Country: '.$country.'';
// echo 'search by number or search by keyword
';
define('START_NUMBER', '1');
$starts = (isset($_GET['starts'])) ? $_GET['starts'] : START_NUMBER;
define('STOP_NUMBER', '10');
$stops = (isset($_GET['stops'])) ? $_GET['stops'] : STOP_NUMBER;
echo '| enter start number | enter stop number | |
';
echo '
';
for ( $search=$starts; $search<=$stops; $search++ ) {
$url = 'http://www.imdb.com/title/tt'.sprintf("%07.0f",$search).'/';
$fh = fopen ($csv_file_to_write, 'a') or die ("can't open file");
//get the page content
$imdb_content = get_data($url);
//parse for product name
$name = get_match('/(.*)<\/title>/isU',$imdb_content);
$director = strip_tags(get_match('/<h5[^>]*>Director:<\/h5>(.*)<\/div>/isU',$imdb_content));
$plot = get_match('/<h5[^>]*>Plot:<\/h5>(.*)<\/div>/isU',$imdb_content);
$genre_original = get_match('/<h5[^>]*>Genre:<\/h5>(.*)<\/div>/isU',$imdb_content);
$genre_processed_once = str_replace('<a href="/Sections/Genres/','<a target="_blank" href="http://www.imdb.com/Sections/Genres/',$genre_original);
$genre = str_replace('href="/title/tt','target="new" href="http://www.imdb.com/title/tt',$genre_processed_once);
$keywords_original = get_match('/<h5[^>]*>Plot Keywords:<\/h5>(.*)<\/div>/isU',$imdb_content);
$keywords = str_replace('<a href="/keyword/','<a target="_blank" href="http://www.imdb.com/keyword/',$keywords_original);
$tagline_original = get_match('/<h5[^>]*>Tagline:<\/h5>(.*)<\/div>/isU',$imdb_content);
$tagline = str_replace('href="/title/tt','target="new" href="http://www.imdb.com/title/tt',$tagline_original);
$awards_original = get_match('/<h5[^>]*>Awards:<\/h5>(.*)<\/div>/isU',$imdb_content);
$awards = str_replace('href="/title/tt','target="new" href="http://www.imdb.com/title/tt',$awards_original);
$release_date = get_match('/<h5[^>]*>Release Date:<\/h5>(.*)<\/div>/isU',$imdb_content);
$mpaa = get_match('/<a href="\/mpaa">MPAA<\/a>:<\/h5>(.*)<\/div>/isU',$imdb_content);
$run_time = get_match('/Runtime:<\/h5>(.*)<\/div>/isU',$imdb_content);
$user_rating = get_match('/<div class="starbar-meta">\n <b>(.*)<\/b>/isU',$imdb_content);
$poster = get_match('/<a name="poster".* src="(.*)".*<\/a>/isU',$imdb_content);
//$img = explode("\" ", $name); $img
//$img = get_match('/title="'.$img[0].'\"" src="(.*)" \\/>/isU',$imdb_content);
//echo "image:".$img;
//build content
$content.= '<h2>Film</h2><p>'.$name.'</p>'."\n";
if ( $poster != '' ) {
$content.= '<h2>Poster</h2><p><img src="'.$poster.'" border="0" alt="'.$name.'"><br/><a href="'.$poster.'" target="_blank">link to poster</a><br/>(not authorized)</p>'."\n";
}
else $content.= '<h2>Poster</h2><p>No Poster available</p>'."\n";
$content.= '<h2>Director</h2><p>'.$director.'</p>'."\n";
$content.= '<h2>Plot</h2><p>'.substr($plot,0,strpos($plot,'<a')).'</p>'."\n";
$content.= '<h2>Genre</h2><p>'.$genre.'</p>'."\n";
$content.= '<h2>Keywords</h2><p>'.$keywords.'</p>'."\n";
$content.= '<h2>Tagline</h2><p>'.$tagline.'</p>'."\n";
$content.= '<h2>Awards</h2><p>'.$awards.'</p>'."\n";
$content.= '<h2>Release Date</h2><p>'.substr($release_date,0,strpos($release_date,'<a')).'</p>'."\n";
$content.= '<h2>MPAA</h2><p>'.$mpaa.'</p>'."\n";
$content.= '<h2>Run Time</h2><p>'.$run_time.'</p>'."\n";
$content.= '<h2>User Rating</h2><p>'.$user_rating.'</p>'."\n";
$content.= '<h2>Full Details</h2><p><a href="'.$url.'" rel="nofollow" target="_blank">'.$url.'</a></p>'."\n";
// echo $content;
//build csv content
$csv_content .= 'tt'.sprintf("%07.0f",$search).'|';
if ( $name != '' ) { $csv_content.= str_replace('|','-',$name)."|"; } else $csv_content.= " |";
if ( $director != '' ) { $csv_content.= ltrim(rtrim(str_replace("\n",'',str_replace('|','-',$director))))."|"; } else $csv_content.= " |";
if ( $plot != '' ) { $csv_content.= ltrim(rtrim(str_replace('"','"',str_replace("\n",'',str_replace('|','-',substr($plot,0,strpos($plot,'<a')))))))."|"; } else $csv_content.= " |";
if ( $release_date != '' ) { $csv_content.= ltrim(rtrim(str_replace("\n",'',str_replace('|','-',substr($release_date,0,strpos($release_date,'<a'))))))."|"; } else $csv_content.= " |";
if ( $run_time != '' ) { $csv_content.= ltrim(rtrim(str_replace("\n",'',str_replace('|','-',$run_time))))."|\n"; } else $csv_content.= "\r\n";
// if ( $run_time != '' ) { $csv_content.= ltrim(rtrim(str_replace("\n",'',str_replace('|','-',$run_time))))."|"; } else $csv_content.= " |";
// if ( $url != '' ) { $csv_content.= str_replace("\n",'',str_replace('|','-',$url))."|\n"; } else $csv_content.= "\r\n";
$csv_preg_replace = preg_replace('#<a.*</a>#','\\1',$csv_content);
$csv_content_clean = str_replace(' »','',str_replace('</span>','',str_replace('<div class="info-content">','',str_replace('<div id="tn15plotkeywords" style="display:inline;"><span>','',$csv_preg_replace))));
$csv_content = '';
fwrite($fh, $csv_content_clean);
fclose($fh);
}
//gets the match content
function get_match($regex,$content)
{
preg_match($regex,$content,$matches);
return $matches[1];
}
//gets the data from a URL
function get_data($url)
{
$ch = curl_init();
$timeout = 5;
curl_setopt($ch,CURLOPT_URL,$url);
curl_setopt($ch,CURLOPT_RETURNTRANSFER,1);
curl_setopt($ch,CURLOPT_CONNECTTIMEOUT,$timeout);
$data = curl_exec($ch);
curl_close($ch);
return $data;
}
?>
</body>
</html>