href = $h2a[0]->href;
$id = substr($href, 0, strrpos($href, '.'));
$fullURL = substr($url, 0, strrpos($url, '/')) . '/' . $href;
$dom->__destruct();
$uri = URI_BASE . $id;
$html = _getHTML($fullURL);
$dom = new simple_html_dom();
$dom->load($html);
// x a http://www.metalex.eu/metalex/2008-05-02#BibliographicWork
// dct:title, dct:description, metalex:fragment/fragmentOf, rdfs:seeAlso, rdfs:label, dcterms:source
// x rdf:type metalex:BibliographicWork
$rdfData[$uri] = array(RDF_TYPE => array(array('type' => 'uri', 'value' => METALEX_BIBWORK)));
// title, label
$h1Element = $dom->find("div[@id='paddingLR12'] div[@class='jnheader'] h1");
$title = html_entity_decode($h1Element[0]->plaintext, ENT_COMPAT, 'ISO-8859-1');
$rdfData[$uri][$vocab['dct'] . 'title'] = array(array('type' => 'literal', 'value' => $title, 'lang' => 'de'));
$rdfData[$uri][$vocab['rdfs'] . 'label'] = array(array('type' => 'literal', 'value' => $title, 'lang' => 'de'));
// token
$pElems = $dom->find("div[@id='paddingLR12'] div[@class='jnheader'] p");
if (isset($pElems[0])) {
$token = html_entity_decode($pElems[0]->plaintext, ENT_COMPAT, 'ISO-8859-1');
$rdfData[$uri][$vocab['nlex'] . 'token'] = array(array('type' => 'literal', 'value' => $token));
}
// seeAlso
$rdfData[$uri][$vocab['rdfs'] . 'seeAlso'] = array(array('type' => 'uri', 'value' => $fullURL));
$i = 0;
$currentChapterURI = null;
$jnNorms = $dom->find("div[@id='paddingLR12'] div[@class='jnnorm']");
foreach ($jnNorms as $data) {
if ($i === 0) {
// We skip the first for now, since it contains only header info
++$i;
continue;
}
++$i;
$result = false;
// If the text contains a h3, it is a fragment
$h3 = $data->find("div[@class='jnheader'] h3 span");
if ($h3) {
if (null !== $currentChapterURI) {
$result = _handleFragmentForURI($data, $currentChapterURI, $fullURL);
} else {
$result = _handleFragmentForURI($data, $uri, $fullURL);
}
} else {
// If we find a h2 this is a chapter
$h2 = $data->find("div[@class='jnheader'] h2");
if ($h2) {
$currentChapterURI = _handleChapterForURI($data, $uri, $fullURL);
$result = true;
// special case...
} else {
return false;
}
}
if (!$result) {
return false;
}
}
$dom->__destruct();
return true;
}
function _handleDetailPage($url, $result)
{
$html = _getHTML($url);
$dom = new simple_html_dom();
$dom->load($html);
$divBoxes = $dom->find("div[@class='featured_box margin_top_fb']");
if (count($divBoxes) < 1) {
return $result;
}
$style = $divBoxes[0]->style;
$styleParts = explode("'", $style);
$imageURL = EUROPA_URL_BASE . $styleParts[1];
$result['logoURL'] = $imageURL;
$addressText = $divBoxes[0]->xmltext;
$addressTextParts = explode('</h3>', $addressText);
if (count($addressTextParts) !== 2) {
return $result;
}
$addressText = $addressTextParts[1];
$addressText = str_replace('<br />', '<br>', $addressText);
$addressTextParts = explode('<br>', $addressText);
if (count($addressTextParts) < 3) {
return $result;
}
$street = trim($addressTextParts[0]);
$matches = array();
$curPos = 1;
preg_match('/^.*[0-9]+.*$/', $street, $matches);
if (count($matches) === 0) {
$street .= ' ' . trim($addressTextParts[$curPos++]);
}
$result['zipCity'] = trim($addressTextParts[$curPos++]);
$result['country'] = trim($addressTextParts[$curPos++]);
for ($i = $curPos; $i < count($addressTextParts); ++$i) {
$val = strtolower(trim($addressTextParts[$i]));
if (substr($val, 0, 4) === 'tel:') {
$result['tel'] = trim(substr($val, 4));
} else {
if (substr($val, 0, 4) === 'fax:') {
$result['fax'] = trim(substr($val, 4));
} else {
if (substr($val, 0, 2) === '<a') {
$parts = explode('"', $val);
$result['mailto'] = trim($parts[1]);
}
}
}
}
$links = array();
$aElements = $dom->find("div[@id='euCenter'] a");
foreach ($aElements as $a) {
if (strpos($a->href, 'http://') !== false) {
$links[] = $a->href;
}
}
if (count($links) > 0) {
if (isset($result['seeAlso'])) {
$result['seeAlso'] = array_merge($result['seeAlso'], $links);
} else {
$result['seeAlso'] = $links;
}
}
$dom->__destruct();
return $result;
}