chr.module in cURL HTTP Request 8

File

chr.module

<?php

/**
 * Performs an HTTP request.
 *
 * This is a flexible and powerful HTTP client implementation. Correctly
 * handles GET, POST, PUT or any other HTTP requests. Handles redirects.
 *
 * @param $url
 *   A string containing a fully qualified URI.
 * @param array $options
 *   (optional) An array that can have one or more of the following elements:
 *   - headers: An array containing request headers to send as name/value pairs.
 *   - method: A string containing the request method. Defaults to 'GET'.
 *   - data: A string containing the request body, formatted as
 *     'param=value&param=value&...'. Defaults to NULL.
 *   - max_redirects: An integer representing how many times a redirect
 *     may be followed. Defaults to 3.
 *   - timeout: A float representing the maximum number of seconds the function
 *     call may take. The default is 30 seconds. If a timeout occurs, the error
 *     code is set to the HTTP_REQUEST_TIMEOUT constant.
 *   - context: A context resource created with stream_context_create().
 *   - verify_ssl: A boolean, to decide whether (TRUE) or not (FALSE) verify the SSL certificate and host.
 *   - verbose: A boolean, to switch on (TRUE) and off (FALSE) the cURL verbose mode.
 *   - cookiefile: A string containing a local path to the cookie file.
 *   - http_proxy: An array that will override the system-wide HTTP proxy settings. Array's elements:
 *   - https_proxy: An array that will override the system-wide HTTPS proxy settings.
 * @return object
 *   An object that can have one or more of the following components:
 *   - request: A string containing the request body that was sent.
 *   - code: An integer containing the response status code, or the error code
 *     if an error occurred.
 *   - protocol: The response protocol (e.g. HTTP/1.1 or HTTP/1.0).
 *   - status_message: The status message from the response, if a response was
 *     received.
 *   - redirect_code: If redirected, an integer containing the initial response
 *     status code.
 *   - redirect_url: If redirected, a string containing the URL of the redirect
 *     target.
 *   - error: If an error occurred, the error message. Otherwise not set.
 *   - errno: If an error occurred, a cURL error number greater than 0. Otherwise set to 0.
 *   - headers: An array containing the response headers as name/value pairs.
 *     HTTP header names are case-insensitive (RFC 2616, section 4.2), so for
 *     easy access the array keys are returned in lower case.
 *   - data: A string containing the response body that was received.
 */
function curl_http_request($url, array $options = array()) {
  $result = new stdClass();

  // Parse the URL and make sure we can handle the schema.
  $uri = @parse_url($url);
  if ($uri == FALSE) {
    $result->error = 'unable to parse URL';
    $result->code = -1001;
    return $result;
  }
  if (!isset($uri['scheme'])) {
    $result->error = 'missing schema';
    $result->code = -1002;
    return $result;
  }
  timer_start(__FUNCTION__);

  // Merge the default options.
  $options += array(
    'headers' => array(),
    'method' => 'GET',
    'data' => NULL,
    'max_redirects' => 3,
    'timeout' => 30.0,
    'context' => NULL,
    'verify_ssl' => FALSE,
    'verbose' => FALSE,
    'cookiefile' => NULL,
    'http_proxy' => variable_get('http_proxy'),
    'https_proxy' => variable_get('https_proxy'),
  );

  // Select the right proxy for the right protocol.
  $proxy = 'https' == $uri['scheme'] ? $options['https_proxy'] : $options['http_proxy'];

  // Nullify the proxy if the host to send the request to is part of the proxy's exceptions.
  if (!empty($proxy['exceptions']) && array_key_exists($uri['host'], $proxy['exceptions'])) {
    $proxy = NULL;
  }
  $curl_opt = array(
    CURLOPT_HEADER => TRUE,
    CURLINFO_HEADER_OUT => TRUE,
    CURLOPT_TIMEOUT => $options['timeout'],
    CURLOPT_FOLLOWLOCATION => TRUE,
    CURLOPT_RETURNTRANSFER => TRUE,
    CURLOPT_SSL_VERIFYPEER => $options['verify_ssl'],
    CURLOPT_SSL_VERIFYHOST => $options['verify_ssl'],
    CURLOPT_MAXREDIRS => $options['max_redirects'],
  );
  if ($options['cookiefile']) {
    $curl_opt += array(
      CURLOPT_COOKIEJAR => $options['cookiefile'],
      CURLOPT_COOKIEFILE => $options['cookiefile'],
    );
  }
  if (!empty($proxy)) {
    $proxy_options = array(
      CURLOPT_PROXY => $proxy['server'],
      CURLOPT_PROXYPORT => $proxy['port'],
    );

