src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2023, the Friendica project
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\PageInfo;
  31 use Friendica\Content\Smilies;
  32 use Friendica\Core\Hook;
  33 use Friendica\Core\Logger;
  34 use Friendica\Core\Protocol;
  35 use Friendica\Core\Renderer;
  36 use Friendica\DI;
  37 use Friendica\Model\Contact;
  38 use Friendica\Model\Event;
  39 use Friendica\Model\Photo;
  40 use Friendica\Model\Post;
  41 use Friendica\Model\Tag;
  42 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
  43 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
  44 use Friendica\Object\Image;
  45 use Friendica\Protocol\Activity;
  46 use Friendica\Util\Images;
  47 use Friendica\Util\Map;
  48 use Friendica\Util\ParseUrl;
  49 use Friendica\Util\Proxy;
  50 use Friendica\Util\Strings;
  51 use Friendica\Util\XML;
  52
  53 class BBCode
  54 {
  55         // Update this value to the current date whenever changes are made to BBCode::convert
  56         const VERSION = '2021-07-28';
  57
  58         const INTERNAL     = 0;
  59         const EXTERNAL     = 1;
  60         const MASTODON_API = 2;
  61         const DIASPORA     = 3;
  62         const CONNECTORS   = 4;
  63         const TWITTER_API  = 5;
  64         const OSTATUS      = 7;
  65         const TWITTER      = 8;
  66         const BACKLINK     = 8;
  67         const ACTIVITYPUB  = 9;
  68
  69         const TOP_ANCHOR = '<br class="top-anchor">';
  70         const BOTTOM_ANCHOR = '<br class="button-anchor">';
  71
  72         const PREVIEW_NONE     = 0;
  73         const PREVIEW_NO_IMAGE = 1;
  74         const PREVIEW_LARGE    = 2;
  75         const PREVIEW_SMALL    = 3;
  76
  77         /**
  78          * Fetches attachment data that were generated the old way
  79          *
  80          * @param string $body Message body
  81          * @return array
  82          *                     'type' -> Message type ('link', 'video', 'photo')
  83          *                     'text' -> Text before the shared message
  84          *                     'after' -> Text after the shared message
  85          *                     'image' -> Preview image of the message
  86          *                     'url' -> Url to the attached message
  87          *                     'title' -> Title of the attachment
  88          *                     'description' -> Description of the attachment
  89          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  90          */
  91         private static function getOldAttachmentData(string $body): array
  92         {
  93                 $post = [];
  94
  95                 // Simplify image codes
  96                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  97
  98                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  99                         foreach ($attached as $data) {
 100                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
 101                                         continue;
 102                                 }
 103
 104                                 $post['type'] = substr($data[1], 5);
 105
 106                                 $pos = strpos($body, $data[0]);
 107                                 if ($pos > 0) {
 108                                         $post['text'] = trim(substr($body, 0, $pos));
 109                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
 110                                 } else {
 111                                         $post['text'] = trim(str_replace($data[0], '', $body));
 112                                         $post['after'] = '';
 113                                 }
 114
 115                                 $attacheddata = $data[2];
 116
 117                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 118
 119                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 120
 121                                         if ($picturedata) {
 122                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 123                                                         $post['image'] = $matches[1];
 124                                                 } else {
 125                                                         $post['preview'] = $matches[1];
 126                                                 }
 127                                         }
 128                                 }
 129
 130                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 131                                         $post['url'] = $matches[1];
 132                                         $post['title'] = $matches[2];
 133                                 }
 134                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 135                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 136                                         $post['url'] = $matches[1];
 137                                 }
 138
 139                                 // Search for description
 140                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 141                                         $post['description'] = $matches[1];
 142                                 }
 143                         }
 144                 }
 145                 return $post;
 146         }
 147
 148         /**
 149          * Fetches attachment data that were generated with the "attachment" element
 150          *
 151          * @param string $body Message body
 152          * @return array
 153          *                     'type' -> Message type ('link', 'video', 'photo')
 154          *                     'text' -> Text before the shared message
 155          *                     'after' -> Text after the shared message
 156          *                     'image' -> Preview image of the message
 157          *                     'url' -> Url to the attached message
 158          *                     'title' -> Title of the attachment
 159          *                     'description' -> Description of the attachment
 160          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 161          */
 162         public static function getAttachmentData(string $body): array
 163         {
 164                 DI::profiler()->startRecording('rendering');
 165                 $data = [
 166                         'type'          => '',
 167                         'text'          => '',
 168                         'after'         => '',
 169                         'image'         => null,
 170                         'url'           => '',
 171                         'author_name'   => '',
 172                         'author_url'    => '',
 173                         'provider_name' => '',
 174                         'provider_url'  => '',
 175                         'title'         => '',
 176                         'description'   => '',
 177                 ];
 178
 179                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 180                         DI::profiler()->stopRecording();
 181                         return self::getOldAttachmentData($body);
 182                 }
 183
 184                 $attributes = $match[2];
 185
 186                 $data['text'] = trim($match[1]);
 187
 188                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 189                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 190                         $value = $matches[2] ?? '';
 191
 192                         if ($value != '') {
 193                                 switch ($field) {
 194                                         case 'publisher_name':
 195                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 196                                                 break;
 197
 198                                         case 'publisher_url':
 199                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 200                                                 break;
 201
 202                                         case 'author_name':
 203                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 204                                                 if ($data['provider_name'] == $data['author_name']) {
 205                                                         $data['author_name'] = '';
 206                                                 }
 207                                                 break;
 208
 209                                         case 'author_url':
 210                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 211                                                 if ($data['provider_url'] == $data['author_url']) {
 212                                                         $data['author_url'] = '';
 213                                                 }
 214                                                 break;
 215
 216                                         case 'title':
 217                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
 218                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 219                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 220                                                 $data['title'] = $value;
 221
 222                                         default:
 223                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 224                                                 break;
 225                                 }
 226                         }
 227                 }
 228
 229                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 230                         DI::profiler()->stopRecording();
 231                         return [];
 232                 }
 233
 234                 $data['description'] = trim($match[3]);
 235
 236                 $data['after'] = trim($match[4]);
 237
 238                 $parts = parse_url($data['url']);
 239                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 240                         if (empty($data['provider_name'])) {
 241                                 $data['provider_name'] = $parts['host'];
 242                         }
 243                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 244                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 245
 246                                 if (!empty($parts['port'])) {
 247                                         $data['provider_url'] .= ':' . $parts['port'];
 248                                 }
 249                         }
 250                 }
 251
 252                 DI::profiler()->stopRecording();
 253                 return $data;
 254         }
 255
 256         public static function getAttachedData(string $body, array $item = []): array
 257         {
 258                 /*
 259                 - text:
 260                 - type: link, video, photo
 261                 - title:
 262                 - url:
 263                 - image:
 264                 - description:
 265                 - (thumbnail)
 266                 */
 267
 268                 DI::profiler()->startRecording('rendering');
 269                 $has_title = !empty($item['title']);
 270                 $plink = $item['plink'] ?? '';
 271                 $post = self::getAttachmentData($body);
 272
 273                 // Get all linked images with alternative image description
 274                 if (preg_match_all("/\[img=(http[^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 275                         foreach ($pictures as $picture) {
 276                                 if ($id = Photo::getIdForName($picture[1])) {
 277                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2], 'id' => $id];
 278                                 } else {
 279                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
 280                                 }
 281                         }
 282                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 283                                 $post['image_description'] = $post['images'][0]['description'];
 284                         }
 285                 }
 286
 287                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 288                         foreach ($pictures as $picture) {
 289                                 if ($id = Photo::getIdForName($picture[1])) {
 290                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => '', 'id' => $id];
 291                                 } else {
 292                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
 293                                 }
 294                         }
 295                 }
 296
 297                 if (!isset($post['type'])) {
 298                         $post['text'] = $body;
 299                 }
 300
 301                 // Simplify image codes
 302                 $post['text'] = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $post['text']);
 303                 $post['text'] = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $post['text']);
 304
 305                 // if nothing is found, it maybe having an image.
 306                 if (!isset($post['type'])) {
 307                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $post['text'], $pictures, PREG_SET_ORDER)) {
 308                                 if ((count($pictures) == 1) && !$has_title) {
 309                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 310                                                 // Replace the preview picture with the real picture
 311                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 312                                                 $data = ['url' => $url, 'type' => 'photo'];
 313                                         } else {
 314                                                 // Checking, if the link goes to a picture
 315                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 316                                         }
 317
 318                                         // Workaround:
 319                                         // Sometimes photo posts to the own album are not detected at the start.
 320                                         // So we seem to cannot use the cache for these cases. That's strange.
