src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2023, the Friendica project
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\PageInfo;
  31 use Friendica\Content\Smilies;
  32 use Friendica\Core\Hook;
  33 use Friendica\Core\Logger;
  34 use Friendica\Core\Protocol;
  35 use Friendica\Core\Renderer;
  36 use Friendica\DI;
  37 use Friendica\Model\Contact;
  38 use Friendica\Model\Event;
  39 use Friendica\Model\Post;
  40 use Friendica\Model\Tag;
  41 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
  42 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
  43 use Friendica\Util\Map;
  44 use Friendica\Util\ParseUrl;
  45 use Friendica\Util\Proxy;
  46 use Friendica\Util\Strings;
  47 use Friendica\Util\XML;
  48
  49 class BBCode
  50 {
  51         // Update this value to the current date whenever changes are made to BBCode::convert
  52         const VERSION = '2021-07-28';
  53
  54         const INTERNAL     = 0;
  55         const EXTERNAL     = 1;
  56         const MASTODON_API = 2;
  57         const DIASPORA     = 3;
  58         const CONNECTORS   = 4;
  59         const TWITTER_API  = 5;
  60         const NPF          = 6;
  61         const OSTATUS      = 7;
  62         const TWITTER      = 8;
  63         const BACKLINK     = 8;
  64         const ACTIVITYPUB  = 9;
  65         const BLUESKY      = 10;
  66
  67         const SHARED_ANCHOR = '<hr class="shared-anchor">';
  68         const TOP_ANCHOR    = '<br class="top-anchor">';
  69         const BOTTOM_ANCHOR = '<br class="button-anchor">';
  70
  71         const PREVIEW_NONE     = 0;
  72         const PREVIEW_NO_IMAGE = 1;
  73         const PREVIEW_LARGE    = 2;
  74         const PREVIEW_SMALL    = 3;
  75
  76         /**
  77          * Fetches attachment data that were generated with the "attachment" element
  78          *
  79          * @param string $body Message body
  80          * @return array
  81          *                     'type' -> Message type ('link', 'video', 'photo')
  82          *                     'text' -> Text before the shared message
  83          *                     'after' -> Text after the shared message
  84          *                     'image' -> Preview image of the message
  85          *                     'url' -> Url to the attached message
  86          *                     'title' -> Title of the attachment
  87          *                     'description' -> Description of the attachment
  88          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  89          */
  90         public static function getAttachmentData(string $body): array
  91         {
  92                 DI::profiler()->startRecording('rendering');
  93                 $data = [
  94                         'type'          => '',
  95                         'text'          => '',
  96                         'after'         => '',
  97                         'image'         => null,
  98                         'url'           => '',
  99                         'author_name'   => '',
 100                         'author_url'    => '',
 101                         'provider_name' => '',
 102                         'provider_url'  => '',
 103                         'title'         => '',
 104                         'description'   => '',
 105                 ];
 106
 107                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 108                         DI::profiler()->stopRecording();
 109                         return [];
 110                 }
 111
 112                 $attributes = $match[2];
 113
 114                 $data['text'] = trim($match[1]);
 115
 116                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 117                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 118                         $value = $matches[2] ?? '';
 119
 120                         if ($value != '') {
 121                                 switch ($field) {
 122                                         case 'publisher_name':
 123                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 124                                                 break;
 125
 126                                         case 'publisher_url':
 127                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 128                                                 break;
 129
 130                                         case 'author_name':
 131                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 132                                                 if ($data['provider_name'] == $data['author_name']) {
 133                                                         $data['author_name'] = '';
 134                                                 }
 135                                                 break;
 136
 137                                         case 'author_url':
 138                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 139                                                 if ($data['provider_url'] == $data['author_url']) {
 140                                                         $data['author_url'] = '';
 141                                                 }
 142                                                 break;
 143
 144                                         case 'title':
 145                                                 $value = self::convertForUriId(0, html_entity_decode($value, ENT_QUOTES, 'UTF-8'), BBCode::EXTERNAL);
 146                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 147                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 148                                                 $data['title'] = $value;
 149
 150                                         default:
 151                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 152                                                 break;
 153                                 }
