| public static withHTML ( $source = null, $selector = null, $options = [] ) |
public static function htmlqp($document = NULL, $selector = NULL, $options = array())
{
self::$qp = false;
// I copied and modified the default options from the htmlqp method to provide a custom version for Amslib
// NOTE: Hmm....I'm not 100% sure this will work in all circumstances....
$document = iconv("ISO8859-1", "UTF-8", $document);
// NOTE: see output buffer trick comment in qp()
ob_start();
self::$qp = QueryPath::withHTML($document, $selector, $options);
$warnings = ob_get_clean();
if (strlen($warnings)) {
Amslib_Debug::log("FAILED TO OBTAIN CLEAN OUTPUT WHEN PROCESSING HTML: error = ", $warnings);
}
return self::$qp;
}
private function extractTheatreMovieShowtimes($pageData, $limit, &$totalPages)
{
$startTime = microtime(true);
\SystemLogger::debug("Begining extraction of data from file, size = ", strlen($pageData));
if ($limit <= 0) {
\SystemLogger::warn("Invalid limit was supplied: ", $limit);
return array();
}
\SystemLogger::debug('Attempting to load into Query Path');
/* @var $moviePage DOMQuery */
$moviePage = \QueryPath::withHTML($pageData, null, array('convert_to_encoding' => "UTF-8", 'convert_from_encoding' => "UTF-8"));
\SystemLogger::debug('Loaded into QueryPath');
/* @var $theatersDom DOMQuery */
$theatersDom = $moviePage->find("div.theater");
//get total pages
$paginationDom = $moviePage->find("#navbar td");
$totalPages = $paginationDom->length ? $paginationDom->length - 2 : 1;
\SystemLogger::debug("Found", $theatersDom->length, "theatres");
$theatreCinemas = array();
$foundTheatres = 0;
\SystemLogger::debug('Loading data from Theatres DOM list');
for ($i = 0; $i < $theatersDom->length && $foundTheatres < $limit; $i++) {
$theatre = array();
$theatreDom = new DOMQuery($theatersDom->get($i));
$theatre['name'] = trim($theatreDom->find("h2.name")->first()->text());
if (!$theatre['name']) {
\SystemLogger::warn("Found no theatre at dom level: ", $i);
continue;
}
\SystemLogger::debug("processing theatre: ", $theatre['name']);
$addressText = $theatreDom->find(".info")->first()->text();
//echo $addressText, "<br>";
$tmp = explode(" - ", trim($addressText));
array_pop($tmp);
$theatre['address'] = join(' ', $tmp);
$theatreCinemas[] = array('theatre' => $theatre, 'movies' => $this->extractMovieShowtimes($theatreDom));
$foundTheatres++;
}
\SystemLogger::info('Extraction done, completed in ', microtime(true) - $startTime, 'ms');
return $theatreCinemas;
}
/**
* A special-purpose version of {@link qp()} designed specifically for HTML.
*
* XHTML (if valid) can be easily parsed by {@link qp()} with no problems. However,
* because of the way that libxml handles HTML, there are several common steps that
* need to be taken to reliably parse non-XML HTML documents. This function is
* a convenience tool for configuring QueryPath to parse HTML.
*
* The following options are automatically set unless overridden:
* - ignore_parser_warnings: TRUE
* - convert_to_encoding: ISO-8859-1 (the best for the HTML parser).
* - convert_from_encoding: auto (autodetect encoding)
* - use_parser: html
*
* Parser warning messages are also suppressed, so if the parser emits a warning,
* the application will not be notified. This is equivalent to
* calling @code@qp()@endcode.
*
* Warning: Character set conversions will only work if the Multi-Byte (mb) library
* is installed and enabled. This is usually enabled, but not always.
*
* @ingroup querypath_core
* @see qp()
*/
function htmlqp($document = NULL, $selector = NULL, $options = array())
{
return QueryPath::withHTML($document, $selector, $options);
}