 321                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], '/photos/')) {
 322                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 323                                         }
 324
 325                                         if ($data['type'] == 'photo') {
 326                                                 $post['type'] = 'photo';
 327                                                 if (isset($data['images'][0])) {
 328                                                         $post['image'] = $data['images'][0]['src'];
 329                                                         $post['url'] = $data['url'];
 330                                                 } else {
 331                                                         $post['image'] = $data['url'];
 332                                                 }
 333
 334                                                 $post['preview'] = $pictures[0][2];
 335                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $post['text']));
 336                                         } else {
 337                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 338                                                 if (($imgdata) && substr($imgdata['mime'], 0, 6) == 'image/') {
 339                                                         $post['type'] = 'photo';
 340                                                         $post['image'] = $pictures[0][1];
 341                                                         $post['preview'] = $pictures[0][2];
 342                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $post['text']));
 343                                                 }
 344                                         }
 345                                 } elseif (count($pictures) > 0) {
 346                                         if (count($pictures) > 4) {
 347                                                 $post['type'] = 'link';
 348                                                 $post['url'] = $plink;
 349                                         } else {
 350                                                 $post['type'] = 'photo';
 351                                         }
 352
 353                                         $post['image'] = $pictures[0][2];
 354
 355                                         foreach ($pictures as $picture) {
 356                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 357                                         }
 358                                 }
 359                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $post['text'], $pictures, PREG_SET_ORDER)) {
 360                                 if ($has_title) {
 361                                         $post['type'] = 'link';
 362                                         $post['url'] = $plink;
 363                                 } else {
 364                                         $post['type'] = 'photo';
 365                                 }
 366
 367                                 $post['image'] = $pictures[0][1];
 368                                 foreach ($pictures as $picture) {
 369                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 370                                 }
 371                         }
 372
 373                         // Test for the external links
 374                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 375                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 376
 377                         $links = array_merge($links1, $links2);
 378
 379                         // If there is only a single one, then use it.
 380                         // This should cover link posts via API.
 381                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 382                                 $post['type'] = 'link';
 383                                 $post['url'] = $links[0][1];
 384                         }
 385
 386                         // Simplify "video" element
 387                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 388
 389                         // Now count the number of external media links
 390                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 391                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 392                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 393                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 394
 395                         // Add them to the other external links
 396                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 397
 398                         // Are there more than one?
 399                         if (count($links) > 1) {
 400                                 // The post will be the type "text", which means a blog post
 401                                 unset($post['type']);
 402                                 $post['url'] = $plink;
 403                         }
 404
 405                         if (!isset($post['type'])) {
 406                                 $post['type'] = 'text';
 407                         }
 408
 409                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
 410                                 $post['images'] = $post['remote_images'];
 411                                 $post['image'] = $post['images'][0]['url'];
 412                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 413                                         $post['image_description'] = $post['images'][0]['description'];
 414                                 }
 415                         }
 416                         unset($post['remote_images']);
 417                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 418                         $data = ParseUrl::getSiteinfoCached($post['url']);
 419
 420                         if (isset($data['images'][0])) {
 421                                 $post['image'] = $data['images'][0]['src'];
 422                         }
 423                 } elseif (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $post['text'], $pictures, PREG_SET_ORDER)) {
 424                         foreach ($pictures as $picture) {
 425                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 426                         }
 427                 }
 428
 429                 DI::profiler()->stopRecording();
 430                 return $post;
 431         }
 432
 433         /**
 434          * Remove [attachment] BBCode and replaces it with a regular [url]
 435          *
 436          * @param string  $body
 437          * @param boolean $no_link_desc No link description
 438          * @return string with replaced body
 439          */
 440         public static function removeAttachment(string $body, bool $no_link_desc = false): string
 441         {
 442                 return preg_replace_callback("/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
 443                         function ($match) use ($body, $no_link_desc) {
 444                                 $attach_data = self::getAttachmentData($match[0]);
 445                                 if (empty($attach_data['url'])) {
 446                                         return $match[0];
 447                                 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
 448                                         return '';
 449                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 450                                         return " \n[url]" . $attach_data['url'] . "[/url]\n";
 451                                 } else {
 452                                         return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 453                                 }
 454                 }, $body);
 455         }
 456
 457         /**
 458          * Converts a BBCode text into plaintext
 459          *
 460          * @param string $text
 461          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 462          * @return string
 463          */
 464         public static function toPlaintext(string $text, bool $keep_urls = true): string
 465         {
 466                 DI::profiler()->startRecording('rendering');
 467                 // Remove pictures in advance to avoid unneeded proxy calls
 468                 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
 469                 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
 470
 471                 // Remove attachment
 472                 $text = self::removeAttachment($text);
 473
 474                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 475
 476                 DI::profiler()->stopRecording();
 477                 return $naked_text;
 478         }
 479
 480         private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
 481         {
 482                 // Only send proxied pictures to API and for internal display
 483                 if (!in_array($simplehtml, [self::INTERNAL, self::MASTODON_API, self::TWITTER_API])) {
 484                         return $image;
 485                 } elseif ($uriid > 0) {
 486                         return Post\Link::getByLink($uriid, $image, $size);
 487                 } else {
 488                         return Proxy::proxifyUrl($image, $size);
 489                 }
 490         }
 491
 492         /**
 493          * This function changing the visual size (not the real size) of images.
 494          * The function does not work for pictures with an alternate text description.
 495          * This could only be changed by using some new "img" BBCode format.
 496          *
 497          * @param string $srctext The body with images
 498          * @return string The body with possibly scaled images
 499          */
 500         public static function scaleExternalImages(string $srctext): string
 501         {
 502                 DI::profiler()->startRecording('rendering');
 503                 $s = $srctext;
 504
 505                 // Simplify image links
 506                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 507
 508                 $matches = null;
 509                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 510                 if ($c) {
 511                         foreach ($matches as $mtch) {
 512                                 Logger::debug('scale_external_image', ['image' => $mtch[1]]);
 513
 514                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 515                                 if (stristr($mtch[1], $hostname)) {
 516                                         continue;
 517                                 }
 518
 519                                 $curlResult = DI::httpClient()->get($mtch[1], HttpClientAccept::IMAGE);
 520                                 if (!$curlResult->isSuccess()) {
 521                                         continue;
 522                                 }
 523
 524                                 Logger::debug('Got picture', ['Content-Type' => $curlResult->getHeader('Content-Type'), 'url' => $mtch[1]]);
 525
 526                                 $i = $curlResult->getBody();
 527                                 $type = $curlResult->getContentType();
 528                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 529
 530                                 if ($i) {
 531                                         $Image = new Image($i, $type);
 532                                         if ($Image->isValid()) {
 533                                                 $orig_width = $Image->getWidth();
 534                                                 $orig_height = $Image->getHeight();
 535
 536                                                 if ($orig_width > 640 || $orig_height > 640) {
 537                                                         $Image->scaleDown(640);
 538                                                         $new_width = $Image->getWidth();
 539                                                         $new_height = $Image->getHeight();
 540                                                         Logger::debug('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 541                                                         $s = str_replace(
 542                                                                 $mtch[0],
 543                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 544                                                                 . "\n",
 545                                                                 $s
 546                                                         );
 547                                                         Logger::debug('New string', ['image' => $s]);
 548                                                 }
 549                                         }
 550                                 }
 551                         }
 552                 }
 553
 554                 DI::profiler()->stopRecording();
 555                 return $s;
 556         }
 557
 558         /**
 559          * Truncates imported message body string length to max_import_size
 560          *
 561          * The purpose of this function is to apply system message length limits to
 562          * imported messages without including any embedded photos in the length
 563          *
 564          * @param string $body
 565          * @return string
 566          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 567          */
 568         public static function limitBodySize(string $body): string
 569         {
