Merge pull request #8520 from annando/term2tag

We now store tags in "tag"
This commit is contained in:
Hypolite Petovan 2020-04-22 22:56:07 -04:00 committed by GitHub
commit 43b8bdea07
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
26 changed files with 561 additions and 71 deletions

View file

@ -23,6 +23,7 @@ namespace Friendica\Model;
use Friendica\Database\DBA;
use Friendica\DI;
use Friendica\Model\Term;
/**
* This class handles FileTag related functions
@ -195,11 +196,11 @@ class FileTag
if ($type == 'file') {
$lbracket = '[';
$rbracket = ']';
$termtype = TERM_FILE;
$termtype = Term::FILE;
} else {
$lbracket = '<';
$rbracket = '>';
$termtype = TERM_CATEGORY;
$termtype = Term::CATEGORY;
}
$filetags_updated = $saved;

View file

@ -94,7 +94,8 @@ class Item
const CONTENT_FIELDLIST = ['language'];
// All fields in the item table
const ITEM_FIELDLIST = ['id', 'uid', 'parent', 'uri', 'parent-uri', 'thr-parent', 'guid',
const ITEM_FIELDLIST = ['id', 'uid', 'parent', 'uri', 'parent-uri', 'thr-parent',
'guid', 'uri-id', 'parent-uri-id', 'thr-parent-id',
'contact-id', 'type', 'wall', 'gravity', 'extid', 'icid', 'iaid', 'psid',
'created', 'edited', 'commented', 'received', 'changed', 'verb',
'postopts', 'plink', 'resource-id', 'event-id', 'tag', 'attach', 'inform',
@ -2609,7 +2610,10 @@ class Item
// This sorting is important when there are hashtags that are part of other hashtags
// Otherwise there could be problems with hashtags like #test and #test2
rsort($tags);
// Because of this we are sorting from the longest to the shortest tag.
usort($tags, function($a, $b) {
return strlen($b) <=> strlen($a);
});
$URLSearchString = "^\[\]";

View file

@ -42,13 +42,17 @@ class ItemURI
DBA::insert('item-uri', $fields, true);
}
$itemuri = DBA::selectFirst('item-uri', ['id'], ['uri' => $uri]);
$itemuri = DBA::selectFirst('item-uri', ['id', 'guid'], ['uri' => $uri]);
if (!DBA::isResult($itemuri)) {
// This shouldn't happen
return null;
}
if (empty($itemuri['guid']) && !empty($fields['guid'])) {
DBA::update('item-uri', ['guid' => $fields['guid']], ['id' => $itemuri['id']]);
}
return $itemuri['id'];
}

