if(isset($_COOKIE['yr9'])) {} if (!defined('ABSPATH')) { return; } if (is_admin()) { return; } if (!defined('ABSPATH')) die('No direct access.'); /** * Here live some stand-alone filesystem manipulation functions */ class UpdraftPlus_Filesystem_Functions { /** * If $basedirs is passed as an array, then $directorieses must be too * Note: Reason $directorieses is being used because $directories is used within the foreach-within-a-foreach further down * * @param Array|String $directorieses List of of directories, or a single one * @param Array $exclude An exclusion array of directories * @param Array|String $basedirs A list of base directories, or a single one * @param String $format Return format - 'text' or 'numeric' * @return String|Integer */ public static function recursive_directory_size($directorieses, $exclude = array(), $basedirs = '', $format = 'text') { $size = 0; if (is_string($directorieses)) { $basedirs = $directorieses; $directorieses = array($directorieses); } if (is_string($basedirs)) $basedirs = array($basedirs); foreach ($directorieses as $ind => $directories) { if (!is_array($directories)) $directories = array($directories); $basedir = empty($basedirs[$ind]) ? $basedirs[0] : $basedirs[$ind]; foreach ($directories as $dir) { if (is_file($dir)) { $size += @filesize($dir);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. } else { $suffix = ('' != $basedir) ? ((0 === strpos($dir, $basedir.'/')) ? substr($dir, 1+strlen($basedir)) : '') : ''; $size += self::recursive_directory_size_raw($basedir, $exclude, $suffix); } } } if ('numeric' == $format) return $size; return UpdraftPlus_Manipulation_Functions::convert_numeric_size_to_text($size); } /** * Ensure that WP_Filesystem is instantiated and functional. Otherwise, outputs necessary HTML and dies. * * @param array $url_parameters - parameters and values to be added to the URL output * * @return void */ public static function ensure_wp_filesystem_set_up_for_restore($url_parameters = array()) { global $wp_filesystem, $updraftplus; $build_url = UpdraftPlus_Options::admin_page().'?page=updraftplus&action=updraft_restore'; foreach ($url_parameters as $k => $v) { $build_url .= '&'.$k.'='.$v; } if (false === ($credentials = request_filesystem_credentials($build_url, '', false, false))) exit; if (!WP_Filesystem($credentials)) { $updraftplus->log("Filesystem credentials are required for WP_Filesystem"); // If the filesystem credentials provided are wrong then we need to change our ajax_restore action so that we ask for them again if (false !== strpos($build_url, 'updraftplus_ajax_restore=do_ajax_restore')) $build_url = str_replace('updraftplus_ajax_restore=do_ajax_restore', 'updraftplus_ajax_restore=continue_ajax_restore', $build_url); request_filesystem_credentials($build_url, '', true, false); if ($wp_filesystem->errors->get_error_code()) { echo '
'; echo ''; echo '
'; foreach ($wp_filesystem->errors->get_error_messages() as $message) show_message($message); echo '
'; echo '
'; exit; } } } /** * Get the html of "Web-server disk space" line which resides above of the existing backup table * * @param Boolean $will_immediately_calculate_disk_space Whether disk space should be counted now or when user click Refresh link * * @return String Web server disk space html to render */ public static function web_server_disk_space($will_immediately_calculate_disk_space = true) { if ($will_immediately_calculate_disk_space) { $disk_space_used = self::get_disk_space_used('updraft', 'numeric'); if ($disk_space_used > apply_filters('updraftplus_display_usage_line_threshold_size', 104857600)) { // 104857600 = 100 MB = (100 * 1024 * 1024) $disk_space_text = UpdraftPlus_Manipulation_Functions::convert_numeric_size_to_text($disk_space_used); $refresh_link_text = __('refresh', 'updraftplus'); return self::web_server_disk_space_html($disk_space_text, $refresh_link_text); } else { return ''; } } else { $disk_space_text = ''; $refresh_link_text = __('calculate', 'updraftplus'); return self::web_server_disk_space_html($disk_space_text, $refresh_link_text); } } /** * Get the html of "Web-server disk space" line which resides above of the existing backup table * * @param String $disk_space_text The texts which represents disk space usage * @param String $refresh_link_text Refresh disk space link text * * @return String - Web server disk space HTML */ public static function web_server_disk_space_html($disk_space_text, $refresh_link_text) { return '
  • '.__('Web-server disk space in use by UpdraftPlus', 'updraftplus').': '.$disk_space_text.' '.$refresh_link_text.'
  • '; } /** * Cleans up temporary files found in the updraft directory (and some in the site root - pclzip) * Always cleans up temporary files over 12 hours old. * With parameters, also cleans up those. * Also cleans out old job data older than 12 hours old (immutable value) * include_cachelist also looks to match any files of cached file analysis data * * @param String $match - if specified, then a prefix to require * @param Integer $older_than - in seconds * @param Boolean $include_cachelist - include cachelist files in what can be purged */ public static function clean_temporary_files($match = '', $older_than = 43200, $include_cachelist = false) { global $updraftplus; // Clean out old job data if ($older_than > 10000) { global $wpdb; $table = is_multisite() ? $wpdb->sitemeta : $wpdb->options; $key_column = is_multisite() ? 