X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FProtocol%2FFeed.php;h=ebd6f5ed3008ddf1f569d55b78ec08501092cbac;hb=852d148efdfb737ea192b296cf07055194ac7706;hp=67baf4b2aebf6412945579c7d3ca8e3cdf654bd2;hpb=fcddf24f8808d18c2f5697994d64bddb88d4120a;p=friendica.git

diff --git a/src/Protocol/Feed.php b/src/Protocol/Feed.php
index 67baf4b2ae..ebd6f5ed30 100644
--- a/src/Protocol/Feed.php
+++ b/src/Protocol/Feed.php
@@ -1,6 +1,6 @@
 <?php
 /**
- * @copyright Copyright (C) 2020, Friendica
+ * @copyright Copyright (C) 2010-2021, the Friendica project
  *
  * @license GNU AGPL version 3 or any later version
  *
@@ -33,6 +33,7 @@ use Friendica\Database\DBA;
 use Friendica\DI;
 use Friendica\Model\Contact;
 use Friendica\Model\Item;
+use Friendica\Model\Post;
 use Friendica\Model\Tag;
 use Friendica\Model\User;
 use Friendica\Util\DateTimeFormat;
@@ -46,65 +47,6 @@ use Friendica\Util\XML;
  */
 class Feed
 {
-	/**
-	 * consume - process atom feed and update anything/everything we might need to update
-	 *
-	 * $xml = the (atom) feed to consume - RSS isn't as fully supported but may work for simple feeds.
-	 *
-	 * $importer = the contact_record (joined to user_record) of the local user who owns this relationship.
-	 *             It is this person's stuff that is going to be updated.
-	 * $contact =  the person who is sending us stuff. If not set, we MAY be processing a "follow" activity
-	 *             from an external network and MAY create an appropriate contact record. Otherwise, we MUST
-	 *             have a contact record.
-	 * $hub = should we find a hub declation in the feed, pass it back to our calling process, who might (or
-	 *        might not) try and subscribe to it.
-	 * $datedir sorts in reverse order
-	 * $pass - by default ($pass = 0) we cannot guarantee that a parent item has been
-	 *      imported prior to its children being seen in the stream unless we are certain
-	 *      of how the feed is arranged/ordered.
-	 * With $pass = 1, we only pull parent items out of the stream.
-	 * With $pass = 2, we only pull children (comments/likes).
-	 *
-	 * So running this twice, first with pass 1 and then with pass 2 will do the right
-	 * thing regardless of feed ordering. This won't be adequate in a fully-threaded
-	 * model where comments can have sub-threads. That would require some massive sorting
-	 * to get all the feed items into a mostly linear ordering, and might still require
-	 * recursion.
-	 *
-	 * @param       $xml
-	 * @param array $importer
-	 * @param array $contact
-	 * @param       $hub
-	 * @throws ImagickException
-	 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
-	 */
-	public static function consume($xml, array $importer, array $contact, &$hub)
-	{
-		if ($contact['network'] === Protocol::OSTATUS) {
-			Logger::info('Consume OStatus messages');
-			OStatus::import($xml, $importer, $contact, $hub);
-
-			return;
-		}
-
-		if ($contact['network'] === Protocol::FEED) {
-			Logger::info('Consume feeds');
-			self::import($xml, $importer, $contact);
-
-			return;
-		}
-
-		if ($contact['network'] === Protocol::DFRN) {
-			Logger::info('Consume DFRN messages');
-			$dfrn_importer = DFRN::getImporter($contact['id'], $importer['uid']);
-			if (!empty($dfrn_importer)) {
-				Logger::info('Now import the DFRN feed');
-				DFRN::import($xml, $dfrn_importer, true);
-				return;
-			}
-		}
-	}
-
 	/**
 	 * Read a RSS/RDF/Atom feed and create an item entry for it
 	 *
@@ -291,6 +233,7 @@ class Feed
 		$header["private"] = Item::PUBLIC;
 		$header["verb"] = Activity::POST;
 		$header["object-type"] = Activity\ObjectType::NOTE;
+		$header["post-type"] = Item::PT_ARTICLE;
 
 		$header["contact-id"] = $contact["id"] ?? 0;
 
@@ -309,6 +252,8 @@ class Feed
 			$total_items = $max_items;
 		}
 
+		$postings = [];
+
 		// Importing older entries first
 		for ($i = $total_items; $i >= 0; --$i) {
 			$entry = $entries->item($i);
@@ -353,8 +298,6 @@ class Feed
 
 			$item["plink"] = DI::httpRequest()->finalUrl($item["plink"]);
 