 570                 DI::profiler()->startRecording('rendering');
 571                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 572
 573                 // If the length of the body, including the embedded images, is smaller
 574                 // than the maximum, then don't waste time looking for the images
 575                 if ($maxlen && (strlen($body) > $maxlen)) {
 576
 577                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 578
 579                         $orig_body = $body;
 580                         $new_body = '';
 581                         $textlen = 0;
 582
 583                         $img_start = strpos($orig_body, '[img');
 584                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 585                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 586                         while (($img_st_close !== false) && ($img_end !== false)) {
 587
 588                                 $img_st_close++; // make it point to AFTER the closing bracket
 589                                 $img_end += $img_start;
 590                                 $img_end += strlen('[/img]');
 591
 592                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 593                                         // This is an embedded image
 594
 595                                         if (($textlen + $img_start) > $maxlen) {
 596                                                 if ($textlen < $maxlen) {
 597                                                         Logger::debug('the limit happens before an embedded image');
 598                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 599                                                         $textlen = $maxlen;
 600                                                 }
 601                                         } else {
 602                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 603                                                 $textlen += $img_start;
 604                                         }
 605
 606                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 607                                 } else {
 608
 609                                         if (($textlen + $img_end) > $maxlen) {
 610                                                 if ($textlen < $maxlen) {
 611                                                         Logger::debug('the limit happens before the end of a non-embedded image');
 612                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 613                                                         $textlen = $maxlen;
 614                                                 }
 615                                         } else {
 616                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 617                                                 $textlen += $img_end;
 618                                         }
 619                                 }
 620                                 $orig_body = substr($orig_body, $img_end);
 621
 622                                 if ($orig_body === false) {
 623                                         // in case the body ends on a closing image tag
 624                                         $orig_body = '';
 625                                 }
 626
 627                                 $img_start = strpos($orig_body, '[img');
 628                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 629                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 630                         }
 631
 632                         if (($textlen + strlen($orig_body)) > $maxlen) {
 633                                 if ($textlen < $maxlen) {
 634                                         Logger::debug('the limit happens after the end of the last image');
 635                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 636                                 }
 637                         } else {
 638                                 Logger::debug('the text size with embedded images extracted did not violate the limit');
 639                                 $new_body = $new_body . $orig_body;
 640                         }
 641
 642                         DI::profiler()->stopRecording();
 643                         return $new_body;
 644                 } else {
 645                         DI::profiler()->stopRecording();
 646                         return $body;
 647                 }
 648         }
 649
 650         /**
 651          * Processes [attachment] tags
 652          *
 653          * Note: Can produce a [bookmark] tag in the returned string
 654          *
 655          * @param string  $text
 656          * @param integer $simplehtml
 657          * @param bool    $tryoembed
 658          * @param array   $data
 659          * @param int     $uriid
 660          * @return string
 661          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 662          */
 663         public static function convertAttachment(string $text, int $simplehtml = self::INTERNAL, bool $tryoembed = true, array $data = [], int $uriid = 0, int $preview_mode = self::PREVIEW_LARGE): string
 664         {
 665                 DI::profiler()->startRecording('rendering');
 666                 $data = $data ?: self::getAttachmentData($text);
 667                 if (empty($data) || empty($data['url'])) {
 668                         DI::profiler()->stopRecording();
 669                         return $text;
 670                 }
 671
 672                 if (isset($data['title'])) {
 673                         $data['title'] = strip_tags($data['title']);
 674                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 675                 } else {
 676                         $data['title'] = '';
 677                 }
 678
 679                 if (((strpos($data['text'], '[img=') !== false) || (strpos($data['text'], '[img]') !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 680                         $data['preview'] = $data['image'];
 681                         $data['image'] = '';
 682                 }
 683
 684                 $return = '';
 685                 try {
 686                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 687                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 688                         } else {
 689                                 throw new Exception('OEmbed is disabled for this attachment.');
 690                         }
 691                 } catch (Exception $e) {
 692                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 693
 694                         if ($simplehtml != self::CONNECTORS) {
 695                                 $return = sprintf('<div class="type-%s">', $data['type']);
 696                         }
 697
 698                         if ($preview_mode == self::PREVIEW_NO_IMAGE) {
 699                                 unset($data['image']);
 700                                 unset($data['preview']);
 701                         }
 702
 703                         if (!empty($data['title']) && !empty($data['url'])) {
 704                                 $preview_class = $preview_mode == self::PREVIEW_LARGE ? 'attachment-image' : 'attachment-preview';
 705                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 706                                         $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="' . $preview_class . '" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
 707                                 } else {
 708                                         if (!empty($data['image'])) {
 709                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="' . $preview_class . '" /></a><br>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
 710                                         } elseif (!empty($data['preview'])) {
 711                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br>', $data['url'], self::proxyUrl($data['preview'], $simplehtml, $uriid), $data['title']);
 712                                         }
 713                                         $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
 714                                 }
 715                         }
 716
 717                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 718                                 // Sanitize the HTML
 719                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 720                         }
 721
 722                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 723                                 if (!empty($data['author_name'])) {
 724                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 725                                 } else {
 726                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 727                                 }
 728                         }
 729
 730                         if ($simplehtml != self::CONNECTORS) {
 731                                 $return .= '</div>';
 732                         }
 733                 }
 734
 735                 DI::profiler()->stopRecording();
 736                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 737         }
 738
 739         public static function removeShareInformation(string $text, bool $plaintext = false, bool $nolink = false): string
 740         {
 741                 DI::profiler()->startRecording('rendering');
 742                 $data = self::getAttachmentData($text);
 743
 744                 if (!$data) {
 745                         DI::profiler()->stopRecording();
 746                         return $text;
 747                 } elseif ($nolink) {
 748                         DI::profiler()->stopRecording();
 749                         return $data['text'] . ($data['after'] ?? '');
 750                 }
 751
 752                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 753                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 754                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 755                         $data['title'] = $data['url'];
 756                 } elseif (($text != '') && strstr($title, $text)) {
 757                         $data['text'] = $data['title'];
 758                         $data['title'] = $data['url'];
 759                 }
 760
 761                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 762                         DI::profiler()->stopRecording();
 763                         return $data['title'] . $data['after'];
 764                 }
 765
 766                 // If the link already is included in the post, don't add it again
 767                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 768                         DI::profiler()->stopRecording();
 769                         return $data['text'] . $data['after'];
 770                 }
 771
 772                 $text = $data['text'];
 773
 774                 if (!empty($data['url']) && !empty($data['title'])) {
 775                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 776                 } elseif (!empty($data['url'])) {
 777                         $text .= "\n[url]" . $data['url'] . '[/url]';
 778                 }
 779
 780                 DI::profiler()->stopRecording();
 781                 return $text . "\n" . $data['after'];
 782         }
 783
 784         /**
 785          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 786          *
 787          * @param array $match Array with the matching values
 788          * @return string reformatted link including HTML codes
 789          */
 790         private static function convertUrlForActivityPubCallback(array $match): string
 791         {
 792                 $url = $match[1];
 793
 794                 if (isset($match[2]) && ($match[1] != $match[2])) {
 795                         return $match[0];
 796                 }
 797
 798                 $parts = parse_url($url);
 799                 if (!isset($parts['scheme'])) {
 800                         return $match[0];
 801                 }
 802
 803                 return self::convertUrlForActivityPub($url);
 804         }
 805
 806         /**
 807          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 808          *
 809          * @param string $url URL that is about to be reformatted
 810          * @return string reformatted link including HTML codes
 811          */
 812         private static function convertUrlForActivityPub(string $url): string
 813         {
 814                 return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, self::getStyledURL($url));
 815         }
 816
 817         /**
 818          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 819          *
 820          * @param string $url URL that is about to be reformatted
 821          * @return string reformatted link
 822          */
 823         private static function getStyledURL(string $url): string
 824         {
 825                 $parts = parse_url($url);
 826                 $scheme = $parts['scheme'] . '://';
 827                 $styled_url = str_replace($scheme, '', $url);
 828
 829                 if (strlen($styled_url) > 30) {
 830                         $styled_url = substr($styled_url, 0, 30) . "…";
 831                 }
 832
 833                 return $styled_url;
 834         }
 835
 836         /*
 837          * [noparse][i]italic[/i][/noparse] turns into
 838          * [noparse][ i ]italic[ /i ][/noparse],
 839          * to hide them from parser.
