From cc75eb5d18a78f68bbe1ea58ffab71bdad42e76a Mon Sep 17 00:00:00 2001 From: Michael Date: Tue, 5 Jul 2022 05:04:05 +0000 Subject: [PATCH] Improved server detection --- src/Console/Relay.php | 6 +- src/Core/Worker/Cron.php | 22 + src/Model/APContact.php | 2 + src/Model/GServer.php | 740 ++++++++++++--------- src/Module/Admin/Federation.php | 2 + src/Module/Api/Mastodon/Instance/Peers.php | 5 +- src/Util/Network.php | 65 +- src/Util/ParseUrl.php | 17 +- static/generator.config.php | 143 ++++ 9 files changed, 660 insertions(+), 342 deletions(-) create mode 100644 static/generator.config.php diff --git a/src/Console/Relay.php b/src/Console/Relay.php index 4f6bbc6b6..dfa39d860 100644 --- a/src/Console/Relay.php +++ b/src/Console/Relay.php @@ -91,8 +91,8 @@ HELP; if ((count($this->args) == 1) && ($this->getArgument(0) == 'list')) { $contacts = $this->dba->select('apcontact', ['url'], - ["`type` = ? AND `url` IN (SELECT `url` FROM `contact` WHERE `uid` = ? AND `rel` = ?)", - 'Application', 0, Contact::FRIEND]); + ["`type` IN (?, ?) AND `url` IN (SELECT `url` FROM `contact` WHERE `uid` = ? AND `rel` = ?)", + 'Application', 'Service', 0, Contact::FRIEND]); while ($contact = $this->dba->fetch($contacts)) { $this->out($contact['url']); } @@ -108,7 +108,7 @@ HELP; $actor = $this->getArgument(1); $apcontact = APContact::getByURL($actor); - if (empty($apcontact) || ($apcontact['type'] != 'Application')) { + if (empty($apcontact) || !in_array($apcontact['type'], ['Application', 'Service'])) { $this->out($actor . ' is no relay actor'); return 1; } diff --git a/src/Core/Worker/Cron.php b/src/Core/Worker/Cron.php index 4bf134f7b..8df88e9e2 100644 --- a/src/Core/Worker/Cron.php +++ b/src/Core/Worker/Cron.php @@ -25,9 +25,11 @@ use Friendica\Core\Logger; use Friendica\Core\Worker; use Friendica\Database\DBA; use Friendica\DI; +use Friendica\Model\Contact; use Friendica\Model\Post; use Friendica\Protocol\ActivityPub; use Friendica\Util\DateTimeFormat; +use Friendica\Util\Strings; /** * Contains the class for jobs that are executed in an interval @@ -179,4 +181,24 @@ class Cron } } } + + /** + * Add missing "intro" records. + * + * @return void + */ + private static function addIntros() + { + $contacts = DBA::p("SELECT `uid`, `id`, `created` FROM `contact` WHERE `rel` = ? AND `pending` AND NOT EXISTS (SELECT `id` FROM `intro` WHERE `contact-id` = `contact`.`id`)", Contact::FOLLOWER); + while ($contact = DBA::fetch($contacts)) { + $fields = [ + 'uid' => $contact['uid'], + 'contact-id' => $contact['id'], + 'datetime' => $contact['created'], + 'hash' => Strings::getRandomHex() + ]; + Logger::notice('Adding missing intro', ['fields' => $fields]); + DBA::insert('intro', $fields); + } + } } diff --git a/src/Model/APContact.php b/src/Model/APContact.php index 0255af7a3..5363ab6e2 100644 --- a/src/Model/APContact.php +++ b/src/Model/APContact.php @@ -164,6 +164,8 @@ class APContact return $fetched_contact; } $url = $apcontact['url']; + } elseif (empty(parse_url($url, PHP_URL_PATH))) { + $apcontact['baseurl'] = $url; } // Detect multiple fast repeating request to the same address diff --git a/src/Model/GServer.php b/src/Model/GServer.php index c0e489992..d04186240 100644 --- a/src/Model/GServer.php +++ b/src/Model/GServer.php @@ -30,16 +30,22 @@ use Friendica\Core\System; use Friendica\Core\Worker; use Friendica\Database\Database; use Friendica\Database\DBA; +use Friendica\Database\DBStructure; use Friendica\DI; use Friendica\Module\Register; use Friendica\Network\HTTPClient\Client\HttpClientAccept; use Friendica\Network\HTTPClient\Client\HttpClientOptions; use Friendica\Network\HTTPClient\Capability\ICanHandleHttpResponses; +use Friendica\Network\Probe; +use Friendica\Protocol\ActivityPub; use Friendica\Protocol\Relay; use Friendica\Util\DateTimeFormat; +use Friendica\Util\JsonLD; use Friendica\Util\Network; use Friendica\Util\Strings; use Friendica\Util\XML; +use GuzzleHttp\Exception\TransferException; +use Friendica\Network\HTTPException; /** * This class handles GServer related functions @@ -57,6 +63,12 @@ class GServer const DETECT_MANUAL = 0; const DETECT_HEADER = 1; const DETECT_BODY = 2; + const DETECT_HOST_META = 3; + const DETECT_CONTACTS = 4; + const DETECT_AP_ACTOR = 5; + const DETECT_AP_COLLECTION = 6; + + const DETECT_UNSPECIFIC = [self::DETECT_MANUAL, self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_HOST_META, self::DETECT_CONTACTS, self::DETECT_AP_ACTOR]; // Implementation specific endpoints const DETECT_FRIENDIKA = 10; @@ -68,12 +80,14 @@ class GServer const DETECT_MASTODON_API = 16; const DETECT_STATUS_PHP = 17; // Nextcloud const DETECT_V1_CONFIG = 18; - const DETECT_PUMPIO = 19; + const DETECT_PUMPIO = 19; // Deprecated + const DETECT_SYSTEM_ACTOR = 20; // Mistpark, Osada, Roadhouse, Zap // Standardized endpoints const DETECT_STATISTICS_JSON = 100; const DETECT_NODEINFO_1 = 101; const DETECT_NODEINFO_2 = 102; + const DETECT_NODEINFO_210 = 103; /** * Check for the existance of a server and adds it in the background if not existant @@ -108,7 +122,7 @@ class GServer $gserver = DBA::selectFirst('gserver', ['id'], ['nurl' => Strings::normaliseLink($url)]); if (DBA::isResult($gserver)) { - Logger::info('Got ID for URL', ['id' => $gserver['id'], 'url' => $url, 'callstack' => System::callstack(20)]); + Logger::debug('Got ID for URL', ['id' => $gserver['id'], 'url' => $url, 'callstack' => System::callstack(20)]); return $gserver['id']; } @@ -268,8 +282,8 @@ class GServer $gserver = DBA::selectFirst('gserver', [], ['nurl' => Strings::normaliseLink($url)]); if (DBA::isResult($gserver)) { $next_update = self::getNextUpdateDate(false, $gserver['created'], $gserver['last_contact']); - DBA::update('gserver', ['failed' => true, 'last_failure' => DateTimeFormat::utcNow(), - 'next_contact' => $next_update, 'detection-method' => null], + DBA::update('gserver', ['url' => $url, 'failed' => true, 'last_failure' => DateTimeFormat::utcNow(), + 'next_contact' => $next_update, 'network' => Protocol::PHANTOM, 'detection-method' => null], ['nurl' => Strings::normaliseLink($url)]); Logger::info('Set failed status for existing server', ['url' => $url]); return; @@ -299,19 +313,6 @@ class GServer return Network::unparseURL($urlparts); } - /** - * Return the base URL - * - * @param string $url - * @return string base URL - */ - private static function getBaseURL(string $url) - { - $urlparts = parse_url(self::cleanURL($url)); - unset($urlparts['path']); - return Network::unparseURL($urlparts); - } - /** * Detect server data (type, protocol, version number, ...) * The detected data is then updated or inserted in the gserver table. @@ -325,184 +326,167 @@ class GServer public static function detect(string $url, string $network = '', bool $only_nodeinfo = false) { Logger::info('Detect server type', ['server' => $url]); - $serverdata = ['detection-method' => self::DETECT_MANUAL]; $original_url = $url; // Remove URL content that is not supposed to exist for a server url - $url = self::cleanURL($url); - - // Get base URL - $baseurl = self::getBaseURL($url); - - // If the URL missmatches, then we mark the old entry as failure - if ($url != $original_url) { - /// @todo What to do with "next_contact" here? - DBA::update('gserver', ['failed' => true, 'last_failure' => DateTimeFormat::utcNow()], - ['nurl' => Strings::normaliseLink($original_url)]); + $url = rtrim(self::cleanURL($url), '/'); + if (!Network::isUrlValid($url)) { + self::setFailure($url); + return false; } + // If the URL missmatches, then we mark the old entry as failure + if (Strings::normaliseLink($url) != Strings::normaliseLink($original_url)) { + self::setFailure($original_url); + self::detect($url, $network, $only_nodeinfo); + return false; + } + + // On a redirect follow the new host but mark the old one as failure + try { + $finalurl = rtrim(DI::httpClient()->finalUrl($url), '/'); + } catch (TransferException $exception) { + Logger::notice('Error fetching final URL.', ['url' => $url, 'exception' => $exception]); + self::setFailure($url); + return false; + } + + // We only follow redirects when the path stays the same or the target url has no path. + // Some systems have got redirects on their landing page to a single account page. This check handles it. + if (((parse_url($url, PHP_URL_HOST) != parse_url($finalurl, PHP_URL_HOST)) && (parse_url($url, PHP_URL_PATH) == parse_url($finalurl, PHP_URL_PATH))) || + (((parse_url($url, PHP_URL_HOST) != parse_url($finalurl, PHP_URL_HOST)) || (parse_url($url, PHP_URL_PATH) != parse_url($finalurl, PHP_URL_PATH))) && empty(parse_url($finalurl, PHP_URL_PATH)))) { + Logger::info('Found redirect. Mark old entry as failure', ['old' => $url, 'new' => $finalurl]); + self::setFailure($url); + self::detect($finalurl, $network, $only_nodeinfo); + return false; + } + + if ((parse_url($url, PHP_URL_HOST) == parse_url($finalurl, PHP_URL_HOST)) && + (parse_url($url, PHP_URL_PATH) == parse_url($finalurl, PHP_URL_PATH)) && + (parse_url($url, PHP_URL_SCHEME) != parse_url($finalurl, PHP_URL_SCHEME))) { + if (!Network::isUrlValid($finalurl)) { + self::setFailure($finalurl); + return false; + } + $url = $finalurl; + } + + $in_webroot = empty(parse_url($url, PHP_URL_PATH)); + // When a nodeinfo is present, we don't need to dig further - $xrd_timeout = DI::config()->get('system', 'xrd_timeout'); - $curlResult = DI::httpClient()->get($url . '/.well-known/nodeinfo', HttpClientAccept::JSON, [HttpClientOptions::TIMEOUT => $xrd_timeout]); + $curlResult = DI::httpClient()->get($url . '/.well-known/x-nodeinfo2', HttpClientAccept::JSON); if ($curlResult->isTimeout()) { self::setFailure($url); return false; } - // On a redirect follow the new host but mark the old one as failure - if ($curlResult->isSuccess() && !empty($curlResult->getRedirectUrl()) && (parse_url($url, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST))) { - $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); - if (!empty($curlResult->getRedirectUrl()) && parse_url($url, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST)) { - Logger::info('Found redirect. Mark old entry as failure', ['old' => $url, 'new' => $curlResult->getRedirectUrl()]); - self::setFailure($url); - self::detect($curlResult->getRedirectUrl(), $network, $only_nodeinfo); - return false; - } + $serverdata = self::parseNodeinfo210($curlResult); + if (empty($serverdata)) { + $curlResult = DI::httpClient()->get($url . '/.well-known/nodeinfo', HttpClientAccept::JSON); + $serverdata = self::fetchNodeinfo($url, $curlResult); } - $nodeinfo = self::fetchNodeinfo($url, $curlResult); - if ($only_nodeinfo && empty($nodeinfo)) { + if ($only_nodeinfo && empty($serverdata)) { Logger::info('Invalid nodeinfo in nodeinfo-mode, server is marked as failure', ['url' => $url]); self::setFailure($url); return false; + } elseif (empty($serverdata)) { + $serverdata = ['detection-method' => self::DETECT_MANUAL, 'network' => Protocol::PHANTOM, 'platform' => '', 'version' => '', 'site_name' => '', 'info' => '']; } - // When nodeinfo isn't present, we use the older 'statistics.json' endpoint - if (empty($nodeinfo)) { - $nodeinfo = self::fetchStatistics($url); - } - - // If that didn't work out well, we use some protocol specific endpoints - // For Friendica and Zot based networks we have to dive deeper to reveal more details - if (empty($nodeinfo['network']) || in_array($nodeinfo['network'], [Protocol::DFRN, Protocol::ZOT])) { - if (!empty($nodeinfo['detection-method'])) { - $serverdata['detection-method'] = $nodeinfo['detection-method']; - - foreach (['registered-users', 'active_users_monthly', 'active-halfyear-users', 'local-posts'] as $field) { - if (!empty($nodeinfo[$field])) { - $serverdata[$field] = $nodeinfo[$field]; - } - } - } - - // Fetch the landing page, possibly it reveals some data - if (empty($nodeinfo['network'])) { - if ($baseurl == $url) { - $basedata = $serverdata; + // When there is no Nodeinfo, then use some protocol specific endpoints + if ($serverdata['network'] == Protocol::PHANTOM) { + if ($in_webroot) { + // Fetch the landing page, possibly it reveals some data + $accept = 'application/activity+json,application/ld+json,application/json,*/*;q=0.9'; + $curlResult = DI::httpClient()->get($url, $accept); + if (!$curlResult->isSuccess() && $curlResult->getReturnCode() == '406') { + $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML); + $html_fetched = true; } else { - $basedata = ['detection-method' => self::DETECT_MANUAL]; + $html_fetched = false; } - $curlResult = DI::httpClient()->get($baseurl, HttpClientAccept::HTML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); if ($curlResult->isSuccess()) { - if (!empty($curlResult->getRedirectUrl()) && (parse_url($baseurl, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST))) { - Logger::info('Found redirect. Mark old entry as failure', ['old' => $url, 'new' => $curlResult->getRedirectUrl()]); - self::setFailure($url); - self::detect($curlResult->getRedirectUrl(), $network, $only_nodeinfo); - return false; + $json = json_decode($curlResult->getBody(), true); + if (!empty($json)) { + $data = self::fetchDataFromSystemActor($json, $serverdata); + $serverdata = $data['server']; + $systemactor = $data['actor']; + if (!$html_fetched && ($serverdata['detection-method'] == self::DETECT_AP_ACTOR)) { + $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML); + } + } elseif (!$html_fetched && (strlen($curlResult->getBody()) < 1000)) { + $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML); } - $basedata = self::analyseRootHeader($curlResult, $basedata); - $basedata = self::analyseRootBody($curlResult, $basedata, $baseurl); + if ($serverdata['detection-method'] != self::DETECT_SYSTEM_ACTOR) { + $serverdata = self::analyseRootHeader($curlResult, $serverdata); + $serverdata = self::analyseRootBody($curlResult, $serverdata); + } } - if (!$curlResult->isSuccess() || empty($curlResult->getBody()) || self::invalidBody($curlResult->getBody())) { + if (!$curlResult->isSuccess() || empty($curlResult->getBody())) { self::setFailure($url); return false; } - if ($baseurl == $url) { - $serverdata = $basedata; - } else { - // When the base path doesn't seem to contain a social network we try the complete path. - // Most detectable system have to be installed in the root directory. - // We checked the base to avoid false positives. - $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); - if ($curlResult->isSuccess()) { - $urldata = self::analyseRootHeader($curlResult, $serverdata); - $urldata = self::analyseRootBody($curlResult, $urldata, $url); - - $comparebase = $basedata; - unset($comparebase['info']); - unset($comparebase['site_name']); - $compareurl = $urldata; - unset($compareurl['info']); - unset($compareurl['site_name']); - - // We assume that no one will install the identical system in the root and a subfolder - if (!empty(array_diff($comparebase, $compareurl))) { - $serverdata = $urldata; - } - } + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectMastodonAlikes($url, $serverdata); } } - if (empty($nodeinfo['network']) && (empty($serverdata['network']) || ($serverdata['network'] == Protocol::ACTIVITYPUB))) { - $serverdata = self::detectMastodonAlikes($url, $serverdata); - } - // All following checks are done for systems that always have got a "host-meta" endpoint. // With this check we don't have to waste time and ressources for dead systems. // Also this hopefully prevents us from receiving abuse messages. - $validHostMeta = self::validHostMeta($url); - - if (empty($serverdata['network']) && !$validHostMeta) { - $serverdata = self::detectFromContacts($url, $serverdata); + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $validHostMeta = self::validHostMeta($url); + } else { + $validHostMeta = false; } - if (empty($serverdata['network']) && !$validHostMeta) { - self::setFailure($url); - return false; + if ($validHostMeta) { + if ($serverdata['detection-method'] == self::DETECT_MANUAL) { + $serverdata['detection-method'] = self::DETECT_HOST_META; + $serverdata['platform'] = ''; + } + + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectFriendica($url, $serverdata); + } + + // The following systems have to be installed in the root directory. + if ($in_webroot) { + // the 'siteinfo.json' is some specific endpoint of Hubzilla and Red + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::fetchSiteinfo($url, $serverdata); + } + + // The 'siteinfo.json' doesn't seem to be present on older Hubzilla installations, so we check other endpoints as well + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectHubzilla($url, $serverdata); + } + + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectPeertube($url, $serverdata); + } + + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectGNUSocial($url, $serverdata); + } + } } - if (empty($serverdata['network']) || in_array($serverdata['network'], [Protocol::DFRN, Protocol::ACTIVITYPUB])) { - $serverdata = self::detectFriendica($url, $serverdata); + if (($serverdata['network'] == Protocol::PHANTOM) || in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata = self::detectNextcloud($url, $serverdata, $validHostMeta); } - // the 'siteinfo.json' is some specific endpoint of Hubzilla and Red - if (empty($serverdata['network']) || ($serverdata['network'] == Protocol::ZOT)) { - $serverdata = self::fetchSiteinfo($url, $serverdata); + // When nodeinfo isn't present, we use the older 'statistics.json' endpoint + // Since this endpoint is only rarely used, we query it at a later time + if (in_array($serverdata['detection-method'], array_merge(self::DETECT_UNSPECIFIC, [self::DETECT_FRIENDICA, self::DETECT_CONFIG_JSON]))) { + $serverdata = self::fetchStatistics($url, $serverdata); } - - // The 'siteinfo.json' doesn't seem to be present on older Hubzilla installations - if (empty($serverdata['network'])) { - $serverdata = self::detectHubzilla($url, $serverdata); - } - - if (empty($serverdata['network']) || in_array($serverdata['detection-method'], [self::DETECT_MANUAL, self::DETECT_BODY])) { - $serverdata = self::detectPeertube($url, $serverdata); - } - - if (empty($serverdata['network'])) { - $serverdata = self::detectNextcloud($url, $serverdata); - } - - if (empty($nodeinfo['network']) && empty($serverdata['network'])) { - $serverdata = self::detectGNUSocial($url, $serverdata); - } - - if (empty($serverdata['network'])) { - $serverdata = self::detectPumpIO($url, $serverdata); - } - - $serverdata = array_merge($nodeinfo, $serverdata); - } else { - $serverdata = $nodeinfo; - } - - // Detect the directory type - $serverdata['directory-type'] = self::DT_NONE; - - $serverdata = self::checkMastodonDirectory($url, $serverdata); - - if ($serverdata['directory-type'] == self::DT_NONE) { - $serverdata = self::checkPoCo($url, $serverdata); - } - - // We can't detect the network type. Possibly it is some system that we don't know yet - if (empty($serverdata['network'])) { - $serverdata['network'] = Protocol::PHANTOM; } // When we hadn't been able to detect the network type, we use the hint from the parameter @@ -510,13 +494,34 @@ class GServer $serverdata['network'] = $network; } + // Most servers aren't installed in a subdirectory, so we declare this entry as failed + if (($serverdata['network'] == Protocol::PHANTOM) && !empty(parse_url($url, PHP_URL_PATH)) && in_array($serverdata['detection-method'], [self::DETECT_MANUAL])) { + self::setFailure($url); + return false; + } + $serverdata['url'] = $url; $serverdata['nurl'] = Strings::normaliseLink($url); - if (in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED])) { + // We have to prevent an endless loop here. + // When a server is new, then there is no gserver entry yet. + // But in "detectNetworkViaContacts" it could happen that a contact is updated, + // and this can call this function here as well. + if (in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED]) && self::getID($url, true)) { $serverdata = self::detectNetworkViaContacts($url, $serverdata); } + // Detect the directory type + $serverdata['directory-type'] = self::DT_NONE; + + if (in_array($serverdata['network'], Protocol::FEDERATED)) { + $serverdata = self::checkMastodonDirectory($url, $serverdata); + + if ($serverdata['directory-type'] == self::DT_NONE) { + $serverdata = self::checkPoCo($url, $serverdata); + } + } + if ($serverdata['network'] == Protocol::ACTIVITYPUB) { $serverdata = self::fetchWeeklyUsage($url, $serverdata); } @@ -535,6 +540,9 @@ class GServer $serverdata['last_contact'] = DateTimeFormat::utcNow(); $serverdata['failed'] = false; + // Limit the length on incoming fields + $serverdata = DBStructure::getFieldsForTable('gserver', $serverdata); + $gserver = DBA::selectFirst('gserver', ['network'], ['nurl' => Strings::normaliseLink($url)]); if (!DBA::isResult($gserver)) { $serverdata['created'] = DateTimeFormat::utcNow(); @@ -565,7 +573,7 @@ class GServer DBA::update('gserver', ['active-month-users' => $contacts], ['id' => $id]); } } - + if (empty($serverdata['active-halfyear-users'])) { $contacts = DBA::count('contact', ["`uid` = ? AND `gsid` = ? AND NOT `failed` AND `last-item` > ?", 0, $id, DateTimeFormat::utc('now - 180 days')]); if ($contacts > 0) { @@ -575,10 +583,15 @@ class GServer } } - if (!empty($serverdata['network']) && in_array($serverdata['network'], [Protocol::DFRN, Protocol::DIASPORA])) { + if (in_array($serverdata['network'], [Protocol::DFRN, Protocol::DIASPORA])) { self::discoverRelay($url); } + if (!empty($systemactor)) { + $contact = Contact::getByURL($systemactor, true, ['gsid', 'baseurl', 'id', 'network', 'url', 'name']); + Logger::debug('Fetched system actor', ['url' => $url, 'gsid' => $id, 'contact' => $contact]); + } + return $ret; } @@ -685,21 +698,30 @@ class GServer * * @return array server data */ - private static function fetchStatistics(string $url) + private static function fetchStatistics(string $url, array $serverdata) { $curlResult = DI::httpClient()->get($url . '/statistics.json', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { - return []; + return $serverdata; } $data = json_decode($curlResult->getBody(), true); if (empty($data)) { - return []; + return $serverdata; } - $serverdata = ['detection-method' => self::DETECT_STATISTICS_JSON]; + // Some AP enabled systems return activity data that we don't expect here. + if (strpos($curlResult->getContentType(), 'application/activity+json') !== false) { + return $serverdata; + } + + $valid = false; + $old_serverdata = $serverdata; + + $serverdata['detection-method'] = self::DETECT_STATISTICS_JSON; if (!empty($data['version'])) { + $valid = true; $serverdata['version'] = $data['version']; // Version numbers on statistics.json are presented with additional info, e.g.: // 0.6.3.0-p1702cc1c, 0.6.99.0-p1b9ab160 or 3.4.3-2-1191. @@ -707,10 +729,12 @@ class GServer } if (!empty($data['name'])) { + $valid = true; $serverdata['site_name'] = $data['name']; } if (!empty($data['network'])) { + $valid = true; $serverdata['platform'] = strtolower($data['network']); if ($serverdata['platform'] == 'diaspora') { @@ -725,18 +749,22 @@ class GServer } if (!empty($data['total_users'])) { + $valid = true; $serverdata['registered-users'] = max($data['total_users'], 1); } if (!empty($data['active_users_monthly'])) { + $valid = true; $serverdata['active-month-users'] = max($data['active_users_monthly'], 0); } if (!empty($data['active_users_halfyear'])) { + $valid = true; $serverdata['active-halfyear-users'] = max($data['active_users_halfyear'], 0); } if (!empty($data['local_posts'])) { + $valid = true; $serverdata['local-posts'] = max($data['local_posts'], 0); } @@ -746,6 +774,10 @@ class GServer $serverdata['register_policy'] = Register::CLOSED; } + if (!$valid) { + return $old_serverdata; + } + return $serverdata; } @@ -791,13 +823,11 @@ class GServer $server = []; - // When the nodeinfo url isn't on the same host, then there is obviously something wrong - if (!empty($nodeinfo2_url) && (parse_url($url, PHP_URL_HOST) == parse_url($nodeinfo2_url, PHP_URL_HOST))) { + if (!empty($nodeinfo2_url)) { $server = self::parseNodeinfo2($nodeinfo2_url); } - // When the nodeinfo url isn't on the same host, then there is obviously something wrong - if (empty($server) && !empty($nodeinfo1_url) && (parse_url($url, PHP_URL_HOST) == parse_url($nodeinfo1_url, PHP_URL_HOST))) { + if (empty($server) && !empty($nodeinfo1_url)) { $server = self::parseNodeinfo1($nodeinfo1_url); } @@ -893,6 +923,10 @@ class GServer return []; } + if (empty($server['network'])) { + $server['network'] = Protocol::PHANTOM; + } + return $server; } @@ -912,7 +946,6 @@ class GServer } $nodeinfo = json_decode($curlResult->getBody(), true); - if (!is_array($nodeinfo)) { return []; } @@ -924,12 +957,12 @@ class GServer $server['register_policy'] = Register::OPEN; } - if (is_array($nodeinfo['software'])) { - if (!empty($nodeinfo['software']['name'])) { + if (!empty($nodeinfo['software'])) { + if (isset($nodeinfo['software']['name'])) { $server['platform'] = strtolower($nodeinfo['software']['name']); } - if (!empty($nodeinfo['software']['version'])) { + if (!empty($nodeinfo['software']['version']) && isset($server['platform'])) { $server['version'] = $nodeinfo['software']['version']; // Version numbers on Nodeinfo are presented with additional info, e.g.: // 0.6.3.0-p1702cc1c, 0.6.99.0-p1b9ab160 or 3.4.3-2-1191. @@ -993,6 +1026,108 @@ class GServer return []; } + if (empty($server['network'])) { + $server['network'] = Protocol::PHANTOM; + } + + return $server; + } + + /** + * Parses NodeInfo2 protocol 1.0 + * + * @see https://github.com/jaywink/nodeinfo2/blob/master/PROTOCOL.md + * @param string $nodeinfo_url address of the nodeinfo path + * @return array Server data + * @throws \Friendica\Network\HTTPException\InternalServerErrorException + */ + private static function parseNodeinfo210(ICanHandleHttpResponses $httpResult) + { + if (!$httpResult->isSuccess()) { + return []; + } + + $nodeinfo = json_decode($httpResult->getBody(), true); + + if (!is_array($nodeinfo)) { + return []; + } + + $server = ['detection-method' => self::DETECT_NODEINFO_210, + 'register_policy' => Register::CLOSED]; + + if (!empty($nodeinfo['openRegistrations'])) { + $server['register_policy'] = Register::OPEN; + } + + if (!empty($nodeinfo['server'])) { + if (!empty($nodeinfo['server']['software'])) { + $server['platform'] = strtolower($nodeinfo['server']['software']); + } + + if (!empty($nodeinfo['server']['version'])) { + $server['version'] = $nodeinfo['server']['version']; + // Version numbers on Nodeinfo are presented with additional info, e.g.: + // 0.6.3.0-p1702cc1c, 0.6.99.0-p1b9ab160 or 3.4.3-2-1191. + $server['version'] = preg_replace('=(.+)-(.