-			$item["parent-uri"] = $item["uri"];
-
 			$item["title"] = XML::getFirstNodeValue($xpath, 'atom:title/text()', $entry);
 
 			if (empty($item["title"])) {
@@ -397,7 +340,7 @@ class Feed
 			if (!$dryRun) {
 				$condition = ["`uid` = ? AND `uri` = ? AND `network` IN (?, ?)",
 					$importer["uid"], $item["uri"], Protocol::FEED, Protocol::DFRN];
-				$previous = Item::selectFirst(['id', 'created'], $condition);
+				$previous = Post::selectFirst(['id', 'created'], $condition);
 				if (DBA::isResult($previous)) {
 					// Use the creation date when the post had been stored. It can happen this date changes in the feed.
 					$creation_dates[] = $previous['created'];
@@ -436,28 +379,22 @@ class Feed
 			$enclosures = $xpath->query("enclosure|atom:link[@rel='enclosure']", $entry);
 			foreach ($enclosures AS $enclosure) {
 				$href = "";
-				$length = "";
-				$type = "";
+				$length = null;
+				$type = null;
 
 				foreach ($enclosure->attributes AS $attribute) {
 					if (in_array($attribute->name, ["url", "href"])) {
 						$href = $attribute->textContent;
 					} elseif ($attribute->name == "length") {
-						$length = $attribute->textContent;
+						$length = (int)$attribute->textContent;
 					} elseif ($attribute->name == "type") {
 						$type = $attribute->textContent;
 					}
 				}
 
-				if (!empty($item["attach"])) {
-					$item["attach"] .= ',';
-				} else {
-					$item["attach"] = '';
+				if (!empty($href)) {
+					$attachments[] = ['type' => Post\Media::DOCUMENT, 'url' => $href, 'mimetype' => $type, 'size' => $length];
 				}
-
-				$attachments[] = ["link" => $href, "type" => $type, "length" => $length];
-
-				$item["attach"] .= '[attach]href="' . $href . '" length="' . $length . '" type="' . $type . '"[/attach]';
 			}
 
 			$taglist = [];
@@ -506,7 +443,10 @@ class Feed
 				$items[] = $item;
 				break;
 			} elseif (!Item::isValid($item)) {
-				Logger::info('Feed is invalid', ['created' => $item['created'], 'uid' => $item['uid'], 'uri' => $item['uri']]);
+				Logger::info('Feed item is invalid', ['created' => $item['created'], 'uid' => $item['uid'], 'uri' => $item['uri']]);
+				continue;
+			} elseif (Item::isTooOld($item)) {
+				Logger::info('Feed is too old', ['created' => $item['created'], 'uid' => $item['uid'], 'uri' => $item['uri']]);
 				continue;
 			}
 
@@ -514,8 +454,8 @@ class Feed
 			if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] < 3)) {
 				// Handle enclosures and treat them as preview picture
 				foreach ($attachments AS $attachment) {
-					if ($attachment["type"] == "image/jpeg") {
-						$preview = $attachment["link"];
+					if ($attachment["mimetype"] == "image/jpeg") {
+						$preview = $attachment["url"];
 					}
 				}
 
@@ -543,7 +483,7 @@ class Feed
 					$item["body"] = trim($item["title"]);
 				}
 
-				$data = ParseUrl::getSiteinfoCached($item['plink'], true);
+				$data = ParseUrl::getSiteinfoCached($item['plink']);
 				if (!empty($data['text']) && !empty($data['title']) && (mb_strlen($item['body']) < mb_strlen($data['text']))) {
 					// When the fetched page info text is longer than the body, we do try to enhance the body
 					if (!empty($item['body']) && (strpos($data['title'], $item['body']) === false) && (strpos($data['text'], $item['body']) === false)) {
@@ -557,25 +497,27 @@ class Feed
 
 				$data = PageInfo::queryUrl($item["plink"], false, $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_denylist"] ?? '');
 
-				// Take the data that was provided by the feed if the query is empty
-				if (($data['type'] == 'link') && empty($data['title']) && empty($data['text'])) {
-					$data['title'] = $saved_title;
-					$item["body"] = $saved_body;
-				}
+				if (!empty($data)) {
+					// Take the data that was provided by the feed if the query is empty
+					if (($data['type'] == 'link') && empty($data['title']) && empty($data['text'])) {
+						$data['title'] = $saved_title;
+						$item["body"] = $saved_body;
+					}
 
-				$data_text = strip_tags(trim($data['text'] ?? ''));
-				$item_body = strip_tags(trim($item['body'] ?? ''));
+					$data_text = strip_tags(trim($data['text'] ?? ''));
+					$item_body = strip_tags(trim($item['body'] ?? ''));
 
-				if (!empty($data_text) && (($data_text == $item_body) || strstr($item_body, $data_text))) {
-					$data['text'] = '';
-				}
+					if (!empty($data_text) && (($data_text == $item_body) || strstr($item_body, $data_text))) {
+						$data['text'] = '';
+					}
 