 840          *
 841          * @param array $match
 842          * @return string
 843          */
 844         private static function escapeNoparseCallback(array $match): string
 845         {
 846                 $whole_match = $match[0];
 847                 $captured = $match[1];
 848                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 849                 $new_str = str_replace($captured, $spacefied, $whole_match);
 850                 return $new_str;
 851         }
 852
 853         /*
 854          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 855          * now turns back and the [noparse] tags are trimed
 856          * returning [i]italic[/i]
 857          *
 858          * @param array $match
 859          * @return string
 860          */
 861         private static function unescapeNoparseCallback(array $match): string
 862         {
 863                 $captured = $match[1];
 864                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 865                 return $unspacefied;
 866         }
 867
 868         /**
 869          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 870          * occurrences
 871          *
 872          * @param string $text        Text to search
 873          * @param string $name        Tag name
 874          * @param int    $occurrences Number of first occurrences to skip
 875          * @return boolean|array
 876          */
 877         public static function getTagPosition(string $text, string $name, int $occurrences = 0)
 878         {
 879                 DI::profiler()->startRecording('rendering');
 880                 if ($occurrences < 0) {
 881                         $occurrences = 0;
 882                 }
 883
 884                 $start_open = -1;
 885                 for ($i = 0; $i <= $occurrences; $i++) {
 886                         if ($start_open !== false) {
 887                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 888                         }
 889                 }
 890
 891                 if ($start_open === false) {
 892                         DI::profiler()->stopRecording();
 893                         return false;
 894                 }
 895
 896                 $start_equal = strpos($text, '=', $start_open);
 897                 $start_close = strpos($text, ']', $start_open);
 898
 899                 if ($start_close === false) {
 900                         DI::profiler()->stopRecording();
 901                         return false;
 902                 }
 903
 904                 $start_close++;
 905
 906                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 907
 908                 if ($end_open === false) {
 909                         DI::profiler()->stopRecording();
 910                         return false;
 911                 }
 912
 913                 $res = [
 914                         'start' => [
 915                                 'open' => $start_open,
 916                                 'close' => $start_close
 917                         ],
 918                         'end' => [
 919                                 'open' => $end_open,
 920                                 'close' => $end_open + strlen('[/' . $name . ']')
 921                         ],
 922                 ];
 923
 924                 if ($start_equal !== false) {
 925                         $res['start']['equal'] = $start_equal + 1;
 926                 }
 927
 928                 DI::profiler()->stopRecording();
 929                 return $res;
 930         }
 931
 932         /**
 933          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 934          *
 935          * @param string $pattern Preg pattern string
 936          * @param string $replace Preg replace string
 937          * @param string $name    BBCode tag name
 938          * @param string $text    Text to search
 939          * @return string
 940          */
 941         public static function pregReplaceInTag(string $pattern, string $replace, string $name, string $text): string
 942         {
 943                 DI::profiler()->startRecording('rendering');
 944                 $occurrences = 0;
 945                 $pos = self::getTagPosition($text, $name, $occurrences);
 946                 while ($pos !== false && $occurrences++ < 1000) {
 947                         $start = substr($text, 0, $pos['start']['open']);
 948                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 949                         $end = substr($text, $pos['end']['close']);
 950                         if ($end === false) {
 951                                 $end = '';
 952                         }
 953
 954                         $subject = preg_replace($pattern, $replace, $subject);
 955                         $text = $start . $subject . $end;
 956
 957                         $pos = self::getTagPosition($text, $name, $occurrences);
 958                 }
 959
 960                 DI::profiler()->stopRecording();
 961                 return $text;
 962         }
 963
 964         private static function extractImagesFromItemBody(string $body): array
 965         {
 966                 $saved_image = [];
 967                 $orig_body = $body;
 968                 $new_body = '';
 969
 970                 $cnt = 0;
 971                 $img_start = strpos($orig_body, '[img');
 972                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 973                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 974                 while (($img_st_close !== false) && ($img_end !== false)) {
 975                         $img_st_close++; // make it point to AFTER the closing bracket
 976                         $img_end += $img_start;
 977
 978                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 979                                 // This is an embedded image
 980                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 981                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 982
 983                                 $cnt++;
 984                         } else {
 985                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 986                         }
 987
 988                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 989
 990                         if ($orig_body === false) {
 991                                 // in case the body ends on a closing image tag
 992                                 $orig_body = '';
 993                         }
 994
 995                         $img_start = strpos($orig_body, '[img');
 996                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 997                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 998                 }
 999
1000                 $new_body = $new_body . $orig_body;
1001
1002                 return ['body' => $new_body, 'images' => $saved_image];
1003         }
1004
1005         private static function interpolateSavedImagesIntoItemBody(int $uriid, string $body, array $images): string
1006         {
1007                 $newbody = $body;
1008
1009                 $cnt = 0;
1010                 foreach ($images as $image) {
1011                         // We're depending on the property of 'foreach' (specified on the PHP website) that
1012                         // it loops over the array starting from the first element and going sequentially
1013                         // to the last element
1014                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1015                                 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
1016                         $cnt++;
1017                 }
1018
1019                 return $newbody;
1020         }
1021
1022         /**
1023          * @param string $text A BBCode string
1024          * @return array Empty array if no share tag is present or the following array, missing attributes end up empty strings:
1025          *               - comment   : Text before the opening share tag
1026          *               - shared    : Text inside the share tags
1027          *               - author    : (Optional) Display name of the shared author
1028          *               - profile   : (Optional) Profile page URL of the shared author
1029          *               - avatar    : (Optional) Profile picture URL of the shared author
1030          *               - link      : (Optional) Canonical URL of the shared post
1031          *               - posted    : (Optional) Date the shared post was initially posted ("Y-m-d H:i:s" in GMT)
1032          *               - message_id: (Optional) Shared post URI if any
1033          *               - guid      : (Optional) Shared post GUID if any
1034          */
1035         public static function fetchShareAttributes(string $text): array
1036         {
1037                 DI::profiler()->startRecording('rendering');
1038                 if (preg_match('~(.*?)\[share](.*)\[/share]~ism', $text, $matches)) {
1039                         DI::profiler()->stopRecording();
1040                         return [
1041                                 'author'     => '',
1042                                 'profile'    => '',
1043                                 'avatar'     => '',
1044                                 'link'       => '',
1045                                 'posted'     => '',
1046                                 'guid'       => '',
1047                                 'message_id' => trim($matches[2]),
1048                                 'comment'    => trim($matches[1]),
1049                                 'shared'     => '',
1050                         ];
1051                 }
1052                 // See Issue https://github.com/friendica/friendica/issues/10454
1053                 // Hashtags in usernames are expanded to links. This here is a quick fix.
1054                 $text = preg_replace('~([@!#])\[url=.*?](.*?)\[/url]~ism', '$1$2', $text);
1055
1056                 if (!preg_match('~(.*?)\[share(.*?)](.*)\[/share]~ism', $text, $matches)) {
1057                         DI::profiler()->stopRecording();
1058                         return [];
1059                 }
1060
1061                 $attributes = self::extractShareAttributes($matches[2]);
1062
1063                 $attributes['comment'] = trim($matches[1]);
1064                 $attributes['shared'] = trim($matches[3]);
1065
1066                 DI::profiler()->stopRecording();
1067                 return $attributes;
1068         }
1069
1070         /**
1071          * @see BBCode::fetchShareAttributes()
1072          * @param string $shareString Internal opening share tag string matched by the regular expression
1073          * @return array A fixed attribute array where missing attribute are represented by empty strings
1074          */
1075         private static function extractShareAttributes(string $shareString): array
1076         {
1077                 $attributes = [];
1078                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid', 'message_id'] as $field) {
1079                         preg_match("/$field=(['\"])(.+?)\\1/ism", $shareString, $matches);
1080                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1081                 }
1082
1083                 return $attributes;
1084         }
1085
1086         /**
1087          * Remove the share block
1088          *
1089          * @param string $body
1090          * @return string
1091          */
1092         public static function removeSharedData(string $body): string
1093         {
1094                 return trim(preg_replace("/\s*\[share.*?\].*?\[\/share\]\s*/ism", '', $body));
1095         }
1096
1097         /**
1098          * This function converts a [share] block to text according to a provided callback function whose signature is:
1099          *
1100          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
1101          *
1102          * Where:
1103          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
1104          * data lookup
1105          * - $author_contact is a contact record array
1106          * - $content is the inner content of the [share] block
1107          * - $is_quote_share indicates whether there's any content before the [share] block
1108          * - Return value is the string that should replace the [share] block in the provided text
1109          *
1110          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
1111          *
1112          * @param  string   $text     A BBCode string
1113          * @param  callable $callback
1114          * @return string The BBCode string with all [share] blocks replaced
1115          */
1116         public static function convertShare(string $text, callable $callback, int $uriid = 0): string
1117         {
1118                 DI::profiler()->startRecording('rendering');
1119                 $return = preg_replace_callback(
1120                         '~(.*?)\[share(.*?)](.*)\[/share]~ism',
1121                         function ($match) use ($callback, $uriid) {
1122                                 $attributes = self::extractShareAttributes($match[2]);
1123
1124                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
1125                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1126                                 $author_contact['addr'] = ($author_contact['addr'] ?? '');
1127
1128                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1129                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1130                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1131
1132                                 if (!empty($author_contact['id'])) {
1133                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
1134                                 } elseif ($attributes['avatar']) {
1135                                         $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
1136                                 }
1137
1138                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1139
1140                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1141                         },
1142                         $text
1143                 );
1144
1145                 DI::profiler()->stopRecording();
1146                 return trim($return);
1147         }
1148
1149         /**
1150          * Convert complex IMG and ZMG elements
1151          *
1152          * @param [type] $text
1153          * @param integer $simplehtml
1154          * @param integer $uriid
1155          * @return string
1156          */
1157         private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
1158         {
1159                 DI::profiler()->startRecording('rendering');
1160                 $return = preg_replace_callback(
1161                         "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
1162                         function ($match) use ($simplehtml, $uriid) {
1163                                 $attribute_string = $match[1];
1164                                 $attributes = [];
1165                                 foreach (['alt', 'width', 'height'] as $field) {
1166                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1167                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1168                                 }
1169
1170                                 $img_str = '<img src="' .
1171                                 self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
1172                                 foreach ($attributes as $key => $value) {
1173                                         if (!empty($value)) {
1174                                                 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
1175                                         }
1176                                 }
1177                                 return $img_str . '>';
1178                         },
1179                         $text
1180                 );
1181
1182                 DI::profiler()->stopRecording();
1183                 return $return;
1184         }
1185
1186         /**
1187          * Default [share] tag conversion callback
1188          *
1189          * Note: Can produce a [bookmark] tag in the output
1190          *
1191          * @see BBCode::convertShare()
1192          * @param array   $attributes     [share] block attribute values
1193          * @param array   $author_contact Contact row of the shared author
1194          * @param string  $content        Inner content of the [share] block
1195          * @param boolean $is_quote_share Whether there is content before the [share] block
1196          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1197          * @return string
1198          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1199          */
1200         private static function convertShareCallback(array $attributes, array $author_contact, string $content, bool $is_quote_share, int $simplehtml): string
1201         {
1202                 DI::profiler()->startRecording('rendering');
1203                 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
1204
1205                 switch ($simplehtml) {
1206                         case self::MASTODON_API:
1207                         case self::TWITTER_API:
1208                                 $text = ($is_quote_share? '<br>' : '') .