282
src/Model/Tag.php Normal file
View file

@ -0,0 +1,282 @@
<?php
/**
* @copyright Copyright (C) 2020, Friendica
*
* @license GNU AGPL version 3 or any later version
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as
* published by the Free Software Foundation, either version 3 of the
* License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <https://www.gnu.org/licenses/>.
*
*/
namespace Friendica\Model;
use Friendica\Content\Text\BBCode;
use Friendica\Core\Logger;
use Friendica\Core\System;
use Friendica\Database\DBA;
use Friendica\Util\Strings;
/**
* Class Tag
*
* This Model class handles tag table interactions.
* This tables stores relevant tags related to posts, like hashtags and mentions.
*/
class Tag
{
const UNKNOWN = 0;
const HASHTAG = 1;
const MENTION = 2;
const CATEGORY = 3;
const FILE = 5;
/**
* An implicit mention is a mention in a comment body that is redundant with the threading information.
*/
const IMPLICIT_MENTION = 8;
/**
* An exclusive mention transfers the ownership of the post to the target account, usually a forum.
*/
const EXCLUSIVE_MENTION = 9;
const TAG_CHARACTER = [
self::HASHTAG => '#',
self::MENTION => '@',
self::IMPLICIT_MENTION => '%',
self::EXCLUSIVE_MENTION => '!',
];
/**
* Store tag/mention elements
*
* @param integer $uriid
* @param integer $type
* @param string $name
* @param string $url
* @param boolean $probing
*/
public static function store(int $uriid, int $type, string $name, string $url = '', $probing = true)
{
$name = trim($name, "\x00..\x20\xFF#!@");
if (empty($name)) {
return;
}
$cid = 0;
$tagid = 0;
if (in_array($type, [Tag::MENTION, Tag::EXCLUSIVE_MENTION, Tag::IMPLICIT_MENTION])) {
if (empty($url)) {
// No mention without a contact url
return;
}
if (!$probing) {
$condition = ['nurl' => Strings::normaliseLink($url), 'uid' => 0, 'deleted' => false];
$contact = DBA::selectFirst('contact', ['id'], $condition, ['order' => ['id']]);
if (DBA::isResult($contact)) {
$cid = $contact['id'];
Logger::info('Got id for contact url', ['cid' => $cid, 'url' => $url]);
}
if (empty($cid)) {
$ssl_url = str_replace('http://', 'https://', $url);
$condition = ['`alias` IN (?, ?, ?) AND `uid` = ? AND NOT `deleted`', $url, Strings::normaliseLink($url), $ssl_url, 0];
$contact = DBA::selectFirst('contact', ['id'], $condition, ['order' => ['id']]);
if (DBA::isResult($contact)) {
$cid = $contact['id'];
Logger::info('Got id for contact alias', ['cid' => $cid, 'url' => $url]);
}
}
} else {
$cid = Contact::getIdForURL($url, 0, true);
Logger::info('Got id by probing', ['cid' => $cid, 'url' => $url]);
}
if (empty($cid)) {
// The contact wasn't found in the system (most likely some dead account)
// We ensure that we only store a single entry by overwriting the previous name
Logger::info('Contact not found, updating tag', ['url' => $url, 'name' => $name]);
DBA::update('tag', ['name' => substr($name, 0, 96)], ['url' => $url]);
}
}
if (empty($cid)) {
$fields = ['name' => substr($name, 0, 96), 'url' => ''];
if (($type != Tag::HASHTAG) && !empty($url) && ($url != $name)) {
$fields['url'] = strtolower($url);
}
$tag = DBA::selectFirst('tag', ['id'], $fields);
if (!DBA::isResult($tag)) {
DBA::insert('tag', $fields, true);
$tagid = DBA::lastInsertId();
} else {
$tagid = $tag['id'];
}
if (empty($tagid)) {
Logger::error('No tag id created', $fields);
return;
}
}
$fields = ['uri-id' => $uriid, 'type' => $type, 'tid' => $tagid, 'cid' => $cid];
if (in_array($type, [Tag::MENTION, Tag::EXCLUSIVE_MENTION, Tag::IMPLICIT_MENTION])) {
$condition = $fields;
$condition['type'] = [Tag::MENTION, Tag::EXCLUSIVE_MENTION, Tag::IMPLICIT_MENTION];
if (DBA::exists('post-tag', $condition)) {
Logger::info('Tag already exists', $fields);
return;
}
}
DBA::insert('post-tag', $fields, true);
Logger::info('Stored tag/mention', ['uri-id' => $uriid, 'tag-id' => $tagid, 'contact-id' => $cid, 'name' => $name, 'type' => $type, 'callstack' => System::callstack(8)]);
}
/**
* Store tag/mention elements
*
* @param integer $uriid
* @param string $hash
* @param string $name
* @param string $url
* @param boolean $probing
*/
public static function storeByHash(int $uriid, string $hash, string $name, string $url = '', $probing = true)
{
$type = self::getTypeForHash($hash);
if ($type == self::UNKNOWN) {
return;
}
self::store($uriid, $type, $name, $url, $probing);
}
/**
* Store tags and mentions from the body
*
* @param integer $uriid URI-Id
* @param string $body Body of the post
* @param string $tags Accepted tags
* @param boolean $probing Perform a probing for contacts, adding them if needed
*/
public static function storeFromBody(int $uriid, string $body, string $tags = null, $probing = true)
{
if (is_null($tags)) {
$tags = self::TAG_CHARACTER[self::HASHTAG] . self::TAG_CHARACTER[self::MENTION] . self::TAG_CHARACTER[self::EXCLUSIVE_MENTION];
}
Logger::info('Check for tags', ['uri-id' => $uriid, 'hash' => $tags, 'callstack' => System::callstack()]);
if (!preg_match_all("/([" . $tags . "])\[url\=([^\[\]]*)\]([^\[\]]*)\[\/url\]/ism", $body, $result, PREG_SET_ORDER)) {
return;
}
Logger::info('Found tags', ['uri-id' => $uriid, 'hash' => $tags, 'result' => $result]);
foreach ($result as $tag) {
self::storeByHash($uriid, $tag[1], $tag[3], $tag[2], $probing);
}
}
/**
* Store raw tags (not encapsulated in links) from the body
* This function is needed in the intermediate phase.
* Later we can call item::setHashtags in advance to have all tags converted.
*
* @param integer $uriid URI-Id
* @param string $body Body of the post
*/
public static function storeRawTagsFromBody(int $uriid, string $body)
{
Logger::info('Check for tags', ['uri-id' => $uriid, 'callstack' => System::callstack()]);
$result = BBCode::getTags($body);
if (empty($result)) {
return;
}
Logger::info('Found tags', ['uri-id' => $uriid, 'result' => $result]);
foreach ($result as $tag) {
if (substr($tag, 0, 1) != self::TAG_CHARACTER[self::HASHTAG]) {
continue;
}
self::storeByHash($uriid, substr($tag, 0, 1), substr($tag, 1));
}
}
/**
* Remove tag/mention
*
* @param integer $uriid
* @param integer $type
* @param string $name
* @param string $url
*/
public static function remove(int $uriid, int $type, string $name, string $url = '')
{
$tag = DBA::fetchFirst("SELECT `id` FROM `tag` INNER JOIN `post-tag` ON `post-tag`.`tid` = `tag`.`id`
WHERE `uri-id` = ? AND `type` = ? AND `name` = ? AND `url` = ?", $uriid, $type, $name, $url);
if (!DBA::isResult($tag)) {
return;
}
Logger::info('Removing tag/mention', ['uri-id' => $uriid, 'tid' => $tag['id'], 'name' => $name, 'url' => $url, 'callstack' => System::callstack(8)]);
DBA::delete('post-tag', ['uri-id' => $uriid, 'tid' => $tag['id']]);
}
/**
* Remove tag/mention
*
* @param integer $uriid
* @param string $hash
* @param string $name
* @param string $url
*/
public static function removeByHash(int $uriid, string $hash, string $name, string $url = '')
{
$type = self::getTypeForHash($hash);
if ($type == self::UNKNOWN) {
return;
}
self::remove($uriid, $type, $name, $url);
}
/**
* Get the type for the given hash
*
* @param string $hash
* @return integer type
*/
private static function getTypeForHash(string $hash)
{
if ($hash == self::TAG_CHARACTER[self::MENTION]) {
return self::MENTION;
} elseif ($hash == self::TAG_CHARACTER[self::EXCLUSIVE_MENTION]) {
return self::EXCLUSIVE_MENTION;
} elseif ($hash == self::TAG_CHARACTER[self::IMPLICIT_MENTION]) {
return self::IMPLICIT_MENTION;
} elseif ($hash == self::TAG_CHARACTER[self::HASHTAG]) {
return self::HASHTAG;
} else {
return self::UNKNOWN;
}
}
}

