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