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 NPF          = 6;
  61         const OSTATUS      = 7;
  62         const TWITTER      = 8;
  63         const BACKLINK     = 8;
  64         const ACTIVITYPUB  = 9;
  65         const BLUESKY      = 10;
  66
  67         const SHARED_ANCHOR = '<hr class="shared-anchor">';
  68         const TOP_ANCHOR    = '<br class="top-anchor">';
  69         const BOTTOM_ANCHOR = '<br class="button-anchor">';
  70
  71         const PREVIEW_NONE     = 0;
  72         const PREVIEW_NO_IMAGE = 1;
  73         const PREVIEW_LARGE    = 2;
  74         const PREVIEW_SMALL    = 3;
  75
  76         /**
  77          * Fetches attachment data that were generated with the "attachment" element
  78          *
  79          * @param string $body Message body
  80          * @return array
  81          *                     'type' -> Message type ('link', 'video', 'photo')
  82          *                     'text' -> Text before the shared message
  83          *                     'after' -> Text after the shared message
  84          *                     'image' -> Preview image of the message
  85          *                     'url' -> Url to the attached message
  86          *                     'title' -> Title of the attachment
  87          *                     'description' -> Description of the attachment
  88          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  89          */
  90         public static function getAttachmentData(string $body): array
  91         {
  92                 DI::profiler()->startRecording('rendering');
  93                 $data = [
  94                         'type'          => '',
  95                         'text'          => '',
  96                         'after'         => '',
  97                         'image'         => null,
  98                         'url'           => '',
  99                         'author_name'   => '',
 100                         'author_url'    => '',
 101                         'provider_name' => '',
 102                         'provider_url'  => '',
 103                         'title'         => '',
 104                         'description'   => '',
 105                 ];
 106
 107                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 108                         DI::profiler()->stopRecording();
 109                         return [];
 110                 }
 111
 112                 $attributes = $match[2];
 113
 114                 $data['text'] = trim($match[1]);
 115
 116                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 117                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 118                         $value = $matches[2] ?? '';
 119
 120                         if ($value != '') {
 121                                 switch ($field) {
 122                                         case 'publisher_name':
 123                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 124                                                 break;
 125
 126                                         case 'publisher_url':
 127                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 128                                                 break;
 129
 130                                         case 'author_name':
 131                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 132                                                 if ($data['provider_name'] == $data['author_name']) {
 133                                                         $data['author_name'] = '';
 134                                                 }
 135                                                 break;
 136
 137                                         case 'author_url':
 138                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 139                                                 if ($data['provider_url'] == $data['author_url']) {
 140                                                         $data['author_url'] = '';
 141                                                 }
 142                                                 break;
 143
 144                                         case 'title':
 145                                                 $value = self::toPlaintext(html_entity_decode($value, ENT_QUOTES, 'UTF-8'));
 146                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 147                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 148                                                 $data['title'] = $value;
 149
 150                                         default:
 151                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 152                                                 break;
 153                                 }
 154                         }
 155                 }
 156
 157                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 158                         DI::profiler()->stopRecording();
 159                         return [];
 160                 }
 161
 162                 $data['description'] = trim($match[3]);
 163
 164                 $data['after'] = trim($match[4]);
 165
 166                 $parts = parse_url($data['url']);
 167                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 168                         if (empty($data['provider_name'])) {
 169                                 $data['provider_name'] = $parts['host'];
 170                         }
 171                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 172                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 173
 174                                 if (!empty($parts['port'])) {
 175                                         $data['provider_url'] .= ':' . $parts['port'];
 176                                 }
 177                         }
 178                 }
 179
 180                 DI::profiler()->stopRecording();
 181                 return $data;
 182         }
 183
 184         /**
 185          * Remove [attachment] BBCode and replaces it with a regular [url]
 186          *
 187          * @param string  $body
 188          * @param boolean $no_link_desc No link description
 189          * @return string with replaced body
 190          */
 191         public static function replaceAttachment(string $body, bool $no_link_desc = false): string
 192         {
 193                 return preg_replace_callback(
 194                         "/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
 195                         function ($match) use ($body, $no_link_desc) {
 196                                 $attach_data = self::getAttachmentData($match[0]);
 197                                 if (empty($attach_data['url'])) {
 198                                         return $match[0];
 199                                 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
 200                                         return '';
 201                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 202                                         return " \n[url]" . $attach_data['url'] . "[/url]\n";
 203                                 } else {
 204                                         return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 205                                 }
 206                         },
 207                         $body
 208                 );
 209         }
 210
 211         /**
 212          * Remove [attachment] BBCode
 213          *
 214          * @param string  $body
 215          * @return string with removed attachment
 216          */
 217         public static function removeAttachment(string $body): string
 218         {
 219                 return trim(preg_replace("/\s*\[attachment .*?\].*?\[\/attachment\]\s*/ism", '', $body));
 220         }
 221
 222         /**
 223          * Converts a BBCode text into plaintext
 224          *
 225          * @param string $text
 226          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 227          * @return string
 228          */
 229         public static function toPlaintext(string $text, bool $keep_urls = true): string
 230         {
 231                 DI::profiler()->startRecording('rendering');
 232                 // Remove pictures in advance to avoid unneeded proxy calls
 233                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $text);
 234                 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
 235                 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
 236
 237                 // Remove attachment
 238                 $text = self::replaceAttachment($text);
 239
 240                 $naked_text = HTML::toPlaintext(self::convert($text, false, BBCode::EXTERNAL, true), 0, !$keep_urls);
 241
 242                 DI::profiler()->stopRecording();
 243                 return $naked_text;
 244         }
 245
 246         private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
 247         {
 248                 // Only send proxied pictures to API and for internal display
 249                 if (!in_array($simplehtml, [self::INTERNAL, self::MASTODON_API, self::TWITTER_API])) {
 250                         return $image;
 251                 } elseif ($uriid > 0) {
 252                         return Post\Link::getByLink($uriid, $image, $size);
 253                 } else {
 254                         return Proxy::proxifyUrl($image, $size);
 255                 }
 256         }
 257
 258         /**
 259          * Truncates imported message body string length to max_import_size
 260          *
 261          * The purpose of this function is to apply system message length limits to
 262          * imported messages without including any embedded photos in the length
 263          *
 264          * @param string $body
 265          * @return string
 266          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 267          */
 268         public static function limitBodySize(string $body): string
 269         {
 270                 DI::profiler()->startRecording('rendering');
 271                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 272
 273                 // If the length of the body, including the embedded images, is smaller
 274                 // than the maximum, then don't waste time looking for the images
 275                 if ($maxlen && (strlen($body) > $maxlen)) {
 276
 277                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 278
 279                         $orig_body = $body;
 280                         $new_body = '';
 281                         $textlen = 0;
 282
 283                         $img_start = strpos($orig_body, '[img');
 284                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 285                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 286                         while (($img_st_close !== false) && ($img_end !== false)) {
 287
 288                                 $img_st_close++; // make it point to AFTER the closing bracket
 289                                 $img_end += $img_start;
 290                                 $img_end += strlen('[/img]');
 291
 292                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 293                                         // This is an embedded image
 294
 295                                         if (($textlen + $img_start) > $maxlen) {
 296                                                 if ($textlen < $maxlen) {
 297                                                         Logger::debug('the limit happens before an embedded image');
 298                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 299                                                         $textlen = $maxlen;
 300                                                 }
 301                                         } else {
 302                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 303                                                 $textlen += $img_start;
 304                                         }
 305
 306                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 307                                 } else {
 308
 309                                         if (($textlen + $img_end) > $maxlen) {
 310                                                 if ($textlen < $maxlen) {
 311                                                         Logger::debug('the limit happens before the end of a non-embedded image');
 312                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 313                                                         $textlen = $maxlen;
 314                                                 }
 315                                         } else {
 316                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 317                                                 $textlen += $img_end;
 318                                         }
 319                                 }
 320                                 $orig_body = substr($orig_body, $img_end);
 321
 322                                 if ($orig_body === false) {
 323                                         // in case the body ends on a closing image tag
 324                                         $orig_body = '';
 325                                 }