-				// We always strip the title since it will be added in the page information
-				$item["title"] = "";
-				$item["body"] = $item["body"] . "\n" . PageInfo::getFooterFromData($data, false);
-				$taglist = $contact["fetch_further_information"] == 2 ? PageInfo::getTagsFromUrl($item["plink"], $preview, $contact["ffi_keyword_denylist"] ?? '') : [];
-				$item["object-type"] = Activity\ObjectType::BOOKMARK;
-				unset($item["attach"]);
+					// We always strip the title since it will be added in the page information
+					$item["title"] = "";
+					$item["body"] = $item["body"] . "\n" . PageInfo::getFooterFromData($data, false);
+					$taglist = $contact["fetch_further_information"] == 2 ? PageInfo::getTagsFromUrl($item["plink"], $preview, $contact["ffi_keyword_denylist"] ?? '') : [];
+					$item["object-type"] = Activity\ObjectType::BOOKMARK;
+					$attachments = [];
+				}
 			} else {
 				if (!empty($summary)) {
 					$item["body"] = '[abstract]' . HTML::toBBCode($summary, $basepath) . "[/abstract]\n" . $item["body"];
@@ -596,6 +538,10 @@ class Feed
 				}
 			}
 
+			if (empty($item['title'])) {
+				$item['post-type'] = Item::PT_NOTE;
+			}
+
 			Logger::info('Stored feed', ['item' => $item]);
 
 			$notify = Item::isRemoteSelf($contact, $item);
@@ -604,22 +550,57 @@ class Feed
 			// Additionally we have to avoid conflicts with identical URI between imported feeds and these items.
 			if ($notify) {
 				$item['guid'] = Item::guidFromUri($orig_plink, DI::baseUrl()->getHostname());
-				unset($item['uri']);
+				$item['uri'] = Item::newURI($item['uid'], $item['guid']);
+				unset($item['thr-parent']);
 				unset($item['parent-uri']);
 
 				// Set the delivery priority for "remote self" to "medium"
 				$notify = PRIORITY_MEDIUM;
 			}
 
-			$id = Item::insert($item, $notify);
+			$condition = ['uid' => $item['uid'], 'uri' => $item['uri']];
+			if (!Post::exists($condition) && !Post\Delayed::exists($item["uri"], $item['uid'])) {
+				if (!$notify) {
+					Post\Delayed::publish($item, $notify, $taglist, $attachments);
+				} else {
+					$postings[] = ['item' => $item, 'notify' => $notify,
+						'taglist' => $taglist, 'attachments' => $attachments];
+				}
+			} else {
+				Logger::info('Post already created or exists in the delayed posts queue', ['uid' => $item['uid'], 'uri' => $item["uri"]]);
+			}
+		}
+
+		if (!empty($postings)) {
+			$min_posting = DI::config()->get('system', 'minimum_posting_interval', 0);
+			$total = count($postings);
+			if ($total > 1) {
+				// Posts shouldn't be delayed more than a day
+				$interval = min(1440, self::getPollInterval($contact));
+				$delay = max(round(($interval * 60) / $total), 60 * $min_posting);
+				Logger::info('Got posting delay', ['delay' => $delay, 'interval' => $interval, 'items' => $total, 'cid' => $contact['id'], 'url' => $contact['url']]);
+			} else {
+				$delay = 0;
+			}
+
+			$post_delay = 0;
 
-			Logger::info("Feed for contact " . $contact["url"] . " stored under id " . $id);
+			foreach ($postings as $posting) {
+				if ($delay > 0) {
+					$publish_time = time() + $post_delay;
+					$post_delay += $delay;
+				} else {
+					$publish_time = time();
+				}
 
-			if (!empty($id) && !empty($taglist)) {
-				$feeditem = Item::selectFirst(['uri-id'], ['id' => $id]);
-				foreach ($taglist as $tag) {
-					Tag::store($feeditem['uri-id'], Tag::HASHTAG, $tag);
+				$last_publish = DI::pConfig()->get($posting['item']['uid'], 'system', 'last_publish', 0, true);
+				$next_publish = max($last_publish + (60 * $min_posting), time());
+				if ($publish_time < $next_publish) {
+					$publish_time = $next_publish;
 				}
+				$publish_at = date(DateTimeFormat::MYSQL, $publish_time);
+
+				Post\Delayed::add($posting['item']['uri'], $posting['item'], $posting['notify'], false, $publish_at, $posting['taglist'], $posting['attachments']);
 			}
 		}
 
@@ -674,7 +655,7 @@ class Feed
 					$oldest = $day;
 					$oldest_date = $date;
 				}
-			
+
 				if ($newest < $day) {
 					$newest = $day;
 					$newest_date = $date;
@@ -747,6 +728,55 @@ class Feed
 		}
 	}
 