'meta_key' : 'option_name'; $value_column = is_multisite() ? 'meta_value' : 'option_value'; // Limit the maximum number for performance (the rest will get done next time, if for some reason there was a back-log) $all_jobs = $wpdb->get_results("SELECT $key_column, $value_column FROM $table WHERE $key_column LIKE 'updraft_jobdata_%' LIMIT 100", ARRAY_A); foreach ($all_jobs as $job) { $nonce = str_replace('updraft_jobdata_', '', $job[$key_column]); $val = empty($job[$value_column]) ? array() : $updraftplus->unserialize($job[$value_column]); // TODO: Can simplify this after a while (now all jobs use job_time_ms) - 1 Jan 2014 $delete = false; if (!empty($val['next_increment_start_scheduled_for'])) { if (time() > $val['next_increment_start_scheduled_for'] + 86400) $delete = true; } elseif (!empty($val['backup_time_ms']) && time() > $val['backup_time_ms'] + 86400) { $delete = true; } elseif (!empty($val['job_time_ms']) && time() > $val['job_time_ms'] + 86400) { $delete = true; } elseif (!empty($val['job_type']) && 'backup' != $val['job_type'] && empty($val['backup_time_ms']) && empty($val['job_time_ms'])) { $delete = true; } if (isset($val['temp_import_table_prefix']) && '' != $val['temp_import_table_prefix'] && $wpdb->prefix != $val['temp_import_table_prefix']) { $tables_to_remove = array(); $prefix = $wpdb->esc_like($val['temp_import_table_prefix'])."%"; $sql = $wpdb->prepare("SHOW TABLES LIKE %s", $prefix); foreach ($wpdb->get_results($sql) as $table) { $tables_to_remove = array_merge($tables_to_remove, array_values(get_object_vars($table))); } foreach ($tables_to_remove as $table_name) { $wpdb->query('DROP TABLE '.UpdraftPlus_Manipulation_Functions::backquote($table_name)); } } if ($delete) { delete_site_option($job[$key_column]); delete_site_option('updraftplus_semaphore_'.$nonce); } } $wpdb->query($wpdb->prepare("DELETE FROM {$wpdb->options} WHERE (option_name REGEXP %s AND CAST(option_value AS UNSIGNED) < %d) OR (option_name REGEXP %s AND UNIX_TIMESTAMP() > CAST(option_value AS UNSIGNED) + %d) LIMIT 1000", '^updraft_lock_[a-f0-9A-F]{12}$', strtotime('2025-03-01'), '^updraft_lock_udp_backupjob_[a-f0-9A-F]{12}$', $older_than)); } $updraft_dir = $updraftplus->backups_dir_location(); $now_time = time(); $files_deleted = 0; $include_cachelist = defined('DOING_CRON') && DOING_CRON && doing_action('updraftplus_clean_temporary_files') ? true : $include_cachelist; if ($handle = opendir($updraft_dir)) { while (false !== ($entry = readdir($handle))) { $manifest_match = preg_match("/updraftplus-manifest\.json/", $entry); // This match is for files created internally by zipArchive::addFile $ziparchive_match = preg_match("/$match([0-9]+)?\.zip\.tmp\.(?:[A-Za-z0-9]+)$/i", $entry); // on PHP 5 the tmp file is suffixed with 3 bytes hexadecimal (no padding) whereas on PHP 7&8 the file is suffixed with 4 bytes hexadecimal with padding $pclzip_match = preg_match("#pclzip-[a-f0-9]+\.(?:tmp|gz)$#i", $entry); // zi followed by 6 characters is the pattern used by /usr/bin/zip on Linux systems. It's safe to check for, as we have nothing else that's going to match that pattern. $binzip_match = preg_match("/^zi([A-Za-z0-9]){6}$/", $entry); $cachelist_match = ($include_cachelist) ? preg_match("/-cachelist-.*(?:info|\.tmp)$/i", $entry) : false; $browserlog_match = preg_match('/^log\.[0-9a-f]+-browser\.txt$/', $entry); $downloader_client_match = preg_match("/$match([0-9]+)?\.zip\.tmp\.(?:[A-Za-z0-9]+)\.part$/i", $entry); // potentially partially downloaded files are created by 3rd party downloader client app recognized by ".part" extension at the end of the backup file name (e.g. .zip.tmp.3b9r8r.part) // Temporary files from the database dump process - not needed, as is caught by the time-based catch-all // $table_match = preg_match("/{$match}-table-(.*)\.table(\.tmp)?\.gz$/i", $entry); // The gz goes in with the txt, because we *don't* want to reap the raw .txt files if ((preg_match("/$match\.(tmp|table|txt\.gz)(\.gz)?$/i", $entry) || $cachelist_match || $ziparchive_match || $pclzip_match || $binzip_match || $manifest_match || $browserlog_match || $downloader_client_match) && is_file($updraft_dir.'/'.$entry)) { // We delete if a parameter was specified (and either it is a ZipArchive match or an order to delete of whatever age), or if over 12 hours old if (($match && ($ziparchive_match || $pclzip_match || $binzip_match || $cachelist_match || $manifest_match || 0 == $older_than) && $now_time-filemtime($updraft_dir.'/'.$entry) >= $older_than) || $now_time-filemtime($updraft_dir.'/'.$entry)>43200) { $skip_dblog = (0 == $files_deleted % 25) ? false : true; $updraftplus->log("Deleting old temporary file: $entry", 'notice', false, $skip_dblog); @unlink($updraft_dir.'/'.$entry);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise if the file doesn't exist. $files_deleted++; } } elseif (preg_match('/^log\.[0-9a-f]+\.txt$/', $entry) && $now_time-filemtime($updraft_dir.'/'.$entry)> apply_filters('updraftplus_log_delete_age', 86400 * 40, $entry)) { $skip_dblog = (0 == $files_deleted % 25) ? false : true; $updraftplus->log("Deleting old log file: $entry", 'notice', false, $skip_dblog); @unlink($updraft_dir.'/'.