 154                         }
 155                 }
 156
 157                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 158                         DI::profiler()->stopRecording();
 159                         return [];
 160                 }
 161
 162                 $data['description'] = trim($match[3]);
 163
 164                 $data['after'] = trim($match[4]);
 165
 166                 $parts = parse_url($data['url']);
 167                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 168                         if (empty($data['provider_name'])) {
 169                                 $data['provider_name'] = $parts['host'];
 170                         }
 171                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 172                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 173
 174                                 if (!empty($parts['port'])) {
 175                                         $data['provider_url'] .= ':' . $parts['port'];
 176                                 }
 177                         }
 178                 }
 179
 180                 DI::profiler()->stopRecording();
 181                 return $data;
 182         }
 183
 184         /**
 185          * Remove [attachment] BBCode and replaces it with a regular [url]
 186          *
 187          * @param string  $body
 188          * @param boolean $no_link_desc No link description
 189          * @return string with replaced body
 190          */
 191         public static function replaceAttachment(string $body, bool $no_link_desc = false): string
 192         {
 193                 return preg_replace_callback(
 194                         "/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
 195                         function ($match) use ($body, $no_link_desc) {
 196                                 $attach_data = self::getAttachmentData($match[0]);
 197                                 if (empty($attach_data['url'])) {
 198                                         return $match[0];
 199                                 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
 200                                         return '';
 201                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 202                                         return " \n[url]" . $attach_data['url'] . "[/url]\n";
 203                                 } else {
 204                                         return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 205                                 }
 206                         },
 207                         $body
 208                 );
 209         }
 210
 211         /**
 212          * Remove [attachment] BBCode
 213          *
 214          * @param string  $body
 215          * @return string with removed attachment
 216          */
 217         public static function removeAttachment(string $body): string
 218         {
 219                 return trim(preg_replace("/\s*\[attachment .*?\].*?\[\/attachment\]\s*/ism", '', $body));
 220         }
 221
 222         /**
 223          * Converts a BBCode text into plaintext
 224          *
 225          * @param string $text
 226          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 227          * @return string
 228          */
 229         public static function toPlaintext(string $text, bool $keep_urls = true): string
 230         {
 231                 DI::profiler()->startRecording('rendering');
 232                 // Remove pictures in advance to avoid unneeded proxy calls
 233                 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
 234                 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
 235
 236                 // Remove attachment
 237                 $text = self::replaceAttachment($text);
 238
 239                 $naked_text = HTML::toPlaintext(self::convert($text, false, BBCode::EXTERNAL, true), 0, !$keep_urls);
 240
 241                 DI::profiler()->stopRecording();
 242                 return $naked_text;
 243         }
 244
 245         private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
 246         {
 247                 // Only send proxied pictures to API and for internal display
 248                 if (!in_array($simplehtml, [self::INTERNAL, self::MASTODON_API, self::TWITTER_API])) {
 249                         return $image;
 250                 } elseif ($uriid > 0) {
 251                         return Post\Link::getByLink($uriid, $image, $size);
 252                 } else {
 253                         return Proxy::proxifyUrl($image, $size);
 254                 }
 255         }
 256
 257         /**
 258          * Truncates imported message body string length to max_import_size
 259          *
 260          * The purpose of this function is to apply system message length limits to
 261          * imported messages without including any embedded photos in the length
 262          *
 263          * @param string $body
 264          * @return string
 265          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 266          */
 267         public static function limitBodySize(string $body): string
 268         {
 269                 DI::profiler()->startRecording('rendering');
 270                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 271
 272                 // If the length of the body, including the embedded images, is smaller
 273                 // than the maximum, then don't waste time looking for the images
 274                 if ($maxlen && (strlen($body) > $maxlen)) {
 275
 276                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 277
 278                         $orig_body = $body;
 279                         $new_body = '';
 280                         $textlen = 0;
 281
 282                         $img_start = strpos($orig_body, '[img');
