From: ArthurHoaro Date: Tue, 23 Jan 2018 17:41:38 +0000 (+0100) Subject: Merge pull request #977 from ArthurHoaro/feature/dl-filter X-Git-Tag: v0.9.4~7 X-Git-Url: https://git.immae.eu/?a=commitdiff_plain;h=d449f79a0d7ca808b891baf73b9e25ce7f7e48fe;hp=-c;p=github%2Fshaarli%2FShaarli.git Merge pull request #977 from ArthurHoaro/feature/dl-filter Extract the title/charset during page download, and check content type --- d449f79a0d7ca808b891baf73b9e25ce7f7e48fe diff --combined application/HttpUtils.php index c9371b55,2edf5ce2..83a4c5e2 --- a/application/HttpUtils.php +++ b/application/HttpUtils.php @@@ -3,9 -3,11 +3,11 @@@ * 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) + * @param string $url URL to get (http://...) + * @param int $timeout network timeout (in seconds) + * @param int $maxBytes maximum downloaded bytes (default: 4 MiB) + * @param callable|string $curlWriteFunction Optional callback called during the download (cURL CURLOPT_WRITEFUNCTION). + * Can be used to add download conditions on the headers (response code, content type, etc.). * * @return array HTTP response headers, downloaded content * @@@ -29,7 -31,7 +31,7 @@@ * @see http://stackoverflow.com/q/9183178 * @see http://stackoverflow.com/q/1462720 */ - function get_http_response($url, $timeout = 30, $maxBytes = 4194304) + function get_http_response($url, $timeout = 30, $maxBytes = 4194304, $curlWriteFunction = null) { $urlObj = new Url($url); $cleanUrl = $urlObj->idnToAscii(); @@@ -75,8 -77,12 +77,12 @@@ curl_setopt($ch, CURLOPT_TIMEOUT, $timeout); curl_setopt($ch, CURLOPT_USERAGENT, $userAgent); + if (is_callable($curlWriteFunction)) { + curl_setopt($ch, CURLOPT_WRITEFUNCTION, $curlWriteFunction); + } + // Max download size management - curl_setopt($ch, CURLOPT_BUFFERSIZE, 1024); + curl_setopt($ch, CURLOPT_BUFFERSIZE, 1024*16); curl_setopt($ch, CURLOPT_NOPROGRESS, false); curl_setopt($ch, CURLOPT_PROGRESSFUNCTION, function($arg0, $arg1, $arg2, $arg3, $arg4 = 0) use ($maxBytes) @@@ -302,13 -308,6 +308,13 @@@ function server_url($server $port = $server['HTTP_X_FORWARDED_PORT']; } + // This is a workaround for proxies that don't forward the scheme properly. + // Connecting over port 443 has to be in HTTPS. + // See https://github.com/shaarli/Shaarli/issues/1022 + if ($port == '443') { + $scheme = 'https'; + } + if (($scheme == 'http' && $port != '80') || ($scheme == 'https' && $port != '443') ) { diff --combined application/LinkUtils.php index e3d95d08,c0dd32a6..3705f7e9 --- a/application/LinkUtils.php +++ b/application/LinkUtils.php @@@ -1,60 -1,81 +1,81 @@@ (.*?)!is', $html, $matches)) { - return trim(str_replace("\n", '', $matches[1])); - } - return false; + /** + * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download). + * + * While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text' + * Then we extract the title and the charset and stop the download when it's done. + * + * @param resource $ch cURL resource + * @param string $data chunk of data being downloaded + * + * @return int|bool length of $data or false if we need to stop the download + */ + return function(&$ch, $data) use ($curlGetInfo, &$charset, &$title) { + $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE); + if (!empty($responseCode) && $responseCode != 200) { + return false; + } + $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE); + if (!empty($contentType) && strpos($contentType, 'text/html') === false) { + return false; + } + if (empty($charset)) { + $charset = header_extract_charset($contentType); + } + if (empty($charset)) { + $charset = html_extract_charset($data); + } + if (empty($title)) { + $title = html_extract_title($data); + } + // We got everything we want, stop the download. + if (!empty($responseCode) && !empty($contentType) && !empty($charset) && !empty($title)) { + return false; + } + + return strlen($data); + }; } /** - * Determine charset from downloaded page. - * Priority: - * 1. HTTP headers (Content type). - * 2. HTML content page (tag ). - * 3. Use a default charset (default: UTF-8). + * Extract title from an HTML document. * - * @param array $headers HTTP headers array. - * @param string $htmlContent HTML content where to look for charset. - * @param string $defaultCharset Default charset to apply if other methods failed. + * @param string $html HTML content where to look for a title. * - * @return string Determined charset. + * @return bool|string Extracted title if found, false otherwise. */ - function get_charset($headers, $htmlContent, $defaultCharset = 'utf-8') + function html_extract_title($html) { - if ($charset = headers_extract_charset($headers)) { - return $charset; - } - - if ($charset = html_extract_charset($htmlContent)) { - return $charset; + if (preg_match('!(.*?)