{4,})=ism', '$1', $server['version']); + } + + if (!empty($nodeinfo['server']['name'])) { + $server['site_name'] = $nodeinfo['server']['name']; + } + } + + if (!empty($nodeinfo['usage']['users']['total'])) { + $server['registered-users'] = max($nodeinfo['usage']['users']['total'], 1); + } + + if (!empty($nodeinfo['usage']['users']['activeMonth'])) { + $server['active-month-users'] = max($nodeinfo['usage']['users']['activeMonth'], 0); + } + + if (!empty($nodeinfo['usage']['users']['activeHalfyear'])) { + $server['active-halfyear-users'] = max($nodeinfo['usage']['users']['activeHalfyear'], 0); + } + + if (!empty($nodeinfo['usage']['localPosts'])) { + $server['local-posts'] = max($nodeinfo['usage']['localPosts'], 0); + } + + if (!empty($nodeinfo['usage']['localComments'])) { + $server['local-comments'] = max($nodeinfo['usage']['localComments'], 0); + } + + if (!empty($nodeinfo['protocols'])) { + $protocols = []; + foreach ($nodeinfo['protocols'] as $protocol) { + if (is_string($protocol)) { + $protocols[$protocol] = true; + } + } + + if (!empty($protocols['dfrn'])) { + $server['network'] = Protocol::DFRN; + } elseif (!empty($protocols['activitypub'])) { + $server['network'] = Protocol::ACTIVITYPUB; + } elseif (!empty($protocols['diaspora'])) { + $server['network'] = Protocol::DIASPORA; + } elseif (!empty($protocols['ostatus'])) { + $server['network'] = Protocol::OSTATUS; + } elseif (!empty($protocols['gnusocial'])) { + $server['network'] = Protocol::OSTATUS; + } elseif (!empty($protocols['zot'])) { + $server['network'] = Protocol::ZOT; + } + } + + if (empty($server) || empty($server['platform'])) { + return []; + } + + if (empty($server['network'])) { + $server['network'] = Protocol::PHANTOM; + } + return $server; } @@ -1016,7 +1151,7 @@ class GServer return $serverdata; } - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_SITEINFO_JSON; } @@ -1081,6 +1216,42 @@ class GServer return $serverdata; } + private static function fetchDataFromSystemActor(array $data, array $serverdata) + { + if (empty($data)) { + return ['server' => $serverdata, 'actor' => '']; + } + + $actor = JsonLD::compact($data); + if (in_array(JsonLD::fetchElement($actor, '@type'), ActivityPub\Receiver::ACCOUNT_TYPES)) { + $serverdata['network'] = Protocol::ACTIVITYPUB; + $serverdata['site_name'] = JsonLD::fetchElement($actor, 'as:name', '@value'); + $serverdata['info'] = JsonLD::fetchElement($actor, 'as:summary', '@value'); + if (!empty($actor['as:generator'])) { + $serverdata['platform'] = JsonLD::fetchElement($actor['as:generator'], 'as:name', '@value'); + $serverdata['detection-method'] = self::DETECT_SYSTEM_ACTOR; + } else { + $serverdata['detection-method'] = self::DETECT_AP_ACTOR; + } + return ['server' => $serverdata, 'actor' => $actor['@id']]; + } elseif ((JsonLD::fetchElement($actor, '@type') == 'as:Collection')) { + // By now only Ktistec seems to provide collections this way + $serverdata['platform'] = 'ktistec'; + $serverdata['network'] = Protocol::ACTIVITYPUB; + $serverdata['detection-method'] = self::DETECT_AP_COLLECTION; + + $actors = JsonLD::fetchElementArray($actor, 'as:items'); + if (!empty($actors) && !empty($actors[0]['@id'])) { + $actor_url = $actor['@id'] . $actors[0]['@id']; + } else { + $actor_url = ''; + } + + return ['server' => $serverdata, 'actor' => $actor_url]; + } + return ['server' => $serverdata, 'actor' => '']; + } + /** * Checks if the server contains a valid host meta file * @@ -1138,13 +1309,15 @@ class GServer { $contacts = []; - $apcontacts = DBA::select('apcontact', ['url'], ['baseurl' => [$url, $serverdata['nurl']]]); + $nurl = Strings::normaliseLink($url); + + $apcontacts = DBA::select('apcontact', ['url'], ['baseurl' => [$url, $nurl]]); while ($apcontact = DBA::fetch($apcontacts)) { $contacts[Strings::normaliseLink($apcontact['url'])] = $apcontact['url']; } DBA::close($apcontacts); - $pcontacts = DBA::select('contact', ['url', 'nurl'], ['uid' => 0, 'baseurl' => [$url, $serverdata['nurl']]]); + $pcontacts = DBA::select('contact', ['url', 'nurl'], ['uid' => 0, 'baseurl' => [$url, $nurl]]); while ($pcontact = DBA::fetch($pcontacts)) { $contacts[$pcontact['nurl']] = $pcontact['url']; } @@ -1156,9 +1329,14 @@ class GServer $time = time(); foreach ($contacts as $contact) { - $probed = Contact::getByURL($contact, true); - if (!empty($probed) && !$probed['failed'] && in_array($probed['network'], Protocol::FEDERATED)) { - $serverdata['network'] = $probed['network']; + // Endlosschleife verhindern wegen gsid! + $data = Probe::uri($contact); + if (in_array($data['network'], Protocol::FEDERATED)) { + $serverdata['network'] = $data['network']; + + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata['detection-method'] = self::DETECT_CONTACTS; + } break; } elseif ((time() - $time) > 10) { // To reduce the stress on remote systems we probe a maximum of 10 seconds @@ -1269,7 +1447,7 @@ class GServer } } - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_V1_CONFIG; } } @@ -1282,10 +1460,11 @@ class GServer * * @param string $url URL of the given server * @param array $serverdata array with server data + * @param bool $validHostMeta * * @return array server data */ - private static function detectNextcloud(string $url, array $serverdata) + private static function detectNextcloud(string $url, array $serverdata, bool $validHostMeta) { $curlResult = DI::httpClient()->get($url . '/status.php', HttpClientAccept::JSON); if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { @@ -1300,9 +1479,12 @@ class GServer if (!empty($data['version'])) { $serverdata['platform'] = 'nextcloud'; $serverdata['version'] = $data['version']; - $serverdata['network'] = Protocol::ACTIVITYPUB; - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if ($validHostMeta) { + $serverdata['network'] = Protocol::ACTIVITYPUB; + } + + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_STATUS_PHP; } } @@ -1331,7 +1513,7 @@ class GServer // Most likely the data is sorted correctly. But we better are safe than sorry if (empty($current_week['week']) || ($current_week['week'] < $week['week'])) { $current_week = $week; - } + } } if (!empty($current_week['logins'])) { @@ -1340,34 +1522,6 @@ class GServer return $serverdata; } - - /** - * Detects the server network type from contacts of that server - * - * @param string $url URL of the given server - * @param array $serverdata array with server data - * - * @return array server data - */ - private static function detectFromContacts(string $url, array $serverdata) - { - $gserver = DBA::selectFirst('gserver', ['id'], ['nurl' => Strings::normaliseLink($url)]); - if (empty($gserver)) { - return $serverdata; - } - - $contact = Contact::selectFirst(['id'], ['uid' => 0, 'failed' => false, 'gsid' => $gserver['id']]); - - // Via probing we can be sure that the server is responding - if (!empty($contact['id']) && Contact::updateFromProbe($contact['id'])) { - $contact = Contact::selectFirst(['network', 'failed'], ['id' => $contact['id']]); - if (!$contact['failed'] && in_array($contact['network'], Protocol::FEDERATED)) { - $serverdata['network'] = $contact['network']; - } - } - - return $serverdata; - } /** * Detects data from a given server url if it was a mastodon alike system @@ -1389,23 +1543,23 @@ class GServer return $serverdata; } - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { - $serverdata['detection-method'] = self::DETECT_MASTODON_API; - } + $valid = false; if (!empty($data['version'])) { $serverdata['platform'] = 'mastodon'; $serverdata['version'] = $data['version'] ?? ''; $serverdata['network'] = Protocol::ACTIVITYPUB; + $valid = true; } if (!empty($data['title'])) { $serverdata['site_name'] = $data['title']; } - if (!empty($data['title']) && empty($serverdata['platform']) && empty($serverdata['network'])) { + if (!empty($data['title']) && empty($serverdata['platform']) && ($serverdata['network'] == Protocol::PHANTOM)) { $serverdata['platform'] = 'mastodon'; $serverdata['network'] = Protocol::ACTIVITYPUB; + $valid = true; } if (!empty($data['description'])) { @@ -1419,16 +1573,23 @@ class GServer if (!empty($serverdata['version']) && preg_match('/.