 283                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 284                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 285                         while (($img_st_close !== false) && ($img_end !== false)) {
 286
 287                                 $img_st_close++; // make it point to AFTER the closing bracket
 288                                 $img_end += $img_start;
 289                                 $img_end += strlen('[/img]');
 290
 291                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 292                                         // This is an embedded image
 293
 294                                         if (($textlen + $img_start) > $maxlen) {
 295                                                 if ($textlen < $maxlen) {
 296                                                         Logger::debug('the limit happens before an embedded image');
 297                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 298                                                         $textlen = $maxlen;
 299                                                 }
 300                                         } else {
 301                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 302                                                 $textlen += $img_start;
 303                                         }
 304
 305                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 306                                 } else {
 307
 308                                         if (($textlen + $img_end) > $maxlen) {
 309                                                 if ($textlen < $maxlen) {
 310                                                         Logger::debug('the limit happens before the end of a non-embedded image');
 311                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 312                                                         $textlen = $maxlen;
 313                                                 }
 314                                         } else {
 315                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 316                                                 $textlen += $img_end;
 317                                         }
 318                                 }
 319                                 $orig_body = substr($orig_body, $img_end);
 320
 321                                 if ($orig_body === false) {
 322                                         // in case the body ends on a closing image tag
 323                                         $orig_body = '';
 324                                 }
 325
 326                                 $img_start = strpos($orig_body, '[img');
 327                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 328                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 329                         }
 330
 331                         if (($textlen + strlen($orig_body)) > $maxlen) {
 332                                 if ($textlen < $maxlen) {
 333                                         Logger::debug('the limit happens after the end of the last image');
 334                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 335                                 }
 336                         } else {
 337                                 Logger::debug('the text size with embedded images extracted did not violate the limit');
 338                                 $new_body = $new_body . $orig_body;
 339                         }
 340
 341                         DI::profiler()->stopRecording();
 342                         return $new_body;
 343                 } else {
 344                         DI::profiler()->stopRecording();
 345                         return $body;
 346                 }
 347         }
 348
 349         /**
 350          * Processes [attachment] tags
 351          *
 352          * Note: Can produce a [bookmark] tag in the returned string
 353          *
 354          * @param string  $text
 355          * @param integer $simplehtml
 356          * @param bool    $tryoembed
 357          * @param array   $data
 358          * @param int     $uriid
 359          * @return string
 360          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 361          */
 362         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
 363         {
 364                 DI::profiler()->startRecording('rendering');
 365                 $data = $data ?: self::getAttachmentData($text);
 366                 if (empty($data) || empty($data['url'])) {
 367                         DI::profiler()->stopRecording();
 368                         return $text;
 369                 }
 370
 371                 if (isset($data['title'])) {
 372                         $data['title'] = strip_tags($data['title']);
 373                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 374                 } else {
 375                         $data['title'] = '';
 376                 }
 377
 378                 if (((strpos($data['text'], '[img=') !== false) || (strpos($data['text'], '[img]') !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 379                         $data['preview'] = $data['image'];
 380                         $data['image'] = '';
 381                 }
 382
 383                 $return = '';
 384                 try {
 385                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 386                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 387                         } else {
 388                                 throw new Exception('OEmbed is disabled for this attachment.');
 389                         }
 390                 } catch (Exception $e) {
 391                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 392
 393                         if ($simplehtml != self::CONNECTORS) {
 394                                 $return = sprintf('<div class="type-%s">', $data['type']);