1209                                 '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
1210                                 '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
1211                                 break;
1212                         case self::DIASPORA:
1213                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1214                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1215                                 } else {
1216                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1217
1218                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1219                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1220                                         }
1221
1222                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1223
1224                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1225                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1226                                         }
1227                                 }
1228
1229                                 break;
1230                         case self::CONNECTORS:
1231                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1232                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1233                                 $headline .= ':</b></p>' . "\n";
1234
1235                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
1236
1237                                 break;
1238                         case self::OSTATUS:
1239                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1240                                 break;
1241                         case self::ACTIVITYPUB:
1242                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1243                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1244                                 break;
1245                         default:
1246                                 $text = ($is_quote_share? "\n" : '');
1247
1248                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1249                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1250
1251                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1252                                 $text .= Renderer::replaceMacros($tpl, [
1253                                         '$profile'      => $attributes['profile'],
1254                                         '$avatar'       => $attributes['avatar'],
1255                                         '$author'       => $attributes['author'],
1256                                         '$link'         => $attributes['link'],
1257                                         '$link_title'   => DI::l10n()->t('Link to source'),
1258                                         '$posted'       => $attributes['posted'],
1259                                         '$guid'         => $attributes['guid'],
1260                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1261                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1262                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
1263                                 ]);
1264                                 break;
1265                 }
1266
1267                 return $text;
1268         }
1269
1270         private static function removePictureLinksCallback(array $match): string
1271         {
1272                 $cache_key = 'remove:' . $match[1];
1273                 $text = DI::cache()->get($cache_key);
1274
1275                 if (is_null($text)) {
1276                         $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1277                         if ($curlResult->isSuccess()) {
1278                                 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1279                         } else {
1280                                 $mimetype = '';
1281                         }
1282
1283                         if (substr($mimetype, 0, 6) == 'image/') {
1284                                 $text = '[url=' . $match[1] . ']' . $match[1] . '[/url]';
1285                         } else {
1286                                 $text = '[url=' . $match[2] . ']' . $match[2] . '[/url]';
1287
1288                                 // if its not a picture then look if its a page that contains a picture link
1289                                 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1290                                 if (empty($body)) {
1291                                         DI::cache()->set($cache_key, $text);
1292                                         return $text;
1293                                 }
1294
1295                                 $doc = new DOMDocument();
1296                                 @$doc->loadHTML($body);
1297                                 $xpath = new DOMXPath($doc);
1298                                 $list = $xpath->query('//meta[@name]');
1299                                 foreach ($list as $node) {
1300                                         $attr = [];
1301
1302                                         if ($node->attributes->length) {
1303                                                 foreach ($node->attributes as $attribute) {
1304                                                         $attr[$attribute->name] = $attribute->value;
1305                                                 }
1306                                         }
1307
1308                                         if (strtolower($attr['name']) == 'twitter:image') {
1309                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1310                                         }
1311                                 }
1312                         }
1313                         DI::cache()->set($cache_key, $text);
1314                 }
1315
1316                 return $text;
1317         }
1318
1319         /**
1320          * Callback: Expands links from given $match array
1321          *
1322          * @param array $match Array with link match
1323          * @return string BBCode
1324          */
1325         private static function expandLinksCallback(array $match): string
1326         {
1327                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1328                         return ($match[1] . '[url]' . $match[2] . '[/url]');
1329                 } else {
1330                         return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1331                 }
1332         }
1333
1334         /**
1335          * Callback: Cleans picture links
1336          *
1337          * @param array $match Array with link match
1338          * @return string BBCode
1339          */
1340         private static function cleanPictureLinksCallback(array $match): string
1341         {
1342                 // When the picture link is the own photo path then we can avoid fetching the link
1343                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1344                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1345                         if (!empty($match[3])) {
1346                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1347                         } else {
1348                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1349                         }
1350                         return $text;
1351                 }
1352
1353                 $cache_key = 'clean:' . $match[1];
1354                 $text = DI::cache()->get($cache_key);
1355                 if (!is_null($text)) {
1356                         return $text;
1357                 }
1358
1359                 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1360                 if ($curlResult->isSuccess()) {
1361                         $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1362                 } else {
1363                         $mimetype = '';
1364                 }
1365
1366                 // if its a link to a picture then embed this picture
1367                 if (substr($mimetype, 0, 6) == 'image/') {
1368                         $text = '[img]' . $match[1] . '[/img]';
1369                 } else {
1370                         if (!empty($match[3])) {
1371                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1372                         } else {
1373                                 $text = '[img]' . $match[2] . '[/img]';
1374                         }
1375
1376                         // if its not a picture then look if its a page that contains a picture link
1377                         $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1378                         if (empty($body)) {
1379                                 DI::cache()->set($cache_key, $text);
1380                                 return $text;
1381                         }
1382
1383                         $doc = new DOMDocument();
1384                         @$doc->loadHTML($body);
1385                         $xpath = new DOMXPath($doc);
1386                         $list = $xpath->query('//meta[@name]');
1387                         foreach ($list as $node) {
1388                                 $attr = [];
1389                                 if ($node->attributes->length) {
1390                                         foreach ($node->attributes as $attribute) {
1391                                                 $attr[$attribute->name] = $attribute->value;
1392                                         }
1393                                 }
1394
1395                                 if (strtolower($attr['name']) == "twitter:image") {
1396                                         if (!empty($match[3])) {
1397                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1398                                         } else {
1399                                                 $text = "[img]" . $attr['content'] . "[/img]";
1400                                         }
1401                                 }
1402                         }
1403                 }
1404                 DI::cache()->set($cache_key, $text);
1405
1406                 return $text;
1407         }
1408
1409         /**
1410          * Cleans picture links
1411          *
1412          * @param string $text HTML/BBCode string
1413          * @return string Cleaned HTML/BBCode
1414          */
1415         public static function cleanPictureLinks(string $text): string
1416         {
1417                 DI::profiler()->startRecording('rendering');
1418                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $text);
1419                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $return);
1420                 DI::profiler()->stopRecording();
1421                 return $return;
1422         }
1423
1424         /**
1425          * Removes links
1426          *
1427          * @param string $text HTML/BBCode string
1428          * @return string Cleaned HTML/BBCode
1429          */
1430         public static function removeLinks(string $bbcode): string
1431         {
1432                 DI::profiler()->startRecording('rendering');
1433                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1434                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1435
1436                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1437                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1438                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1439                 DI::profiler()->stopRecording();
1440                 return $bbcode;
1441         }
1442
1443         /**
1444          * Replace names in mentions with nicknames
1445          *
1446          * @param string $body HTML/BBCode
1447          * @return string Body with replaced mentions
1448          */
1449         public static function setMentionsToNicknames(string $body): string
1450         {
1451                 DI::profiler()->startRecording('rendering');
1452                 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1453                 $body = preg_replace_callback($regexp, [self::class, 'mentionCallback'], $body);
1454                 DI::profiler()->stopRecording();
1455                 return $body;
1456         }
1457
1458         /**
1459          * Callback function to replace a Friendica style mention in a mention with the nickname
1460          *
1461          * @param array $match Matching values for the callback
1462          * @return string Replaced mention or empty string
1463          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1464          */
1465         private static function mentionCallback(array $match): string
1466         {
1467                 if (empty($match[2])) {
1468                         return '';
1469                 }
1470
1471                 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1472                 if (empty($data['nick'])) {
1473                         return $match[0];
1474                 }
1475
1476                 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1477         }
1478
1479         /**
1480          * Converts a BBCode message for a given URI-ID to a HTML message
1481          *
1482          * BBcode 2 HTML was written by WAY2WEB.net
1483          * extended to work with Mistpark/Friendica - Mike Macgirvin
1484          *
1485          * Simple HTML values meaning:
1486          * - 0: Friendica display
1487          * - 1: Unused
1488          * - 2: Used for Windows Phone push, Friendica API
1489          * - 3: Used before converting to Markdown in bb2diaspora.php
1490          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1491          * - 5: Unused
1492          * - 6: Unused
1493          * - 7: Used for dfrn, OStatus
1494          * - 8: Used for Twitter, WP backlink text setting
1495          * - 9: ActivityPub
1496          *
1497          * @param int    $uriid
1498          * @param string $text
1499          * @param int    $simple_html
1500          * @return string
1501          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1502          */
1503         public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1504         {
1505                 $try_oembed = ($simple_html == self::INTERNAL);
1506
1507                 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1508         }
1509
1510         /**
1511          * Converts a BBCode message to HTML message
1512          *
1513          * BBcode 2 HTML was written by WAY2WEB.net
1514          * extended to work with Mistpark/Friendica - Mike Macgirvin
1515          *
1516          * Simple HTML values meaning:
1517          * - 0: Friendica display
1518          * - 1: Unused
1519          * - 2: Used for Windows Phone push, Friendica API
1520          * - 3: Used before converting to Markdown in bb2diaspora.php
1521          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1522          * - 5: Unused
1523          * - 6: Unused
1524          * - 7: Used for dfrn, OStatus
1525          * - 8: Used for Twitter, WP backlink text setting
1526          * - 9: ActivityPub
1527          *
1528          * @param string $text
1529          * @param bool   $try_oembed
1530          * @param int    $simple_html
1531          * @param bool   $for_plaintext
1532          * @param int    $uriid
1533          * @return string Converted code or empty string
1534          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1535          */
1536         public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1537         {
1538                 // Accounting for null default column values
1539                 if (is_null($text) || $text === '') {
1540                         return '';
1541                 }
1542
1543                 DI::profiler()->startRecording('rendering');
1544
1545                 Hook::callAll('bbcode', $text);
1546
1547                 $a = DI::app();
1548
1549                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1550                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1551                                 /*
1552                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1553                                  *
1554                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1555                                  * $match[1] = $url
1556                                  * $match[2] = $title or absent
1557                                  */
1558                                 $try_oembed_callback = function (array $match)
1559                                 {
1560                                         $url = $match[1];
1561                                         $title = $match[2] ?? '';
1562
1563                                         try {
1564                                                 $return = OEmbed::getHTML($url, $title);
1565                                         } catch (Exception $ex) {
1566                                                 $return = $match[0];
1567                                         }
1568
1569                                         return $return;
1570                                 };
1571
1572                                 // Remove the abstract element. It is a non visible element.
