Curl wrapper

Curl.php

<?php

/**
 * Curl is class you will probably want to use
 * in order to fetch remote content.
 *
 * It is faster than `file_get_contents()` so using it is preferred.
 *
 * @author Zarko Stankovic <stankovic.zarko@gmail.com>
 */
class Curl extends CComponent
{

    /**
     * @var string target url to visit
     */
    public $url;

    /**
     * @var array cURL options
     */
    public $options = array();

    /**
     * @var array default cURL options
     */
    public $defaultOptions = array(
        CURLOPT_HEADER => true,
        CURLOPT_RETURNTRANSFER => true,
        CURLOPT_USERAGENT => 'Mozilla/5.0 (X11; U; Linux i686; en-US) AppleWebKit/534.16 (KHTML, like Gecko) Chrome/10.0.648.205 Safari/534.16',
    );

    /**
     * @var boolean process cURL request after object creation. Defaults to true.
     */
    public $autoOpen = true;

    /**
     * @var boolean close cURL session when request is finished. Defaults to true.
     */
    public $autoClose = true;

    /**
     * @var string cURL response body
     */
    public $response;

    /**
     * @var string current url that is processed
     */
    public $currentUrl;

    /**
     * @var integer maximum execution timeout (seconds)
     */
    public $maxExecutionTime = 20;

    /**
     * @var integer cURL start time
     */
    private $startTime = 0;

    /**
     * @var resource cURL handle
     */
    private $_handle;

    /**
     * @var integer keeps track of cURL redirections. Used internally.
     */
    private static $_loops = 0;

    /**
     * @var integer max redirections allowed. Used internally.
     */
    private static $_maxLoops = 20;

    /**
     * The Constructor
     *
     * @param string $url url to process
     * @param array $options cURL options (will be merged with default options).
     */
    public function __construct($url, $options = array())
    {
        $this->url = $url;
        $this->defaultOptions[CURLOPT_URL] = $url;
        $this->defaultOptions[CURLOPT_TIMEOUT] = $this->maxExecutionTime;
        $this->options = $this->defaultOptions + $options;

        $this->init();
    }

    /**
     * Initializes cURL session.
     * If {@link $autoOpen} is set to `true` connection is automatically made,
     * otherwise you need to fire {@link open()) by yourself.
     */
    public function init()
    {
        $this->_handle = curl_init();

        if ($this->autoOpen)
        {
            $this->response = $this->open();
        }
    }

    /**
     * Open a cURL sesion.
     *
     * @return string cURL response body
     * @throws CException if fail to set cURL options or to start cURL session
     */
    public function open()
    {
        $optionsSet = curl_setopt_array($this->_handle, $this->options);

        if (!$optionsSet)
        {
            throw new CException('Failed to set cURL options: ' . print_r($this->options, true));
        }

        $this->startTime = time();
        $response = $this->curl_redir_exec($this->_handle);

        if (curl_errno($this->_handle))
        {
            throw new CException(curl_error($this->_handle));
        }
        else
        {
            if ($this->autoClose)
            {
                $this->close();
            }
            return $response;
        }
    }

    /**
     * Close a cURL session
     */
    public function close()
    {
        curl_close($this->_handle);
    }

    /**
     * Allows you to use `Curl` instance as a string. {@link $response} will be used.
     * @return string returns cURL response body
     */
    public function __toString()
    {
        return $this->response;
    }

    /**
     * Credits:
     * http://php.net/manual/en/function.curl-setopt.php#71313
     * with little modifications by `Zarko Stankovic <stankovic.zarko@gmail.com>`
     *
     * @return string content fetched from remote host
     */
    private function curl_redir_exec()
    {
        $remoteHost = curl_getinfo($this->_handle);
        $remoteHost = $remoteHost['url'];
        $message = "cURL got no response from $remoteHost.";

        if (self::$_loops++ >= self::$_maxLoops)
        {
            $message .= " Reason: Maximum redirections (".self::$_loops.").";
            throw new CException($message);
        }

        $data = curl_exec($this->_handle);

        if (!$data)
        {
            $timeouted = (time() - $this->startTime) >= $this->maxExecutionTime;
            if ($timeouted)
            {
                $message .= " Reason: Timeouted ({$this->maxExecutionTime} seconds).";
                throw new CException($message);
            }
        }

        $responseArray = explode("\r\n\r\n", $data, 2);

        if (empty($responseArray) || empty($data))
        {
            $message .= " Reason: Couldn't parse response.";
            throw new CException($message);
        }

        $headers = $responseArray[0];
        $data = $responseArray[1];

        $http_code = curl_getinfo($this->_handle, CURLINFO_HTTP_CODE);

        if ($http_code == 301 || $http_code == 302)
        {
            $matches = array();
            preg_match('/Location:(.*)/', $headers, $matches);
            $url = @parse_url(trim(array_pop($matches)));

            if (!$url)
            {
                // couldn't process the url to redirect to
                self::$_loops = 0;
                return $data;
            }

            $last_url = parse_url(curl_getinfo($this->_handle, CURLINFO_EFFECTIVE_URL));

            if (!isset($url['scheme']))
            {
                $url['scheme'] = $last_url['scheme'];
            }
            if (!isset($url['host']))
            {
                $url['host'] = $last_url['host'];
            }
            if (!isset($url['path']))
            {
                $url['path'] = $last_url['path'];
            }

            $new_url = $url['scheme'] . '://' . $url['host'] . $url['path'] . (isset($url['query']) ? '?' . $url['query'] : '');
            curl_setopt($this->_handle, CURLOPT_URL, $new_url);

            // redirect to new url
            return $this->curl_redir_exec($this->_handle);
        }
        else
        {
            $info = curl_getinfo($this->_handle);
            $this->currentUrl = $info['url'];

            self::$_loops = 0;
            return $data;
        }
    }

}