 326
 327                                 $img_start = strpos($orig_body, '[img');
 328                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 329                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 330                         }
 331
 332                         if (($textlen + strlen($orig_body)) > $maxlen) {
 333                                 if ($textlen < $maxlen) {
 334                                         Logger::debug('the limit happens after the end of the last image');
 335                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 336                                 }
 337                         } else {
 338                                 Logger::debug('the text size with embedded images extracted did not violate the limit');
 339                                 $new_body = $new_body . $orig_body;
 340                         }
 341
 342                         DI::profiler()->stopRecording();
 343                         return $new_body;
 344                 } else {
 345                         DI::profiler()->stopRecording();
 346                         return $body;
 347                 }
 348         }
 349
 350         /**
 351          * Processes [attachment] tags
 352          *
 353          * Note: Can produce a [bookmark] tag in the returned string
 354          *
 355          * @param string  $text
 356          * @param integer $simplehtml
 357          * @param bool    $tryoembed
 358          * @param array   $data
 359          * @param int     $uriid
 360          * @return string
 361          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 362          */
 363         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
 364         {
 365                 DI::profiler()->startRecording('rendering');
 366                 $data = $data ?: self::getAttachmentData($text);
 367                 if (empty($data) || empty($data['url'])) {
 368                         DI::profiler()->stopRecording();
 369                         return $text;
 370                 }
 371
 372                 if (isset($data['title'])) {
 373                         $data['title'] = strip_tags($data['title']);
 374                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 375                 } else {
 376                         $data['title'] = '';
 377                 }
 378
 379                 if (((strpos($data['text'], '[img=') !== false) || (strpos($data['text'], '[img]') !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 380                         $data['preview'] = $data['image'];
 381                         $data['image'] = '';
 382                 }
 383
 384                 $return = '';
 385                 try {
 386                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 387                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 388                         } else {
 389                                 throw new Exception('OEmbed is disabled for this attachment.');
 390                         }
 391                 } catch (Exception $e) {
 392                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 393
 394                         if ($simplehtml != self::CONNECTORS) {
 395                                 $return = sprintf('<div class="type-%s">', $data['type']);
 396                         }
 397
 398                         if ($preview_mode == self::PREVIEW_NO_IMAGE) {
 399                                 unset($data['image']);
 400                                 unset($data['preview']);
 401                         }
 402
 403                         if (!empty($data['title']) && !empty($data['url'])) {
 404                                 $preview_class = $preview_mode == self::PREVIEW_LARGE ? 'attachment-image' : 'attachment-preview';
 405                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 406                                         $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), $data['title']);
 407                                 } else {
 408                                         if (!empty($data['image'])) {
 409                                                 $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), $data['title']);
 410                                         } elseif (!empty($data['preview'])) {
 411                                                 $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']);
 412                                         }
 413                                         $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
 414                                 }
 415                         }
 416
 417                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 418                                 // Sanitize the HTML
 419                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 420                         }
 421
 422                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 423                                 if (!empty($data['author_name'])) {
 424                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 425                                 } else {
 426                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 427                                 }
 428                         }
 429
 430                         if ($simplehtml != self::CONNECTORS) {
 431                                 $return .= '</div>';
 432                         }
 433                 }
 434
 435                 DI::profiler()->stopRecording();
 436                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 437         }
 438
 439         public static function removeShareInformation(string $text, bool $plaintext = false, bool $nolink = false): string
 440         {
 441                 DI::profiler()->startRecording('rendering');
 442                 $data = self::getAttachmentData($text);
 443
 444                 if (!$data) {
 445                         DI::profiler()->stopRecording();
 446                         return $text;
 447                 } elseif ($nolink) {
 448                         DI::profiler()->stopRecording();
 449                         return $data['text'] . ($data['after'] ?? '');
 450                 }
 451
 452                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 453                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 454                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 455                         $data['title'] = $data['url'];
 456                 } elseif (($text != '') && strstr($title, $text)) {
 457                         $data['text'] = $data['title'];
 458                         $data['title'] = $data['url'];
 459                 }
 460
 461                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 462                         DI::profiler()->stopRecording();
 463                         return $data['title'] . $data['after'];
 464                 }
 465
 466                 // If the link already is included in the post, don't add it again
 467                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 468                         DI::profiler()->stopRecording();
 469                         return $data['text'] . $data['after'];
 470                 }
 471
 472                 $text = $data['text'];
 473
 474                 if (!empty($data['url']) && !empty($data['title'])) {
 475                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 476                 } elseif (!empty($data['url'])) {
 477                         $text .= "\n[url]" . $data['url'] . '[/url]';
 478                 }
 479
 480                 DI::profiler()->stopRecording();
 481                 return $text . "\n" . $data['after'];
 482         }
 483
 484         /**
 485          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 486          *
 487          * @param array $match Array with the matching values
 488          * @return string reformatted link including HTML codes
 489          */
 490         private static function convertUrlForActivityPubCallback(array $match): string
 491         {
 492                 $url = $match[1];
 493
 494                 if (isset($match[2]) && ($match[1] != $match[2])) {
 495                         return $match[0];
 496                 }
 497
 498                 $parts = parse_url($url);
 499                 if (!isset($parts['scheme'])) {
 500                         return $match[0];
 501                 }
 502
 503                 return self::convertUrlForActivityPub($url);
 504         }
 505
 506         /**
 507          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 508          *
 509          * @param string $url URL that is about to be reformatted
 510          * @return string reformatted link including HTML codes
 511          */
 512         private static function convertUrlForActivityPub(string $url): string
 513         {
 514                 return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, Strings::getStyledURL($url));
 515         }
 516
 517         /*
 518          * [noparse][i]italic[/i][/noparse] turns into
 519          * [noparse][ i ]italic[ /i ][/noparse],
 520          * to hide them from parser.
 521          *
 522          * @param array $match
 523          * @return string
 524          */
 525         private static function escapeNoparseCallback(array $match): string
 526         {
 527                 $whole_match = $match[0];
 528                 $captured = $match[1];
 529                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 530                 $new_str = str_replace($captured, $spacefied, $whole_match);
 531                 return $new_str;
 532         }
 533
 534         /*
 535          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 536          * now turns back and the [noparse] tags are trimmed
 537          * returning [i]italic[/i]
 538          *
 539          * @param array $match
 540          * @return string
 541          */
 542         private static function unescapeNoparseCallback(array $match): string
 543         {
 544                 $captured = $match[1];
 545                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 546                 return $unspacefied;
 547         }
 548
 549         /**
 550          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 551          * occurrences
 552          *
 553          * @param string $text        Text to search
 554          * @param string $name        Tag name
 555          * @param int    $occurrences Number of first occurrences to skip
 556          * @return boolean|array
 557          */
 558         public static function getTagPosition(string $text, string $name, int $occurrences = 0)
 559         {
 560                 DI::profiler()->startRecording('rendering');
 561                 if ($occurrences < 0) {
 562                         $occurrences = 0;
 563                 }
 564
 565                 $start_open = -1;
 566                 for ($i = 0; $i <= $occurrences; $i++) {
 567                         if ($start_open !== false) {
 568                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 569                         }
 570                 }
 571
 572                 if ($start_open === false) {
 573                         DI::profiler()->stopRecording();
 574                         return false;
 575                 }
 576
 577                 $start_equal = strpos($text, '=', $start_open);
 578                 $start_close = strpos($text, ']', $start_open);
 579
 580                 if ($start_close === false) {
 581                         DI::profiler()->stopRecording();
 582                         return false;
 583                 }
 584
 585                 $start_close++;
 586
 587                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 588
 589                 if ($end_open === false) {
 590                         DI::profiler()->stopRecording();
 591                         return false;
 592                 }
 593
 594                 $res = [
 595                         'start' => [
 596                                 'open' => $start_open,
 597                                 'close' => $start_close
 598                         ],
 599                         'end' => [
 600                                 'open' => $end_open,
 601                                 'close' => $end_open + strlen('[/' . $name . ']')
 602                         ],
 603                 ];
 604
 605                 if ($start_equal !== false) {
 606                         $res['start']['equal'] = $start_equal + 1;
 607                 }
 608
 609                 DI::profiler()->stopRecording();
 610                 return $res;
 611         }
 612
 613         /**
 614          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 615          *
 616          * @param string $pattern Preg pattern string
 617          * @param string $replace Preg replace string
 618          * @param string $name    BBCode tag name
 619          * @param string $text    Text to search
 620          * @return string
 621          */
 622         public static function pregReplaceInTag(string $pattern, string $replace, string $name, string $text): string
 623         {
 624                 DI::profiler()->startRecording('rendering');