+	/**
+	 * Get the poll interval for the given contact array
+	 *
+	 * @param array $contact
+	 * @return int Poll interval in minutes
+	 */
+	public static function getPollInterval(array $contact)
+	{
+		if (in_array($contact['network'], [Protocol::MAIL, Protocol::FEED])) {
+			$ratings = [0, 3, 7, 8, 9, 10];
+			if (DI::config()->get('system', 'adjust_poll_frequency') && ($contact['network'] == Protocol::FEED)) {
+				$rating = $contact['rating'];
+			} elseif (array_key_exists($contact['priority'], $ratings)) {
+				$rating = $ratings[$contact['priority']];
+			} else {
+				$rating = -1;
+			}
+		} else {
+			// Check once a week per default for all other networks
+			$rating = 9;
+		}
+
+		// Friendica and OStatus are checked once a day
+		if (in_array($contact['network'], [Protocol::DFRN, Protocol::OSTATUS])) {
+			$rating = 8;
+		}
+
+		// Check archived contacts or contacts with unsupported protocols once a month
+		if ($contact['archive'] || in_array($contact['network'], [Protocol::ZOT, Protocol::PHANTOM])) {
+			$rating = 10;
+		}
+
+		if ($rating < 0) {
+			return 0;
+		}
+		/*
+		 * Based on $contact['priority'], should we poll this site now? Or later?
+		 */
+
+		$min_poll_interval = max(1, DI::config()->get('system', 'min_poll_interval'));
+
+		$poll_intervals = [$min_poll_interval, 15, 30, 60, 120, 180, 360, 720 ,1440, 10080, 43200];
+
+		//$poll_intervals = [$min_poll_interval . ' minute', '15 minute', '30 minute',
+		//	'1 hour', '2 hour', '3 hour', '6 hour', '12 hour' ,'1 day', '1 week', '1 month'];
+
+		return $poll_intervals[$rating];
+	}
+
 	/**
 	 * Convert a tag array to a tag string
 	 *
@@ -761,7 +791,7 @@ class Feed
 			if ($tagstr != "") {
 				$tagstr .= ", ";
 			}
-	
+
 			$tagstr .= "#[url=" . DI::baseUrl() . "/search?tag=" . urlencode($tag) . "]" . $tag . "[/url]";
 		}
 
@@ -860,12 +890,12 @@ class Feed
 		$params = ['order' => ['received' => true], 'limit' => $max_items];
 
 		if ($filter === 'posts') {
-			$ret = Item::selectThread([], $condition, $params);
+			$ret = Post::selectThread(Item::DELIVER_FIELDLIST, $condition, $params);
 		} else {
-			$ret = Item::select([], $condition, $params);
+			$ret = Post::select(Item::DELIVER_FIELDLIST, $condition, $params);
 		}
 
-		$items = Item::inArray($ret);
+		$items = Post::toArray($ret);
 
 		$doc = new DOMDocument('1.0', 'utf-8');
 		$doc->formatOutput = true;
@@ -1014,7 +1044,7 @@ class Feed
 
 		$condition = ['uid' => $owner["uid"], 'guid' => $repeated_guid, 'private' => [Item::PUBLIC, Item::UNLISTED],
 			'network' => Protocol::FEDERATED];
-		$repeated_item = Item::selectFirst([], $condition);
+		$repeated_item = Post::selectFirst(Item::DELIVER_FIELDLIST, $condition);
 		if (!DBA::isResult($repeated_item)) {
 			return false;
 		}
@@ -1106,10 +1136,9 @@ class Feed
 		$mentioned = [];
 
 		if ($item['gravity'] != GRAVITY_PARENT) {
-			$parent = Item::selectFirst(['guid', 'author-link', 'owner-link'], ['id' => $item['parent']]);
-			$parent_item = (($item['thr-parent']) ? $item['thr-parent'] : $item['parent-uri']);
+			$parent = Post::selectFirst(['guid', 'author-link', 'owner-link'], ['id' => $item['parent']]);
 
-			$thrparent = Item::selectFirst(['guid', 'author-link', 'owner-link', 'plink'], ['uid' => $owner["uid"], 'uri' => $parent_item]);
+			$thrparent = Post::selectFirst(['guid', 'author-link', 'owner-link', 'plink'], ['uid' => $owner["uid"], 'uri' => $item['thr-parent']]);
 
 			if (DBA::isResult($thrparent)) {
 				$mentioned[$thrparent["author-link"]] = $thrparent["author-link"];
@@ -1122,7 +1151,7 @@ class Feed
 			}
 
 			$attributes = [
-					"ref" => $parent_item,
+					"ref" => $item['thr-parent'],
 					"href" => $parent_plink];
 			XML::addElement($doc, $entry, "thr:in-reply-to", "", $attributes);