X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FModel%2FGServer.php;h=c0e4899923d86cb375109e9b2240675d166cc5eb;hb=a662245c744f4d2faa1b533977f7d21b4de6a724;hp=ca939c176c00fe5b7e83ffed4525ca8165db70b9;hpb=251bd30d778af3ac250c63f8096a838a72d54f2c;p=friendica.git diff --git a/src/Model/GServer.php b/src/Model/GServer.php index ca939c176c..c0e4899923 100644 --- a/src/Model/GServer.php +++ b/src/Model/GServer.php @@ -1,6 +1,6 @@ '\_', '%' => '\%', '?' => '_', '*' => '%']); + + // The SUBSTRING_INDEX returns everything before the eventual third /, which effectively trims an + // eventual server path and keep only the server domain which we're matching against the pattern. + $sql = "SELECT `gserver`.*, COUNT(*) AS `contacts` + FROM `gserver` + LEFT JOIN `contact` ON `gserver`.`id` = `contact`.`gsid` + WHERE SUBSTRING_INDEX(`gserver`.`nurl`, '/', 3) LIKE ? + AND NOT `gserver`.`failed` + GROUP BY `gserver`.`id`"; + + $stmt = DI::dba()->p($sql, $likePattern); + + return DI::dba()->toArray($stmt); + } + /** * Checks if the given server is reachable * @@ -171,7 +201,7 @@ class GServer if (($now - $contact_time) < (60 * 60 * 24)) { return DateTimeFormat::utc('now +1 day'); } - + // If the last contact was less than a week before then try again in a week if (($now - $contact_time) < (60 * 60 * 24 * 7)) { return DateTimeFormat::utc('now +1 week'); @@ -314,16 +344,16 @@ class GServer // When a nodeinfo is present, we don't need to dig further $xrd_timeout = DI::config()->get('system', 'xrd_timeout'); - $curlResult = DI::httpRequest()->get($url . '/.well-known/nodeinfo', ['timeout' => $xrd_timeout]); + $curlResult = DI::httpClient()->get($url . '/.well-known/nodeinfo', HttpClientAccept::JSON, [HttpClientOptions::TIMEOUT => $xrd_timeout]); if ($curlResult->isTimeout()) { self::setFailure($url); return false; } // On a redirect follow the new host but mark the old one as failure - if ($curlResult->isSuccess() && (parse_url($url, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST))) { - $curlResult = DI::httpRequest()->get($url, ['timeout' => $xrd_timeout]); - if (parse_url($url, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST)) { + if ($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, $only_nodeinfo); @@ -348,6 +378,12 @@ class GServer if (empty($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 @@ -358,9 +394,9 @@ class GServer $basedata = ['detection-method' => self::DETECT_MANUAL]; } - $curlResult = DI::httpRequest()->get($baseurl, ['timeout' => $xrd_timeout]); + $curlResult = DI::httpClient()->get($baseurl, HttpClientAccept::HTML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); if ($curlResult->isSuccess()) { - if ((parse_url($baseurl, PHP_URL_HOST) != parse_url($curlResult->getRedirectUrl(), PHP_URL_HOST))) { + 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); @@ -382,7 +418,7 @@ class GServer // 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. - $curlResult = DI::httpRequest()->get($url, ['timeout' => $xrd_timeout]); + $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); if ($curlResult->isSuccess()) { $urldata = self::analyseRootHeader($curlResult, $serverdata); $urldata = self::analyseRootBody($curlResult, $urldata, $url); @@ -402,14 +438,20 @@ class GServer } } - if (empty($serverdata['network']) || ($serverdata['network'] == Protocol::ACTIVITYPUB)) { + if (empty($nodeinfo['network']) && (empty($serverdata['network']) || ($serverdata['network'] == Protocol::ACTIVITYPUB))) { $serverdata = self::detectMastodonAlikes($url, $serverdata); } // All following checks are done for systems that always have got a "host-meta" endpoint. // With this check we don't have to waste time and ressources for dead systems. // Also this hopefully prevents us from receiving abuse messages. - if (empty($serverdata['network']) && !self::validHostMeta($url)) { + $validHostMeta = self::validHostMeta($url); + + if (empty($serverdata['network']) && !