 625                 $occurrences = 0;
 626                 $pos = self::getTagPosition($text, $name, $occurrences);
 627                 while ($pos !== false && $occurrences++ < 1000) {
 628                         $start = substr($text, 0, $pos['start']['open']);
 629                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 630                         $end = substr($text, $pos['end']['close']);
 631                         if ($end === false) {
 632                                 $end = '';
 633                         }
 634
 635                         $subject = preg_replace($pattern, $replace, $subject);
 636                         $text = $start . $subject . $end;
 637
 638                         $pos = self::getTagPosition($text, $name, $occurrences);
 639                 }
 640
 641                 DI::profiler()->stopRecording();
 642                 return $text;
 643         }
 644
 645         private static function extractImagesFromItemBody(string $body): array
 646         {
 647                 $saved_image = [];
 648                 $orig_body = $body;
 649                 $new_body = '';
 650
 651                 $cnt = 0;
 652                 $img_start = strpos($orig_body, '[img');
 653                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 654                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 655                 while (($img_st_close !== false) && ($img_end !== false)) {
 656                         $img_st_close++; // make it point to AFTER the closing bracket
 657                         $img_end += $img_start;
 658
 659                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 660                                 // This is an embedded image
 661                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 662                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 663
 664                                 $cnt++;
 665                         } else {
 666                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 667                         }
 668
 669                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 670
 671                         if ($orig_body === false) {
 672                                 // in case the body ends on a closing image tag
 673                                 $orig_body = '';
 674                         }
 675
 676                         $img_start = strpos($orig_body, '[img');
 677                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 678                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 679                 }
 680
 681                 $new_body = $new_body . $orig_body;
 682
 683                 return ['body' => $new_body, 'images' => $saved_image];
 684         }
 685
 686         private static function interpolateSavedImagesIntoItemBody(int $uriid, string $body, array $images): string
 687         {
 688                 $newbody = $body;
 689
 690                 $cnt = 0;
 691                 foreach ($images as $image) {
 692                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 693                         // it loops over the array starting from the first element and going sequentially
 694                         // to the last element
 695                         $newbody = str_replace(
 696                                 '[$#saved_image' . $cnt . '#$]',
 697                                 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />',
 698                                 $newbody
 699                         );
 700                         $cnt++;
 701                 }
 702
 703                 return $newbody;
 704         }
 705
 706         /**
 707          * @param string $text A BBCode string
 708          * @return array Empty array if no share tag is present or the following array, missing attributes end up empty strings:
 709          *               - comment   : Text before the opening share tag
 710          *               - shared    : Text inside the share tags
 711          *               - author    : (Optional) Display name of the shared author
 712          *               - profile   : (Optional) Profile page URL of the shared author
 713          *               - avatar    : (Optional) Profile picture URL of the shared author
 714          *               - link      : (Optional) Canonical URL of the shared post
 715          *               - posted    : (Optional) Date the shared post was initially posted ("Y-m-d H:i:s" in GMT)
 716          *               - message_id: (Optional) Shared post URI if any
 717          *               - guid      : (Optional) Shared post GUID if any
 718          */
 719         public static function fetchShareAttributes(string $text): array
 720         {
 721                 DI::profiler()->startRecording('rendering');
 722                 if (preg_match('~(.*?)\[share](.*)\[/share]~ism', $text, $matches)) {
 723                         DI::profiler()->stopRecording();
 724                         return [
 725                                 'author'     => '',
 726                                 'profile'    => '',
 727                                 'avatar'     => '',
 728                                 'link'       => '',
 729                                 'posted'     => '',
 730                                 'guid'       => '',
 731                                 'message_id' => trim($matches[2]),
 732                                 'comment'    => trim($matches[1]),
 733                                 'shared'     => '',
 734                         ];
 735                 }
 736                 // See Issue https://github.com/friendica/friendica/issues/10454
 737                 // Hashtags in usernames are expanded to links. This here is a quick fix.
 738                 $text = preg_replace('~([@!#])\[url=.*?](.*?)\[/url]~ism', '$1$2', $text);
 739
 740                 if (!preg_match('~(.*?)\[share(.*?)](.*)\[/share]~ism', $text, $matches)) {
 741                         DI::profiler()->stopRecording();
 742                         return [];
 743                 }
 744
 745                 $attributes = self::extractShareAttributes($matches[2]);
 746
 747                 $attributes['comment'] = trim($matches[1]);
 748                 $attributes['shared'] = trim($matches[3]);
 749
 750                 DI::profiler()->stopRecording();
 751                 return $attributes;
 752         }
 753
 754         /**
 755          * @see BBCode::fetchShareAttributes()
 756          * @param string $shareString Internal opening share tag string matched by the regular expression
 757          * @return array A fixed attribute array where missing attribute are represented by empty strings
 758          */
 759         private static function extractShareAttributes(string $shareString): array
 760         {
 761                 $attributes = [];
 762                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid', 'message_id'] as $field) {
 763                         preg_match("/$field=(['\"])(.+?)\\1/ism", $shareString, $matches);
 764                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 765                 }
 766
 767                 return $attributes;
 768         }
 769
 770         /**
 771          * Remove the share block
 772          *
 773          * @param string $body
 774          * @return string
 775          */
 776         public static function removeSharedData(string $body): string
 777         {
 778                 return trim(preg_replace("/\s*\[share.*?\].*?\[\/share\]\s*/ism", '', $body));
 779         }
 780
 781         /**
 782          * This function converts a [share] block to text according to a provided callback function whose signature is:
 783          *
 784          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 785          *
 786          * Where:
 787          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 788          * data lookup
 789          * - $author_contact is a contact record array
 790          * - $content is the inner content of the [share] block
 791          * - $is_quote_share indicates whether there's any content before the [share] block
 792          * - Return value is the string that should replace the [share] block in the provided text
 793          *
 794          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 795          *
 796          * @param  string   $text     A BBCode string
 797          * @param  callable $callback
 798          * @return string The BBCode string with all [share] blocks replaced
 799          */
 800         public static function convertShare(string $text, callable $callback, int $uriid = 0): string
 801         {
 802                 DI::profiler()->startRecording('rendering');
 803                 $return = preg_replace_callback(
 804                         '~(.*?)\[share(.*?)](.*)\[/share]~ism',
 805                         function ($match) use ($callback, $uriid) {
 806                                 $attributes = self::extractShareAttributes($match[2]);
 807
 808                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
 809                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
 810                                 $author_contact['addr'] = ($author_contact['addr'] ?? '');
 811
 812                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 813                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 814                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 815
 816                                 if (!empty($author_contact['id'])) {
 817                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
 818                                 } elseif ($attributes['avatar']) {
 819                                         $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
 820                                 }
 821
 822                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
 823
 824                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
 825                         },
 826                         $text
 827                 );
 828
 829                 DI::profiler()->stopRecording();
 830                 return trim($return);
 831         }
 832
 833         /**
 834          * Convert complex IMG and ZMG elements
 835          *
 836          * @param [type] $text
 837          * @param integer $simplehtml
 838          * @param integer $uriid
 839          * @return string
 840          */
 841         private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
 842         {
 843                 DI::profiler()->startRecording('rendering');
 844                 $return = preg_replace_callback(
 845                         "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
 846                         function ($match) use ($simplehtml, $uriid) {
 847                                 $attribute_string = $match[1];
 848                                 $attributes = [];
 849                                 foreach (['alt', 'width', 'height'] as $field) {
 850                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 851                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 852                                 }
 853
 854                                 $img_str = '<img src="' . self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
 855                                 foreach ($attributes as $key => $value) {
 856                                         if (!empty($value)) {
 857                                                 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
 858                                         }
 859                                 }
 860                                 return $img_str . '>';
 861                         },
 862                         $text
 863                 );
 864
 865                 DI::profiler()->stopRecording();
 866                 return $return;
 867         }
 868
 869         /**
 870          * Default [share] tag conversion callback
 871          *
 872          * Note: Can produce a [bookmark] tag in the output
 873          *
 874          * @see BBCode::convertShare()
 875          * @param array   $attributes     [share] block attribute values
 876          * @param array   $author_contact Contact row of the shared author
 877          * @param string  $content        Inner content of the [share] block
 878          * @param boolean $is_quote_share Whether there is content before the [share] block
 879          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 880          * @return string
 881          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 882          */
 883         private static function convertShareCallback(array $attributes, array $author_contact, string $content, bool $is_quote_share, int $simplehtml): string
 884         {
 885                 DI::profiler()->startRecording('rendering');
 886                 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
 887
 888                 switch ($simplehtml) {
 889                         case self::MASTODON_API:
 890                         case self::TWITTER_API:
 891                                 $text = ($is_quote_share ? '<br>' : '') .