    // For the time being let's just support HTTP proxies with basic authentication.
    if (isset($proxy['username']) && isset($proxy['password'])) {
      $proxy_options += array(
        CURLOPT_PROXYUSERPWD => implode(':', array(
          $proxy['username'],
          $proxy['password'],
        )),
        CURLOPT_PROXYTYPE => CURLPROXY_HTTP,
        CURLOPT_PROXYAUTH => CURLAUTH_BASIC,
      );
    }
    $curl_opt += $proxy_options;
  }
  $default_ports = array(
    'http' => 80,
    'feed' => 80,
    'https' => 443,
  );
  if (array_key_exists($uri['scheme'], $default_ports)) {
    if (!isset($uri['port'])) {
      $uri['port'] = $default_ports[$uri['scheme']];
    }

    // RFC 2616: "non-standard ports MUST, default ports MAY be included".
    // We don't add the standard port to prevent from breaking rewrite rules
    // checking the host that do not take into account the port number.
    $options['headers']['Host'] = $uri['host'] . ($uri['port'] != 80 ? ':' . $uri['port'] : '');
  }
  else {
    $result->error = 'invalid schema ' . $uri['scheme'];
    $result->code = -1003;
    return $result;
  }

  // Merge the default headers.
  $options['headers'] += array(
    'User-Agent' => 'Drupal (+http://drupal.org/)',
  );

  // Only add Content-Length if we actually have any content or if it is a POST
  // or PUT request. Some non-standard servers get confused by Content-Length in
  // at least HEAD/GET requests, and Squid always requires Content-Length in
  // POST/PUT requests.
  $content_length = strlen($options['data']);
  if ($content_length > 0 || $options['method'] == 'POST' || $options['method'] == 'PUT') {
    $options['headers']['Content-Length'] = $content_length;
  }

  // If the server URL has a user then attempt to use basic authentication.
  if (isset($uri['user'])) {
    $options['headers']['Authorization'] = 'Basic ' . base64_encode($uri['user'] . (isset($uri['pass']) ? ':' . $uri['pass'] : ''));
  }

  // If the database prefix is being used by SimpleTest to run the tests in a copied
  // database then set the user-agent header to the database prefix so that any
  // calls to other Drupal pages will run the SimpleTest prefixed database. The
  // user-agent is used to ensure that multiple testing sessions running at the
  // same time won't interfere with each other as they would if the database
  // prefix were stored statically in a file or database variable.
  $test_info =& $GLOBALS['drupal_test_info'];
  if (!empty($test_info['test_run_id'])) {
    $options['headers']['User-Agent'] = drupal_generate_test_ua($test_info['test_run_id']);
  }

  // Set all the headers.
  $curl_opt[CURLOPT_HTTPHEADER] = array();
  foreach ($options['headers'] as $name => $value) {
    $curl_opt[CURLOPT_HTTPHEADER][] = $name . ": " . trim($value);
  }

  // Set the request method.
  switch ($options['method']) {
    case 'GET':
      $curl_opt[CURLOPT_HTTPGET] = TRUE;
      break;
    case 'POST':
      $curl_opt[CURLOPT_POST] = TRUE;
      if (!empty($options['data'])) {
        $curl_opt[CURLOPT_POSTFIELDS] = drupal_http_build_query($options['data']);
      }
      break;
    case 'PUT':
      $curl_opt[CURLOPT_PUT] = TRUE;
      break;
    default:
      $result->error = 'invalid method ' . $options['method'];
      $result->code = -1004;
      return $result;
  }

  // Make the request.
  $ch = curl_init($url);
  curl_setopt_array($ch, $curl_opt);

  // Full response stored. To be parsed later on and split in protocol, code, status message, and response headers.
  $result->data = trim(curl_exec($ch));
  $result->error = curl_error($ch);
  $result->errno = curl_errno($ch);

  // If there's been an error, do not continue.
  if ($result->error) {

    // Request timed out.
    if (CURLE_OPERATION_TIMEOUTED == $result->errno) {
      $result->code = HTTP_REQUEST_TIMEOUT;
      $result->error = 'request timed out';
      return $result;
    }
    $result->code = $result->errno;
    return $result;
  }

