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