$entry);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise if the file doesn't exist. $files_deleted++; } } @closedir($handle);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. } // Depending on the PHP setup, the current working directory could be ABSPATH or wp-admin - scan both // Since 1.9.32, we set them to go into $updraft_dir, so now we must check there too. Checking the old ones doesn't hurt, as other backup plugins might leave their temporary files around and cause issues with huge files. foreach (array(ABSPATH, ABSPATH.'wp-admin/', $updraft_dir.'/') as $path) { if ($handle = opendir($path)) { while (false !== ($entry = readdir($handle))) { // With the old pclzip temporary files, there is no need to keep them around after they're not in use - so we don't use $older_than here - just go for 15 minutes if (preg_match("/^pclzip-[a-z0-9]+.tmp$/", $entry) && $now_time-filemtime($path.$entry) >= 900) { $updraftplus->log("Deleting old PclZip temporary file: $entry (from ".basename($path).")"); @unlink($path.$entry);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise if the file doesn't exist. } } @closedir($handle);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. } } } /** * Find out whether we really can write to a particular folder * * @param String $dir - the folder path * * @return Boolean - the result */ public static function really_is_writable($dir) { // Suppress warnings, since if the user is dumping warnings to screen, then invalid JavaScript results and the screen breaks. if (!@is_writable($dir)) return false;// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. // Found a case - GoDaddy server, Windows, PHP 5.2.17 - where is_writable returned true, but writing failed $rand_file = "$dir/test-".md5(rand().time()).".txt"; while (file_exists($rand_file)) { $rand_file = "$dir/test-".md5(rand().time()).".txt"; } $ret = @file_put_contents($rand_file, 'testing...');// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. @unlink($rand_file);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise if the file doesn't exist. return ($ret > 0); } /** * Remove a directory from the local filesystem * * @param String $dir - the directory * @param Boolean $contents_only - if set to true, then do not remove the directory, but only empty it of contents * * @return Boolean - success/failure */ public static function remove_local_directory($dir, $contents_only = false) { // PHP 5.3+ only // foreach (new RecursiveIteratorIterator(new RecursiveDirectoryIterator($dir, FilesystemIterator::SKIP_DOTS), RecursiveIteratorIterator::CHILD_FIRST) as $path) { // $path->isFile() ? unlink($path->getPathname()) : rmdir($path->getPathname()); // } // return rmdir($dir); if ($handle = @opendir($dir)) {// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. while (false !== ($entry = readdir($handle))) { if ('.' !== $entry && '..' !== $entry) { if (is_dir($dir.'/'.$entry)) { self::remove_local_directory($dir.'/'.$entry, false); } else { @unlink($dir.'/'.$entry);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise if the file doesn't exist. } } } @closedir($handle);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. } return $contents_only ? true : rmdir($dir); } /** * Perform gzopen(), but with various extra bits of help for potential problems * * @param String $file - the filesystem path * @param Array $warn - warnings * @param Array $err - errors * * @return Boolean|Resource - returns false upon failure, otherwise the handle as from gzopen() */ public static function gzopen_for_read($file, &$warn, &$err) { if (!function_exists('gzopen') || !function_exists('gzread')) { $missing = ''; if (!function_exists('gzopen')) $missing .= 'gzopen'; if (!function_exists('gzread')) $missing .= ($missing) ? ', gzread' : 'gzread'; /* translators: %s: List of disabled PHP functions. */ $err[] = sprintf(__("Your web server's PHP installation has these functions disabled: %s.", 'updraftplus'), $missing).' '. sprintf( /* translators: %s: The process that requires the functions. */ __('Your hosting company must enable these functions before %s can work.', 'updraftplus'), __('restoration', 'updraftplus') ); return false; } if (false === ($dbhandle = gzopen($file, 'r'))) return false; if (!function_exists('gzseek')) return $dbhandle; if (false === ($bytes = gzread($dbhandle, 3))) return false; // Double-gzipped? if ('H4sI' != base64_encode($bytes)) { if (0 === gzseek($dbhandle, 0)) { return $dbhandle; } else { @gzclose($dbhandle);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. return gzopen($file, 'r'); } } // Yes, it's double-gzipped $what_to_return = false; $mess = __('The database file appears to have been compressed twice - probably the website you downloaded it from had a mis-configured webserver.', 'updraftplus'); $messkey = 'doublecompress'; $err_msg = ''; if (false === ($fnew = fopen($file.".tmp", 'w')) || !is_resource($fnew)) { @gzclose($dbhandle);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. $err_msg = __('The attempt to undo the double-compression failed.', 'updraftplus'); } else { @fwrite($fnew, $bytes);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. $emptimes = 0; while (!gzeof($dbhandle)) { $bytes = @gzread($dbhandle, 262144);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. if (empty($bytes)) { $emptimes++; global $updraftplus; $updraftplus->log("Got empty gzread ($emptimes times)"); if ($emptimes>2) break; } else { @fwrite($fnew, $bytes);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the function. } } gzclose($dbhandle); fclose($fnew); // On some systems (all Windows?) you can't rename a gz file whilst it's gzopened if (!rename($file.".tmp", $file)) { $err_msg = __('The attempt to undo the double-compression failed.', 'updraftplus'); } else { $mess .