 892                                         '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
 893                                         '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
 894                                 break;
 895                         case self::DIASPORA:
 896                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 897                                         $text = ($is_quote_share ? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 898                                 } else {
 899                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
 900
 901                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
 902                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
 903                                         }
 904
 905                                         $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 906
 907                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
 908                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
 909                                         }
 910                                 }
 911
 912                                 break;
 913                         case self::CONNECTORS:
 914                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 915                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 916                                 $headline .= ':</b></p>' . "\n";
 917
 918                                 $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
 919
 920                                 break;
 921                         case self::OSTATUS:
 922                                 $text = ($is_quote_share ? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 923                                 break;
 924                         case self::ACTIVITYPUB:
 925                                 $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>';
 926                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
 927                                 break;
 928                         default:
 929                                 $text = ($is_quote_share ? "\n" : '');
 930
 931                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
 932                                 $network = $contact['network'] ?? Protocol::PHANTOM;
 933
 934                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
 935                                 $text .= BBCode::SHARED_ANCHOR . Renderer::replaceMacros($tpl, [
 936                                         '$profile'      => $attributes['profile'],
 937                                         '$avatar'       => $attributes['avatar'],
 938                                         '$author'       => $attributes['author'],
 939                                         '$link'         => $attributes['link'],
 940                                         '$link_title'   => DI::l10n()->t('Link to source'),
 941                                         '$posted'       => $attributes['posted'],
 942                                         '$guid'         => $attributes['guid'],
 943                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
 944                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
 945                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
 946                                 ]);
 947                                 break;
 948                 }
 949
 950                 return $text;
 951         }
 952
 953         private static function removePictureLinksCallback(array $match): string
 954         {
 955                 $cache_key = 'remove:' . $match[1];
 956                 $text = DI::cache()->get($cache_key);
 957
 958                 if (is_null($text)) {
 959                         $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
 960                         if ($curlResult->isSuccess()) {
 961                                 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
 962                         } else {
 963                                 $mimetype = '';
 964                         }
 965
 966                         if (substr($mimetype, 0, 6) == 'image/') {
 967                                 $text = '[url=' . $match[1] . ']' . $match[1] . '[/url]';
 968                         } else {
 969                                 $text = '[url=' . $match[2] . ']' . $match[2] . '[/url]';
 970
 971                                 // if its not a picture then look if its a page that contains a picture link
 972                                 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
 973                                 if (empty($body)) {
 974                                         DI::cache()->set($cache_key, $text);
 975                                         return $text;
 976                                 }
 977
 978                                 $doc = new DOMDocument();
 979                                 @$doc->loadHTML($body);
 980                                 $xpath = new DOMXPath($doc);
 981                                 $list = $xpath->query('//meta[@name]');
 982                                 foreach ($list as $node) {
 983                                         $attr = [];
 984
 985                                         if ($node->attributes->length) {
 986                                                 foreach ($node->attributes as $attribute) {
 987                                                         $attr[$attribute->name] = $attribute->value;
 988                                                 }
 989                                         }
 990
 991                                         if (strtolower($attr['name']) == 'twitter:image') {
 992                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
 993                                         }
 994                                 }
 995                         }
 996                         DI::cache()->set($cache_key, $text);
 997                 }
 998
 999                 return $text;
1000         }
1001
1002         /**
1003          * Callback: Expands links from given $match array
1004          *
1005          * @param array $match Array with link match
1006          * @return string BBCode
1007          */
1008         private static function expandLinksCallback(array $match): string
1009         {
1010                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1011                         return ($match[1] . '[url]' . $match[2] . '[/url]');
1012                 } else {
1013                         return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1014                 }
1015         }
1016
1017         /**
1018          * Callback: Cleans picture links
1019          *
1020          * @param array $match Array with link match
1021          * @return string BBCode
1022          */
1023         private static function cleanPictureLinksCallback(array $match): string
1024         {
1025                 // When the picture link is the own photo path then we can avoid fetching the link
1026                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()) . '/photos/');
1027                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1028                         if (!empty($match[3])) {
1029                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1030                         } else {
1031                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1032                         }
1033                         return $text;
1034                 }
1035
1036                 $cache_key = 'clean:' . $match[1];
1037                 $text = DI::cache()->get($cache_key);
1038                 if (!is_null($text)) {
1039                         return $text;
1040                 }
1041
1042                 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1043                 if ($curlResult->isSuccess()) {
1044                         $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1045                 } else {
1046                         $mimetype = '';
1047                 }
1048
1049                 // if its a link to a picture then embed this picture
1050                 if (substr($mimetype, 0, 6) == 'image/') {
1051                         $text = '[img]' . $match[1] . '[/img]';
1052                 } else {
1053                         if (!empty($match[3])) {
1054                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1055                         } else {
1056                                 $text = '[img]' . $match[2] . '[/img]';
1057                         }
1058
1059                         // if its not a picture then look if its a page that contains a picture link
1060                         $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1061                         if (empty($body)) {
1062                                 DI::cache()->set($cache_key, $text);
1063                                 return $text;
1064                         }
1065
1066                         $doc = new DOMDocument();
1067                         @$doc->loadHTML($body);
1068                         $xpath = new DOMXPath($doc);
1069                         $list = $xpath->query('//meta[@name]');
1070                         foreach ($list as $node) {
1071                                 $attr = [];
1072                                 if ($node->attributes->length) {
1073                                         foreach ($node->attributes as $attribute) {
1074                                                 $attr[$attribute->name] = $attribute->value;
1075                                         }
1076                                 }
1077
1078                                 if (strtolower($attr['name']) == "twitter:image") {
1079                                         if (!empty($match[3])) {
1080                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1081                                         } else {
1082                                                 $text = "[img]" . $attr['content'] . "[/img]";
1083                                         }
1084                                 }
1085                         }
1086                 }
1087                 DI::cache()->set($cache_key, $text);
1088
1089                 return $text;
1090         }
1091
1092         /**
1093          * Cleans picture links
1094          *
1095          * @param string $text HTML/BBCode string
1096          * @return string Cleaned HTML/BBCode
1097          */
1098         public static function cleanPictureLinks(string $text): string
1099         {
1100                 DI::profiler()->startRecording('rendering');
1101                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $text);
1102                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $return);
1103                 DI::profiler()->stopRecording();
1104                 return $return;
1105         }
1106
1107         /**
1108          * Removes links
1109          *
1110          * @param string $text HTML/BBCode string
1111          * @return string Cleaned HTML/BBCode
1112          */
1113         public static function removeLinks(string $bbcode): string
1114         {
1115                 DI::profiler()->startRecording('rendering');
1116                 $bbcode = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $bbcode);
1117                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1118                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1119
1120                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1121                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1122                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1123                 DI::profiler()->stopRecording();
1124                 return $bbcode;
1125         }
1126
1127         /**
1128          * Replace names in mentions with nicknames
1129          *
1130          * @param string $body HTML/BBCode
1131          * @return string Body with replaced mentions
1132          */
1133         public static function setMentionsToNicknames(string $body): string
1134         {
1135                 DI::profiler()->startRecording('rendering');
1136                 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1137                 $body = preg_replace_callback($regexp, [self::class, 'mentionCallback'], $body);
1138                 DI::profiler()->stopRecording();
1139                 return $body;
1140         }
1141
1142         /**
1143          * Callback function to replace a Friendica style mention in a mention with the nickname
1144          *
1145          * @param array $match Matching values for the callback
1146          * @return string Replaced mention or empty string
1147          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1148          */
1149         private static function mentionCallback(array $match): string
1150         {
1151                 if (empty($match[2])) {
1152                         return '';
1153                 }
1154
1155                 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1156                 if (empty($data['nick'])) {
1157                         return $match[0];
1158                 }
1159
1160                 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1161         }
1162
1163         /**
1164          * Normalize links to Youtube and Vimeo to a unified format.