*?\(compatible;\s(.*)\s(.*)\)/ism', $serverdata['version'], $matches)) { $serverdata['platform'] = strtolower($matches[1]); $serverdata['version'] = $matches[2]; + $valid = true; } if (!empty($serverdata['version']) && strstr(strtolower($serverdata['version']), 'pleroma')) { $serverdata['platform'] = 'pleroma'; $serverdata['version'] = trim(str_ireplace('pleroma', '', $serverdata['version'])); + $valid = true; } if (!empty($serverdata['platform']) && strstr($serverdata['platform'], 'pleroma')) { $serverdata['version'] = trim(str_ireplace('pleroma', '', $serverdata['platform'])); $serverdata['platform'] = 'pleroma'; + $valid = true; + } + + if ($valid && in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { + $serverdata['detection-method'] = self::DETECT_MASTODON_API; } return $serverdata; @@ -1505,8 +1666,7 @@ class GServer $serverdata['register_policy'] = Register::CLOSED; } - if (!empty($serverdata['network']) && in_array($serverdata['detection-method'], - [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (($serverdata['network'] != Protocol::PHANTOM) && in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_CONFIG_JSON; } @@ -1531,56 +1691,6 @@ class GServer return $val; } - /** - * Detect if the URL belongs to a pump.io server - * - * @param string $url URL of the given server - * @param array $serverdata array with server data - * - * @return array server data - */ - private static function detectPumpIO(string $url, array $serverdata) - { - $curlResult = DI::httpClient()->get($url . '/.well-known/host-meta.json', HttpClientAccept::JSON); - if (!$curlResult->isSuccess()) { - return $serverdata; - } - - $data = json_decode($curlResult->getBody(), true); - if (empty($data['links'])) { - return $serverdata; - - } - - // We are looking for some endpoints that are typical for pump.io - $trust = 0; - foreach ($data['links'] as $link) { - if (empty($link['rel'])) { - continue; - } - if (in_array($link['rel'], ['registration_endpoint', 'dialback', 'http://apinamespace.org/activitypub/whoami'])) { - ++$trust; - } - } - - if ($trust == 3) { - $serverdata['detection-method'] = self::DETECT_PUMPIO; - - $serverdata['platform'] = 'pumpio'; - $serverdata['version'] = ''; - $serverdata['network'] = Protocol::PUMPIO; - - $servers = $curlResult->getHeader('Server'); - foreach ($servers as $server) { - if (preg_match("#pump.io/(.*)\s#U", $server, $matches)) { - $serverdata['version'] = $matches[1]; - } - } - } - - return $serverdata; - } - /** * Detect if the URL belongs to a GNU Social server * @@ -1602,7 +1712,7 @@ class GServer $serverdata['version'] = trim($serverdata['version'], '"'); $serverdata['network'] = Protocol::OSTATUS; - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_GNUSOCIAL; } @@ -1628,7 +1738,7 @@ class GServer $serverdata['network'] = Protocol::OSTATUS; } - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = self::DETECT_STATUSNET; } } @@ -1667,7 +1777,7 @@ class GServer return $serverdata; } - if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { + if (in_array($serverdata['detection-method'], self::DETECT_UNSPECIFIC)) { $serverdata['detection-method'] = $friendika ? self::DETECT_FRIENDIKA : self::DETECT_FRIENDICA; } @@ -1716,24 +1826,34 @@ class GServer * * @param object $curlResult result of curl execution * @param array $serverdata array with server data - * @param string $url Server URL * * @return array server data */ - private static function analyseRootBody($curlResult, array $serverdata, string $url) + private static function analyseRootBody($curlResult, array $serverdata) { if (empty($curlResult->getBody())) { return $serverdata; } - // Using only body information we cannot safely detect a lot of systems. - // So we define a list of platforms that we can detect safely. - $valid_platforms = ['friendica', 'friendika', 'diaspora', 'mastodon', 'hubzilla', 'misskey', 'peertube', 'wordpress', 'write.as']; + if (file_exists(__DIR__ . '/../../static/generator.config.php')) { + require __DIR__ . '/../../static/generator.config.php'; + } else { + throw new HTTPException\InternalServerErrorException('Invalid generator file'); + } + + $platforms = array_merge($ap_platforms, $dfrn_platforms, $zap_platforms, $platforms); + $valid_platforms = array_values($platforms); $doc = new DOMDocument(); @$doc->loadHTML($curlResult->getBody()); $xpath = new DOMXPath($doc); + // We can only detect honk via some HTML element on their page + if ($xpath->query('//div[@id="honksonpage"]')->count() == 1) { + $serverdata['platform'] = 'honk'; + $serverdata['network'] = Protocol::ACTIVITYPUB; + } + $title = trim(XML::getFirstNodeValue($xpath, '//head/title/text()')); if (!empty($title)) { $serverdata['site_name'] = $title; @@ -1763,34 +1883,28 @@ class GServer } if (in_array($attr['name'], ['application-name', 'al:android:app_name', 'al:ios:app_name', - 'twitter:app:name:googleplay', 'twitter:app:name:iphone', 'twitter:app:name:ipad'])) { - $serverdata['platform'] = strtolower($attr['content']); - if (in_array($attr['content'], ['Misskey', 'Write.as'])) { - $serverdata['network'] = Protocol::ACTIVITYPUB; - } - } - if (($attr['name'] == 'generator') && (empty($serverdata['platform']) || (substr(strtolower($attr['content']), 0, 9) == 'wordpress'))) { - $serverdata['platform'] = strtolower($attr['content']); - $version_part = explode(' ', $attr['content']); + 'twitter:app:name:googleplay', 'twitter:app:name:iphone', 'twitter:app:name:ipad', 'generator'])) { + $platform = str_replace(array_keys($platforms), array_values($platforms), $attr['content']); + $platform = strtolower(str_replace('/', ' ', $platform)); + $version_part = explode(' ', $platform); - if (count($version_part) == 2) { - if (in_array($version_part[0], ['WordPress'])) { - $serverdata['platform'] = 'wordpress'; - $serverdata['version'] = $version_part[1]; - - // We still do need a reliable test if some AP plugin is activated - // By now we just check in a later process for some known contacts - $serverdata['network'] = Protocol::FEED; - - if ($serverdata['detection-method'] == self::DETECT_MANUAL) { - $serverdata['detection-method'] = self::DETECT_BODY; - } - } - if (in_array($version_part[0], ['Friendika', 'Friendica'])) { - $serverdata['platform'] = strtolower($version_part[0]); - $serverdata['version'] = $version_part[1]; + if (count($version_part) >= 2) { + if (in_array($version_part[0], array_values($dfrn_platforms))) { $serverdata['network'] = Protocol::DFRN; + } elseif (in_array($version_part[0], array_values($ap_platforms))) { + $serverdata['network'] = Protocol::ACTIVITYPUB; + } elseif (in_array($version_part[0], array_values($zap_platforms))) { + $serverdata['network'] = Protocol::ZOT; } + if (in_array(strtolower($version_part[0]), $valid_platforms)) { + $platform = strtolower($version_part[0]); + $serverdata['version'] = $version_part[1]; + } + } + if (in_array($platform, array_values($platforms))) { + $serverdata['platform'] = $platform; + } elseif (empty($serverdata['platform'])) { + print_r($attr); } } } @@ -1822,29 +1936,22 @@ class GServer $serverdata['info'] = $attr['content']; } - if ($attr['property'] == 'og:platform') { - $serverdata['platform'] = strtolower($attr['content']); - - if (in_array($attr['content'], ['PeerTube'])) { - $serverdata['network'] = Protocol::ACTIVITYPUB; + if (in_array($attr['property'], ['og:platform', 'generator'])) { + if (in_array($attr['content'], array_keys($platforms))) { + $serverdata['platform'] = $platforms[$attr['content']]; + } else { + print_r($attr); } - } - if ($attr['property'] == 'generator') { - $serverdata['platform'] = strtolower($attr['content']); - - if (in_array($attr['content'], ['hubzilla'])) { - // We later check which compatible protocol modules are loaded. + if (in_array($attr['content'], array_keys($ap_platforms))) { + $serverdata['network'] = Protocol::ACTIVITYPUB; + } elseif (in_array($attr['content'], array_values($zap_platforms))) { $serverdata['network'] = Protocol::ZOT; } } } - if (!empty($serverdata['platform']) && in_array($serverdata['detection-method'], [self::DETECT_MANUAL, self::DETECT_BODY]) && !in_array($serverdata['platform'], $valid_platforms)) { - $serverdata['network'] = Protocol::PHANTOM; - $serverdata['version'] = ''; - $serverdata['detection-method'] = self::DETECT_MANUAL; - } elseif (!empty($serverdata['network']) && ($serverdata['detection-method'] == self::DETECT_MANUAL)) { + if (in_array($serverdata['platform'], $valid_platforms) && ($serverdata['detection-method'] == self::DETECT_MANUAL)) { $serverdata['detection-method'] = self::DETECT_BODY; } @@ -1883,19 +1990,6 @@ class GServer return $serverdata; } - /** - * Test if the body contains valid content - * - * @param string $body - * @return boolean - */ - private static function invalidBody(string $body) - { - // Currently we only test for a HTML element. - // Possibly we enhance this in the future. - return !strpos($body, '>'); - } - /** * Update GServer entries */ diff --git a/src/Module/Admin/Federation.php b/src/Module/Admin/Federation.php index f0f3bd6e5..ccde3b4de 100644 --- a/src/Module/Admin/Federation.php +++ b/src/Module/Admin/Federation.php @@ -40,12 +40,14 @@ class Federation extends BaseAdmin 'friendica' => ['name' => 'Friendica', 'color' => '#ffc018'], // orange from the logo 'birdsitelive' => ['name' => 'BirdsiteLIVE', 'color' => '#1b6ec2'], // Color from the page 'bookwyrm' => ['name' => 'BookWyrm', 'color' => '#00d1b2'], // Color from the page + 'castopod' => ['name' => 'Castopod', 'color' => '#00564a'], // Background color from the page 'diaspora' => ['name' => 'Diaspora', 'color' => '#a1a1a1'], // logo is black and white, makes a gray 'funkwhale' => ['name' => 'Funkwhale', 'color' => '#4082B4'], // From the homepage 'gnusocial' => ['name' => 'GNU Social/Statusnet', 'color' => '#a22430'], // dark red from the logo 'gotosocial' => ['name' => 'GoToSocial', 'color' => '#df8958'], // Some color from their mascot 'hometown' => ['name' => 'Hometown', 'color' => '#1f70c1'], // Color from the Patreon page 'hubzilla' => ['name' => 'Hubzilla/Red Matrix', 'color' => '#43488a'], // blue from the logo + 'hugo' => ['name' => 'Hugo', 'color' => '#0a1922'], // Color from the homepage background 'lemmy' => ['name' => 'Lemmy', 'color' => '#00c853'], // Green from the page 'mastodon' => ['name' => 'Mastodon', 'color' => '#1a9df9'], // blue from the Mastodon logo 'misskey' => ['name' => 'Misskey', 'color' => '#ccfefd'], // Font color of the homepage diff --git a/src/Module/Api/Mastodon/Instance/Peers.php b/src/Module/Api/Mastodon/Instance/Peers.php index 796523d7e..3511f9a83 100644 --- a/src/Module/Api/Mastodon/Instance/Peers.php +++ b/src/Module/Api/Mastodon/Instance/Peers.php @@ -24,6 +24,7 @@ namespace Friendica\Module\Api\Mastodon\Instance; use Friendica\Core\Protocol; use Friendica\Core\System; use Friendica\Database\DBA; +use Friendica\Model\GServer; use Friendica\Module\BaseApi; use Friendica\Network\HTTPException; use Friendica\Util\Network; @@ -41,7 +42,9 @@ class Peers extends BaseApi $return = []; // We only select for Friendica and ActivityPub servers, since it is expected to only deliver AP compatible systems here. - $instances = DBA::select('gserver', ['url'], ["`network` in (?, ?) AND NOT `failed`", Protocol::DFRN, Protocol::ACTIVITYPUB]); + $instances = DBA::select('gserver', ['url'], ["`network` in (?, ?) AND NOT `failed` AND NOT `detection-method` IN (?, ?, ?, ?)", + Protocol::DFRN, Protocol::ACTIVITYPUB, + GServer::DETECT_MANUAL, GServer::DETECT_HEADER, GServer::DETECT_BODY, GServer::DETECT_HOST_META]); while ($instance = DBA::fetch($instances)) { $urldata = parse_url($instance['url']); unset($urldata['scheme']); diff --git a/src/Util/Network.php b/src/Util/Network.php index af4d18b4a..92b577849 100644 --- a/src/Util/Network.php +++ b/src/Util/Network.php @@ -25,6 +25,7 @@ use Friendica\Core\Hook; use Friendica\Core\Logger; use Friendica\DI; use Friendica\Model\Contact; +use Friendica\Network\HTTPClient\Client\HttpClientAccept; use Friendica\Network\HTTPException\NotModifiedException; use GuzzleHttp\Psr7\Uri; @@ -66,14 +67,58 @@ class Network $url = 'http://' . $url; } - /// @TODO Really suppress function outcomes? Why not find them + debug them? - $h = @parse_url($url); + $xrd_timeout = DI::config()->get('system', 'xrd_timeout'); + $host = parse_url($url, PHP_URL_HOST); - if (!empty($h['host']) && (@dns_get_record($h['host'], DNS_A + DNS_CNAME) || filter_var($h['host'], FILTER_VALIDATE_IP))) { - return $url; + if (empty($host) || !(@dns_get_record($host . '.', DNS_A + DNS_AAAA + DNS_CNAME) || filter_var($host, FILTER_VALIDATE_IP))) { + return false; } - return false; + // Check if the certificate is valid for this hostname + if (parse_url($url, PHP_URL_SCHEME) == 'https') { + $port = parse_url($url, PHP_URL_PORT) ?? 