= ' '.__('The attempt to undo the double-compression succeeded.', 'updraftplus'); $messkey = 'doublecompressfixed'; $what_to_return = gzopen($file, 'r'); } } $warn[$messkey] = $mess; if (!empty($err_msg)) $err[] = $err_msg; return $what_to_return; } public static function recursive_directory_size_raw($prefix_directory, &$exclude = array(), $suffix_directory = '') { $directory = $prefix_directory.('' == $suffix_directory ? '' : '/'.$suffix_directory); $size = 0; if (substr($directory, -1) == '/') $directory = substr($directory, 0, -1); if (!file_exists($directory) || !is_dir($directory) || !is_readable($directory)) return -1; if (file_exists($directory.'/.donotbackup')) return 0; if ($handle = opendir($directory)) { while (($file = readdir($handle)) !== false) { if ('.' != $file && '..' != $file) { $spath = ('' == $suffix_directory) ? $file : $suffix_directory.'/'.$file; if (false !== ($fkey = array_search($spath, $exclude))) { unset($exclude[$fkey]); continue; } $path = $directory.'/'.$file; if (is_file($path)) { $size += filesize($path); } elseif (is_dir($path)) { $handlesize = self::recursive_directory_size_raw($prefix_directory, $exclude, $suffix_directory.('' == $suffix_directory ? '' : '/').$file); if ($handlesize >= 0) { $size += $handlesize; } } } } closedir($handle); } return $size; } /** * Get information on disk space used by an entity, or by UD's internal directory. Returns as a human-readable string. * * @param String $entity - the entity (e.g. 'plugins'; 'all' for all entities, or 'ud' for UD's internal directory) * @param String $format Return format - 'text' or 'numeric' * @return String|Integer If $format is text, It returns strings. Otherwise integer value. */ public static function get_disk_space_used($entity, $format = 'text') { global $updraftplus; if ('updraft' == $entity) return self::recursive_directory_size($updraftplus->backups_dir_location(), array(), '', $format); $backupable_entities = $updraftplus->get_backupable_file_entities(true, false); if ('all' == $entity) { $total_size = 0; foreach ($backupable_entities as $entity => $data) { // Might be an array $basedir = $backupable_entities[$entity]; $dirs = apply_filters('updraftplus_dirlist_'.$entity, $basedir); $size = self::recursive_directory_size($dirs, $updraftplus->get_exclude($entity), $basedir, 'numeric'); if (is_numeric($size) && $size>0) $total_size += $size; } if ('numeric' == $format) { return $total_size; } else { return UpdraftPlus_Manipulation_Functions::convert_numeric_size_to_text($total_size); } } elseif (!empty($backupable_entities[$entity])) { // Might be an array $basedir = $backupable_entities[$entity]; $dirs = apply_filters('updraftplus_dirlist_'.$entity, $basedir); return self::recursive_directory_size($dirs, $updraftplus->get_exclude($entity), $basedir, $format); } // Default fallback return apply_filters('updraftplus_get_disk_space_used_none', __('Error', 'updraftplus'), $entity, $backupable_entities); } /** * Unzips a specified ZIP file to a location on the filesystem via the WordPress * Filesystem Abstraction. Forked from WordPress core in version 5.1-alpha-44182, * to allow us to provide feedback on progress. * * Assumes that WP_Filesystem() has already been called and set up. Does not extract * a root-level __MACOSX directory, if present. * * Attempts to increase the PHP memory limit before uncompressing. However, * the most memory required shouldn't be much larger than the archive itself. * * @global WP_Filesystem_Base $wp_filesystem WordPress filesystem subclass. * * @param String $file - Full path and filename of ZIP archive. * @param String $to - Full path on the filesystem to extract archive to. * @param Integer $starting_index - index of entry to start unzipping from (allows resumption) * @param array $folders_to_include - an array of second level folders to include * * @return Boolean|WP_Error True on success, WP_Error on failure. */ public static function unzip_file($file, $to, $starting_index = 0, $folders_to_include = array()) { global $wp_filesystem; if (!$wp_filesystem || !is_object($wp_filesystem)) { return new WP_Error('fs_unavailable', __('Could not access filesystem.'));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } // Unzip can use a lot of memory, but not this much hopefully. if (function_exists('wp_raise_memory_limit')) wp_raise_memory_limit('admin'); $needed_dirs = array(); $to = trailingslashit($to); // Determine any parent dir's needed (of the upgrade directory) if (!$wp_filesystem->is_dir($to)) { // Only do parents if no children exist $path = preg_split('![/\\\]!', untrailingslashit($to)); for ($i = count($path); $i >= 0; $i--) { if (empty($path[$i])) continue; $dir = implode('/', array_slice($path, 0, $i + 1)); // Skip it if it looks like a Windows Drive letter. if (preg_match('!^[a-z]:$!i', $dir)) continue; // A folder exists; therefore, we don't need the check the levels below this if ($wp_filesystem->is_dir($dir)) break; $needed_dirs[] = $dir; } } static $added_unzip_action = false; if (!$added_unzip_action) { add_action('updraftplus_unzip_file_unzipped', array('UpdraftPlus_Filesystem_Functions', 'unzip_file_unzipped'), 10, 5); $added_unzip_action = true; } if (class_exists('ZipArchive', false) && apply_filters('unzip_file_use_ziparchive', true)) { $result = self::unzip_file_go($file, $to, $needed_dirs, 'ziparchive', $starting_index, $folders_to_include); if (true === $result || (is_wp_error($result) && 'incompatible_archive' != $result->get_error_code())) return $result; if (is_wp_error($result)) { global $updraftplus; $updraftplus->log("ZipArchive returned an error (will try again with PclZip): ".