1165          *
1166          * @param string $text
1167          * @return string
1168          */
1169         private static function normalizeVideoLinks(string $text): string
1170         {
1171                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1172                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1173                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1174                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1175
1176                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1177                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1178
1179                 return $text;
1180         }
1181
1182         /**
1183          * Expand Youtube and Vimeo links to
1184          *
1185          * @param string $text
1186          * @return string
1187          */
1188         public static function expandVideoLinks(string $text): string
1189         {
1190                 $text = self::normalizeVideoLinks($text);
1191                 $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '[url=https://www.youtube.com/watch?v=$1]https://www.youtube.com/watch?v=$1[/url]', $text);
1192                 $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '[url=https://vimeo.com/$1]https://vimeo.com/$1[/url]', $text);
1193
1194                 return $text;
1195         }
1196
1197         /**
1198          * Converts a BBCode message for a given URI-ID to a HTML message
1199          *
1200          * BBcode 2 HTML was written by WAY2WEB.net
1201          * extended to work with Mistpark/Friendica - Mike Macgirvin
1202          *
1203          * Simple HTML values meaning:
1204          * - 0: Friendica display
1205          * - 1: Unused
1206          * - 2: Used for Windows Phone push, Friendica API
1207          * - 3: Used before converting to Markdown in bb2diaspora.php
1208          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1209          * - 5: Unused
1210          * - 6: Unused
1211          * - 7: Used for dfrn, OStatus
1212          * - 8: Used for Twitter, WP backlink text setting
1213          * - 9: ActivityPub
1214          *
1215          * @param int    $uriid
1216          * @param string $text
1217          * @param int    $simple_html
1218          * @return string
1219          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1220          */
1221         public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1222         {
1223                 $try_oembed = ($simple_html == self::INTERNAL);
1224
1225                 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1226         }
1227
1228         /**
1229          * Converts a BBCode message to HTML message
1230          *
1231          * BBcode 2 HTML was written by WAY2WEB.net
1232          * extended to work with Mistpark/Friendica - Mike Macgirvin
1233          *
1234          * Simple HTML values meaning:
1235          * - 0: Friendica display
1236          * - 1: Unused
1237          * - 2: Used for Windows Phone push, Friendica API
1238          * - 3: Used before converting to Markdown in bb2diaspora.php
1239          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1240          * - 5: Unused
1241          * - 6: Unused
1242          * - 7: Used for dfrn, OStatus
1243          * - 8: Used for Twitter, WP backlink text setting
1244          * - 9: ActivityPub
1245          *
1246          * @param string $text
1247          * @param bool   $try_oembed
1248          * @param int    $simple_html
1249          * @param bool   $for_plaintext
1250          * @param int    $uriid
1251          * @return string Converted code or empty string
1252          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1253          */
1254         public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1255         {
1256                 // Accounting for null default column values
1257                 if (is_null($text) || $text === '') {
1258                         return '';
1259                 }
1260
1261                 DI::profiler()->startRecording('rendering');
1262
1263                 Hook::callAll('bbcode', $text);
1264
1265                 $a = DI::app();
1266
1267                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1268                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1269                                 /*
1270                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1271                                  *
1272                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1273                                  * $match[1] = $url
1274                                  * $match[2] = $title or absent
1275                                  */
1276                                 $try_oembed_callback = function (array $match) {
1277                                         $url = $match[1];
1278                                         $title = $match[2] ?? '';
1279
1280                                         try {
1281                                                 $return = OEmbed::getHTML($url, $title);
1282                                         } catch (Exception $ex) {
1283                                                 $return = $match[0];
1284                                         }
1285
1286                                         return $return;
1287                                 };
1288
1289                                 // Remove the abstract element. It is a non visible element.
1290                                 $text = self::stripAbstract($text);
1291
1292                                 // Line ending normalisation
1293                                 $text = str_replace("\r\n", "\n", $text);
1294
1295                                 // Move new lines outside of tags
1296                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1297                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1298
1299                                 // Extract the private images which use data urls since preg has issues with
1300                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1301                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1302
1303                                 $extracted = self::extractImagesFromItemBody($text);
1304                                 $text = $extracted['body'];
1305                                 $saved_image = $extracted['images'];
1306
1307                                 // If we find any event code, turn it into an event.
1308                                 // After we're finished processing the bbcode we'll
1309                                 // replace all of the event code with a reformatted version.
1310
1311                                 $ev = Event::fromBBCode($text);
1312
1313                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1314                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1315
1316                                 $text = str_replace("<", "&lt;", $text);
1317                                 $text = str_replace(">", "&gt;", $text);
1318
1319                                 // remove some newlines before the general conversion
1320                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1]$2[/share]\n", $text);
1321                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "\n[quote$1]$2[/quote]\n", $text);
1322
1323                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1324                                 if (!$try_oembed) {
1325                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1326                                 }
1327
1328                                 // Remove linefeeds inside of the table elements. See issue #6799
1329                                 $search = [
1330                                         "\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1331                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1332                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1333                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1334                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1335                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] ",
1336                                         " \n", "\t\n", "[/li]\n", "\n[li]", "\n[*]",
1337                                 ];
1338                                 $replace = [
1339                                         "[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1340                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1341                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1342                                         "[hr]", "[hr]", "[hr]", "[hr]",
1343                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1344                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]",
1345                                         "\n", "\n", "[/li]", "[li]", "[*]",
1346                                 ];
1347                                 do {
1348                                         $oldtext = $text;
1349                                         $text = str_replace($search, $replace, $text);
1350                                 } while ($oldtext != $text);
1351
1352                                 // Replace these here only once
1353                                 $search = ["\n[table]", "[/table]\n"];
1354                                 $replace = ["[table]", "[/table]"];
1355                                 $text = str_replace($search, $replace, $text);
1356
1357                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1358                                 $text = trim($text, "\n");
1359
1360                                 // removing multiplicated newlines
1361                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1362                                         $search = [
1363                                                 "\n\n\n", "[/quote]\n\n", "\n[/quote]", "\n[ul]", "[/ul]\n", "\n[ol]", "[/ol]\n", "\n\n[share ", "[/attachment]\n",
1364                                                 "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"
1365                                         ];
1366                                         $replace = [
1367                                                 "\n\n", "[/quote]\n", "[/quote]", "[ul]", "[/ul]", "[ol]", "[/ol]", "\n[share ", "[/attachment]",
1368                                                 "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"
1369                                         ];
1370                                         do {
1371                                                 $oldtext = $text;
1372                                                 $text = str_replace($search, $replace, $text);
1373                                         } while ($oldtext != $text);
1374                                 }
1375
1376                                 /// @todo Have a closer look at the different html modes
1377                                 // Handle attached links or videos
1378                                 if ($simple_html == self::NPF) {
1379                                         $text = self::removeAttachment($text);
1380                                 } elseif (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1381                                         $text = self::replaceAttachment($text);
1382                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1383                                         $text = self::replaceAttachment($text, true);
1384                                 } else {
1385                                         $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1386                                 }
1387
1388                                 $nosmile = strpos($text, '[nosmile]') !== false;
1389                                 $text = str_replace('[nosmile]', '', $text);
1390
1391                                 // Replace non graphical smilies for external posts
1392                                 if (!$nosmile) {
1393                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1394                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1395                                         });
1396                                 }
1397
1398                                 // leave open the possibility of [map=something]
1399                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1400                                 if (strpos($text, '[/map]') !== false) {
1401                                         $text = preg_replace_callback(
1402                                                 "/\[map\](.*?)\[\/map\]/ism",
1403                                                 function ($match) use ($simple_html) {
1404                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1405                                                 },
1406                                                 $text
1407                                         );
1408                                 }
1409
1410                                 if (strpos($text, '[map=') !== false) {
1411                                         $text = preg_replace_callback(
1412                                                 "/\[map=(.*?)\]/ism",
1413                                                 function ($match) use ($simple_html) {
1414                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1415                                                 },
1416                                                 $text
1417                                         );
1418                                 }
1419
1420                                 if (strpos($text, '[map]') !== false) {
1421                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1422                                 }
1423
1424                                 // Check for headers
1425
1426                                 if ($simple_html == self::INTERNAL) {
1427                                         //Ensure to always start with <h4> if possible
1428                                         $heading_count = 0;
1429                                         for ($level = 6; $level > 0; $level--) {
1430                                                 if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1431                                                         $heading_count++;
1432                                                 }
1433                                         }
1434                                         if ($heading_count > 0) {
1435                                                 $heading = min($heading_count + 3, 6);
1436                                                 for ($level = 6; $level > 0; $level--) {
1437                                                         if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1438                                                                 $text = preg_replace("(\[h$level\](.*?)\[\/h$level\])ism", "</p><h$heading>$1</h$heading><p>", $text);
1439                                                                 $heading--;
1440                                                         }
1441                                                 }
1442                                         }
1443                                 } else {
1444                                         $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '</p><h1>$1</h1><p>', $text);
1445                                         $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '</p><h2>$1</h2><p>', $text);
1446                                         $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '</p><h3>$1</h3><p>', $text);
1447                                         $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '</p><h4>$1</h4><p>', $text);
1448                                         $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '</p><h5>$1</h5><p>', $text);
1449                                         $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '</p><h6>$1</h6><p>', $text);
1450                                 }
1451
1452                                 // Check for paragraph
1453                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1454
1455                                 // Check for bold text
1456                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1457
1458                                 // Check for Italics text
1459                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1460
1461                                 // Check for Underline text
1462                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1463
1464                                 // Check for strike-through text
1465                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1466
1467                                 // Check for over-line text
1468                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1469
1470                                 // Check for colored text
1471                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1472
1473                                 // Check for sized text
1474                                 // [size=50] --> font-size: 50px (with the unit).