$validHostMeta) { + $serverdata = self::detectFromContacts($url, $serverdata); + } + + if (empty($serverdata['network']) && !$validHostMeta) { self::setFailure($url); return false; } @@ -436,10 +478,14 @@ class GServer $serverdata = self::detectNextcloud($url, $serverdata); } - if (empty($serverdata['network'])) { + 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; @@ -447,9 +493,13 @@ class GServer // Detect the directory type $serverdata['directory-type'] = self::DT_NONE; - $serverdata = self::checkPoCo($url, $serverdata); + $serverdata = self::checkMastodonDirectory($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; @@ -463,17 +513,21 @@ class GServer $serverdata['url'] = $url; $serverdata['nurl'] = Strings::normaliseLink($url); - // We take the highest number that we do find - $registeredUsers = $serverdata['registered-users'] ?? 0; + if (in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED])) { + $serverdata = self::detectNetworkViaContacts($url, $serverdata); + } - // On an active server there has to be at least a single user - if (($serverdata['network'] != Protocol::PHANTOM) && ($registeredUsers == 0)) { - $registeredUsers = 1; + if ($serverdata['network'] == Protocol::ACTIVITYPUB) { + $serverdata = self::fetchWeeklyUsage($url, $serverdata); } - if ($serverdata['network'] == Protocol::PHANTOM) { - $serverdata['registered-users'] = max($registeredUsers, 1); - $serverdata = self::detectNetworkViaContacts($url, $serverdata); + $serverdata['registered-users'] = $serverdata['registered-users'] ?? 0; + + // On an active server there has to be at least a single user + if (!in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED]) && ($serverdata['registered-users'] == 0)) { + $serverdata['registered-users'] = 1; + } elseif (in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED])) { + $serverdata['registered-users'] = 0; } $serverdata['next_contact'] = self::getNextUpdateDate(true); @@ -487,11 +541,6 @@ class GServer $ret = DBA::insert('gserver', $serverdata); $id = DBA::lastInsertId(); } else { - // Don't override the network with 'unknown' when there had been a valid entry before - if (($serverdata['network'] == Protocol::PHANTOM) && !empty($gserver['network'])) { - unset($serverdata['network']); - } - $ret = DBA::update('gserver', $serverdata, ['nurl' => $serverdata['nurl']]); $gserver = DBA::selectFirst('gserver', ['id'], ['nurl' => $serverdata['nurl']]); if (DBA::isResult($gserver)) { @@ -499,14 +548,31 @@ class GServer } } - if (!empty($serverdata['network']) && !empty($id) && ($serverdata['network'] != Protocol::PHANTOM)) { + // Count the number of known contacts from this server + if (!empty($id) && !in_array($serverdata['network'], [Protocol::PHANTOM, Protocol::FEED])) { $apcontacts = DBA::count('apcontact', ['gsid' => $id]); - $contacts = DBA::count('contact', ['uid' => 0, 'gsid' => $id]); - $max_users = max($apcontacts, $contacts, $registeredUsers, 1); - if ($max_users > $registeredUsers) { + $contacts = DBA::count('contact', ['uid' => 0, 'gsid' => $id, 'failed' => false]); + $max_users = max($apcontacts, $contacts); + if ($max_users > $serverdata['registered-users']) { Logger::info('Update registered users', ['id' => $id, 'url' => $serverdata['nurl'], 'registered-users' => $max_users]); DBA::update('gserver', ['registered-users' => $max_users], ['id' => $id]); } + + if (empty($serverdata['active-month-users'])) { + $contacts = DBA::count('contact', ["`uid` = ? AND `gsid` = ? AND NOT `failed` AND `last-item` > ?", 0, $id, DateTimeFormat::utc('now - 30 days')]); + if ($contacts > 0) { + Logger::info('Update monthly users', ['id' => $id, 'url' => $serverdata['nurl'], 'monthly-users' => $contacts]); + 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) { + Logger::info('Update halfyear users', ['id' => $id, 