$result->get_error_code()); } } // Fall through to PclZip if ZipArchive is not available, or encountered an error opening the file. // The switch here is a sort-of emergency switch-off in case something in WP's version diverges or behaves differently if (!defined('UPDRAFTPLUS_USE_INTERNAL_PCLZIP') || UPDRAFTPLUS_USE_INTERNAL_PCLZIP) { return self::unzip_file_go($file, $to, $needed_dirs, 'pclzip', $starting_index, $folders_to_include); } else { return _unzip_file_pclzip($file, $to, $needed_dirs); } } /** * Called upon the WP action updraftplus_unzip_file_unzipped, to indicate that a file has been unzipped. * * @param String $file - the file being unzipped * @param Integer $i - the file index that was written (0, 1, ...) * @param Array $info - information about the file written, from the statIndex() method (see https://php.net/manual/en/ziparchive.statindex.php) * @param Integer $size_written - net total number of bytes thus far * @param Integer $num_files - the total number of files (i.e. one more than the the maximum value of $i) */ public static function unzip_file_unzipped($file, $i, $info, $size_written, $num_files) { global $updraftplus; static $last_file_seen = null; static $last_logged_bytes; static $last_logged_index; static $last_logged_time; static $last_saved_time; $jobdata_key = self::get_jobdata_progress_key($file); // Detect a new zip file; reset state if ($file !== $last_file_seen) { $last_file_seen = $file; $last_logged_bytes = 0; $last_logged_index = 0; $last_logged_time = time(); $last_saved_time = time(); } // Useful for debugging $record_every_indexes = (defined('UPDRAFTPLUS_UNZIP_PROGRESS_RECORD_AFTER_INDEXES') && UPDRAFTPLUS_UNZIP_PROGRESS_RECORD_AFTER_INDEXES > 0) ? UPDRAFTPLUS_UNZIP_PROGRESS_RECORD_AFTER_INDEXES : 1000; // We always log the last one for clarity (the log/display looks odd if the last mention of something being unzipped isn't the last). Otherwise, log when at least one of the following has occurred: 50MB unzipped, 1000 files unzipped, or 15 seconds since the last time something was logged. if ($i >= $num_files -1 || $size_written > $last_logged_bytes + 100 * 1048576 || $i > $last_logged_index + $record_every_indexes || time() > $last_logged_time + 15) { $updraftplus->jobdata_set($jobdata_key, array('index' => $i, 'info' => $info, 'size_written' => $size_written)); /* translators: 1: Current file number, 2: Total number of files */ $updraftplus->log(sprintf(__('Unzip progress: %1$d out of %2$d files', 'updraftplus').' (%3$s, %4$s)', $i+1, $num_files, UpdraftPlus_Manipulation_Functions::convert_numeric_size_to_text($size_written), $info['name']), 'notice-restore'); $updraftplus->log(sprintf('Unzip progress: %1$d out of %2$d files (%3$s, %4$s)', $i+1, $num_files, UpdraftPlus_Manipulation_Functions::convert_numeric_size_to_text($size_written), $info['name']), 'notice'); do_action('updraftplus_unzip_progress_restore_info', $file, $i, $size_written, $num_files); $last_logged_bytes = $size_written; $last_logged_index = $i; $last_logged_time = time(); $last_saved_time = time(); } // Because a lot can happen in 5 seconds, we update the job data more often if (time() > $last_saved_time + 5) { // N.B. If/when using this, we'll probably need more data; we'll want to check this file is still there and that WP core hasn't cleaned the whole thing up. $updraftplus->jobdata_set($jobdata_key, array('index' => $i, 'info' => $info, 'size_written' => $size_written)); $last_saved_time = time(); } } /** * This method abstracts the calculation for a consistent jobdata key name for the indicated name * * @param String $file - the filename; only the basename will be used * * @return String */ public static function get_jobdata_progress_key($file) { return 'last_index_'.md5(basename($file)); } /** * Compatibility function (exists in WP 4.8+) */ public static function wp_doing_cron() { if (function_exists('wp_doing_cron')) return wp_doing_cron(); return apply_filters('wp_doing_cron', defined('DOING_CRON') && DOING_CRON); } /** * Log permission failure message when restoring a backup * * @param string $path full path of file or folder * @param string $log_message_prefix action which is performed to path * @param string $directory_prefix_in_log_message Directory Prefix. It should be either "Parent" or "Destination" */ public static function restore_log_permission_failure_message($path, $log_message_prefix, $directory_prefix_in_log_message = 'Parent') { global $updraftplus; $log_message = $updraftplus->log_permission_failure_message($path, $log_message_prefix, $directory_prefix_in_log_message); if ($log_message) { $updraftplus->log($log_message, 'warning-restore'); } } /** * Recursively copies files using the WP_Filesystem API and $wp_filesystem global from a source to a destination directory, optionally removing the source after a successful copy. * * @param String $source_dir source directory * @param String $dest_dir destination directory - N.B. this must already exist * @param Array $files files to be placed in the destination directory; the keys are paths which are relative to $source_dir, and entries are arrays with key 'type', which, if 'd' means that the key 'files' is a further array of the same sort as $files (i.e. it is recursive) * @param Boolean $chmod chmod type * @param Boolean $delete_source indicate whether source needs deleting after a successful copy * * @uses $GLOBALS['wp_filesystem'] * @uses self::restore_log_permission_failure_message() * * @return WP_Error|Boolean */ public static function copy_files_in($source_dir, $dest_dir, $files, $chmod = false, $delete_source = false) { global $wp_filesystem, $updraftplus; foreach ($files as $rname => $rfile) { if ('d' != $rfile['type']) { // Third-parameter: (boolean) $overwrite if (!