* @author Sterling Hughes * @author Tomas V.V.Cox * @author Richard Heyes * @author Philippe Jausions * @author Michael Wallner * @copyright 2002-2008 The Authors * @license http://www.opensource.org/licenses/bsd-license.php New BSD License * @link http://pear.php.net/package/HTTP2 */ /** * Load exception class */ require_once 'HTTP2/Exception.php'; /** * Miscellaneous HTTP Utilities * * PEAR::HTTP2 provides shorthand methods for generating HTTP dates, * issueing HTTP HEAD requests, building absolute URIs, firing redirects and * negotiating user preferred language. * * @category HTTP * @package HTTP2 * @author Stig Bakken * @author Sterling Hughes * @author Tomas V.V.Cox * @author Richard Heyes * @author Philippe Jausions * @author Michael Wallner * @license http://www.opensource.org/licenses/bsd-license.php New BSD License * @version Release: 1.1.2 * @link http://pear.php.net/package/HTTP2 */ class HTTP2 { /** * Formats a RFC compliant GMT date HTTP header. This function honors the * "y2k_compliance" php.ini directive and formats the GMT date corresponding * to either RFC850 or RFC822. * * @param mixed $time unix timestamp or date (default = current time) * * @return string|boolean GMT date string, or FALSE for an invalid * $time parameter */ public function date($time = null) { if (!isset($time)) { $time = time(); } elseif (!is_numeric($time) && (-1 === $time = strtotime($time))) { return false; } // RFC822 or RFC850 $format = ini_get('y2k_compliance') ? 'D, d M Y' : 'l, d-M-y'; return gmdate($format .' H:i:s \G\M\T', $time); } /** * Negotiates language with the user's browser through the Accept-Language * HTTP header or the user's host address. Language codes are generally in * the form "ll" for a language spoken in only one country, or "ll-CC" for a * language spoken in a particular country. For example, U.S. English is * "en-US", while British English is "en-UK". Portugese as spoken in * Portugal is "pt-PT", while Brazilian Portugese is "pt-BR". * * Quality factors in the Accept-Language: header are supported, e.g.: * Accept-Language: en-UK;q=0.7, en-US;q=0.6, no, dk;q=0.8 * *


     *  require_once 'HTTP2.php';
     *  $http = new HTTP2();
     *  $langs = array(
     *      'en'    => 'locales/en',
     *      'en-US' => 'locales/en',
     *      'en-UK' => 'locales/en',
     *      'de'    => 'locales/de',
     *      'de-DE' => 'locales/de',
     *      'de-AT' => 'locales/de',
     *  );
     *  $neg = $http->negotiateLanguage($langs);
     *  $dir = $langs[$neg];
     *

* * @param array $supported An associative array of supported languages, * whose values must evaluate to true. * @param string $default The default language to use if none is found. * * @return string The negotiated language result or the supplied default. */ public function negotiateLanguage($supported, $default = 'en-US') { $supp = array(); foreach ($supported as $lang => $isSupported) { if ($isSupported) { $supp[strtolower($lang)] = $lang; } } if (!count($supp)) { return $default; } if (isset($_SERVER['HTTP_ACCEPT_LANGUAGE'])) { $match = $this->matchAccept( $_SERVER['HTTP_ACCEPT_LANGUAGE'], $supp ); if (!is_null($match)) { return $match; } } if (isset($_SERVER['REMOTE_HOST'])) { $lang = strtolower(end($h = explode('.', $_SERVER['REMOTE_HOST']))); if (isset($supp[$lang])) { return $supp[$lang]; } } return $default; } /** * Negotiates charset with the user's browser through the Accept-Charset * HTTP header. * * Quality factors in the Accept-Charset: header are supported, e.g.: * Accept-Language: en-UK;q=0.7, en-US;q=0.6, no, dk;q=0.8 * *


     *  require_once 'HTTP2.php';
     *  $http = new HTTP2();
     *  $charsets = array(
     *      'UTF-8',
     *      'ISO-8859-1',
     *  );
     *  $charset = $http->negotiateCharset($charsets);
     *