  // The last effective URL should correspond to the Redirect URL.
  $result->redirect_url = curl_getinfo($ch, CURLINFO_EFFECTIVE_URL);

  // Save the request sent into the result object.
  $result->request = curl_getinfo($ch, CURLINFO_HEADER_OUT);

  // Parse response headers from the response body.
  // Be tolerant of malformed HTTP responses that separate header and body with
  // \n\n or \r\r instead of \r\n\r\n.
  list($response, $result->data) = preg_split("/\r\n\r\n|\n\n|\r\r/", $result->data, 2);

  // Sometimes when making an HTTP request via proxy using cURL, you end up with a multiple set of headers:
  // from the web server being the actual target, from the proxy itself, etc.
  // The following 'if' statement is to check for such a situation and make sure we get a proper split between
  // actual response body and actual response headers both coming from the web server.
  while ('HTTP/' == substr($result->data, 0, 5)) {
    list($response, $result->data) = preg_split("/\r\n\r\n|\n\n|\r\r/", $result->data, 2);
  }
  $response = preg_split("/\r\n|\n|\r/", $response);

  // Parse the response status line.
  list($protocol, $code, $status_message) = explode(' ', trim(array_shift($response)), 3);
  $result->protocol = $protocol;
  $result->status_message = $status_message;
  $result->headers = array();

  // Parse the response headers.
  while ($line = trim(array_shift($response))) {
    list($name, $value) = explode(':', $line, 2);
    $name = strtolower($name);
    if (isset($result->headers[$name]) && $name == 'set-cookie') {

      // RFC 2109: the Set-Cookie response header comprises the token Set-
      // Cookie:, followed by a comma-separated list of one or more cookies.
      $result->headers[$name] .= ',' . trim($value);
    }
    else {
      $result->headers[$name] = trim($value);
    }
  }
  $responses = array(
    100 => 'Continue',
    101 => 'Switching Protocols',
    200 => 'OK',
    201 => 'Created',
    202 => 'Accepted',
    203 => 'Non-Authoritative Information',
    204 => 'No Content',
    205 => 'Reset Content',
    206 => 'Partial Content',
    300 => 'Multiple Choices',
    301 => 'Moved Permanently',
    302 => 'Found',
    303 => 'See Other',
    304 => 'Not Modified',
    305 => 'Use Proxy',
    307 => 'Temporary Redirect',
    400 => 'Bad Request',
    401 => 'Unauthorized',
    402 => 'Payment Required',
    403 => 'Forbidden',
    404 => 'Not Found',
    405 => 'Method Not Allowed',
    406 => 'Not Acceptable',
    407 => 'Proxy Authentication Required',
    408 => 'Request Time-out',
    409 => 'Conflict',
    410 => 'Gone',
    411 => 'Length Required',
    412 => 'Precondition Failed',
    413 => 'Request Entity Too Large',
    414 => 'Request-URI Too Large',
    415 => 'Unsupported Media Type',
    416 => 'Requested range not satisfiable',
    417 => 'Expectation Failed',
    500 => 'Internal Server Error',
    501 => 'Not Implemented',
    502 => 'Bad Gateway',
    503 => 'Service Unavailable',
    504 => 'Gateway Time-out',
    505 => 'HTTP Version not supported',
  );

  // RFC 2616 states that all unknown HTTP codes must be treated the same as the
  // base code in their class.
  if (!isset($responses[$code])) {
    $code = floor($code / 100) * 100;
  }
  $result->code = $code;
  switch ($code) {
    case 200:

    // OK
    case 304:

      // Not modified
      break;
    case 301:

    // Moved permanently
    case 302:

    // Moved temporarily
    case 307:

      // Moved temporarily
      $location = $result->headers['location'];
      $options['timeout'] -= timer_read(__FUNCTION__) / 1000;
      if ($options['timeout'] <= 0) {
        $result->code = HTTP_REQUEST_TIMEOUT;
        $result->error = 'request timed out';
      }
      elseif ($options['max_redirects']) {

        // Redirect to the new location.
        $options['max_redirects']--;
        $result = curl_http_request($location, $options);
        $result->redirect_code = $code;
      }
      if (!isset($result->redirect_url)) {
        $result->redirect_url = $location;
      }
      break;
    default:
      $result->error = $status_message;
  }
  curl_close($ch);
  return $result;
}

Functions

Name	Description
curl_http_request	Performs an HTTP request.