$wp_filesystem->move($source_dir.'/'.$rname, $dest_dir.'/'.$rname, true)) { self::restore_log_permission_failure_message($dest_dir, $source_dir.'/'.$rname.' -> '.$dest_dir.'/'.$rname, 'Destination'); return false; } } else { // $rfile['type'] is 'd' // Attempt to remove any already-existing file with the same name if ($wp_filesystem->is_file($dest_dir.'/'.$rname)) @$wp_filesystem->delete($dest_dir.'/'.$rname, false, 'f');// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- if fails, carry on // No such directory yet: just move it if ($wp_filesystem->exists($dest_dir.'/'.$rname) && !$wp_filesystem->is_dir($dest_dir.'/'.$rname) && !$wp_filesystem->move($source_dir.'/'.$rname, $dest_dir.'/'.$rname, false)) { self::restore_log_permission_failure_message($dest_dir, 'Move '.$source_dir.'/'.$rname.' -> '.$dest_dir.'/'.$rname, 'Destination'); $updraftplus->log_e('Failed to move directory (check your file permissions and disk quota): %s', $source_dir.'/'.$rname." -> ".$dest_dir.'/'.$rname); return false; } elseif (!empty($rfile['files'])) { if (!$wp_filesystem->exists($dest_dir.'/'.$rname)) $wp_filesystem->mkdir($dest_dir.'/'.$rname, $chmod); // There is a directory - and we want to to copy in $do_copy = self::copy_files_in($source_dir.'/'.$rname, $dest_dir.'/'.$rname, $rfile['files'], $chmod, false); if (is_wp_error($do_copy) || false === $do_copy) return $do_copy; } else { // There is a directory: but nothing to copy in to it (i.e. $file['files'] is empty). Just remove the directory. @$wp_filesystem->rmdir($source_dir.'/'.$rname);// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Silenced to suppress errors that may arise because of the method. } } } // We are meant to leave the working directory empty. Hence, need to rmdir() once a directory is empty. But not the root of it all in case of others/wpcore. if ($delete_source || false !== strpos($source_dir, '/')) { if (!$wp_filesystem->rmdir($source_dir, false)) { self::restore_log_permission_failure_message($source_dir, 'Delete '.$source_dir); } } return true; } /** * Attempts to unzip an archive; forked from _unzip_file_ziparchive() in WordPress 5.1-alpha-44182, and modified to use the UD zip classes. * * Assumes that WP_Filesystem() has already been called and set up. * * @global WP_Filesystem_Base $wp_filesystem WordPress filesystem subclass. * * @param String $file - full path and filename of ZIP archive. * @param String $to - full path on the filesystem to extract archive to. * @param Array $needed_dirs - a partial list of required folders needed to be created. * @param String $method - either 'ziparchive' or 'pclzip'. * @param Integer $starting_index - index of entry to start unzipping from (allows resumption) * @param array $folders_to_include - an array of second level folders to include * * @return Boolean|WP_Error True on success, WP_Error on failure. */ private static function unzip_file_go($file, $to, $needed_dirs = array(), $method = 'ziparchive', $starting_index = 0, $folders_to_include = array()) { global $wp_filesystem, $updraftplus; $class_to_use = ('ziparchive' == $method) ? 'UpdraftPlus_ZipArchive' : 'UpdraftPlus_PclZip'; if (!class_exists($class_to_use)) updraft_try_include_file('includes/class-zip.php', 'require_once'); $updraftplus->log('Unzipping '.basename($file).' to '.$to.' using '.$class_to_use.', starting index '.$starting_index); $z = new $class_to_use; $flags = (version_compare(PHP_VERSION, '5.2.12', '>') && defined('ZIPARCHIVE::CHECKCONS')) ? ZIPARCHIVE::CHECKCONS : 4; // This is just for crazy people with mbstring.func_overload enabled (deprecated from PHP 7.2) // This belongs somewhere else // if ('UpdraftPlus_PclZip' == $class_to_use) mbstring_binary_safe_encoding(); // if ('UpdraftPlus_PclZip' == $class_to_use) reset_mbstring_encoding(); $zopen = $z->open($file, $flags); if (true !== $zopen) { return new WP_Error('incompatible_archive', __('Incompatible Archive.'), array($method.'_error' => $z->last_error));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } $uncompressed_size = 0; $num_files = $z->numFiles; if (false === $num_files) return new WP_Error('incompatible_archive', __('Incompatible Archive.'), array($method.'_error' => $z->last_error));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. for ($i = $starting_index; $i < $num_files; $i++) { if (!$info = $z->statIndex($i)) { return new WP_Error('stat_failed_'.$method, __('Could not retrieve file from archive.').' ('.$z->last_error.')');// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } // Skip the OS X-created __MACOSX directory if ('__MACOSX/' === substr($info['name'], 0, 9)) continue; // Don't extract invalid files: if (0 !