!is', $html, $matches)) { + return trim(str_replace("\n", '', $matches[1])); } - - return $defaultCharset; + return false; } /** - * Extract charset from HTTP headers if it's defined. + * Extract charset from HTTP header if it's defined. * - * @param array $headers HTTP headers array. + * @param string $header HTTP header Content-Type line. * * @return bool|string Charset string if found (lowercase), false otherwise. */ - function headers_extract_charset($headers) + function header_extract_charset($header) { - if (! empty($headers['Content-Type']) && strpos($headers['Content-Type'], 'charset=') !== false) { - preg_match('/charset="?([^; ]+)/i', $headers['Content-Type'], $match); - if (! empty($match[1])) { - return strtolower(trim($match[1])); - } + preg_match('/charset="?([^; ]+)/i', $header, $match); + if (! empty($match[1])) { + return strtolower(trim($match[1])); } return false; @@@ -102,15 -123,14 +123,15 @@@ function count_private($links * * @param string $text input string. * @param string $redirector if a redirector is set, use it to gerenate links. + * @param bool $urlEncode Use `urlencode()` on the URL after the redirector or not. * * @return string returns $text with all links converted to HTML links. * * @see Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722 */ -function text2clickable($text, $redirector = '') +function text2clickable($text, $redirector = '', $urlEncode = true) { - $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[[:alnum:]]/?)!si'; + $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[a-z0-9\(\)]/?)!si'; if (empty($redirector)) { return preg_replace($regex, '$1', $text); @@@ -118,9 -138,8 +139,9 @@@ // Redirector is set, urlencode the final URL. return preg_replace_callback( $regex, - function ($matches) use ($redirector) { - return ''. $matches[1] .''; + function ($matches) use ($redirector, $urlEncode) { + $url = $urlEncode ? urlencode($matches[1]) : $matches[1]; + return ''. $matches[1] .''; }, $text ); @@@ -166,13 -185,12 +187,13 @@@ function space2nbsp($text * * @param string $description shaare's description. * @param string $redirector if a redirector is set, use it to gerenate links. + * @param bool $urlEncode Use `urlencode()` on the URL after the redirector or not. * @param string $indexUrl URL to Shaarli's index. - * + * @return string formatted description. */ -function format_description($description, $redirector = '', $indexUrl = '') { - return nl2br(space2nbsp(hashtag_autolink(text2clickable($description, $redirector), $indexUrl))); +function format_description($description, $redirector = '', $urlEncode = true, $indexUrl = '') { + return nl2br(space2nbsp(hashtag_autolink(text2clickable($description, $redirector, $urlEncode), $indexUrl))); } /** diff --combined index.php index 27335a36,ac51038d..d57789e6 --- a/index.php +++ b/index.php @@@ -64,6 -64,7 +64,6 @@@ require_once 'application/FeedBuilder.p require_once 'application/FileUtils.php'; require_once 'application/History.php'; require_once 'application/HttpUtils.php'; -require_once 'application/Languages.php'; require_once 'application/LinkDB.php'; require_once 'application/LinkFilter.php'; require_once 'application/LinkUtils.php'; @@@ -75,10 -76,8 +75,10 @@@ require_once 'application/Utils.php' require_once 'application/PluginManager.php'; require_once 'application/Router.php'; require_once 'application/Updater.php'; +use \Shaarli\Languages; use \Shaarli\ThemeUtils; use \Shaarli\Config\ConfigManager; +use \Shaarli\SessionManager; // Ensure the PHP version is supported try { @@@ -89,7 -88,7 +89,7 @@@ exit; } -define('shaarli_version', ApplicationUtils::getVersion(__DIR__ .'/'. ApplicationUtils::$VERSION_FILE)); +define('SHAARLI_VERSION', ApplicationUtils::getVersion(__DIR__ .'/'. ApplicationUtils::$VERSION_FILE)); // Force cookie path (but do not change lifetime) $cookie = session_get_cookie_params(); @@@ -116,23 -115,14 +116,23 @@@ if (session_id() == '') } // Regenerate session ID if invalid or not defined in cookie. -if (isset($_COOKIE['shaarli']) && !is_session_id_valid($_COOKIE['shaarli'])) { +if (isset($_COOKIE['shaarli']) && !SessionManager::checkId($_COOKIE['shaarli'])) { session_regenerate_id(true); $_COOKIE['shaarli'] = session_id(); } $conf = new ConfigManager(); +$sessionManager = new SessionManager($_SESSION, $conf); + +// Sniff browser language and set date format accordingly. +if (isset($_SERVER['HTTP_ACCEPT_LANGUAGE'])) { + autoLocale($_SERVER['HTTP_ACCEPT_LANGUAGE']); +} + +new Languages(setlocale(LC_MESSAGES, 0), $conf); + $conf->setEmpty('general.timezone', date_default_timezone_get()); -$conf->setEmpty('general.title', 'Shared links on '. escape(index_url($_SERVER))); +$conf->setEmpty('general.title', t('Shared links on '). escape(index_url($_SERVER))); RainTPL::$tpl_dir = $conf->get('resource.raintpl_tpl').'/'.$conf->get('resource.theme').'/'; // template directory RainTPL::$cache_dir = $conf->get('resource.raintpl_tmp'); // cache directory @@@ -154,7 -144,7 +154,7 @@@ if (! is_file($conf->getConfigFileExt() $errors = ApplicationUtils::checkResourcePermissions($conf); if ($errors != array()) { - $message = '

Insufficient permissions: