X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FProtocol%2FActivityPub%2FProcessor.php;h=f142b7e91d148070b0234ef5de97c2923900c028;hb=548b7f43a5f22fa4eea917de1ed3de7588ac294d;hp=db8f28aa0cabcac4d6f2927a99c9e115db9126ca;hpb=86105635ca3b0e096d128718b778b313b5eaa88e;p=friendica.git diff --git a/src/Protocol/ActivityPub/Processor.php b/src/Protocol/ActivityPub/Processor.php index db8f28aa0c..f142b7e91d 100644 --- a/src/Protocol/ActivityPub/Processor.php +++ b/src/Protocol/ActivityPub/Processor.php @@ -24,6 +24,7 @@ namespace Friendica\Protocol\ActivityPub; use Friendica\Content\Text\BBCode; use Friendica\Content\Text\HTML; use Friendica\Content\Text\Markdown; +use Friendica\Core\Cache\Enum\Duration; use Friendica\Core\Logger; use Friendica\Core\Protocol; use Friendica\Core\System; @@ -56,6 +57,34 @@ use Friendica\Worker\Delivery; */ class Processor { + const CACHEKEY_FETCH_ACTIVITY = 'processor:fetchMissingActivity:'; + const CACHEKEY_JUST_FETCHED = 'processor:isJustFetched:'; + + /** + * Add an object id to the list of processed ids + * + * @param string $id + * + * @return void + */ + private static function addActivityId(string $id) + { + DBA::delete('processed-activity', ["`received` < ?", DateTimeFormat::utc('now - 5 minutes')]); + DBA::insert('processed-activity', ['object-id' => $id, 'received' => DateTimeFormat::utcNow()]); + } + + /** + * Checks if the given object id has just been processed + * + * @param string $id + * + * @return boolean + */ + private static function isProcessed(string $id): bool + { + return DBA::exists('processed-activity', ['object-id' => $id]); + } + /** * Extracts the tag character (#, @, !) from mention links * @@ -197,11 +226,12 @@ class Processor */ public static function updateItem(array $activity) { - $item = Post::selectFirst(['uri', 'uri-id', 'thr-parent', 'gravity', 'post-type'], ['uri' => $activity['id']]); + $item = Post::selectFirst(['uri', 'uri-id', 'thr-parent', 'gravity', 'post-type', 'private'], ['uri' => $activity['id']]); if (!DBA::isResult($item)) { Logger::warning('No existing item, item will be created', ['uri' => $activity['id']]); - $item = self::createItem($activity); + $item = self::createItem($activity, false); if (empty($item)) { + Queue::remove($activity); return; } @@ -213,14 +243,14 @@ class Processor $item['edited'] = DateTimeFormat::utc($activity['updated']); $item = self::processContent($activity, $item); - - self::storeAttachments($activity, $item); - self::storeQuestion($activity, $item); - if (empty($item)) { + Queue::remove($activity); return; } + self::storeAttachments($activity, $item); + self::storeQuestion($activity, $item); + Post\History::add($item['uri-id'], $item); Item::update($item, ['uri' => $activity['id']]); @@ -263,13 +293,23 @@ class Processor /** * Prepares data for a message * - * @param array $activity Activity array + * @param array $activity Activity array + * @param bool $fetch_parents + * * @return array Internal item + * * @throws \Friendica\Network\HTTPException\InternalServerErrorException * @throws \ImagickException */ - public static function createItem(array $activity): array + public static function createItem(array $activity, bool $fetch_parents): array { + if (self::isProcessed($activity['id']) && !Post::exists(['uri' => $activity['id']])) { + Logger::info('Id is already processed', ['id' => $activity['id']]); + return []; + } + + self::addActivityId($activity['id']); + $item = []; $item['verb'] = Activity::POST; $item['thr-parent'] = $activity['reply-to-id']; @@ -282,10 +322,10 @@ class Processor $item['object-type'] = Activity\ObjectType::COMMENT; } - if (!empty($activity['context'])) { - $item['conversation'] = $activity['context']; - } elseif (!empty($activity['conversation'])) { + if (!empty($activity['conversation'])) { $item['conversation'] = $activity['conversation']; + } elseif (!empty($activity['context'])) { + $item['conversation'] = $activity['context']; } if (!empty($item['conversation'])) { @@ -298,42 +338,25 @@ class Processor $conversation = []; } - if (empty($activity['directmessage']) && ($activity['id'] != $activity['reply-to-id']) && !Post::exists(['uri' => $activity['reply-to-id']])) { - $recursion_depth = $activity['recursion-depth'] ?? 