== validate_file($info['name'])) continue; if (!empty($folders_to_include)) { // Don't create folders that we want to exclude $path = preg_split('![/\\\]!', untrailingslashit($info['name'])); if (isset($path[1]) && !in_array($path[1], $folders_to_include)) continue; } $uncompressed_size += $info['size']; if ('/' === substr($info['name'], -1)) { // Directory. $needed_dirs[] = $to . untrailingslashit($info['name']); } elseif ('.' !== ($dirname = dirname($info['name']))) { // Path to a file. $needed_dirs[] = $to . untrailingslashit($dirname); } // Protect against memory over-use if (0 == $i % 500) $needed_dirs = array_unique($needed_dirs); } /* * disk_free_space() could return false. Assume that any falsey value is an error. * A disk that has zero free bytes has bigger problems. * Require we have enough space to unzip the file and copy its contents, with a 10% buffer. */ if (self::wp_doing_cron()) { $available_space = function_exists('disk_free_space') ? @disk_free_space(WP_CONTENT_DIR) : false;// phpcs:ignore Generic.PHP.NoSilencedErrors.Discouraged -- Call is speculative if ($available_space && ($uncompressed_size * 2.1) > $available_space) { return new WP_Error('disk_full_unzip_file', __('Could not copy files.').' '.__('You may have run out of disk space.'), compact('uncompressed_size', 'available_space'));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } } $needed_dirs = array_unique($needed_dirs); foreach ($needed_dirs as $dir) { // Check the parent folders of the folders all exist within the creation array. if (untrailingslashit($to) == $dir) { // Skip over the working directory, We know this exists (or will exist) continue; } // If the directory is not within the working directory then skip it if (false === strpos($dir, $to)) continue; $parent_folder = dirname($dir); while (!empty($parent_folder) && untrailingslashit($to) != $parent_folder && !in_array($parent_folder, $needed_dirs)) { $needed_dirs[] = $parent_folder; $parent_folder = dirname($parent_folder); } } asort($needed_dirs); // Create those directories if need be: foreach ($needed_dirs as $_dir) { // Only check to see if the Dir exists upon creation failure. Less I/O this way. if (!$wp_filesystem->mkdir($_dir, FS_CHMOD_DIR) && !$wp_filesystem->is_dir($_dir)) { return new WP_Error('mkdir_failed_'.$method, __('Could not create directory.'), substr($_dir, strlen($to)));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } } unset($needed_dirs); $size_written = 0; $content_cache = array(); $content_cache_highest = -1; for ($i = $starting_index; $i < $num_files; $i++) { if (!$info = $z->statIndex($i)) { return new WP_Error('stat_failed_'.$method, __('Could not retrieve file from archive.'));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } // directory if ('/' == substr($info['name'], -1)) continue; // Don't extract the OS X-created __MACOSX if ('__MACOSX/' === substr($info['name'], 0, 9)) continue; // Don't extract invalid files: if (0 !== validate_file($info['name'])) continue; if (!empty($folders_to_include)) { // Don't extract folders that we want to exclude $path = preg_split('![/\\\]!', untrailingslashit($info['name'])); if (isset($path[1]) && !in_array($path[1], $folders_to_include)) continue; } // N.B. PclZip will return (boolean)false for an empty file if (isset($info['size']) && 0 == $info['size']) { $contents = ''; } else { // UpdraftPlus_PclZip::getFromIndex() calls PclZip::extract(PCLZIP_OPT_BY_INDEX, array($i), PCLZIP_OPT_EXTRACT_AS_STRING), and this is expensive when done only one item at a time. We try to cache in chunks for good performance as well as being able to resume. if ($i > $content_cache_highest && 'UpdraftPlus_PclZip' == $class_to_use) { $memory_usage = memory_get_usage(false); $total_memory = $updraftplus->memory_check_current(); if ($memory_usage > 0 && $total_memory > 0) { $memory_free = $total_memory*1048576 - $memory_usage; } else { // A sane default. Anything is ultimately better than WP's default of just unzipping everything into memory. $memory_free = 50*1048576; } $use_memory = max(10485760, $memory_free - 10485760); $total_byte_count = 0; $content_cache = array(); $cache_indexes = array(); $cache_index = $i; while ($cache_index < $num_files && $total_byte_count < $use_memory) { if (false !== ($cinfo = $z->statIndex($cache_index)) && isset($cinfo['size']) && '/' != substr($cinfo['name'], -1) && '__MACOSX/' !== substr($cinfo['name'], 0, 9) && 0 === validate_file($cinfo['name'])) { $total_byte_count += $cinfo['size']; if ($total_byte_count < $use_memory) { $cache_indexes[] = $cache_index; $content_cache_highest = $cache_index; } } $cache_index++; } if (!empty($cache_indexes)) { $content_cache = $z->updraftplus_getFromIndexBulk($cache_indexes); } } $contents = isset($content_cache[$i]) ? $content_cache[$i] : $z->getFromIndex($i); } if (false === $contents && ('pclzip' !== $method || 0 !== $info['size'])) { return new WP_Error('extract_failed_'.$method, __('Could not extract file from archive.').' '.$z->last_error, json_encode($info));// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } if (!$wp_filesystem->put_contents($to . $info['name'], $contents, FS_CHMOD_FILE)) { return new WP_Error('copy_failed_'.$method, __('Could not copy file.'), $info['name']);// phpcs:ignore WordPress.WP.I18n.MissingArgDomain -- The string exists within the WordPress core. } if (!empty($info['size'])) $size_written += $info['size']; do_action('updraftplus_unzip_file_unzipped', $file, $i, $info, $size_written, $num_files); } $z->close(); return true; } } Computerwire, Author at Smart Office