1475                                 if ($simple_html != self::DIASPORA) {
1476                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1477                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1478                                 } else {
1479                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1480                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1481                                 }
1482
1483
1484                                 // Check for centered text
1485                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1486
1487                                 // Check for list text
1488                                 $text = str_replace("[*]", "<li>", $text);
1489
1490                                 // Check for block-level custom CSS
1491                                 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1492
1493                                 // Check for inline custom CSS
1494                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1495
1496                                 // Mastodon Emoji (internal tag, do not document for users)
1497                                 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1498
1499                                 // Check for CSS classes
1500                                 // @deprecated since 2021.12, left for backward-compatibility reasons
1501                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1502                                 // Add HTML new lines
1503                                 $text = str_replace("\n\n", '</p><p>', $text);
1504                                 $text = str_replace("\n", '<br>', $text);
1505
1506                                 // handle nested lists
1507                                 $endlessloop = 0;
1508
1509                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1510                                         ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1511                                         ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1512                                         ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1513                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '</p><ul class="listbullet" style="list-style-type: circle;">$1</ul><p>', $text);
1514                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '</p><ul class="listnone" style="list-style-type: none;">$1</ul><p>', $text);
1515                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '</p><ul class="listdecimal" style="list-style-type: decimal;">$1</ul><p>', $text);
1516                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '</p><ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul><p>', $text);
1517                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '</p><ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul><p>', $text);
1518                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '</p><ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul><p>', $text);
1519                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '</p><ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul><p>', $text);
1520                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '</p><ul>$1</ul><p>', $text);
1521                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '</p><ol>$1</ol><p>', $text);
1522                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1523                                 }
1524
1525                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1526                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1527                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1528                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '</p><table>$1</table><p>', $text);
1529
1530                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '</p><table border="1" >$1</table><p>', $text);
1531                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '</p><table border="0" >$1</table><p>', $text);
1532
1533                                 $text = str_replace('[hr]', '</p><hr /><p>', $text);
1534
1535                                 if (!$for_plaintext) {
1536                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1537                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1538                                         });
1539                                 }
1540
1541                                 // Check for font change text
1542                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1543
1544                                 // Declare the format for [spoiler] layout
1545                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1546
1547                                 // Check for [spoiler] text
1548                                 // handle nested quotes
1549                                 $endlessloop = 0;
1550                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1551                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1552                                 }
1553
1554                                 // Check for [spoiler=Title] text
1555
1556                                 // handle nested quotes
1557                                 $endlessloop = 0;
1558                                 while ((strpos($text, "[/spoiler]") !== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1559                                         $text = preg_replace(
1560                                                 "/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1561                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1562                                                 $text
1563                                         );
1564                                 }
1565
1566                                 // Declare the format for [quote] layout
1567                                 $QuoteLayout = '</p><blockquote>$1</blockquote><p>';
1568
1569                                 // Check for [quote] text
1570                                 // handle nested quotes
1571                                 $endlessloop = 0;
1572                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1573                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1574                                 }
1575
1576                                 // Check for [quote=Author] text
1577
1578                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1579
1580                                 // handle nested quotes
1581                                 $endlessloop = 0;
1582                                 while ((strpos($text, "[/quote]") !== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1583                                         $text = preg_replace(
1584                                                 "/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1585                                                 "<p><strong class=" . '"author"' . ">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1586                                                 $text
1587                                         );
1588                                 }
1589
1590
1591                                 // [img=widthxheight]image source[/img]
1592                                 $text = preg_replace_callback(
1593                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1594                                         function ($matches) use ($simple_html, $uriid) {
1595                                                 if (strpos($matches[3], "data:image/") === 0) {
1596                                                         return $matches[0];
1597                                                 }
1598
1599                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1600                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1601                                         },
1602                                         $text
1603                                 );
1604
1605                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1606                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1607
1608                                 $text = preg_replace_callback(
1609                                         "/\[[iz]mg\=(.*?)\](.*?)\[\/[iz]mg\]/ism",
1610                                         function ($matches) use ($simple_html, $uriid) {
1611                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1612                                                 $alt = htmlspecialchars($matches[2], ENT_COMPAT);
1613                                                 // Fix for Markdown problems with Diaspora, see issue #12701
1614                                                 if (($simple_html != self::DIASPORA) || strpos($matches[2], '"') === false) {
1615                                                         return '<img src="' . $matches[1] . '" alt="' . $alt . '" title="' . $alt . '">';
1616                                                 } else {
1617                                                         return '<img src="' . $matches[1] . '" alt="' . $alt . '">';
1618                                                 }
1619                                         },
1620                                         $text
1621                                 );
1622
1623                                 // Images
1624                                 // [img]pathtoimage[/img]
1625                                 $text = preg_replace_callback(
1626                                         "/\[[iz]mg\](.*?)\[\/[iz]mg\]/ism",
1627                                         function ($matches) use ($simple_html, $uriid) {
1628                                                 if (strpos($matches[1], "data:image/") === 0) {
1629                                                         return $matches[0];
1630                                                 }
1631
1632                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1633                                                 return "[img]" . $matches[1] . "[/img]";
1634                                         },
1635                                         $text
1636                                 );
1637
1638                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1639                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1640
1641                                 $text = self::convertImages($text, $simple_html, $uriid);
1642
1643                                 $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);
1644                                 $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>', $text);
1645                                 //$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);
1646
1647                                 // Simplify "video" element
1648                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1649
1650                                 if ($simple_html == self::NPF) {
1651                                         $text = preg_replace(
1652                                                 "/\[video\](.*?)\[\/video\]/ism",
1653                                                 '</p><video src="$1" controls width="100%" height="auto">$1</video><p>',
1654                                                 $text
1655                                         );
1656                                         $text = preg_replace(
1657                                                 "/\[audio\](.*?)\[\/audio\]/ism",
1658                                                 '</p><audio src="$1" controls>$1">$1</audio><p>',
1659                                                 $text
1660                                         );
1661                                 } elseif ($try_oembed) {
1662                                         // html5 video and audio
1663                                         $text = preg_replace(
1664                                                 "/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1665                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>',
1666                                                 $text
1667                                         );
1668
1669                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1670                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1671
1672                                         $text = preg_replace(
1673                                                 "/\[video\](.*?)\[\/video\]/ism",
1674                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1675                                                 $text
1676                                         );
1677                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1678                                 } else {
1679                                         $text = preg_replace(
1680                                                 "/\[video\](.*?)\[\/video\]/ism",
1681                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1682                                                 $text
1683                                         );
1684                                         $text = preg_replace(
1685                                                 "/\[audio\](.*?)\[\/audio\]/ism",
1686                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1687                                                 $text
1688                                         );
1689                                 }
1690
1691                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1692                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1693
1694                                 $text = self::normalizeVideoLinks($text);
1695
1696                                 // Youtube extensions
1697                                 if ($try_oembed) {
1698                                         $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);
1699                                 } else {
1700                                         $text = preg_replace(
1701                                                 "/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1702                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>',
1703                                                 $text
1704                                         );
1705                                 }
1706
1707                                 // Vimeo extensions
1708                                 if ($try_oembed) {
1709                                         $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);
1710                                 } else {
1711                                         $text = preg_replace(
1712                                                 "/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1713                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>',
1714                                                 $text
1715                                         );
1716                                 }
1717
1718                                 // oembed tag
1719                                 $text = OEmbed::BBCode2HTML($text);
1720
1721                                 // Avoid triple linefeeds through oembed
1722                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1723
1724                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1725                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1726                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1727                                 // start which is always required). Allow desc with a missing summary for compatibility.