'url' => $serverdata['nurl'], 'halfyear-users' => $contacts]); + DBA::update('gserver', ['active-halfyear-users' => $contacts], ['id' => $id]); + } + } } if (!empty($serverdata['network']) && in_array($serverdata['network'], [Protocol::DFRN, Protocol::DIASPORA])) { @@ -526,7 +592,7 @@ class GServer { Logger::info('Discover relay data', ['server' => $server_url]); - $curlResult = DI::httpRequest()->get($server_url . '/.well-known/x-social-relay'); + $curlResult = DI::httpClient()->get($server_url . '/.well-known/x-social-relay', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return; } @@ -621,7 +687,7 @@ class GServer */ private static function fetchStatistics(string $url) { - $curlResult = DI::httpRequest()->get($url . '/statistics.json'); + $curlResult = DI::httpClient()->get($url . '/statistics.json', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return []; } @@ -658,6 +724,21 @@ class GServer } } + if (!empty($data['total_users'])) { + $serverdata['registered-users'] = max($data['total_users'], 1); + } + + if (!empty($data['active_users_monthly'])) { + $serverdata['active-month-users'] = max($data['active_users_monthly'], 0); + } + + if (!empty($data['active_users_halfyear'])) { + $serverdata['active-halfyear-users'] = max($data['active_users_halfyear'], 0); + } + + if (!empty($data['local_posts'])) { + $serverdata['local-posts'] = max($data['local_posts'], 0); + } if (!empty($data['registrations_open'])) { $serverdata['register_policy'] = Register::OPEN; @@ -671,18 +752,19 @@ class GServer /** * Detect server type by using the nodeinfo data * - * @param string $url address of the server - * @param CurlResult $curlResult + * @param string $url address of the server + * @param ICanHandleHttpResponses $httpResult + * * @return array Server data * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ - private static function fetchNodeinfo(string $url, CurlResult $curlResult) + private static function fetchNodeinfo(string $url, ICanHandleHttpResponses $httpResult) { - if (!$curlResult->isSuccess()) { + if (!$httpResult->isSuccess()) { return []; } - $nodeinfo = json_decode($curlResult->getBody(), true); + $nodeinfo = json_decode($httpResult->getBody(), true); if (!is_array($nodeinfo) || empty($nodeinfo['links'])) { return []; @@ -731,8 +813,7 @@ class GServer */ private static function parseNodeinfo1(string $nodeinfo_url) { - $curlResult = DI::httpRequest()->get($nodeinfo_url); - + $curlResult = DI::httpClient()->get($nodeinfo_url, HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return []; } @@ -771,6 +852,22 @@ class GServer $server['registered-users'] = max($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($nodeinfo['protocols']['inbound']) && is_array($nodeinfo['protocols']['inbound'])) { $protocols = []; foreach ($nodeinfo['protocols']['inbound'] as $protocol) { @@ -802,13 +899,14 @@ class GServer /** * Parses Nodeinfo 2 * + * @see https://git.feneas.org/jaywink/nodeinfo2 * @param string $nodeinfo_url address of the nodeinfo path * @return array Server data * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ private static function parseNodeinfo2(string $nodeinfo_url) { - $curlResult = DI::httpRequest()->get($nodeinfo_url); + $curlResult = DI::httpClient()->get($nodeinfo_url, HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return []; } @@ -836,6 +934,11 @@ class GServer // 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', $server['version']); + + // qoto advertises itself as Mastodon + if (($server['platform'] == 'mastodon') && substr($nodeinfo['software']['version'], -5) == '-qoto') { + $server['platform'] = 'qoto'; + } } } @@ -847,10 +950,28 @@ class GServer $server['registered-users'] = max($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) { - $protocols[$protocol] = true; + if (is_string($protocol)) { + $protocols[$protocol] = true; + } } if (!empty($protocols['dfrn'])) { @@ -885,7 +1006,7 @@ class GServer */ private static function fetchSiteinfo(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/siteinfo.json'); + $curlResult = DI::httpClient()->get($url . '/siteinfo.json', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return $serverdata; } @@ -924,6 +1045,22 @@ class GServer $serverdata['registered-users'] = max($data['channels_total'], 1); } + if (!empty($data['channels_active_monthly'])) { + $serverdata['active-month-users'] = max($data['channels_active_monthly'], 0); + } + + if (!empty($data['channels_active_halfyear'])) { + $serverdata['active-halfyear-users'] = max($data['channels_active_halfyear'], 0); + } + + if (!empty($data['local_posts'])) { + $serverdata['local-posts'] = max($data['local_posts'], 0); + } + + if (!empty($data['local_comments'])) { + $serverdata['local-comments'] = max($data['local_comments'], 0); + } + if (!empty($data['register_policy'])) { switch ($data['register_policy']) { case 'REGISTER_OPEN': @@ -954,7 +1091,7 @@ class GServer private static function validHostMeta(string $url) { $xrd_timeout = DI::config()->get('system', 'xrd_timeout'); - $curlResult = DI::httpRequest()->get($url . '/.well-known/host-meta', ['timeout' => $xrd_timeout]); + $curlResult = DI::httpClient()->get($url . '/.well-known/host-meta', HttpClientAccept::XRD_XML, [HttpClientOptions::TIMEOUT => $xrd_timeout]); if (!$curlResult->isSuccess()) { return false; } @@ -1017,16 +1154,18 @@ class GServer return $serverdata; } + $time = time(); foreach ($contacts as $contact) { - $probed = Contact::getByURL($contact); - if (!empty($probed) && in_array($probed['network'], Protocol::FEDERATED)) { + $probed = Contact::getByURL($contact, true); + if (!empty($probed) && !$probed['failed'] && in_array($probed['network'], Protocol::FEDERATED)) { $serverdata['network'] = $probed['network']; break; + } elseif ((time() - $time) > 10) { + // To reduce the stress on remote systems we probe a maximum of 10 seconds + break; } } - $serverdata['registered-users'] = max($serverdata['registered-users'], count($contacts), 1); - return $serverdata; } @@ -1044,7 +1183,7 @@ class GServer { $serverdata['poco'] = ''; - $curlResult = DI::httpRequest()->get($url . '/poco'); + $curlResult = DI::httpClient()->get($url . '/poco', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return $serverdata; } @@ -1074,7 +1213,7 @@ class GServer */ public static function checkMastodonDirectory(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/api/v1/directory?limit=1'); + $curlResult = DI::httpClient()->get($url . '/api/v1/directory?limit=1', HttpClientAccept::JSON); if (!$curlResult->isSuccess()) { return $serverdata; } @@ -1101,8 +1240,7 @@ class GServer */ private static function detectPeertube(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/api/v1/config'); - + $curlResult = DI::httpClient()->get($url . '/api/v1/config', HttpClientAccept::JSON); if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { return $serverdata; } @@ -1149,8 +1287,7 @@ class GServer */ private static function detectNextcloud(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/status.php'); - + $curlResult = DI::httpClient()->get($url . '/status.php', HttpClientAccept::JSON); if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { return $serverdata; } @@ -1173,6 +1310,65 @@ class GServer return $serverdata; } + private static function fetchWeeklyUsage(string $url, array $serverdata) { + $curlResult = DI::httpClient()->get($url . '/api/v1/instance/activity', HttpClientAccept::JSON); + if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { + return $serverdata; + } + + $data = json_decode($curlResult->getBody(), true); + if (empty($data)) { + return $serverdata; + } + + $current_week = []; + foreach ($data as $week) { + // Use only data from a full week + if (empty($week['week']) || (time() - $week['week']) < 