public function testHeadersParseValueHasColon()
{
$raw = <<<EOD
HTTP/1.1 200 OK
Content-type: text/html
Accept-Ranges: bytes
ETag: "1805770918"
Last-Modified: Tue, 20 Nov 2012 02:11:45 GMT
Content-Length: 611
nnCoection: close
Date: Thu, 29 May 2014 20:53:13 GMT
Connection: Keep-alive
Content-Location: http://somwhere.else
EOD;
$headers = AmberNetworkUtils::extract_headers($raw);
$this->assertEquals($headers["Content-Location"], "http://somwhere.else");
}
public function testHeadersParseCaseSensitive()
{
$raw = <<<EOD
HTTP/1.1 200 OK
Content-type: text/html
Accept-Ranges: bytes
ETag: "1805770918"
Last-Modified: Tue, 20 Nov 2012 02:11:45 GMT
Content-Length: 611
nnCoection: close
Date: Thu, 29 May 2014 20:53:13 GMT
Connection: Keep-alive
EOD;
$headers = AmberNetworkUtils::extract_headers($raw);
$this->assertEquals($headers["Content-Type"], "text/html");
}
/**
* Open a single URL, and return an array with dictionary of header information and the contents
* of the URL. Handle redirects ourselves, rather than using CURLOPT_FOLLOWLOCATION
* Adapted from http://slopjong.de/2012/03/31/curl-follow-locations-with-safe_mode-enabled-or-open_basedir-set/
* @param $url string of resource to download
* @return array dictionary of header information and a stream to the contents of the URL
*/
public static function open_single_url($url, $additional_options = array(), $follow_redirects = TRUE)
{
$options = array(CURLOPT_FAILONERROR => TRUE, CURLOPT_FOLLOWLOCATION => FALSE, CURLOPT_CONNECTTIMEOUT => 5, CURLOPT_TIMEOUT => 10, CURLOPT_RETURNTRANSFER => 1, CURLOPT_HEADER => TRUE, CURLOPT_USERAGENT => AmberNetworkUtils::get_user_agent_string(), CURLOPT_ENCODING => '');
$max_redirects = 5;
try {
$ch = curl_init($url);
if (curl_setopt_array($ch, $additional_options + $options) === FALSE) {
throw new RuntimeException(join(":", array(__FILE__, __METHOD__, "Error setting CURL options", $url, curl_error($ch))));
}
$original_url = curl_getinfo($ch, CURLINFO_EFFECTIVE_URL);
$newurl = $original_url;
do {
curl_setopt($ch, CURLOPT_URL, $newurl);
$response = curl_exec($ch);
$response_info = curl_getinfo($ch);
if ($response_info['http_code'] == 301 || $response_info['http_code'] == 302) {
$newurl = $response_info['redirect_url'];
} else {
if ($meta = AmberNetworkUtils::find_meta_redirect($response)) {
$newurl = $meta;
} else {
break;
// Not a redirect, so we're done
}
}
// if no scheme is present then the new url is a relative path and thus needs some extra care
if (!preg_match("/^https?:/i", $newurl)) {
$last_slash = strrpos($original_url, "/", 9);
// Starting at position 9 starts search past http://
if ($last_slash == strlen($original_url) - 1) {
$newurl = $original_url . $newurl;
} else {
if ($last_slash === FALSE) {
$newurl = join("/", array($original_url, $newurl));
} else {
$newurl = join("/", array(substr($original_url, 0, $last_slash), $newurl));
}
}
}
} while (--$max_redirects && $follow_redirects);
curl_close($ch);
} catch (RuntimeException $e) {
error_log($e->getMessage());
curl_close($ch);
return FALSE;
}
if (!$max_redirects) {
return FALSE;
// We ran out of redirects without getting a result
} else {
/* Split into header and body */
$header_size = $response_info['header_size'];
$header = substr($response, 0, $header_size - 1);
$body = substr($response, $header_size);
$headers = AmberNetworkUtils::extract_headers($header);
return array("headers" => $headers, "body" => $body, "info" => $response_info);
}
}