1728
1729                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1730                                         $sub = Event::getHTML($ev, $simple_html, $uriid);
1731
1732                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1733                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1734                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1735                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1736                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1737                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1738                                 }
1739
1740                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1741                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1742                                         // Emojis are always 4 byte Unicode characters
1743                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1744                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1745                                         }
1746                                 }
1747
1748                                 // Handle mentions and hashtag links
1749                                 if ($simple_html == self::DIASPORA) {
1750                                         // The ! is converted to @ since Diaspora only understands the @
1751                                         $text = preg_replace(
1752                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1753                                                 '@<a href="$2">$3</a>',
1754                                                 $text
1755                                         );
1756                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1757                                         $text = preg_replace(
1758                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1759                                                 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1760                                                 $text
1761                                         );
1762                                         $text = preg_replace(
1763                                                 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1764                                                 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1765                                                 $text
1766                                         );
1767                                 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::TWITTER_API])) {
1768                                         $text = preg_replace(
1769                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1770                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1771                                                 $text
1772                                         );
1773                                 } elseif ($simple_html == self::MASTODON_API) {
1774                                         $text = preg_replace(
1775                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1776                                                 '<a class="u-url mention status-link" href="$2" rel="nofollow noopener noreferrer" target="_blank" title="$3">$1<span>$3</span></a>',
1777                                                 $text
1778                                         );
1779                                         $text = preg_replace(
1780                                                 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1781                                                 '<a class="mention hashtag status-link" href="$2" rel="tag">$1<span>$3</span></a>',
1782                                                 $text
1783                                         );
1784                                 } else {
1785                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1786                                 }
1787
1788                                 if (!$for_plaintext) {
1789                                         if (in_array($simple_html, [self::OSTATUS, self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1790                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1791                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1792                                         }
1793                                 } else {
1794                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1795                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'removePictureLinksCallback'], $text);
1796                                 }
1797
1798                                 // Bookmarks in red - will be converted to bookmarks in friendica
1799                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1800                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1801                                 $text = preg_replace(
1802                                         "/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1803                                         "[bookmark=$1]$2[/bookmark]",
1804                                         $text
1805                                 );
1806
1807                                 if (in_array($simple_html, [self::OSTATUS, self::TWITTER, self::BLUESKY])) {
1808                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", [self::class, 'expandLinksCallback'], $text);
1809                                         //$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
1810                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
1811                                 }
1812
1813                                 // Perform URL Search
1814                                 if ($try_oembed) {
1815                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1816                                 }
1817
1818                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1819
1820                                 // Handle Diaspora posts
1821                                 $text = preg_replace_callback(
1822                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1823                                         function ($match) {
1824                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1825                                         },
1826                                         $text
1827                                 );
1828
1829                                 $text = preg_replace_callback(
1830                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1831                                         function ($match) {
1832                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1833                                         },
1834                                         $text
1835                                 );
1836
1837                                 // Server independent link to posts and comments
1838                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1839                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1840                                 $text = preg_replace($expression, DI::baseUrl() . "/display/$1", $text);
1841
1842                                 /* Tag conversion
1843                                  * Supports:
1844                                  * - #[url=<anything>]<term>[/url]
1845                                  * - [url=<anything>]#<term>[/url]
1846                                  */
1847                                 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
1848                                         $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function ($matches) use ($simple_html) {
1849                                                 if ($simple_html == self::ACTIVITYPUB) {
1850                                                         return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1851                                                                 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1852                                                                 . XML::escape($matches[1]) . '</a>';
1853                                                 } else {
1854                                                         return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1855                                                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1856                                                                 . XML::escape($matches[1]) . '</a>';
1857                                                 }
1858                                         }, $text);
1859                                         return $text;
1860                                 });
1861
1862                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1863                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1864                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1865                                 $text = preg_replace("/\[url\](" . $escapedBaseUrl . ".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1866                                 $text = preg_replace("/\[url\=(" . $escapedBaseUrl . ".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1867
1868                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1869                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1870
1871                                 // Red compatibility, though the link can't be authenticated on Friendica
1872                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1873
1874
1875                                 // we may need to restrict this further if it picks up too many strays
1876                                 // link acct:user@host to a webfinger profile redirector
1877
1878                                 $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);
1879
1880                                 // Perform MAIL Search
1881                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1882                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1883
1884                                 /// @todo What is the meaning of these lines?
1885                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1886                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1887
1888                                 // Currently deactivated, it made problems with " inside of alt texts.
1889                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1890
1891                                 // fix any escaped ampersands that may have been converted into links
1892                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1893
1894                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1895                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'contact/redir/', 'cid:'];
1896
1897                                 array_walk($allowed_src_protocols, function (&$value) {
1898                                         $value = preg_quote($value, '#');
1899                                 });
1900
1901                                 $text = preg_replace(
1902                                         '#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1903                                         '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">',
1904                                         $text
1905                                 );
1906
1907                                 // sanitize href attributes (only allowlisted protocols URLs)
1908                                 // default value for backward compatibility
1909                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1910
1911                                 // Always allowed protocol even if config isn't set or not including it
1912                                 $allowed_link_protocols[] = '//';
1913                                 $allowed_link_protocols[] = 'http://';
1914                                 $allowed_link_protocols[] = 'https://';
1915                                 $allowed_link_protocols[] = 'contact/redir/';
1916
1917                                 array_walk($allowed_link_protocols, function (&$value) {
1918                                         $value = preg_quote($value, '#');
1919                                 });
1920
1921                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1922                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1923
1924                                 // Shared content
1925                                 $text = self::convertShare(
1926                                         $text,
1927                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1928                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1929                                         },
1930                                         $uriid
1931                                 );
1932
1933                                 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
1934
1935                                 return $text;
1936                         }); // Escaped noparse, nobb, pre
1937
1938                         // Remove escaping tags and replace new lines that remain
1939                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1940                                 return str_replace("\n", "<br>", $match[2]);
1941                         }, $text);
1942
1943                         // Additionally, [pre] tags preserve spaces
1944                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1945                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES, 'UTF-8'));
1946                         }, $text);
1947
1948                         return $text;
1949                 }); // Escaped code
1950
1951                 $text = preg_replace_callback(
1952                         "#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1953                         function ($matches) {
1954                                 if (strpos($matches[2], "\n") !== false) {
1955                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1956                                 } else {
1957                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1958                                 }
1959
1960                                 return $return;
1961                         },
1962                         $text
1963                 );
1964
1965                 // Default iframe allowed domains/path
1966                 $allowedIframeDomains = [
1967                         DI::baseUrl()->getHost()
1968                                 . (DI::baseUrl()->getPath() ? '/' . DI::baseUrl()->getPath() : '')
1969                                 . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1970                         'www.youtube.com/embed/',
1971                         'player.vimeo.com/video/',
1972                 ];
1973
1974                 $allowedIframeDomains = array_merge(
1975                         $allowedIframeDomains,
1976                         DI::config()->get('system', 'allowed_oembed') ?
1977                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1978                                 : []
1979                 );
1980
1981                 if (strpos($text, '<p>') !== false || strpos($text, '</p>') !== false) {
1982                         $text = '<p>' . $text . '</p>';
1983                 }
1984
1985                 $text = HTML::purify($text, $allowedIframeDomains);
1986                 DI::profiler()->stopRecording();
1987
1988                 return trim($text);
1989         }
1990
1991         /**
1992          * Strips the "abstract" tag from the provided text
1993          *
1994          * @param string $text The text with BBCode
1995          * @return string The same text - but without "abstract" element
1996          */
1997         public static function stripAbstract(string $text): string
1998         {
1999                 DI::profiler()->startRecording('rendering');
2000
2001                 $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
2002                         $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
2003                         $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
2004                         return $text;
2005                 });
2006
2007                 DI::profiler()->stopRecording();
2008                 return $text;
2009         }
2010
2011         /**
2012          * Returns the value of the "abstract" element
2013          *
2014          * @param string $text  The text that maybe contains the element
2015          * @param string $addon The addon for which the abstract is meant for
2016          * @return string The abstract
2017          */
2018         public static function getAbstract(string $text, string $addon = ''): string
2019         {
2020                 DI::profiler()->startRecording('rendering');
2021                 $addon = strtolower($addon);
2022
2023                 $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
2024                         if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
2025                                 return $matches[1];
2026                         }
2027
2028                         if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
2029                                 return $matches[1];
2030                         }
2031
2032                         return '';
2033                 });
2034
2035                 DI::profiler()->stopRecording();
2036                 return $abstract;
2037         }
2038
2039         /**
2040          * Callback function to replace a Friendica style mention in a mention for Diaspora
2041          *
2042          * @param array $match Matching values for the callback
2043          *                     [1] = Mention type (! or @)
2044          *                     [2] = Name
2045          *                     [3] = Address
2046          * @return string Replaced mention
2047          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2048          * @throws \ImagickException
2049          */
2050         private static function bbCodeMention2DiasporaCallback(array $match): string
2051         {
2052                 $contact = Contact::getByURL($match[3], false, ['addr']);
2053                 if (empty($contact['addr'])) {
2054                         return $match[0];
2055                 }
2056
2057                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2058                 return $mention;
2059         }
2060
2061         /**
2062          * Converts a BBCode text into Markdown
2063          *
2064          * This function converts a BBCode item body to be sent to Markdown-enabled
2065          * systems like Diaspora and Libertree
2066          *
2067          * @param string $text
2068          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2069          * @return string
2070          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2071          */
2072         public static function toMarkdown(string $text, bool $for_diaspora = true): string
2073         {
2074                 DI::profiler()->startRecording('rendering');
2075                 $original_text = $text;
2076
2077                 // Since Diaspora is creating a summary for links, this function removes them before posting
2078                 if ($for_diaspora) {
2079                         $text = self::removeShareInformation($text);
2080                 }
2081
2082                 /**
2083                  * Transform #tags, strip off the [url] and replace spaces with underscore
2084                  */
2085                 $url_search_string = "^\[\]";
2086                 $text = preg_replace_callback(
2087                         "/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2088                         function ($matches) {
2089                                 return '#' . str_replace(' ', '_', $matches[2]);
2090                         },
2091                         $text
2092                 );
2093
2094                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2095                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2096
2097                 // Convert it to HTML - don't try oembed
2098                 if ($for_diaspora) {
2099                         $text = self::convertForUriId(0, $text, self::DIASPORA);
2100
2101                         // Add all tags that maybe were removed
2102                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2103                                 $tagline = '';
2104                                 foreach ($tags[2] as $tag) {
2105                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2106                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2107                                                 $tagline .= '#' . $tag . ' ';
2108                                         }
2109                                 }
2110                                 $text = $text . ' ' . $tagline;
2111                         }
2112                 } else {
2113                         $text = self::convertForUriId(0, $text, self::CONNECTORS);
2114                 }
2115
2116                 // If a link is followed by a quote then there should be a newline before it
2117                 // Maybe we should make this newline at every time before a quote.