0; - Logger::notice('Parent not found. Try to refetch it.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); - if ($recursion_depth < 10) { - $result = self::fetchMissingActivity($activity['reply-to-id'], $activity, '', Receiver::COMPLETION_AUTO); - if (empty($result) && self::isActivityGone($activity['reply-to-id'])) { - // Recursively delete this and all depending entries - Queue::deleteById($activity['entry-id']); - return []; - } - $fetch_by_worker = empty($result); - } else { - Logger::notice('Recursion level is too high.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); - $fetch_by_worker = true; - } + Logger::debug('Create Item', ['id' => $activity['id'], 'conversation' => $item['conversation'] ?? '']); + if (empty($activity['author']) && empty($activity['actor'])) { + Logger::notice('Missing author and actor. We quit here.', ['activity' => $activity]); + Queue::remove($activity); + return []; + } - if ($fetch_by_worker && Queue::hasWorker($activity)) { - Logger::notice('There is already a worker task to fetch the post.', ['id' => $activity['id'], 'parent' => $activity['reply-to-id']]); - $fetch_by_worker = false; - if (!empty($conversation)) { - return []; - } - } + if (!in_array(0, $activity['receiver']) || !DI::config()->get('system', 'fetch_parents')) { + $fetch_parents = false; + } - if ($fetch_by_worker) { - Logger::notice('Fetching is done by worker.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); - $activity['recursion-depth'] = 0; - $wid = Worker::add(PRIORITY_HIGH, 'FetchMissingActivity', $activity['reply-to-id'], $activity, '', Receiver::COMPLETION_AUTO); - Queue::setWorkerId($activity, $wid); - if (!empty($conversation)) { - return []; - } - } elseif (!empty($result)) { + if ($fetch_parents && empty($activity['directmessage']) && ($activity['id'] != $activity['reply-to-id']) && !Post::exists(['uri' => $activity['reply-to-id']])) { + $result = self::fetchParent($activity); + if (!empty($result)) { if (($item['thr-parent'] != $result) && Post::exists(['uri' => $result])) { $item['thr-parent'] = $result; } + } elseif (empty($conversation)) { + return []; } } @@ -341,6 +364,9 @@ class Processor if (empty($conversation) && empty($activity['directmessage']) && ($item['gravity'] != GRAVITY_PARENT) && !Post::exists(['uri' => $item['thr-parent']])) { Logger::info('Parent not found, message will be discarded.', ['thr-parent' => $item['thr-parent']]); + if (!$fetch_parents) { + Queue::remove($activity); + } return []; } @@ -360,11 +386,12 @@ class Processor if (!empty($activity['raw'])) { $item['source'] = $activity['raw']; - $item['protocol'] = Conversation::PARCEL_ACTIVITYPUB; + } - if (isset($activity['push'])) { - $item['direction'] = $activity['push'] ? Conversation::PUSH : Conversation::PULL; - } + $item['protocol'] = Conversation::PARCEL_ACTIVITYPUB; + + if (isset($activity['push'])) { + $item['direction'] = $activity['push'] ? Conversation::PUSH : Conversation::PULL; } if (!empty($activity['from-relay'])) { @@ -433,6 +460,7 @@ class Processor $item = self::processContent($activity, $item); if (empty($item)) { Logger::info('Message was not processed'); + Queue::remove($activity); return []; } @@ -457,6 +485,72 @@ class Processor return $item; } + /** + * Fetch and process parent posts for the given activity + * + * @param array $activity + * + * @return string + */ + private static function fetchParent(array $activity): string + { + $recursion_depth = $activity['recursion-depth'] ?? 0; + + if ($recursion_depth < DI::config()->get('system', 'max_recursion_depth')) { + Logger::notice('Parent not found. Try to refetch it.