src/Content/Text/BBCode.php

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