2118                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2119
2120                 // Now convert HTML to Markdown
2121                 $text = HTML::toMarkdown($text);
2122
2123                 // Libertree has a problem with escaped hashtags.
2124                 $text = str_replace(['\#'], ['#'], $text);
2125
2126                 // Remove any leading or trailing whitespace, as this will mess up
2127                 // the Diaspora signature verification and cause the item to disappear
2128                 $text = trim($text);
2129
2130                 if ($for_diaspora) {
2131                         $url_search_string = "^\[\]";
2132                         $text = preg_replace_callback(
2133                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2134                                 [self::class, 'bbCodeMention2DiasporaCallback'],
2135                                 $text
2136                         );
2137                 }
2138
2139                 Hook::callAll('bb2diaspora', $text);
2140
2141                 DI::profiler()->stopRecording();
2142                 return $text;
2143         }
2144
2145         /**
2146          * Pull out all #hashtags and @person tags from $string.
2147          *
2148          * We also get @person@domain.com - which would make
2149          * the regex quite complicated as tags can also
2150          * end a sentence. So we'll run through our results
2151          * and strip the period from any tags which end with one.
2152          * Returns array of tags found, or empty array.
2153          *
2154          * @param string $string Post content
2155          * @return array List of tag and person names
2156          */
2157         public static function getTags(string $string): array
2158         {
2159                 DI::profiler()->startRecording('rendering');
2160                 $ret = [];
2161
2162                 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img', 'attachment'], function ($string) use (&$ret) {
2163                         // Convert hashtag links to hashtags
2164                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2165
2166                         // Force line feeds at bbtags
2167                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2168
2169                         // ignore anything in a bbtag
2170                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2171
2172                         // Match full names against @tags including the space between first and last
2173                         // We will look these up afterward to see if they are full names or not recognisable.
2174
2175                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2176                                 foreach ($matches[1] as $match) {
2177                                         if (strstr($match, ']')) {
2178                                                 // we might be inside a bbcode color tag - leave it alone
2179                                                 continue;
2180                                         }
2181
2182                                         if (substr($match, -1, 1) === '.') {
2183                                                 $ret[] = substr($match, 0, -1);
2184                                         } else {
2185                                                 $ret[] = $match;
2186                                         }
2187                                 }
2188                         }
2189
2190                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2191                         // and #hash tags.
2192
2193                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2194                                 foreach ($matches[1] as $match) {
2195                                         if (strstr($match, ']')) {
2196                                                 // we might be inside a bbcode color tag - leave it alone
2197                                                 continue;
2198                                         }
2199
2200                                         // try not to catch url fragments
2201                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2202                                                 continue;
2203                                         }
2204
2205                                         $ret[] = $match;
2206                                 }
2207                         }
2208                 });
2209
2210                 DI::profiler()->stopRecording();
2211                 return array_unique($ret);
2212         }
2213
2214         /**
2215          * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2216          *
2217          * @param string $body HTML/BBCode
2218          * @return string body with expanded tags
2219          */
2220         public static function expandTags(string $body): string
2221         {
2222                 return preg_replace_callback(
2223                         "/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2224                         function (array $match) {
2225                                 switch ($match[1]) {
2226                                         case '!':
2227                                         case '@':
2228                                                 $contact = Contact::getByURL($match[2]);
2229                                                 if (!empty($contact)) {
2230                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2231                                                 } else {
2232                                                         return $match[1] . $match[2];
2233                                                 }
2234                                                 break;
2235
2236                                         case '#':
2237                                         default:
2238                                                 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2239                                 }
2240                         },
2241                         $body
2242                 );
2243         }
2244
2245         /**
2246          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2247          *
2248          * @param string   $text HTML/BBCode
2249          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2250          * @param callable $callback
2251          * @return string
2252          * @see Strings::performWithEscapedBlocks
2253          */
2254         public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2255         {
2256                 $tagList = array_map('preg_quote', $tagList);
2257
2258                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2259         }
2260
2261         /**
2262          * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2263          *
2264          * @param string $body HTML/BBCode
2265          * @param int $profile_uid Profile user id
2266          * @param string $network Network name
2267          * @return string HTML/BBCode with inserted images
2268          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2269          * @throws \ImagickException
2270          */
2271         public static function setMentions(string $body, $profile_uid = 0, $network = '')
2272         {
2273                 DI::profiler()->startRecording('rendering');
2274                 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2275                         $tags = self::getTags($body);
2276
2277                         $tagged = [];
2278
2279                         foreach ($tags as $tag) {
2280                                 $tag_type = substr($tag, 0, 1);
2281
2282                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2283                                         continue;
2284                                 }
2285
2286                                 /*
2287                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2288                                  * Robert Johnson should be first in the $tags array
2289                                  */
2290                                 foreach ($tagged as $nextTag) {
2291                                         if (stristr($nextTag, $tag . ' ')) {
2292                                                 continue 2;
2293                                         }
2294                                 }
2295
2296                                 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2297                                         $tagged[] = $tag;
2298                                 }
2299                         }
2300
2301                         return $body;
2302                 });
2303
2304                 DI::profiler()->stopRecording();
2305                 return $body;
2306         }
2307
2308         /**
2309          * @param string      $author  Author display name
2310          * @param string      $profile Author profile URL
2311          * @param string      $avatar  Author profile picture URL
2312          * @param string      $link    Post source URL
2313          * @param string      $posted  Post created date
2314          * @param string|null $guid    Post guid (if any)
2315          * @param string|null $uri     Post uri (if any)
2316          * @return string
2317          * @TODO Rewrite to handle over whole record array
2318          */
2319         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null, string $uri = null): string
2320         {
2321                 DI::profiler()->startRecording('rendering');
2322                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2323                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2324                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2325                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2326                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2327
2328                 if ($guid) {
2329                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2330                 }
2331
2332                 if ($uri) {
2333                         $header .= "' message_id='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $uri);
2334                 }
2335
2336                 $header  .= "']";
2337
2338                 DI::profiler()->stopRecording();
2339                 return $header;
2340         }
2341
2342         /**
2343          * Returns the BBCode relevant to embed the provided URL in a post body.
2344          * For media type, it will return [img], [video] and [audio] tags.
2345          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2346          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2347          *
2348          * @param string      $url
2349          * @param bool        $tryAttachment
2350          * @param string|null $title
2351          * @param string|null $description
2352          * @param string|null $tags
2353          * @return string
2354          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2355          * @see ParseUrl::getSiteinfoCached
2356          */
2357         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2358         {
2359                 DI::profiler()->startRecording('rendering');
2360                 DI::logger()->info($url);
2361
2362                 // If there is already some content information submitted we don't
2363                 // need to parse the url for content.
2364                 if (!empty($title) && !empty($description)) {
2365                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2366
2367                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2368
2369                         $str_tags = '';
2370                         if (!empty($tags)) {
2371                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2372                                 if (count($arr_tags)) {
2373                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2374                                 }
2375                         }
2376
2377                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2378
2379                         DI::logger()->info('(unparsed): returns: ' . $result);
2380
2381                         DI::profiler()->stopRecording();
2382                         return $result;
2383                 }
2384
2385                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2386
2387                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2388                         switch ($siteinfo['type']) {
2389                                 case 'video':
2390                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2391                                         break;
2392                                 case 'audio':
2393                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2394                                         break;
2395                                 default:
2396                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2397                                         break;
2398                         }
2399
2400                         DI::profiler()->stopRecording();
2401                         return $bbcode;
2402                 }
2403
2404                 unset($siteinfo['keywords']);
2405
2406                 // Bypass attachment if parse url for a comment
2407                 if (!$tryAttachment) {
2408                         DI::profiler()->stopRecording();
2409                         return "\n" . '[url=' . $url . ']' . ($siteinfo['title'] ?? $url) . '[/url]';
2410                 }
2411
2412                 // Format it as BBCode attachment
2413                 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2414                 DI::profiler()->stopRecording();
2415                 return $bbcode;
2416         }
2417 }