Created
June 29, 2015 19:40
-
-
Save AliceWonderMiscreations/6f3d0a331cc7ed1e4a6f to your computer and use it in GitHub Desktop.
RTALabel Checker
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
//needs more debug and error handle and stuff | |
//called by rtaCheckHeader() | |
function rtaCheckMeta($dom) { | |
$headList = $dom->getElementsByTagName('head'); | |
if($headList->length > 0) { | |
$head = $headList->item(0); | |
$metaList = $head->getElementsByTagName('meta'); | |
$j = $metaList->length; | |
for($i=0; $i<$j; $i++) { | |
$meta = $metaList->item($i); | |
if($meta->hasAttribute('name')) { | |
$name = strtoupper(trim($meta->getAttribute('name'))); | |
if(strcmp($name, 'RATING') == 0) { | |
if($meta->hasAttribute('content')) { | |
$content = strtoupper(trim($meta->getAttribute('content'))); | |
if(strcmp($content,'RTA-5042-1996-1400-1577-RTA') == 0) { | |
return TRUE; | |
} | |
} | |
} | |
} | |
} | |
} | |
return FALSE; | |
} | |
function rtaCheckHeader($link) { | |
$return = new stdClass(); | |
$return->rta = FALSE; | |
$return->scheme = 'default'; | |
$return->mime = 'foo/bar'; | |
$return->httpcode = 0; | |
$parsed = parse_url($link); | |
if(isset($parsed['scheme'])) { | |
$return->scheme = strtolower(trim($parsed['scheme'])); | |
} | |
if(isset($parsed['port'])) { | |
$port = 0 + $parsed['port']; | |
switch($return->scheme) { | |
case 'http': | |
if($port == 80) { | |
$port = 0; | |
} | |
break; | |
case 'https': | |
if($port == 443) { | |
$port = 0; | |
} | |
break; | |
case 'ftp': | |
if($port = 21) { | |
$port = 0; | |
} | |
} //end case | |
if($port > 0) { | |
$return->port = $port; | |
} | |
} | |
if(isset($parsed['host'])) { | |
$host = $parsed['host']; | |
if(function_exists('idn_to_ascii')) { | |
$return->host = trim(strtolower(idn_to_ascii($host))); | |
} else { | |
$return->host = trim(strtolower($host)); | |
} | |
} | |
if(isset($parsed['path'])) { | |
$return->path = trim($parsed['path']); | |
} else { | |
$return->path = '/'; | |
} | |
if(isset($parsed['query'])) { | |
$return->query = $parsed['query']; | |
} | |
$hypertext = FALSE; | |
if(strcmp($return->scheme, 'http')) { | |
$hypertext = TRUE; | |
} elseif(strcmp($return->scheme, 'https')) { | |
$hypertext = TRUE; | |
} | |
if(! isset($return->host)) { | |
$hypertext == FALSE; | |
} | |
if($hypertext) { | |
$link = $return->scheme . '://' . $return->host; | |
if(isset($return->port)) { | |
$link .= ':' . $return->port; | |
} | |
$link .= $return->path; | |
if(isset($return->query)) { | |
$link .= '?' . $return->query; | |
} | |
if(filter_var($link, FILTER_VALIDATE_URL)) { | |
$return->link = $link; | |
} | |
} | |
if(isset($return->link)) { | |
$ch = curl_init(); | |
curl_setopt($ch, CURLOPT_URL, $return->link); | |
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); | |
curl_setopt($ch, CURLOPT_USERAGENT, 'Godzilla/5.0 (AmigaOS 3.1; OOH64) BananaSplit/1337.69 (KSGML, like Anguid) Aluminum/43.7.2774.666 Party/445.22'); | |
curl_setopt($ch, CURLOPT_HEADER, 1); | |
curl_setopt($ch, CURLOPT_NOBODY, true); | |
$data = explode("\n", trim(curl_exec($ch))); | |
$j = count($data); | |
if($j > 0) { | |
$response = preg_replace('/\s\s+/', ' ', trim($data[0])); | |
//fixme check for integer | |
$return->httpcode = 0 + explode(' ', $response)[1]; | |
for($i=0; $i<$j; $i++) { | |
$data[$i] = preg_replace('/\s\s+/', ' ', trim($data[$i])); | |
//$string = strtoupper($data[$i]); | |
if(strcmp(strtoupper($data[$i]), 'RATING: RTA-5042-1996-1400-1577-RTA') == 0) { | |
$return->rta = TRUE; | |
} | |
$substring = substr(strtoupper($data[$i]), 0, 13); | |
if(strcmp($substring, 'CONTENT-TYPE:') == 0) { | |
$return->mime = preg_replace('/;/', '', strtoupper(explode(' ', $data[$i])[1])); | |
} | |
} | |
if(! $return->rta) { | |
$scrape = FALSE; | |
if($return->httpcode = 200) { | |
switch($return->mime) { | |
case 'TEXT/HTML': | |
$scrape = TRUE; | |
break; | |
case 'APPLICATION/XHTML+XML': | |
$scrape = TRUE; | |
break; | |
} //end switch | |
if($scrape) { | |
$return->foo = 'bar'; | |
curl_setopt($ch, CURLOPT_HEADER, 0); | |
curl_setopt($ch, CURLOPT_NOBODY, false); | |
$data = trim(curl_exec($ch)); | |
$config = array('output-xhtml' => true); | |
$tidy = new tidy; | |
$tidy->parseString($data, $config, 'utf8'); | |
$tidy->cleanRepair(); | |
$data = $tidy->value; | |
$dom = new DOMDocument("1.0", "UTF-8"); | |
$dom->loadHTML($data); | |
$return->rta = rtaCheckMeta($dom); | |
} | |
} | |
} | |
} | |
//var_dump($data); | |
curl_close($ch); | |
} | |
//var_dump($parsed); | |
//var_dump($return); | |
return $return; | |
} | |
?> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment