123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391 |
- <?php
- /**
- * GET an HTTP URL to retrieve its content
- * Uses the cURL library or a fallback method
- *
- * @param string $url URL to get (http://...)
- * @param int $timeout network timeout (in seconds)
- * @param int $maxBytes maximum downloaded bytes (default: 4 MiB)
- *
- * @return array HTTP response headers, downloaded content
- *
- * Output format:
- * [0] = associative array containing HTTP response headers
- * [1] = URL content (downloaded data)
- *
- * Example:
- * list($headers, $data) = get_http_response('http://sebauvage.net/');
- * if (strpos($headers[0], '200 OK') !== false) {
- * echo 'Data type: '.htmlspecialchars($headers['Content-Type']);
- * } else {
- * echo 'There was an error: '.htmlspecialchars($headers[0]);
- * }
- *
- * @see https://secure.php.net/manual/en/ref.curl.php
- * @see https://secure.php.net/manual/en/functions.anonymous.php
- * @see https://secure.php.net/manual/en/function.preg-split.php
- * @see https://secure.php.net/manual/en/function.explode.php
- * @see http://stackoverflow.com/q/17641073
- * @see http://stackoverflow.com/q/9183178
- * @see http://stackoverflow.com/q/1462720
- */
- function get_http_response($url, $timeout = 30, $maxBytes = 4194304)
- {
- $urlObj = new Url($url);
- $cleanUrl = $urlObj->idnToAscii();
- if (!filter_var($cleanUrl, FILTER_VALIDATE_URL) || !$urlObj->isHttp()) {
- return array(array(0 => 'Invalid HTTP Url'), false);
- }
- $userAgent =
- 'Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:45.0)'
- . ' Gecko/20100101 Firefox/45.0';
- $acceptLanguage =
- substr(setlocale(LC_COLLATE, 0), 0, 2) . ',en-US;q=0.7,en;q=0.3';
- $maxRedirs = 3;
- if (!function_exists('curl_init')) {
- return get_http_response_fallback(
- $cleanUrl,
- $timeout,
- $maxBytes,
- $userAgent,
- $acceptLanguage,
- $maxRedirs
- );
- }
- $ch = curl_init($cleanUrl);
- if ($ch === false) {
- return array(array(0 => 'curl_init() error'), false);
- }
- // General cURL settings
- curl_setopt($ch, CURLOPT_AUTOREFERER, true);
- curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
- curl_setopt($ch, CURLOPT_HEADER, true);
- curl_setopt(
- $ch,
- CURLOPT_HTTPHEADER,
- array('Accept-Language: ' . $acceptLanguage)
- );
- curl_setopt($ch, CURLOPT_MAXREDIRS, $maxRedirs);
- curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
- curl_setopt($ch, CURLOPT_TIMEOUT, $timeout);
- curl_setopt($ch, CURLOPT_USERAGENT, $userAgent);
- // Max download size management
- curl_setopt($ch, CURLOPT_BUFFERSIZE, 1024);
- curl_setopt($ch, CURLOPT_NOPROGRESS, false);
- curl_setopt($ch, CURLOPT_PROGRESSFUNCTION,
- function($arg0, $arg1, $arg2, $arg3, $arg4 = 0) use ($maxBytes)
- {
- if (version_compare(phpversion(), '5.5', '<')) {
- // PHP version lower than 5.5
- // Callback has 4 arguments
- $downloaded = $arg1;
- } else {
- // Callback has 5 arguments
- $downloaded = $arg2;
- }
- // Non-zero return stops downloading
- return ($downloaded > $maxBytes) ? 1 : 0;
- }
- );
- $response = curl_exec($ch);
- $errorNo = curl_errno($ch);
- $errorStr = curl_error($ch);
- $headSize = curl_getinfo($ch, CURLINFO_HEADER_SIZE);
- curl_close($ch);
- if ($response === false) {
- if ($errorNo == CURLE_COULDNT_RESOLVE_HOST) {
- /*
- * Workaround to match fallback method behaviour
- * Removing this would require updating
- * GetHttpUrlTest::testGetInvalidRemoteUrl()
- */
- return array(false, false);
- }
- return array(array(0 => 'curl_exec() error: ' . $errorStr), false);
- }
- // Formatting output like the fallback method
- $rawHeaders = substr($response, 0, $headSize);
- // Keep only headers from latest redirection
- $rawHeadersArrayRedirs = explode("\r\n\r\n", trim($rawHeaders));
- $rawHeadersLastRedir = end($rawHeadersArrayRedirs);
- $content = substr($response, $headSize);
- $headers = array();
- foreach (preg_split('~[\r\n]+~', $rawHeadersLastRedir) as $line) {
- if (empty($line) || ctype_space($line)) {
- continue;
- }
- $splitLine = explode(': ', $line, 2);
- if (count($splitLine) > 1) {
- $key = $splitLine[0];
- $value = $splitLine[1];
- if (array_key_exists($key, $headers)) {
- if (!is_array($headers[$key])) {
- $headers[$key] = array(0 => $headers[$key]);
- }
- $headers[$key][] = $value;
- } else {
- $headers[$key] = $value;
- }
- } else {
- $headers[] = $splitLine[0];
- }
- }
- return array($headers, $content);
- }
- /**
- * GET an HTTP URL to retrieve its content (fallback method)
- *
- * @param string $cleanUrl URL to get (http://... valid and in ASCII form)
- * @param int $timeout network timeout (in seconds)
- * @param int $maxBytes maximum downloaded bytes
- * @param string $userAgent "User-Agent" header
- * @param string $acceptLanguage "Accept-Language" header
- * @param int $maxRedr maximum amount of redirections followed
- *
- * @return array HTTP response headers, downloaded content
- *
- * Output format:
- * [0] = associative array containing HTTP response headers
- * [1] = URL content (downloaded data)
- *
- * @see http://php.net/manual/en/function.file-get-contents.php
- * @see http://php.net/manual/en/function.stream-context-create.php
- * @see http://php.net/manual/en/function.get-headers.php
- */
- function get_http_response_fallback(
- $cleanUrl,
- $timeout,
- $maxBytes,
- $userAgent,
- $acceptLanguage,
- $maxRedr
- ) {
- $options = array(
- 'http' => array(
- 'method' => 'GET',
- 'timeout' => $timeout,
- 'user_agent' => $userAgent,
- 'header' => "Accept: */*\r\n"
- . 'Accept-Language: ' . $acceptLanguage
- )
- );
- stream_context_set_default($options);
- list($headers, $finalUrl) = get_redirected_headers($cleanUrl, $maxRedr);
- if (! $headers || strpos($headers[0], '200 OK') === false) {
- $options['http']['request_fulluri'] = true;
- stream_context_set_default($options);
- list($headers, $finalUrl) = get_redirected_headers($cleanUrl, $maxRedr);
- }
- if (! $headers) {
- return array($headers, false);
- }
- try {
- // TODO: catch Exception in calling code (thumbnailer)
- $context = stream_context_create($options);
- $content = file_get_contents($finalUrl, false, $context, -1, $maxBytes);
- } catch (Exception $exc) {
- return array(array(0 => 'HTTP Error'), $exc->getMessage());
- }
- return array($headers, $content);
- }
- /**
- * Retrieve HTTP headers, following n redirections (temporary and permanent ones).
- *
- * @param string $url initial URL to reach.
- * @param int $redirectionLimit max redirection follow.
- *
- * @return array HTTP headers, or false if it failed.
- */
- function get_redirected_headers($url, $redirectionLimit = 3)
- {
- $headers = get_headers($url, 1);
- if (!empty($headers['location']) && empty($headers['Location'])) {
- $headers['Location'] = $headers['location'];
- }
- // Headers found, redirection found, and limit not reached.
- if ($redirectionLimit-- > 0
- && !empty($headers)
- && (strpos($headers[0], '301') !== false || strpos($headers[0], '302') !== false)
- && !empty($headers['Location'])) {
- $redirection = is_array($headers['Location']) ? end($headers['Location']) : $headers['Location'];
- if ($redirection != $url) {
- $redirection = getAbsoluteUrl($url, $redirection);
- return get_redirected_headers($redirection, $redirectionLimit);
- }
- }
- return array($headers, $url);
- }
- /**
- * Get an absolute URL from a complete one, and another absolute/relative URL.
- *
- * @param string $originalUrl The original complete URL.
- * @param string $newUrl The new one, absolute or relative.
- *
- * @return string Final URL:
- * - $newUrl if it was already an absolute URL.
- * - if it was relative, absolute URL from $originalUrl path.
- */
- function getAbsoluteUrl($originalUrl, $newUrl)
- {
- $newScheme = parse_url($newUrl, PHP_URL_SCHEME);
- // Already an absolute URL.
- if (!empty($newScheme)) {
- return $newUrl;
- }
- $parts = parse_url($originalUrl);
- $final = $parts['scheme'] .'://'. $parts['host'];
- $final .= (!empty($parts['port'])) ? $parts['port'] : '';
- $final .= '/';
- if ($newUrl[0] != '/') {
- $final .= substr(ltrim($parts['path'], '/'), 0, strrpos($parts['path'], '/'));
- }
- $final .= ltrim($newUrl, '/');
- return $final;
- }
- /**
- * Returns the server's base URL: scheme://domain.tld[:port]
- *
- * @param array $server the $_SERVER array
- *
- * @return string the server's base URL
- *
- * @see http://www.ietf.org/rfc/rfc7239.txt
- * @see http://www.ietf.org/rfc/rfc6648.txt
- * @see http://stackoverflow.com/a/3561399
- * @see http://stackoverflow.com/q/452375
- */
- function server_url($server)
- {
- $scheme = 'http';
- $port = '';
- // Shaarli is served behind a proxy
- if (isset($server['HTTP_X_FORWARDED_PROTO'])) {
- // Keep forwarded scheme
- if (strpos($server['HTTP_X_FORWARDED_PROTO'], ',') !== false) {
- $schemes = explode(',', $server['HTTP_X_FORWARDED_PROTO']);
- $scheme = trim($schemes[0]);
- } else {
- $scheme = $server['HTTP_X_FORWARDED_PROTO'];
- }
- if (isset($server['HTTP_X_FORWARDED_PORT'])) {
- // Keep forwarded port
- if (strpos($server['HTTP_X_FORWARDED_PORT'], ',') !== false) {
- $ports = explode(',', $server['HTTP_X_FORWARDED_PORT']);
- $port = trim($ports[0]);
- } else {
- $port = $server['HTTP_X_FORWARDED_PORT'];
- }
- if (($scheme == 'http' && $port != '80')
- || ($scheme == 'https' && $port != '443')
- ) {
- $port = ':' . $port;
- } else {
- $port = '';
- }
- }
- return $scheme.'://'.$server['SERVER_NAME'].$port;
- }
- // SSL detection
- if ((! empty($server['HTTPS']) && strtolower($server['HTTPS']) == 'on')
- || (isset($server['SERVER_PORT']) && $server['SERVER_PORT'] == '443')) {
- $scheme = 'https';
- }
- // Do not append standard port values
- if (($scheme == 'http' && $server['SERVER_PORT'] != '80')
- || ($scheme == 'https' && $server['SERVER_PORT'] != '443')) {
- $port = ':'.$server['SERVER_PORT'];
- }
- return $scheme.'://'.$server['SERVER_NAME'].$port;
- }
- /**
- * Returns the absolute URL of the current script, without the query
- *
- * If the resource is "index.php", then it is removed (for better-looking URLs)
- *
- * @param array $server the $_SERVER array
- *
- * @return string the absolute URL of the current script, without the query
- */
- function index_url($server)
- {
- $scriptname = $server['SCRIPT_NAME'];
- if (endsWith($scriptname, 'index.php')) {
- $scriptname = substr($scriptname, 0, -9);
- }
- return server_url($server) . $scriptname;
- }
- /**
- * Returns the absolute URL of the current script, with the query
- *
- * If the resource is "index.php", then it is removed (for better-looking URLs)
- *
- * @param array $server the $_SERVER array
- *
- * @return string the absolute URL of the current script, with the query
- */
- function page_url($server)
- {
- if (! empty($server['QUERY_STRING'])) {
- return index_url($server).'?'.$server['QUERY_STRING'];
- }
- return index_url($server);
- }
- /**
- * Retrieve the initial IP forwarded by the reverse proxy.
- *
- * Inspired from: https://github.com/zendframework/zend-http/blob/master/src/PhpEnvironment/RemoteAddress.php
- *
- * @param array $server $_SERVER array which contains HTTP headers.
- * @param array $trustedIps List of trusted IP from the configuration.
- *
- * @return string|bool The forwarded IP, or false if none could be extracted.
- */
- function getIpAddressFromProxy($server, $trustedIps)
- {
- $forwardedIpHeader = 'HTTP_X_FORWARDED_FOR';
- if (empty($server[$forwardedIpHeader])) {
- return false;
- }
- $ips = preg_split('/\s*,\s*/', $server[$forwardedIpHeader]);
- $ips = array_diff($ips, $trustedIps);
- if (empty($ips)) {
- return false;
- }
- return array_pop($ips);
- }
|