1573                                 $text = self::stripAbstract($text);
1574
1575                                 // Line ending normalisation
1576                                 $text = str_replace("\r\n", "\n", $text);
1577
1578                                 // Move new lines outside of tags
1579                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1580                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1581
1582                                 // Extract the private images which use data urls since preg has issues with
1583                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1584                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1585
1586                                 $extracted = self::extractImagesFromItemBody($text);
1587                                 $text = $extracted['body'];
1588                                 $saved_image = $extracted['images'];
1589
1590                                 // If we find any event code, turn it into an event.
1591                                 // After we're finished processing the bbcode we'll
1592                                 // replace all of the event code with a reformatted version.
1593
1594                                 $ev = Event::fromBBCode($text);
1595
1596                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1597                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1598
1599                                 $text = str_replace("<", "&lt;", $text);
1600                                 $text = str_replace(">", "&gt;", $text);
1601
1602                                 // remove some newlines before the general conversion
1603                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1]$2[/share]\n", $text);
1604                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "\n[quote$1]$2[/quote]\n", $text);
1605
1606                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1607                                 if (!$try_oembed) {
1608                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1609                                 }
1610
1611                                 // Remove linefeeds inside of the table elements. See issue #6799
1612                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1613                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1614                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1615                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1616                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1617                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1618                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1619                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1620                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1621                                         "[hr]", "[hr]", "[hr]", "[hr]",
1622                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1623                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1624                                 do {
1625                                         $oldtext = $text;
1626                                         $text = str_replace($search, $replace, $text);
1627                                 } while ($oldtext != $text);
1628
1629                                 // Replace these here only once
1630                                 $search = ["\n[table]", "[/table]\n"];
1631                                 $replace = ["[table]", "[/table]"];
1632                                 $text = str_replace($search, $replace, $text);
1633
1634                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1635                                 $text = trim($text, "\n");
1636
1637                                 // removing multiplicated newlines
1638                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1639                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[*]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1640                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1641                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1642                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1643                                         do {
1644                                                 $oldtext = $text;
1645                                                 $text = str_replace($search, $replace, $text);
1646                                         } while ($oldtext != $text);
1647                                 }
1648
1649                                 /// @todo Have a closer look at the different html modes
1650                                 // Handle attached links or videos
1651                                 if (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1652                                         $text = self::removeAttachment($text);
1653                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1654                                         $text = self::removeAttachment($text, true);
1655                                 } else {
1656                                         $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1657                                 }
1658
1659                                 $nosmile = strpos($text, '[nosmile]') !== false;
1660                                 $text = str_replace('[nosmile]', '', $text);
1661
1662                                 // Replace non graphical smilies for external posts
1663                                 if (!$nosmile) {
1664                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1665                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1666                                         });
1667                                 }
1668
1669                                 // leave open the posibility of [map=something]
1670                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1671                                 if (strpos($text, '[/map]') !== false) {
1672                                         $text = preg_replace_callback(
1673                                                 "/\[map\](.*?)\[\/map\]/ism",
1674                                                 function ($match) use ($simple_html) {
1675                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1676                                                 },
1677                                                 $text
1678                                         );
1679                                 }
1680
1681                                 if (strpos($text, '[map=') !== false) {
1682                                         $text = preg_replace_callback(
1683                                                 "/\[map=(.*?)\]/ism",
1684                                                 function ($match) use ($simple_html) {
1685                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1686                                                 },
1687                                                 $text
1688                                         );
1689                                 }
1690
1691                                 if (strpos($text, '[map]') !== false) {
1692                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1693                                 }
1694
1695                                 // Check for headers
1696                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1697                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1698                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1699                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1700                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1701                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1702
1703                                 // Check for paragraph
1704                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1705
1706                                 // Check for bold text
1707                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1708
1709                                 // Check for Italics text
1710                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1711
1712                                 // Check for Underline text
1713                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1714
1715                                 // Check for strike-through text
1716                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1717
1718                                 // Check for over-line text
1719                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1720
1721                                 // Check for colored text
1722                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1723
1724                                 // Check for sized text
1725                                 // [size=50] --> font-size: 50px (with the unit).
1726                                 if ($simple_html != self::DIASPORA) {
1727                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1728                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1729                                 } else {
1730                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1731                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1732                                 }
1733
1734
1735                                 // Check for centered text
1736                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1737
1738                                 // Check for list text
1739                                 $text = str_replace("[*]", "<li>", $text);
1740
1741                                 // Check for block-level custom CSS
1742                                 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1743
1744                                 // Check for inline custom CSS
1745                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1746
1747                                 // Mastodon Emoji (internal tag, do not document for users)
1748                                 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1749
1750                                 // Check for CSS classes
1751                                 // @deprecated since 2021.12, left for backward-compatibility reasons
1752                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1753                                 // Add HTML new lines
1754                                 $text = str_replace("\n", '<br>', $text);
1755
1756                                 // handle nested lists
1757                                 $endlessloop = 0;
1758
1759                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1760                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1761                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1762                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1763                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1764                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1765                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1766                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1767                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1768                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1769                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1770                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1771                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1772                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1773                                 }
1774
1775                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1776                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1777                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1778                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1779
1780                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1781                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1782
1783                                 $text = str_replace('[hr]', '<hr />', $text);
1784
1785                                 if (!$for_plaintext) {
1786                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1787                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1788                                         });
1789                                 }
1790
1791                                 // Check for font change text
1792                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1793
1794                                 // Declare the format for [spoiler] layout
1795                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1796
1797                                 // Check for [spoiler] text
1798                                 // handle nested quotes
1799                                 $endlessloop = 0;
1800                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1801                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1802                                 }
1803
1804                                 // Check for [spoiler=Title] text
1805
1806                                 // handle nested quotes
1807                                 $endlessloop = 0;
1808                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1809                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1810                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1811                                                 $text);
1812                                 }
1813
1814                                 // Declare the format for [quote] layout
1815                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1816
1817                                 // Check for [quote] text
1818                                 // handle nested quotes
1819                                 $endlessloop = 0;
1820                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1821                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1822                                 }
1823
1824                                 // Check for [quote=Author] text
1825
1826                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1827
1828                                 // handle nested quotes
1829                                 $endlessloop = 0;
1830                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1831                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1832                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1833                                                 $text);
1834                                 }
1835
1836
1837                                 // [img=widthxheight]image source[/img]
1838                                 $text = preg_replace_callback(
1839                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1840                                         function ($matches) use ($simple_html, $uriid) {
1841                                                 if (strpos($matches[3], "data:image/") === 0) {
1842                                                         return $matches[0];
1843                                                 }
1844
1845                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1846                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1847                                         },
1848                                         $text
1849                                 );
1850
1851                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1852                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1853
1854                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1855                                         function ($matches) use ($simple_html, $uriid) {
1856                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1857                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1858                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1859                                         },
1860                                         $text);
1861
1862                                 // Images
1863                                 // [img]pathtoimage[/img]
1864                                 $text = preg_replace_callback(
1865                                         "/\[img\](.*?)\[\/img\]/ism",
1866                                         function ($matches) use ($simple_html, $uriid) {
1867                                                 if (strpos($matches[1], "data:image/") === 0) {
1868                                                         return $matches[0];
1869                                                 }
1870
1871                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1872                                                 return "[img]" . $matches[1] . "[/img]";
1873                                         },
1874                                         $text
1875                                 );
1876
1877                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1878                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1879
1880                                 $text = self::convertImages($text, $simple_html, $uriid);
1881
1882                                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1883                                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1884                                 //$text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1885
1886                                 // Simplify "video" element
1887                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1888
1889                                 if ($try_oembed) {
1890                                         // html5 video and audio
1891                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1892                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1893
1894                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1895                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1896
1897                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1898                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1899                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1900                                 } else {
1901                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1902                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1903                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1904                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1905                                 }
1906
1907                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1908                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1909
1910                                 // Youtube extensions
1911                                 if ($try_oembed) {
1912                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1913                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1914                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/shorts\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1915                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1916                                 }
1917
1918                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1919                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1920                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1921                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1922
1923                                 if ($try_oembed) {
1924                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1925                                 } else {
1926                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1927                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1928                                 }
1929
1930                                 if ($try_oembed) {
1931                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1932                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1933                                 }
1934
1935                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1936                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1937
1938                                 if ($try_oembed) {
1939                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1940                                 } else {
1941                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1942                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1943                                 }
1944
1945                                 // oembed tag
1946                                 $text = OEmbed::BBCode2HTML($text);
1947
1948                                 // Avoid triple linefeeds through oembed
1949                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1950
1951                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1952                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1953                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1954                                 // start which is always required). Allow desc with a missing summary for compatibility.
