{
$httpResponse = array('raw_data' => $data, 'headers'=> array(), 'cookies' => array());
- // Support "web-proxy-tunelling" connections for https through proxies
+ // Support "web-proxy-tunnelling" connections for https through proxies
if (preg_match('/^HTTP\/1\.[0-1] 200 Connection established/', $data)) {
// Look for CR/LF or simple LF as line separator,
// (even though it is not valid http)
}
$data = substr($data, $pos);
}
+
+ // When using Curl to query servers using Digest Auth, we get back a double set of http headers.
+ // We strip out the 1st...
+ if ($headersProcessed && preg_match('/^HTTP\/[0-9.]+ 401 /', $data)) {
+ if (preg_match('/(\r?\n){2}HTTP\/[0-9.]+ 200 /', $data)) {
+ $data = preg_replace('/^HTTP\/[0-9.]+ 401 .+?(?:\r?\n){2}(HTTP\/[0-9.]+ 200 )/s', '$1', $data, 1);
+ }
+ }
+
if (!preg_match('/^HTTP\/[0-9.]+ 200 /', $data)) {
$errstr = substr($data, 0, strpos($data, "\n") - 1);
error_log('XML-RPC: ' . __METHOD__ . ': HTTP error, got response: ' . $errstr);
$bd = 0;
}
}
+
// be tolerant to line endings, and extra empty lines
$ar = preg_split("/\r?\n/", trim(substr($data, 0, $pos)));
+
foreach($ar as $line) {
// take care of multi-line headers and cookies
$arr = explode(':', $line, 2);
// if CURL was used for the call, http headers have been processed,
// and dechunking + reinflating have been carried out
if (!$headersProcessed) {
+
// Decode chunked encoding sent by http 1.1 servers
if (isset($httpResponse['headers']['transfer-encoding']) && $httpResponse['headers']['transfer-encoding'] == 'chunked') {
if (!$data = Http::decodeChunked($data)) {