public function testMetaTagExtraction()
{
$this->assertFalse(AmberNetworkUtils::find_meta_redirect(""));
$this->assertFalse(AmberNetworkUtils::find_meta_redirect("bogus string"));
$this->assertFalse(AmberNetworkUtils::find_meta_redirect(<<<EOD
<html>
<head><title>bad man</title></head>
<body>
Use META tags like this: <meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</body>
</html>
EOD
));
$this->assertEquals("http://www.example.org/login", AmberNetworkUtils::find_meta_redirect(<<<EOD
<html>
<head><title>bad man</title>
<meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</head>
<body>
Use META tags like this: <meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</body>
</html>
EOD
));
$this->assertEquals("http://www.example.org/login", AmberNetworkUtils::find_meta_redirect(<<<EOD
<html>
<head><title>bad man</title>
<meta http-equiv="REFRESH" content="0; url=http://www.example.org/login">
</head>
<body>
Use META tags like this: <meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</body>
</html>
EOD
));
$this->assertFalse(AmberNetworkUtils::find_meta_redirect(<<<EOD
<html>
<head><title>bad man</title>
<meta http-equiv="refresh" content="5">
</head>
<body>
Use META tags like this: <meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</body>
</html>
EOD
));
$this->assertEquals("http://www.example.org/login", AmberNetworkUtils::find_meta_redirect(<<<EOD
<html>
<head><title>bad man</title>
<meta http-equiv="REFRESH" content='0;url = http://www.example.org/login'>
</head>
<body>
Use META tags like this: <meta http-equiv="refresh" content="30; URL=http://www.example.org/login">
</body>
</html>
EOD
));
}
/**
* Look at the results from a lookup using curl_multi, and identify urls that we need
* to query again, because a redirect is needed but was not followed
* @param $urls associative array of url lookups, keyed by url
* @return associative array of the subset of items that need to be queried again
*/
public static function find_urls_requiring_redirects($urls)
{
$result = array();
foreach ($urls as $url => $data) {
if ($data['info']['http_code'] == 301 || $data['info']['http_code'] == 302) {
$result[$url] = $data;
} else {
if (AmberNetworkUtils::find_meta_redirect($data['body'])) {
$result[$url] = $data;
}
}
}
return $result;
}