 395                         }
 396
 397                         if ($preview_mode == self::PREVIEW_NO_IMAGE) {
 398                                 unset($data['image']);
 399                                 unset($data['preview']);
 400                         }
 401
 402                         if (!empty($data['title']) && !empty($data['url'])) {
 403                                 $preview_class = $preview_mode == self::PREVIEW_LARGE ? 'attachment-image' : 'attachment-preview';
 404                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 405                                         $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), $data['title']);
 406                                 } else {
 407                                         if (!empty($data['image'])) {
 408                                                 $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), $data['title']);
 409                                         } elseif (!empty($data['preview'])) {
 410                                                 $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']);
 411                                         }
 412                                         $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
 413                                 }
 414                         }
 415
 416                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 417                                 // Sanitize the HTML
 418                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 419                         }
 420
 421                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 422                                 if (!empty($data['author_name'])) {
 423                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 424                                 } else {
 425                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 426                                 }
 427                         }
 428
 429                         if ($simplehtml != self::CONNECTORS) {
 430                                 $return .= '</div>';
 431                         }
 432                 }
 433
 434                 DI::profiler()->stopRecording();
 435                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 436         }
 437
 438         public static function removeShareInformation(string $text, bool $plaintext = false, bool $nolink = false): string
 439         {
 440                 DI::profiler()->startRecording('rendering');
 441                 $data = self::getAttachmentData($text);
 442
 443                 if (!$data) {
 444                         DI::profiler()->stopRecording();
 445                         return $text;
 446                 } elseif ($nolink) {
 447                         DI::profiler()->stopRecording();
 448                         return $data['text'] . ($data['after'] ?? '');
 449                 }
 450
 451                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 452                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 453                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 454                         $data['title'] = $data['url'];
 455                 } elseif (($text != '') && strstr($title, $text)) {
 456                         $data['text'] = $data['title'];
 457                         $data['title'] = $data['url'];
 458                 }
 459
 460                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 461                         DI::profiler()->stopRecording();
 462                         return $data['title'] . $data['after'];
 463                 }
 464
 465                 // If the link already is included in the post, don't add it again
 466                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 467                         DI::profiler()->stopRecording();
 468                         return $data['text'] . $data['after'];
 469                 }
 470
 471                 $text = $data['text'];
 472
 473                 if (!empty($data['url']) && !empty($data['title'])) {
 474                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 475                 } elseif (!empty($data['url'])) {
 476                         $text .= "\n[url]" . $data['url'] . '[/url]';
 477                 }
 478
 479                 DI::profiler()->stopRecording();
 480                 return $text . "\n" . $data['after'];
 481         }
 482
 483         /**
 484          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 485          *
 486          * @param array $match Array with the matching values
 487          * @return string reformatted link including HTML codes
 488          */
 489         private static function convertUrlForActivityPubCallback(array $match): string
 490         {
 491                 $url = $match[1];
 492
 493                 if (isset($match[2]) && ($match[1] != $match[2])) {
 494                         return $match[0];
 495                 }
 496
 497                 $parts = parse_url($url);
 498                 if (!isset($parts['scheme'])) {
 499                         return $match[0];
 500                 }
 501
 502                 return self::convertUrlForActivityPub($url);
 503         }
 504
 505         /**
 506          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 507          *
 508          * @param string $url URL that is about to be reformatted
 509          * @return string reformatted link including HTML codes
 510          */
 511         private static function convertUrlForActivityPub(string $url): string
 512         {
 513                 return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, Strings::getStyledURL($url));
 514         }
 515
 516         /*
 517          * [noparse][i]italic[/i][/noparse] turns into
 518          * [noparse][ i ]italic[ /i ][/noparse],
 519          * to hide them from parser.