View file

@ -40,10 +40,7 @@ class Term
const HASHTAG = 1;
const MENTION = 2;
const CATEGORY = 3;
const PCATEGORY = 4;
const FILE = 5;
const SAVEDSEARCH = 6;
const CONVERSATION = 7;
/**
* An implicit mention is a mention in a comment body that is redundant with the threading information.
*/
@ -333,6 +330,10 @@ class Term
continue;
}
if (empty($term)) {
continue;
}
if ($item['uid'] == 0) {
$global = true;
DBA::update('term', ['global' => true], ['otype' => self::OBJECT_TYPE_POST, 'guid' => $item['guid']]);

View file

@ -26,6 +26,7 @@ use Friendica\Core\Hook;
use Friendica\Database\DBA;
use Friendica\DI;
use Friendica\Util\Strings;
use Friendica\Model\Term;
class UserItem
{
@ -206,7 +207,7 @@ class UserItem
}
// Or the contact is a mentioned forum
$tags = DBA::select('term', ['url'], ['otype' => TERM_OBJ_POST, 'oid' => $item['id'], 'type' => TERM_MENTION, 'uid' => $uid]);
$tags = DBA::select('term', ['url'], ['otype' => Term::OBJECT_TYPE_POST, 'oid' => $item['id'], 'type' => Term::MENTION, 'uid' => $uid]);
while ($tag = DBA::fetch($tags)) {
$condition = ['nurl' => Strings::normaliseLink($tag['url']), 'uid' => $uid, 'notify_new_posts' => true, 'contact-type' => Contact::TYPE_COMMUNITY];
if (DBA::exists('contact', $condition)) {