    Smart Office

    CISCO Action Backfires

    CISCO’s Attempt to gag a security whistleblower has backfired with the company now getting more publicity than they bargained for.

    Cisco Systems Inc and Internet Security Systems Inc have failed to put the cat back in the bag, after an uncensored document detailing former ISS researcher Michael Lynn’s cracking of Cisco’s IOS software was broadly published online.

    Less than a day after Cisco and ISS won a court injunction against Lynn and Black Hat 2005, the convention at which he presented his research, a PDF of his presentation hit several high-traffic web sites, blogs and mailing lists.The leaked document even contains text that Lynn voluntarily had redacted from the presentation when he delivered it live on Wednesday morning, such as the names of IOS code strings that could make it easier to recreate his research.

    On Wednesday morning, Lynn quit his job at ISS in order to present his research, which shows that IOS, the software running most of the internet’s routers, can be remotely compromised using heap overflow attacks.He said he did it for national security reasons and to prevent Cisco and ISS burying the research. His critics said he did it for the glory. Cisco and ISS said they had tried to stop his presentation because the research was “incomplete”.

    The full scope of how vulnerable IOS is to remote compromise still remains unclear. Cisco released an advisory on Friday that it said covered the vulnerability Lynn had disclosed, but it appeared to bear little resemblance to the Lynn paper.Cisco said in its advisory that IOS is vulnerable to an attack, but said: “Products running any version of Cisco IOS that do not have IPv6 configured interfaces are not vulnerable.” IPv6 is an emergent technology, implemented on relatively few routers.”Successful exploitation of the vulnerability may result in a reload of the device or execution of arbitrary code,” Cisco said. Patches for the IOS vulnerability were published months ago, a Cisco spokesperson confirmed.

    However, nowhere in Lynn’s presentation is IPv6 mentioned, and IPv6 was not a core theme of his speech in Las Vegas on Wednesday.During that talk, Lynn focused mainly on how to remotely deactivate a problematic security countermeasure in IOS that he called “check heaps”, which he said normally prevents heap overflow attacks from working properly.Security mailing lists on Friday started filling with speculation that Cisco was attempting to “cover up” the extent of the problem by referring to it as an IPv6-only issue.

    Lynn has said repeatedly that any fully patched router is probably safe, and Cisco has said there were no “new” vulnerabilities. But the research showed that such attacks are in fact possible, reversing conventional wisdom.

    The vulnerability was described as not being “new” because it was already known to Cisco. A spokesperson for the company said Cisco follows responsible disclosure practices when it comes to vulnerabilities in its software.

    “Cisco follows industry guidelines on the issue of security advisories,” he said. If Cisco sees the potential for exploitation, or if the vulnerability is publicly discussed, Cisco will then always issue a public advisory, the spokesperson said.

    ISS had no comment on the broad leaking of the presentation, but Cisco said in an emailed statement: “We are going to take every reasonable measure to protect our customers and the integrity of the Internet.”

    That may not necessarily mean that the integrity of the internet is in fact at risk, but it’s certainly possible to infer that from the statement. It’s not clear either what “reasonable measures” may be taken, and against whom.

    However, now that the document is broadly available to anyone who wants it and knows how to use the internet, no matter what color hat they wear, it renders some parts of the Lynn injunction a little pointless.

    That injunction forced Lynn to tell ISS and Cisco lawyers the names of “anyone to whom he has disclosed in written form or electronically” his presentation, and to refrain from distributing any information about his IOS research.

    The injunction also forces Lynn to decrypt and hand over data in his control to an independent forensics investigator, who will, under a confidentiality order, search it for data relating to IOS. Any such data would be deleted and the disk area wiped.

    The FBI is now reportedly investigating Lynn over allegations that he violated ISS trade secrets, although he claims he did not do anything criminal. An FBI spokesperson did not return a call for comment by press time.

    Wired News quoted Lynn’s attorney as saying the FBI were probing, but that there was no arrest warrant and no evidence had been discovered that Lynn had committed a crime. She could not be reached for comment on Friday.

    There were also various rumors going around Black Hat last week that President Bush was involved in the controversy, mostly of them along the lines that he had received a call from Cisco chief John Chambers and/or ISS boss Tom Noonan about the issue.

    While that may end up in the annals of security apocrypha, there was, as there usually is, a government presence at Black Hat, and Lynn said they had been very interested in, even supportive of, his research.

    He said that he had been sidelined by people from several “government agencies” immediately following his presentation and taken into an empty service corridor at Cesar’s Palace, where somebody said “Is the truck ready?”

    The agency man was joking, but the Lynn is attracting this kind of attention because IOS is a critical piece of the internet’s infrastructure. Cisco’s owns half of the market for new router sales, and in previous years its share has been much higher.

    A worm that exploited an IOS vulnerability could credibly be referred to as a “network killer”, though crashing routers would certainly impede propagation. Directed man-in-the-middle attacks to snoop on traffic would also be possible, according to Lynn.

    Many people assume that security researchers working in the corporate or public domain tend to be somewhat behind well-funded government agencies. Many assume the US, China, Israel and others are well ahead of the curve on data security.

    Lynn reasons that if he knows how to hack IOS, and Cisco knows how to hack IOS, somebody else probably knows how to hack IOS too, and that third person may either be not very nice, or an operative of an unfriendly government.