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); + $result = self::fetchMissingActivity($activity['reply-to-id'], $activity, '', Receiver::COMPLETION_AUTO); + if (empty($result) && self::isActivityGone($activity['reply-to-id'])) { + Logger::notice('The activity is gone, the queue entry will be deleted', ['parent' => $activity['reply-to-id']]); + if (!empty($activity['entry-id'])) { + Queue::deleteById($activity['entry-id']); + } + return ''; + } elseif (!empty($result)) { + $exists = Post::exists(['uri' => [$result, $activity['reply-to-id']]]); + if ($exists) { + Logger::notice('The activity has been fetched and created.', ['parent' => $result]); + return $result; + } elseif (DI::config()->get('system', 'fetch_by_worker') || DI::config()->get('system', 'decoupled_receiver')) { + Logger::notice('The activity has been fetched and will hopefully be created later.', ['parent' => $result]); + } else { + Logger::notice('The activity exists but has not been created, the queue entry will be deleted.', ['parent' => $result]); + if (!empty($activity['entry-id'])) { + Queue::deleteById($activity['entry-id']); + } + } + return ''; + } + if (empty($result) && !DI::config()->get('system', 'fetch_by_worker')) { + return ''; + } + } elseif (self::isActivityGone($activity['reply-to-id'])) { + Logger::notice('The activity is gone. We will not spawn a worker. The queue entry will be deleted', ['parent' => $activity['reply-to-id']]); + if (!empty($activity['entry-id'])) { + Queue::deleteById($activity['entry-id']); + } + return ''; + } else { + Logger::notice('Recursion level is too high.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); + } + + if (Queue::hasWorker($activity['worker-id'] ?? 0)) { + Logger::notice('There is already a worker task to fetch the post.', ['id' => $activity['id'], 'parent' => $activity['reply-to-id']]); + return ''; + } + + if (!Fetch::hasWorker($activity['reply-to-id'])) { + Logger::notice('Fetching is done by worker.', ['parent' => $activity['reply-to-id'], 'recursion-depth' => $recursion_depth]); + Fetch::add($activity['reply-to-id']); + $activity['recursion-depth'] = 0; + $wid = Worker::add(PRIORITY_HIGH, 'FetchMissingActivity', $activity['reply-to-id'], $activity, '', Receiver::COMPLETION_AUTO); + Fetch::setWorkerId($activity['reply-to-id'], $wid); + } else { + Logger::debug('Activity will already be fetched via a worker.', ['url' => $activity['reply-to-id']]); + } + + return ''; + } + /** * Check if a given activity is no longer available * @@ -473,16 +567,23 @@ class Processor } // @todo To ensure that the remote system is working correctly, we can check if the "Content-Type" contains JSON - if (in_array($curlResult->getReturnCode(), [404])) { + if (in_array($curlResult->getReturnCode(), [401, 404])) { return true; } - $object = json_decode($curlResult->getBody(), true); - if (!empty($object)) { - $activity = JsonLD::compact($object); - if (JsonLD::fetchElement($activity, '@type') == 'as:Tombstone') { + if ($curlResult->isSuccess()) { + $object = json_decode($curlResult->getBody(), true); + if (!empty($object)) { + $activity = JsonLD::compact($object); + if (JsonLD::fetchElement($activity, '@type') == 'as:Tombstone') { + return true; + } + } + } elseif ($curlResult->getReturnCode() == 0) { + $host = parse_url($url, PHP_URL_HOST); + if (!(filter_var($host, FILTER_VALIDATE_IP) || @dns_get_record($host . '.', DNS_A + DNS_AAAA))) { return true; - } + } } return false; @@ -544,7 +645,7 @@ class Processor public static function createActivity(array $activity, string $verb) { $activity['reply-to-id'] = $activity['object_id']; - $item = self::createItem($activity); + $item = self::createItem($activity, false); if (empty($item)) { return; } @@ -727,7 +828,7 @@ class Processor Logger::warning('Unknown parent item.', ['uri' => $parent_uri]); return false; } - if (($item['private'] == Item::PRIVATE) && ($parent['private'] != Item::PRIVATE)) { + if (!empty($activity['type']) && in_array($activity['type'], Receiver::CONTENT_TYPES) && ($item['private'] == Item::PRIVATE) && ($parent['private'] != Item::PRIVATE)) { Logger::warning('Item is private but the parent is not. Dropping.', ['item-uri' => $item['uri'], 'thr-parent' => $item['thr-parent']]); return false; } @@ -857,6 +958,9 @@ class Processor if (!self::isSolicitedMessage($activity, $item)) { DBA::delete('item-uri', ['id' => $item['uri-id']]); + if (!empty($activity['entry-id'])) { + Queue::deleteById($activity['entry-id']); + } return; } @@ -894,10 +998,16 @@ class Processor $item['post-reason'] = Item::PR_NONE; } - if (!empty($activity['from-relay'])) { - $item['post-reason'] = Item::PR_RELAY; - } elseif (!empty($activity['thread-completion'])) { - $item['post-reason'] = Item::PR_FETCHED; + $item['post-reason'] = Item::getPostReason($item); + + if (in_array($item['post-reason'], [Item::PR_GLOBAL, Item::PR_NONE])) { + if (!empty($activity['from-relay'])) { + $item['post-reason'] = Item::PR_RELAY; + } elseif (!empty($activity['thread-completion'])) { + $item['post-reason'] = Item::PR_FETCHED; + } elseif (!empty($activity['push'])) { + $item['post-reason'] = Item::PR_PUSHED; + } } if ($item['isForum'] ?? false) { @@ -915,41 +1025,31 @@ class Processor continue; } - if (!($item['isForum'] ?? false) && ($receiver != 0) && ($item['gravity'] == GRAVITY_PARENT) && !Contact::isSharingByURL($activity['author'], $receiver)) { - if ($item['post-reason'] == Item::PR_BCC) { - Logger::info('Top level post via BCC from a non sharer, ignoring', ['uid' => $receiver, 'contact' => $item['contact-id']]); - continue; - } + if (($receiver != 0) && ($item['gravity'] == GRAVITY_PARENT) && !in_array($item['post-reason'], [Item::PR_FOLLOWER, Item::PR_TAG, item::PR_TO, Item::PR_CC])) { + if (!($item['isForum'] ?? false)) { + if ($item['post-reason'] == Item::PR_BCC) { + Logger::info('Top level post via BCC from a non sharer, ignoring', ['uid' => $receiver, 'contact' => $item['contact-id'], 'url' => $item['uri']]); + continue; + } - if ( - !empty($activity['thread-children-type']) - && in_array($activity['thread-children-type'], Receiver::ACTIVITY_TYPES) - && DI::pConfig()->get($receiver, 'system', 'accept_only_sharer') != Item::COMPLETION_LIKE - ) { - Logger::info('Top level post from thread completion from a non sharer had been initiated via an activity, ignoring', - ['type' => $activity['thread-children-type'], 'user' => $item['uid'], 'causer' => $item['causer-link'], 'author' => $activity['author'], 'url' => $item['uri']]); - continue; + if ((DI::pConfig()->get($receiver, 'system', 'accept_only_sharer') != Item::COMPLETION_LIKE) + && in_array($activity['thread-children-type'] ?? '', Receiver::ACTIVITY_TYPES)) { + Logger::info('Top level post from thread completion from a non sharer had been initiated via an activity, ignoring', + ['type' => $activity['thread-children-type'], 'user' => $item['uid'], 'causer' => $item['causer-link'], 'author' => $activity['author'], 'url' => $item['uri']]); + continue; + } } - } - - $is_forum = false; - if ($receiver != 0) { + $is_forum = false; $user = User::getById($receiver, ['account-type']); if (!empty($user['account-type'])) { $is_forum = ($user['account-type'] == User::ACCOUNT_TYPE_COMMUNITY); } - } - - if (!$is_forum && DI::pConfig()->get($receiver, 'system', 'accept_only_sharer') == Item::COMPLETION_NONE && ($receiver != 0) && ($item['gravity'] == GRAVITY_PARENT)) { - $skip = !Contact::isSharingByURL($activity['author'], $receiver); - - if ($skip && (($activity['type'] == 'as:Announce') || ($item['isForum'] ?? false))) { - $skip = !Contact::isSharingByURL($activity['actor'], $receiver); - } - if ($skip) { - Logger::info('Skipping post', ['uid' => $receiver, 'url' => $item['uri']]); + if ((DI::pConfig()->get($receiver, 'system', 'accept_only_sharer') == Item::COMPLETION_NONE) + && ((!$is_forum && !($item['isForum'] ?? false) && ($activity['type'] != 'as:Announce')) + || !Contact::isSharingByURL($activity['actor'], $receiver))) { + Logger::info('Actor is a non sharer, is no forum or it is no announce', ['uid' => $receiver, 'actor' => $activity['actor'], 'url' => $item['uri'], 'type' => $activity['type']]); continue; } @@ -968,8 +1068,9 @@ class Processor $success = true; } else { Logger::notice('Item insertion aborted', ['uri' => $item['uri'], 'uid' => $item['uid']]); - if (Item::isTooOld($item) || !Item::isValid($item)) { - Queue::remove($activity); + if (($item['uid'] == 0) && (count($activity['receiver']) > 1)) { + Logger::info('Public item was aborted. We skip for all users.', ['uri' => $item['uri']]); + break; } } @@ -978,8 +1079,9 @@ class Processor } } - if ($success) { - Queue::remove($activity); + Queue::remove($activity); + + if ($success && Queue::hasChildren($item['uri']) && Post::exists(['uri' => $item['uri']])) { Queue::processReplyByUri($item['uri']); } @@ -1196,6 +1298,39 @@ class Processor Logger::info('Fetched featured posts', ['new' => $new, 'old' => $old, 'contact' => $url]); } + public static function fetchCachedActivity(string $url, int $uid): array + { + $cachekey = self::CACHEKEY_FETCH_ACTIVITY . $uid . ':' . $url; + $object = DI::cache()->get($cachekey); + + if (!is_null($object)) { + if (!empty($object)) { + Logger::debug('Fetch from cache', ['url' => $url, 'uid' => $uid]); + } else { + Logger::debug('Fetch from negative cache', ['url' => $url, 'uid' => $uid]); + } + return $object; + } + + $object = ActivityPub::fetchContent($url, $uid); + if (empty($object)) { + Logger::notice('Activity was not fetchable, aborting.', ['url' => $url, 'uid' => $uid]); + // We perform negative caching. + DI::cache()->set($cachekey, [], Duration::FIVE_MINUTES); + return []; + } + + if (empty($object['id'])) { + Logger::notice('Activity has got not id, aborting. ', ['url' => $url, 'object' => $object]); + return []; + } + DI::cache()->set($cachekey, $object, Duration::FIVE_MINUTES); + + Logger::debug('Activity was fetched successfully', ['url' => $url, 'uid' => $uid]); + + return $object; + } + /** * Fetches missing posts * @@ -1209,20 +1344,8 @@ class Processor */ public static function fetchMissingActivity(string $url, array $child = [], string $relay_actor = '', int $completion = Receiver::COMPLETION_MANUAL): string { - if (!empty($child['receiver'])) { - $uid = ActivityPub\Receiver::getFirstUserFromReceivers($child['receiver']); - } else { - $uid = 0; - } - - $object = ActivityPub::fetchContent($url, $uid); + $object = self::fetchCachedActivity($url, 0); if (empty($object)) { - Logger::notice('Activity was not fetchable, aborting.', ['url' => $url, 'uid' => $uid]); - return ''; - } - - if (empty($object['id'])) { - Logger::notice('Activity has got not id, aborting. ', ['url' => $url, 'object' => $object]); return ''; } @@ -1234,7 +1357,7 @@ class Processor $compacted = JsonLD::compact($object); $attributed_to = JsonLD::fetchElement($compacted, 'as:attributedTo', '@id'); } - $signer[] = $attributed_to; + $signer[] = $attributed_to; } if (!empty($object['actor'])) { @@ -1276,7 +1399,7 @@ class Processor $ldactivity = JsonLD::compact($activity); - $ldactivity['recursion-depth'] = !empty($child['recursion-depth']) ? $child['recursion-depth'] + 1 : 1; + $ldactivity['recursion-depth'] = !empty($child['recursion-depth']) ? $child['recursion-depth'] + 1 : 0; if (!empty($relay_actor)) { $ldactivity['thread-completion'] = $ldactivity['from-relay'] = Contact::getIdForURL($relay_actor); @@ -1289,17 +1412,25 @@ class Processor $ldactivity['completion-mode'] = $completion; } - if (!empty($child['type'])) { + if (!empty($child['thread-children-type'])) { + $ldactivity['thread-children-type'] = $child['thread-children-type']; + } elseif (!empty($child['type'])) { $ldactivity['thread-children-type'] = $child['type']; + } else { + $ldactivity['thread-children-type'] = 'as:Create'; } if (!empty($relay_actor) && !self::acceptIncomingMessage($ldactivity, $object['id'])) { return ''; } - ActivityPub\Receiver::processActivity($ldactivity, json_encode($activity), $uid, true, false, $signer); - - Logger::notice('Activity had been fetched and processed.', ['url' => $url, 'object' => $activity['id']]); + if (($completion == Receiver::COMPLETION_RELAY) && Queue::exists($url, 'as:Create')) { + Logger::notice('Activity has already been queued.', ['url' => $url, 'object' => $activity['id']]); + } elseif (ActivityPub\Receiver::processActivity($ldactivity, json_encode($activity), 0, true, false, $signer, '', $completion)) { + Logger::notice('Activity had been fetched and processed.', ['url' => $url, 'entry' => $child['entry-id'] ?? 0, 'completion' => $completion, 'object' => $activity['id']]); + } else { + Logger::notice('Activity had been fetched and will be processed later.', ['url' => $url, 'entry' => $child['entry-id'] ?? 0, 'completion' => $completion, 'object' => $activity['id']]); + } return $activity['id']; }