1955
1956                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1957                                         $sub = Event::getHTML($ev, $simple_html, $uriid);
1958
1959                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1960                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1961                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1962                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1963                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1964                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1965                                 }
1966
1967                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1968                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1969                                         // Emojis are always 4 byte Unicode characters
1970                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1971                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1972                                         }
1973                                 }
1974
1975                                 // Handle mentions and hashtag links
1976                                 if ($simple_html == self::DIASPORA) {
1977                                         // The ! is converted to @ since Diaspora only understands the @
1978                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1979                                                 '@<a href="$2">$3</a>',
1980                                                 $text);
1981                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1982                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1983                                                 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1984                                                 $text);
1985                                         $text = preg_replace("/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1986                                                 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1987                                                 $text);
1988                                 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::TWITTER_API])) {
1989                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1990                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1991                                                 $text);
1992                                 } elseif ($simple_html == self::MASTODON_API) {
1993                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1994                                                 '<a class="u-url mention status-link" href="$2" rel="nofollow noopener noreferrer" target="_blank" title="$3">$1<span>$3</span></a>',
1995                                                 $text);
1996                                         $text = preg_replace("/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1997                                                 '<a class="mention hashtag status-link" href="$2" rel="tag">$1<span>$3</span></a>',
1998                                                 $text);
1999                                 } else {
2000                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
2001                                 }
2002
2003                                 if (!$for_plaintext) {
2004                                         if (in_array($simple_html, [self::OSTATUS, self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
2005                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
2006                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
2007                                         }
2008                                 } else {
2009                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
2010                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'removePictureLinksCallback'], $text);
2011                                 }
2012
2013                                 // Bookmarks in red - will be converted to bookmarks in friendica
2014                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
2015                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
2016                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
2017                                                         "[bookmark=$1]$2[/bookmark]", $text);
2018
2019                                 if (in_array($simple_html, [self::OSTATUS, self::TWITTER])) {
2020                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", [self::class, 'expandLinksCallback'], $text);
2021                                         //$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
2022                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
2023                                 }
2024
2025                                 // Perform URL Search
2026                                 if ($try_oembed) {
2027                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
2028                                 }
2029
2030                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
2031
2032                                 // Handle Diaspora posts
2033                                 $text = preg_replace_callback(
2034                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
2035                                         function ($match) {
2036                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
2037                                         }, $text
2038                                 );
2039
2040                                 $text = preg_replace_callback(
2041                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
2042                                         function ($match) {
2043                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
2044                                         }, $text
2045                                 );
2046
2047                                 // Server independent link to posts and comments
2048                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
2049                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
2050                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
2051
2052                                 /* Tag conversion
2053                                  * Supports:
2054                                  * - #[url=<anything>]<term>[/url]
2055                                  * - [url=<anything>]#<term>[/url]
2056                                  */
2057                                 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
2058                                         $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
2059                                                 if ($simple_html == self::ACTIVITYPUB) {
2060                                                         return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
2061                                                                 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
2062                                                                 . XML::escape($matches[1]) . '</a>';
2063                                                 } else {
2064                                                         return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
2065                                                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
2066                                                                 . XML::escape($matches[1]) . '</a>';
2067                                                 }
2068                                         }, $text);
2069                                         return $text;
2070                                 });
2071
2072                                 // We need no target="_blank" rel="noopener noreferrer" for local links
2073                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
2074                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
2075                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
2076                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
2077
2078                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
2079                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2080
2081                                 // Red compatibility, though the link can't be authenticated on Friendica
2082                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2083
2084
2085                                 // we may need to restrict this further if it picks up too many strays
2086                                 // link acct:user@host to a webfinger profile redirector
2087
2088                                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
2089
2090                                 // Perform MAIL Search
2091                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
2092                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
2093
2094                                 /// @todo What is the meaning of these lines?
2095                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
2096                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
2097
2098                                 // Currently deactivated, it made problems with " inside of alt texts.
2099                                 //$text = preg_replace('/\&quot\;/', '"', $text);
2100
2101                                 // fix any escaped ampersands that may have been converted into links
2102                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
2103
2104                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
2105                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'contact/redir/', 'cid:'];
2106
2107                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
2108
2109                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
2110                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
2111
2112                                 // sanitize href attributes (only allowlisted protocols URLs)
2113                                 // default value for backward compatibility
2114                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
2115
2116                                 // Always allowed protocol even if config isn't set or not including it
2117                                 $allowed_link_protocols[] = '//';
2118                                 $allowed_link_protocols[] = 'http://';
2119                                 $allowed_link_protocols[] = 'https://';
2120                                 $allowed_link_protocols[] = 'contact/redir/';
2121
2122                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
2123
2124                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
2125                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
2126
2127                                 // Shared content
2128                                 $text = self::convertShare(
2129                                         $text,
2130                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
2131                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
2132                                         }, $uriid
2133                                 );
2134
2135                                 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
2136
2137                                 return $text;
2138                         }); // Escaped noparse, nobb, pre
2139
2140                         // Remove escaping tags and replace new lines that remain
2141                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
2142                                 return str_replace("\n", "<br>", $match[2]);
2143                         }, $text);
2144
2145                         // Additionally, [pre] tags preserve spaces
2146                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
2147                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
2148                         }, $text);
2149
2150                         return $text;
2151                 }); // Escaped code
2152
2153                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
2154                         function ($matches) {
2155                                 if (strpos($matches[2], "\n") !== false) {
2156                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
2157                                 } else {
2158                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
2159                                 }
2160
2161                                 return $return;
2162                         },
2163                         $text
2164                 );
2165
2166                 // Default iframe allowed domains/path
2167                 $allowedIframeDomains = [
2168                         DI::baseUrl()->getHostname()
2169                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
2170                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
2171                         'www.youtube.com/embed/',
2172                         'player.vimeo.com/video/',
2173                 ];
2174
2175                 $allowedIframeDomains = array_merge(
2176                         $allowedIframeDomains,
2177                         DI::config()->get('system', 'allowed_oembed') ?
2178                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
2179                                 : []
2180                 );
2181
2182                 $text = HTML::purify($text, $allowedIframeDomains);
2183                 DI::profiler()->stopRecording();
2184
2185                 return trim($text);
2186         }
2187
2188         /**
2189          * Strips the "abstract" tag from the provided text
2190          *
2191          * @param string $text The text with BBCode
2192          * @return string The same text - but without "abstract" element
2193          */
2194         public static function stripAbstract(string $text): string
2195         {
2196                 DI::profiler()->startRecording('rendering');
2197
2198                 $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
2199                         $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
2200                         $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
2201                         return $text;
2202                 });
2203
2204                 DI::profiler()->stopRecording();
2205                 return $text;
2206         }
2207
2208         /**
2209          * Returns the value of the "abstract" element
2210          *
2211          * @param string $text  The text that maybe contains the element
2212          * @param string $addon The addon for which the abstract is meant for
2213          * @return string The abstract
2214          */
2215         public static function getAbstract(string $text, string $addon = ''): string
2216         {
2217                 DI::profiler()->startRecording('rendering');
2218                 $addon = strtolower($addon);
2219
2220                 $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
2221                         if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
2222                                 return $matches[1];
2223                         }
2224
2225                         if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
2226                                 return $matches[1];
2227                         }
2228
2229                         return '';
2230                 });
2231
2232                 DI::profiler()->stopRecording();
2233                 return $abstract;
2234         }
2235
2236         /**
2237          * Callback function to replace a Friendica style mention in a mention for Diaspora
2238          *
2239          * @param array $match Matching values for the callback
2240          *                     [1] = Mention type (! or @)
2241          *                     [2] = Name
2242          *                     [3] = Address
2243          * @return string Replaced mention
2244          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2245          * @throws \ImagickException
2246          */
2247         private static function bbCodeMention2DiasporaCallback(array $match): string
2248         {
2249                 $contact = Contact::getByURL($match[3], false, ['addr']);
2250                 if (empty($contact['addr'])) {
2251                         return $match[0];
2252                 }
2253
2254                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2255                 return $mention;
2256         }
2257
2258         /**
2259          * Converts a BBCode text into Markdown
2260          *
2261          * This function converts a BBCode item body to be sent to Markdown-enabled
2262          * systems like Diaspora and Libertree
2263          *
2264          * @param string $text
2265          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2266          * @return string
2267          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2268          */
2269         public static function toMarkdown(string $text, bool $for_diaspora = true): string
2270         {
2271                 DI::profiler()->startRecording('rendering');
2272                 $original_text = $text;
2273
2274                 // Since Diaspora is creating a summary for links, this function removes them before posting
2275                 if ($for_diaspora) {
2276                         $text = self::removeShareInformation($text);
2277                 }
2278
2279                 /**
2280                  * Transform #tags, strip off the [url] and replace spaces with underscore
2281                  */
2282                 $url_search_string = "^\[\]";
2283                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2284                         function ($matches) {
2285                                 return '#' . str_replace(' ', '_', $matches[2]);
2286                         },
2287                         $text
2288                 );
2289
2290                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2291                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2292
2293                 // Convert it to HTML - don't try oembed
2294                 if ($for_diaspora) {
2295                         $text = self::convert($text, false, self::DIASPORA);
2296
2297                         // Add all tags that maybe were removed
2298                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2299                                 $tagline = '';
2300                                 foreach ($tags[2] as $tag) {
2301                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2302                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2303                                                 $tagline .= '#' . $tag . ' ';
2304                                         }
2305                                 }
2306                                 $text = $text . ' ' . $tagline;
2307                         }
2308                 } else {
2309                         $text = self::convert($text, false, self::CONNECTORS);
2310                 }
2311
2312                 // If a link is followed by a quote then there should be a newline before it
2313                 // Maybe we should make this newline at every time before a quote.