* * @param array $supported An array of supported charsets * @param string $default The default charset to use if none is found. * * @return string The negotiated language result or the supplied default. * @author Philippe Jausions */ public function negotiateCharset($supported, $default = 'ISO-8859-1') { $supp = array(); foreach ($supported as $charset) { $supp[strtolower($charset)] = $charset; } if (!count($supp)) { return $default; } if (isset($_SERVER['HTTP_ACCEPT_CHARSET'])) { $match = $this->matchAccept( $_SERVER['HTTP_ACCEPT_CHARSET'], $supp ); if (!is_null($match)) { return $match; } } return $default; } /** * Negotiates content type with the user's browser through the Accept * HTTP header. * * Quality factors in the Accept: header are supported, e.g.: * Accept: application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8 * *


     *  require_once 'HTTP2.php';
     *  $http = new HTTP2();
     *  $contentType = array(
     *      'application/xhtml+xml',
     *      'application/xml',
     *      'text/html',
     *      'text/plain',
     *  );
     *  $mime = $http->negotiateMimeType($contentType);
     *

* * @param array $supported An associative array of supported MIME types. * @param string $default The default type to use if none match. * * @return string The negotiated MIME type result or the supplied default. * @author Philippe Jausions */ public function negotiateMimeType($supported, $default) { $supp = array(); foreach ($supported as $type) { $supp[strtolower($type)] = $type; } if (!count($supp)) { return $default; } if (isset($_SERVER['HTTP_ACCEPT'])) { $accepts = $this->sortAccept($_SERVER['HTTP_ACCEPT']); foreach ($accepts as $type => $q) { if (substr($type, -2) != '/*') { if (isset($supp[$type])) { return $supp[$type]; } continue; } if ($type == '*/*') { return array_shift($supp); } list($general, $specific) = explode('/', $type); $general .= '/'; $len = strlen($general); foreach ($supp as $mime => $t) { if (strncasecmp($general, $mime, $len) == 0) { return $t; } } } } return $default; } /** * Parses a weighed "Accept" HTTP header and matches it against a list * of supported options * * @param string $header The HTTP "Accept" header to parse * @param array $supported A list of supported values * * @return string|NULL a matched option, or NULL if no match */ protected function matchAccept($header, $supported) { $matches = $this->sortAccept($header); foreach ($matches as $key => $q) { if (isset($supported[$key])) { return $supported[$key]; } } // If any (i.e. "*") is acceptable, return the first supported format if (isset($matches['*'])) { return array_shift($supported); } return null; } /** * Parses and sorts a weighed "Accept" HTTP header * * @param string $header The HTTP "Accept" header to parse * * @return array Sorted list of "accept" options */ protected function sortAccept($header) { $matches = array(); foreach (explode(',', $header) as $option) { $option = array_map('trim', explode(';', $option)); $l = strtolower($option[0]); if (isset($option[1])) { $q = (float) str_replace('q=', '', $option[1]); } else { $q = null; // Assign default low weight for generic values if ($l == '*/*') { $q = 0.01; } elseif (substr($l, -1) == '*') { $q = 0.02; } } // Unweighted values, get high weight by their position in the // list $matches[$l] = isset($q) ? $q : 1000 - count($matches); } arsort($matches, SORT_NUMERIC); return $matches; } /** * Sends a "HEAD" HTTP command to a server and returns the headers * as an associative array. * * Example output could be: *


     *     Array
     *     (
     *         [response_code] => 200          // The HTTP response code
     *         [response] => HTTP/1.1 200 OK   // The full HTTP response string
     *         [Date] => Fri, 11 Jan 2002 01:41:44 GMT
     *         [Server] => Apache/1.3.20 (Unix) PHP/4.1.1
     *         [X-Powered-By] => PHP/4.1.1
     *         [Connection] => close
     *         [Content-Type] => text/html
     *     )
     *

* * @param string $url A valid URL, e.g.: http://pear.php.net/credits.php * @param integer $timeout Timeout in seconds (default = 10) * * @return array Returns associative array of response headers on success * * @throws HTTP2_Exception When connecting fails * @throws InvalidArgumentException When the protocol is not uspported * * @see HTTP_Client::head() * @see HTTP_Request */ public function head($url, $timeout = 10) { $p = parse_url($url); if (!isset($p['scheme'])) { $p = parse_url($this->absoluteURI($url)); } elseif ($p['scheme'] != 'http') { throw new InvalidArgumentException( 'Unsupported protocol: '. $p['scheme'] ); } $port = isset($p['port']) ? $p['port'] : 80; if (!$fp = @fsockopen($p['host'], $port, $eno, $estr, $timeout)) { throw new HTTP2_Exception("Connection error: $estr ($eno)"); } $path = !empty($p['path']) ? $p['path'] : '/'; $path .= !empty($p['query']) ? '?' . $p['query'] : ''; fputs($fp, "HEAD $path HTTP/1.0\r\n"); fputs($fp, 'Host: ' . $p['host'] . ':' . $port . "\r\n"); fputs($fp, "Connection: close\r\n\r\n"); $response = rtrim(fgets($fp, 4096)); if (preg_match("|^HTTP/[^\s]*\s(.*?)\s|", $response, $status)) { $headers['response_code'] = $status[1]; } $headers['response'] = $response; while ($line = fgets($fp, 4096)) { if (!trim($line)) { break; } if (($pos = strpos($line, ':')) !== false) { $header = substr($line, 0, $pos); $value = trim(substr($line, $pos + 1)); $headers[$header] = $value; } } fclose($fp); return $headers; } /** * This function redirects the client. This is done by issuing * a "Location" header and exiting if wanted. If you set $rfc2616 to true * HTTP will output a hypertext note with the location of the redirect. * * @param string $url URL where the redirect should go to. * @param bool $exit Whether to exit immediately after redirection. * @param bool $rfc2616 Wheter to output a hypertext note where we're * redirecting to (Redirecting to * ....) * * @return boolean Returns TRUE on succes (or exits) or FALSE if headers * have already been sent. */ function redirect($url, $exit = true, $rfc2616 = false) { if (headers_sent()) { return false; } $url = $this->absoluteURI($url); header('Location: '. $url); if ($rfc2616 && isset($_SERVER['REQUEST_METHOD']) && $_SERVER['REQUEST_METHOD'] != 'HEAD' ) { echo '

Redirecting to: ' .htmlspecialchars($url).'.

'; } if ($exit) { exit; } return true; } /** * This function returns the absolute URI for the partial URL passed. * The current scheme (HTTP/HTTPS), host server, port, current script * location are used if necessary to resolve any relative URLs. * * Offsets potentially created by PATH_INFO are taken care of to resolve * relative URLs to the current script. * * You can choose a new protocol while resolving the URI. This is * particularly useful when redirecting a web browser using relative URIs * and to switch from HTTP to HTTPS, or vice-versa, at the same time. * * @param string $url Absolute or relative URI the redirect should * go to. * @param string $protocol Protocol to use when redirecting URIs. * @param integer $port A new port number. * * @return string The absolute URI. * @author Philippe Jausions */ public function absoluteURI($url = null, $protocol = null, $port = null) { // filter CR/LF $url = str_replace(array("\r", "\n"), ' ', $url); // Mess around protocol and port with already absolute URIs if (preg_match('!^([a-z0-9]+)://!i', $url)) { if (empty($protocol) && empty($port)) { return $url; } if (!empty($protocol)) { $url = $protocol .':'. end($array = explode(':', $url, 2)); } if (!empty($port)) { $url = preg_replace( '!^(([a-z0-9]+)://[^/:]+)(:[\d]+)?!i', '\1:'. $port, $url ); } return $url; } $host = 'localhost'; if (!empty($_SERVER['HTTP_HOST'])) { list($host) = explode(':', $_SERVER['HTTP_HOST']); } elseif (!empty($_SERVER['SERVER_NAME'])) { list($host) = explode(':', $_SERVER['SERVER_NAME']); } if (empty($protocol)) { if (isset($_SERVER['HTTPS']) && !strcasecmp($_SERVER['HTTPS'], 'on')) { $protocol = 'https'; } else { $protocol = 'http'; } if (!isset($port) || $port != intval($port)) { $port = 80; if (isset($_SERVER['SERVER_PORT'])) { $port = $_SERVER['SERVER_PORT']; } } } if ($protocol == 'http' && $port == 80) { unset($port); } if ($protocol == 'https' && $port == 443) { unset($port); } $server = $protocol.'://'.$host.(isset($port) ? ':'.$port : ''); $uriAll = isset($_SERVER['REQUEST_URI']) ? $_SERVER['REQUEST_URI'] : $_SERVER['PHP_SELF']; if (false !== ($q = strpos($uriAll, '?'))) { $uriBase = substr($uriAll, 0, $q); } else { $uriBase = $uriAll; } if (!strlen($url) || $url{0} == '#') { $url = $uriAll.$url; } elseif ($url{0} == '?') { $url = $uriBase.$url; } if ($url{0} == '/') { return $server . $url; } // Adjust for PATH_INFO if needed if (isset($_SERVER['PATH_INFO']) && strlen($_SERVER['PATH_INFO'])) { $path = dirname(substr($uriBase, 0, -strlen($_SERVER['PATH_INFO']))); } else { /** * Fixes bug #12672 PHP_SELF ending on / causes incorrect redirects * * @link http://pear.php.net/bugs/12672 */ $path = dirname($uriBase.'-'); } if (substr($path = strtr($path, '\\', '/'), -1) != '/') { $path .= '/'; } return $server . $path . $url; } /** * Parses a HTTP "Link" header value as specified by RFC 5988. * * When errors occur during parsing, all information collected to that * point will be returned used and the rest of the line skipped. * * @param string|array $lines HTTP "Link:" header value (without "Link:"), * or array of such strings * * @return array Array of parsed links * (one single line may contain multiple links) * Only params existing in the link are in the array. * URI is available in key "_uri". * * @link http://tools.ietf.org/html/rfc5988 */ public function parseLinks($lines) { $lines = (array) $lines; $extracted = array(); foreach ($lines as $line) { $state = 'uri'; $pos = 0; $len = strlen($line); while ($pos < $len - 1) { $link = array(); $pos = $len - strlen(ltrim(substr($line, $pos))); if ($line{$pos} == ',') { ++$pos; continue; } else if ($line{$pos} != '<') { break; } $end = strpos($line, '>', $pos + 1); if ($end === false) { break; } $link['_uri'] = substr($line, $pos + 1, $end - $pos - 1); $pos = $end + 1; while ($pos < $len - 1) { if ($line{$pos} == ',') { $extracted[] = $link; $link = array(); ++$pos; continue; } else if ($line{$pos} != ';') { break; } ++$pos; $end = strpos($line, '=', $pos + 1); if ($end === false) { break; } $pname = trim(substr($line, $pos, $end - $pos)); $pos = $end + 1; $rest = trim(substr($line, $pos)); if ($rest{0} == '"') { $pos = strpos($line, '"', $pos) + 1; $end = strpos($line, '"', $pos + 1); $pval = substr($line, $pos, $end - $pos); $pos = $end + 1; } else { $end1 = strpos($line, ';', $pos + 1); $end2 = strpos($line, ',', $pos + 1); if ($end1 === false && $end2 === false) { $end = $len; } else if ($end1 === false) { $end = $end2; } else if ($end2 === false) { $end = $end1; } else { $end = $end1 < $end2 ? $end1 : $end2; } $pval = trim(substr($line, $pos, $end - $pos)); $pos = $end; } if ($pname == 'title*') { $parts = explode("'", $pval, 3); if (count($parts) != 3) { continue; } list($charset, $lang, $val) = $parts; $link[$pname][$lang] = $this->convertCharset( $charset, 'utf-8', urldecode($val) ); } else if ($pname == 'rel' || $pname == 'rev') { if (!isset($link[$pname])) { $link[$pname] = array(); } $link[$pname] = array_merge( $link[$pname], array_map('trim', explode(' ', $pval)) ); } else if (!isset($link[$pname])) { $link[$pname] = $pval; } if ($end >= $len || $line{$end} == ',') { break; } } if (count($link) > 0) { $extracted[] = $link; } } } return $extracted; } /** * Convert the given string from one into another charset. * Uses mb_convert_encoding or iconv if available. * * @param string $from Source charset the string is in * @param string $to Target character set * @param string $str String to convert * * @return string converted string */ protected function convertCharset($from, $to, $str) { if (function_exists('mb_convert_encoding')) { return mb_convert_encoding($str, $to, $from); } else if (function_exists('iconv')) { return iconv($from, $to, $str); } return $str; } } ?>