443; + + $context = stream_context_create(["ssl" => ['capture_peer_cert' => true]]); + + $resource = @stream_socket_client('ssl://' . $host . ':' . $port, $errno, $errstr, $xrd_timeout, STREAM_CLIENT_CONNECT, $context); + if (empty($resource)) { + Logger::notice('Invalid certificate', ['host' => $host]); + return false; + } + + $cert = stream_context_get_params($resource); + if (empty($cert)) { + Logger::notice('Invalid certificate params', ['host' => $host]); + return false; + } + + $certinfo = openssl_x509_parse($cert['options']['ssl']['peer_certificate']); + if (empty($certinfo)) { + Logger::notice('Invalid certificate information', ['host' => $host]); + return false; + } + + $valid_from = date(DATE_RFC2822,$certinfo['validFrom_time_t']); + $valid_to = date(DATE_RFC2822,$certinfo['validTo_time_t']); + + if ($certinfo['validFrom_time_t'] > time()) { + Logger::notice('Certificate validity starts after current date', ['host' => $host, 'from' => $valid_from, 'to' => $valid_to]); + return false; + } + + if ($certinfo['validTo_time_t'] < time()) { + Logger::notice('Certificate validity ends before current date', ['host' => $host, 'from' => $valid_from, 'to' => $valid_to]); + return false; + } + } + if (in_array(parse_url($url, PHP_URL_SCHEME), ['https', 'http'])) { + if (!ParseUrl::getContentType($url, HttpClientAccept::DEFAULT, $xrd_timeout)) { + Logger::notice('Url not reachable', ['host' => $host, 'url' => $url]); + return false; + } + } + + return $url; } /** @@ -95,7 +140,7 @@ class Network $h = substr($addr, strpos($addr, '@') + 1); // Concerning the @ see here: https://stackoverflow.com/questions/36280957/dns-get-record-a-temporary-server-error-occurred - if ($h && (@dns_get_record($h, DNS_A + DNS_MX) || filter_var($h, FILTER_VALIDATE_IP))) { + if ($h && (@dns_get_record($h, DNS_A + DNS_AAAA + DNS_MX) || filter_var($h, FILTER_VALIDATE_IP))) { return true; } if ($h && @dns_get_record($h, DNS_CNAME + DNS_MX)) { @@ -577,8 +622,8 @@ class Network /** * Check if the given URL is a local link * - * @param string $url - * @return bool + * @param string $url + * @return bool */ public static function isLocalLink(string $url) { @@ -588,8 +633,8 @@ class Network /** * Check if the given URL is a valid HTTP/HTTPS URL * - * @param string $url - * @return bool + * @param string $url + * @return bool */ public static function isValidHttpUrl(string $url) { diff --git a/src/Util/ParseUrl.php b/src/Util/ParseUrl.php index abd9154da..8d8329675 100644 --- a/src/Util/ParseUrl.php +++ b/src/Util/ParseUrl.php @@ -57,15 +57,22 @@ class ParseUrl * Fetch the content type of the given url * @param string $url URL of the page * @param string $accept content-type to accept + * @param int $timeout * @return array content type */ - public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT) + public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT, int $timeout = 0) { - $curlResult = DI::httpClient()->head($url, [HttpClientOptions::ACCEPT_CONTENT => $accept]); + if (!empty($timeout)) { + $options = [HttpClientOptions::TIMEOUT => $timeout]; + } else { + $options = []; + } - // Workaround for systems that can't handle a HEAD request - if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() == 405)) { - $curlResult = DI::httpClient()->get($url, $accept, [HttpClientOptions::CONTENT_LENGTH => 1000000]); + $curlResult = DI::httpClient()->head($url, array_merge([HttpClientOptions::ACCEPT_CONTENT => $accept], $options)); + + // Workaround for systems that can't handle a HEAD request. Don't retry on timeouts. + if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() >= 400) && !in_array($curlResult->getReturnCode(), [408, 504])) { + $curlResult = DI::httpClient()->get($url, $accept, array_merge([HttpClientOptions::CONTENT_LENGTH => 1000000], $options)); } if (!$curlResult->isSuccess()) { diff --git a/static/generator.config.php b/static/generator.config.php new file mode 100644 index 000000000..6a5c7c5ef --- /dev/null +++ b/static/generator.config.php @@ -0,0 +1,143 @@ +. + * + */ + +$platforms = [ + 'AlphaChat' => 'alphachat', + 'AV Arcade' => 'av-arcade', + 'BaseKit' => 'basekit', + 'BBEdit' => 'bbedit', + 'Big Cartel' => 'big-cartel', + 'blogger' => 'blogger', + 'Bloom' => 'bloom', + 'Bludit' => 'bludit', + 'BunnyPress' => 'bunnypress', + 'Contao Open Source CMS' => 'contao', + 'castopod' => 'castopod', + 'cgit' => 'cgit', + 'Chevereto' => 'chevereto', + 'ClanSphere' => 'clansphere', + 'Claudia Content' => 'claudia-content', + 'diaspora' => 'diaspora', + 'DIMV' => 'dimv', + 'Discourse' => 'discourse', + 'Divi' => 'divi', + 'Django-cypress' => 'django-cypress', + 'Docutils' => 'docutils', + 'DokuWiki' => 'dokuwiki', + 'Drupal' => 'drupal', + 'Element' => 'element', + 'Eleventy' => 'eleventy', + 'filerun' => 'filerun', + 'FlatPress' => 'flatpress', + 'Gatsby' => 'gatsby', + 'gitweb' => 'gitweb', + 'gnusocial' => 'gnusocial', + 'GravCMS' => 'gravcms', + 'grocy' => 'grocy', + 'Gruta' => 'gruta', + 'hakyll' => 'hakyll', + 'HedgeDoc - Collaborative markdown notes' => 'hedgedoc', + 'helloworld' => 'helloworld', + 'Hexo' => 'hexo', + 'honk' => 'honk', + 'ian' => 'ian', + 'InterRed' => 'interred', + 'Ikiwiki' => 'ikiwiki', + 'Joomla!' => 'joomla', + 'KeyHelp' => 'keyhelp', + 'KONTEXT-CMS (c) WARENFORM [www.warenform.net]' => 'kontext-cms', + 'ktistec' => 'ktistec', + 'LibreOffice' => 'libreoffice', + 'Magazine News Byte' => 'magazine-news-byte', + 'mastodon' => 'mastodon', + 'Mattermost' => 'mattermost', + 'MediaWiki' => 'mediawiki', + 'Medium' => 'medium', + 'microblog' => 'microblog', + 'Misskey' => 'misskey', + 'Mobirise' => 'mobirise', + 'Movable Type Pro' => 'movable-type', + 'Movim' => 'movim', + 'MyArcadePlugin Pro' => 'my-arcade-plugin', + 'Nanoc' => 'nanoc', + 'nb3 system v1.6' => 'nb3', + 'Nikola (getnikola.com)' => 'nikola', + 'Odoo' => 'odoo', + 'Org mode' => 'org-mode', + 'Org-mode' => 'org-mode', + 'Org Mode' => 'org-mode', + 'orig4' => 'orig4', + 'Osclass' => 'osclass', + 'pamphlets/vinyl-press' => 'pamphlets', + 'peertube' => 'peertube', + 'phpMyFAQ' => 'phpmyfaq', + 'Pelican' => 'pelican', + 'pH7CMS,' => 'ph7cms', + 'Polr' => 'polr', + 'Publii Open-Source CMS for Static Site' => 'publii', + 'pygruta' => 'pygruta', + 'Riot' => 'riot', + 'Satelito' => 'satelito', + 'Scully' => 'scully', + 'Sedo' => 'sedo', + 'sitebaker' => 'sitebaker', + 'SitePad' => 'sitepad', + 'SPIP' => 'spip', + 'STUDIO' => 'studio', + 'Synology - Synology DiskStation' => 'synology', + 'Thufie' => 'thufie', + 'TiddlyWiki' => 'tiddlywiki', + 'Typesetter CMS' => 'typesetter', + 'TYPO3 CMS' => 'typo3', + 'Tumblr' => 'tumblr', + 'Typecho' => 'typecho', + 'http://txt2tags.org' => 'txt2tags', + 'UISP' => 'uisp', + 'VuePress' => 'vuepress', + 'webEdition CMS' => 'webedition', + 'Webflow' => 'webflow', + 'WikkaWiki' => 'wikkawiki', + 'Wix.com' => 'wix.com', + 'WordPress' => 'wordpress', + 'WordPress.com' => 'wordpress', + 'Write.as' => 'write.as', + 'XAG/CMS' => 'xagcms', + 'Zim' => 'zim', +]; + +$ap_platforms = [ + 'PeerTube' => 'peertube', + 'Hugo' => 'hugo', + 'lemoncurry' => 'lemoncurry', + 'Ghost' => 'ghost', + 'Jekyll' => 'jekyll', + 'Known https://withknown.com' => 'known', +]; + +$dfrn_platforms = [ + 'Friendika' => 'friendika', + 'Friendica' => 'friendica', +]; + +$zap_platforms = [ + 'hubzilla' => 'hubzilla', + 'osada' => 'osada', +];