src/Content/Text/BBCode.php

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