2314                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2315
2316                 // Now convert HTML to Markdown
2317                 $text = HTML::toMarkdown($text);
2318
2319                 // Libertree has a problem with escaped hashtags.
2320                 $text = str_replace(['\#'], ['#'], $text);
2321
2322                 // Remove any leading or trailing whitespace, as this will mess up
2323                 // the Diaspora signature verification and cause the item to disappear
2324                 $text = trim($text);
2325
2326                 if ($for_diaspora) {
2327                         $url_search_string = "^\[\]";
2328                         $text = preg_replace_callback(
2329                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2330                                 [self::class, 'bbCodeMention2DiasporaCallback'],
2331                                 $text
2332                         );
2333                 }
2334
2335                 Hook::callAll('bb2diaspora', $text);
2336
2337                 DI::profiler()->stopRecording();
2338                 return $text;
2339         }
2340
2341         /**
2342          * Pull out all #hashtags and @person tags from $string.
2343          *
2344          * We also get @person@domain.com - which would make
2345          * the regex quite complicated as tags can also
2346          * end a sentence. So we'll run through our results
2347          * and strip the period from any tags which end with one.
2348          * Returns array of tags found, or empty array.
2349          *
2350          * @param string $string Post content
2351          * @return array List of tag and person names
2352          */
2353         public static function getTags(string $string): array
2354         {
2355                 DI::profiler()->startRecording('rendering');
2356                 $ret = [];
2357
2358                 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img', 'attachment'], function ($string) use (&$ret) {
2359                         // Convert hashtag links to hashtags
2360                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2361
2362                         // Force line feeds at bbtags
2363                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2364
2365                         // ignore anything in a bbtag
2366                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2367
2368                         // Match full names against @tags including the space between first and last
2369                         // We will look these up afterward to see if they are full names or not recognisable.
2370
2371                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2372                                 foreach ($matches[1] as $match) {
2373                                         if (strstr($match, ']')) {
2374                                                 // we might be inside a bbcode color tag - leave it alone
2375                                                 continue;
2376                                         }
2377
2378                                         if (substr($match, -1, 1) === '.') {
2379                                                 $ret[] = substr($match, 0, -1);
2380                                         } else {
2381                                                 $ret[] = $match;
2382                                         }
2383                                 }
2384                         }
2385
2386                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2387                         // and #hash tags.
2388
2389                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2390                                 foreach ($matches[1] as $match) {
2391                                         if (strstr($match, ']')) {
2392                                                 // we might be inside a bbcode color tag - leave it alone
2393                                                 continue;
2394                                         }
2395
2396                                         // try not to catch url fragments
2397                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2398                                                 continue;
2399                                         }
2400
2401                                         $ret[] = $match;
2402                                 }
2403                         }
2404                 });
2405
2406                 DI::profiler()->stopRecording();
2407                 return array_unique($ret);
2408         }
2409
2410         /**
2411          * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2412          *
2413          * @param string $body HTML/BBCode
2414          * @return string body with expanded tags
2415          */
2416         public static function expandTags(string $body): string
2417         {
2418                 return preg_replace_callback("/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2419                         function (array $match) {
2420                                 switch ($match[1]) {
2421                                         case '!':
2422                                         case '@':
2423                                                 $contact = Contact::getByURL($match[2]);
2424                                                 if (!empty($contact)) {
2425                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2426                                                 } else {
2427                                                         return $match[1] . $match[2];
2428                                                 }
2429                                                 break;
2430
2431                                         case '#':
2432                                         default:
2433                                                 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2434                                 }
2435                         }, $body);
2436         }
2437
2438         /**
2439          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2440          *
2441          * @param string   $text HTML/BBCode
2442          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2443          * @param callable $callback
2444          * @return string
2445          * @see Strings::performWithEscapedBlocks
2446          */
2447         public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2448         {
2449                 $tagList = array_map('preg_quote', $tagList);
2450
2451                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2452         }
2453
2454         /**
2455          * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2456          *
2457          * @param string $body HTML/BBCode
2458          * @param int $profile_uid Profile user id
2459          * @param string $network Network name
2460          * @return string HTML/BBCode with inserted images
2461          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2462          * @throws \ImagickException
2463          */
2464         public static function setMentions(string $body, $profile_uid = 0, $network = '')
2465         {
2466                 DI::profiler()->startRecording('rendering');
2467                 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2468                         $tags = self::getTags($body);
2469
2470                         $tagged = [];
2471
2472                         foreach ($tags as $tag) {
2473                                 $tag_type = substr($tag, 0, 1);
2474
2475                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2476                                         continue;
2477                                 }
2478
2479                                 /*
2480                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2481                                  * Robert Johnson should be first in the $tags array
2482                                  */
2483                                 foreach ($tagged as $nextTag) {
2484                                         if (stristr($nextTag, $tag . ' ')) {
2485                                                 continue 2;
2486                                         }
2487                                 }
2488
2489                                 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2490                                         $tagged[] = $tag;
2491                                 }
2492                         }
2493
2494                         return $body;
2495                 });
2496
2497                 DI::profiler()->stopRecording();
2498                 return $body;
2499         }
2500
2501         /**
2502          * @param string      $author  Author display name
2503          * @param string      $profile Author profile URL
2504          * @param string      $avatar  Author profile picture URL
2505          * @param string      $link    Post source URL
2506          * @param string      $posted  Post created date
2507          * @param string|null $guid    Post guid (if any)
2508          * @param string|null $uri     Post uri (if any)
2509          * @return string
2510          * @TODO Rewrite to handle over whole record array
2511          */
2512         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null, string $uri = null): string
2513         {
2514                 DI::profiler()->startRecording('rendering');
2515                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2516                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2517                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2518                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2519                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2520
2521                 if ($guid) {
2522                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2523                 }
2524
2525                 if ($uri) {
2526                         $header .= "' message_id='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $uri);
2527                 }
2528
2529                 $header  .= "']";
2530
2531                 DI::profiler()->stopRecording();
2532                 return $header;
2533         }
2534
2535         /**
2536          * Returns the BBCode relevant to embed the provided URL in a post body.
2537          * For media type, it will return [img], [video] and [audio] tags.
2538          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2539          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2540          *
2541          * @param string      $url
2542          * @param bool        $tryAttachment
2543          * @param string|null $title
2544          * @param string|null $description
2545          * @param string|null $tags
2546          * @return string
2547          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2548          * @see ParseUrl::getSiteinfoCached
2549          */
2550         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2551         {
2552                 DI::profiler()->startRecording('rendering');
2553                 DI::logger()->info($url);
2554
2555                 // If there is already some content information submitted we don't
2556                 // need to parse the url for content.
2557                 if (!empty($title) && !empty($description)) {
2558                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2559
2560                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2561
2562                         $str_tags = '';
2563                         if (!empty($tags)) {
2564                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2565                                 if (count($arr_tags)) {
2566                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2567                                 }
2568                         }
2569
2570                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2571
2572                         DI::logger()->info('(unparsed): returns: ' . $result);
2573
2574                         DI::profiler()->stopRecording();
2575                         return $result;
2576                 }
2577
2578                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2579
2580                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2581                         switch ($siteinfo['type']) {
2582                                 case 'video':
2583                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2584                                         break;
2585                                 case 'audio':
2586                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2587                                         break;
2588                                 default:
2589                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2590                                         break;
2591                         }
2592
2593                         DI::profiler()->stopRecording();
2594                         return $bbcode;
2595                 }
2596
2597                 unset($siteinfo['keywords']);
2598
2599                 // Bypass attachment if parse url for a comment
2600                 if (!$tryAttachment) {
2601                         DI::profiler()->stopRecording();
2602                         return "\n" . '[url=' . $url . ']' . ($siteinfo['title'] ?? $url) . '[/url]';
2603                 }
2604
2605                 // Format it as BBCode attachment
2606                 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2607                 DI::profiler()->stopRecording();
2608                 return $bbcode;
2609         }
2610 }