7 * 24 * 60 * 60) { + continue; + } + + // Most likely the data is sorted correctly. But we better are safe than sorry + if (empty($current_week['week']) || ($current_week['week'] < $week['week'])) { + $current_week = $week; + } + } + + if (!empty($current_week['logins'])) { + $serverdata['active-week-users'] = max($current_week['logins'], 0); + } + + 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 * @@ -1183,8 +1379,7 @@ class GServer */ private static function detectMastodonAlikes(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/api/v1/instance'); - + $curlResult = DI::httpClient()->get($url . '/api/v1/instance', HttpClientAccept::JSON); if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { return $serverdata; } @@ -1249,7 +1444,7 @@ class GServer */ private static function detectHubzilla(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/api/statusnet/config.json'); + $curlResult = DI::httpClient()->get($url . '/api/statusnet/config.json', HttpClientAccept::JSON); if (!$curlResult->isSuccess() || ($curlResult->getBody() == '')) { return $serverdata; } @@ -1334,7 +1529,57 @@ 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 @@ -1347,7 +1592,7 @@ class GServer private static function detectGNUSocial(string $url, array $serverdata) { // Test for GNU Social - $curlResult = DI::httpRequest()->get($url . '/api/gnusocial/version.json'); + $curlResult = DI::httpClient()->get($url . '/api/gnusocial/version.json', HttpClientAccept::JSON); if ($curlResult->isSuccess() && ($curlResult->getBody() != '{"error":"not implemented"}') && ($curlResult->getBody() != '') && (strlen($curlResult->getBody()) < 30)) { $serverdata['platform'] = 'gnusocial'; @@ -1360,12 +1605,12 @@ class GServer if (in_array($serverdata['detection-method'], [self::DETECT_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { $serverdata['detection-method'] = self::DETECT_GNUSOCIAL; } - + return $serverdata; } // Test for Statusnet - $curlResult = DI::httpRequest()->get($url . '/api/statusnet/version.json'); + $curlResult = DI::httpClient()->get($url . '/api/statusnet/version.json', HttpClientAccept::JSON); if ($curlResult->isSuccess() && ($curlResult->getBody() != '{"error":"not implemented"}') && ($curlResult->getBody() != '') && (strlen($curlResult->getBody()) < 30)) { @@ -1401,9 +1646,11 @@ class GServer */ private static function detectFriendica(string $url, array $serverdata) { - $curlResult = DI::httpRequest()->get($url . '/friendica/json'); + // There is a bug in some versions of Friendica that will return an ActivityStream actor when the content type "application/json" is requested. + // Because of this me must not use ACCEPT_JSON here. + $curlResult = DI::httpClient()->get($url . '/friendica/json'); if (!$curlResult->isSuccess()) { - $curlResult = DI::httpRequest()->get($url . '/friendika/json'); + $curlResult = DI::httpClient()->get($url . '/friendika/json'); $friendika = true; $platform = 'Friendika'; } else { @@ -1420,7 +1667,7 @@ 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_HEADER, self::DETECT_BODY, self::DETECT_MANUAL])) { $serverdata['detection-method'] = $friendika ? self::DETECT_FRIENDIKA : self::DETECT_FRIENDICA; } @@ -1475,6 +1722,14 @@ class GServer */ private static function analyseRootBody($curlResult, array $serverdata, string $url) { + if (empty($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']; + $doc = new DOMDocument(); @$doc->loadHTML($curlResult->getBody()); $xpath = new DOMXPath($doc); @@ -1524,11 +1779,8 @@ class GServer $serverdata['version'] = $version_part[1]; // We still do need a reliable test if some AP plugin is activated - if (DBA::exists('apcontact', ['baseurl' => $url])) { - $serverdata['network'] = Protocol::ACTIVITYPUB; - } else { - $serverdata['network'] = Protocol::FEED; - } + // By now we just check in a later process for some known contacts + $serverdata['network'] = Protocol::FEED; if ($serverdata['detection-method'] == self::DETECT_MANUAL) { $serverdata['detection-method'] = self::DETECT_BODY; @@ -1588,7 +1840,11 @@ class GServer } } - if (!empty($serverdata['network']) && ($serverdata['detection-method'] == self::DETECT_MANUAL)) { + if (!empty($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)) { $serverdata['detection-method'] = self::DETECT_BODY; } @@ -1611,11 +1867,11 @@ class GServer } elseif ($curlResult->inHeader('x-diaspora-version')) { $serverdata['platform'] = 'diaspora'; $serverdata['network'] = Protocol::DIASPORA; - $serverdata['version'] = $curlResult->getHeader('x-diaspora-version'); + $serverdata['version'] = $curlResult->getHeader('x-diaspora-version')[0] ?? ''; } elseif ($curlResult->inHeader('x-friendica-version')) { $serverdata['platform'] = 'friendica'; $serverdata['network'] = Protocol::DFRN; - $serverdata['version'] = $curlResult->getHeader('x-friendica-version'); + $serverdata['version'] = $curlResult->getHeader('x-friendica-version')[0] ?? ''; } else { return $serverdata; } @@ -1658,13 +1914,9 @@ class GServer $last_update = date('c', time() - (60 * 60 * 24 * $requery_days)); - $gservers = DBA::p("SELECT `id`, `url`, `nurl`, `network`, `poco`, `directory-type` - FROM `gserver` - WHERE NOT `failed` - AND `directory-type` != ? - AND `last_poco_query` < ? - ORDER BY RAND()", self::DT_NONE, $last_update - ); + $gservers = DBA::select('gserver', ['id', 'url', 'nurl', 'network', 'poco', 'directory-type'], + ["NOT `failed` AND `directory-type` != ? AND `last_poco_query` < ?", GServer::DT_NONE, $last_update], + ['order' => ['RAND()']]); while ($gserver = DBA::fetch($gservers)) { Logger::info('Update peer list', ['server' => $gserver['url'], 'id' => $gserver['id']]); @@ -1675,7 +1927,7 @@ class GServer $fields = ['last_poco_query' => DateTimeFormat::utcNow()]; DBA::update('gserver', $fields, ['nurl' => $gserver['nurl']]); - + if (--$no_of_queries == 0) { break; } @@ -1703,7 +1955,7 @@ class GServer $protocols = ['activitypub', 'diaspora', 'dfrn', 'ostatus']; foreach ($protocols as $protocol) { $query = '{nodes(protocol:"' . $protocol . '"){host}}'; - $curlResult = DI::httpRequest()->fetch('https://the-federation.info/graphql?query=' . urlencode($query)); + $curlResult = DI::httpClient()->fetch('https://the-federation.info/graphql?query=' . urlencode($query), HttpClientAccept::JSON); if (!empty($curlResult)) { $data = json_decode($curlResult, true); if (!empty($data['data']['nodes'])) { @@ -1720,9 +1972,7 @@ class GServer if (!empty($accesstoken)) { $api = 'https://instances.social/api/1.0/instances/list?count=0'; - $header = ['Authorization: Bearer '.$accesstoken]; - $curlResult = DI::httpRequest()->get($api, ['header' => $header]); - + $curlResult = DI::httpClient()->get($api, HttpClientAccept::JSON, [HttpClientOptions::HEADERS => ['Authorization' => ['Bearer ' . $accesstoken]]]); if ($curlResult->isSuccess()) { $servers = json_decode($curlResult->getBody(), true); @@ -1741,8 +1991,8 @@ class GServer * * @param int $gsid Server id * @param int $protocol Protocol id - * @return void - * @throws Exception + * @return void + * @throws Exception */ public static function setProtocol(int $gsid, int $protocol) { @@ -1793,7 +2043,7 @@ class GServer } } - Logger::info('Protocol for server', ['protocol' => $protocol, 'old' => $old, 'id' => $gsid, 'url' => $gserver['url']]); + Logger::info('Protocol for server', ['protocol' => $protocol, 'old' => $old, 'id' => $gsid, 'url' => $gserver['url'], 'callstack' => System::callstack(20)]); DBA::update('gserver', ['protocol' => $protocol], ['id' => $gsid]); } @@ -1801,8 +2051,8 @@ class GServer * Fetch the protocol of the given server * * @param int $gsid Server id - * @return int - * @throws Exception + * @return int + * @throws Exception */ public static function getProtocol(int $gsid) {