 520          *
 521          * @param array $match
 522          * @return string
 523          */
 524         private static function escapeNoparseCallback(array $match): string
 525         {
 526                 $whole_match = $match[0];
 527                 $captured = $match[1];
 528                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 529                 $new_str = str_replace($captured, $spacefied, $whole_match);
 530                 return $new_str;
 531         }
 532
 533         /*
 534          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 535          * now turns back and the [noparse] tags are trimmed
 536          * returning [i]italic[/i]
 537          *
 538          * @param array $match
 539          * @return string
 540          */
 541         private static function unescapeNoparseCallback(array $match): string
 542         {
 543                 $captured = $match[1];
 544                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 545                 return $unspacefied;
 546         }
 547
 548         /**
 549          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 550          * occurrences
 551          *
 552          * @param string $text        Text to search
 553          * @param string $name        Tag name
 554          * @param int    $occurrences Number of first occurrences to skip
 555          * @return boolean|array
 556          */
 557         public static function getTagPosition(string $text, string $name, int $occurrences = 0)
 558         {
 559                 DI::profiler()->startRecording('rendering');
 560                 if ($occurrences < 0) {
 561                         $occurrences = 0;
 562                 }
 563
 564                 $start_open = -1;
 565                 for ($i = 0; $i <= $occurrences; $i++) {
 566                         if ($start_open !== false) {
 567                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 568                         }
 569                 }
 570
 571                 if ($start_open === false) {
 572                         DI::profiler()->stopRecording();
 573                         return false;
 574                 }
 575
 576                 $start_equal = strpos($text, '=', $start_open);
 577                 $start_close = strpos($text, ']', $start_open);
 578
 579                 if ($start_close === false) {
 580                         DI::profiler()->stopRecording();
 581                         return false;
 582                 }
 583
 584                 $start_close++;
 585
 586                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 587
 588                 if ($end_open === false) {
 589                         DI::profiler()->stopRecording();
 590                         return false;
 591                 }
 592
 593                 $res = [
 594                         'start' => [
 595                                 'open' => $start_open,
 596                                 'close' => $start_close
 597                         ],
 598                         'end' => [
 599                                 'open' => $end_open,
 600                                 'close' => $end_open + strlen('[/' . $name . ']')
 601                         ],
 602                 ];
 603
 604                 if ($start_equal !== false) {
 605                         $res['start']['equal'] = $start_equal + 1;
 606                 }
 607
 608                 DI::profiler()->stopRecording();
 609                 return $res;
 610         }
 611
 612         /**
 613          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 614          *
 615          * @param string $pattern Preg pattern string
 616          * @param string $replace Preg replace string
 617          * @param string $name    BBCode tag name
 618          * @param string $text    Text to search
 619          * @return string
 620          */
 621         public static function pregReplaceInTag(string $pattern, string $replace, string $name, string $text): string
 622         {
 623                 DI::profiler()->startRecording('rendering');
 624                 $occurrences = 0;
 625                 $pos = self::getTagPosition($text, $name, $occurrences);
 626                 while ($pos !== false && $occurrences++ < 1000) {
 627                         $start = substr($text, 0, $pos['start']['open']);
 628                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 629                         $end = substr($text, $pos['end']['close']);
 630                         if ($end === false) {
 631                                 $end = '';
 632                         }
 633
 634                         $subject = preg_replace($pattern, $replace, $subject);
 635                         $text = $start . $subject . $end;
 636
 637                         $pos = self::getTagPosition($text, $name, $occurrences);
 638                 }
 639
 640                 DI::profiler()->stopRecording();
 641                 return $text;
 642         }
 643
 644         private static function extractImagesFromItemBody(string $body): array
 645         {
 646                 $saved_image = [];
 647                 $orig_body = $body;
 648                 $new_body = '';
 649
 650                 $cnt = 0;
 651                 $img_start = strpos($orig_body, '[img');
 652                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 653                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 654                 while (($img_st_close !== false) && ($img_end !== false)) {
 655                         $img_st_close++; // make it point to AFTER the closing bracket
 656                         $img_end += $img_start;
 657
 658                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 659                                 // This is an embedded image
 660                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 661                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 662
 663                                 $cnt++;
 664                         } else {
 665                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 666                         }
 667
 668                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 669
 670                         if ($orig_body === false) {
 671                                 // in case the body ends on a closing image tag
 672                                 $orig_body = '';
 673                         }
 674
 675                         $img_start = strpos($orig_body, '[img');
 676                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 677                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 678                 }
 679
 680                 $new_body = $new_body . $orig_body;
 681
 682                 return ['body' => $new_body, 'images' => $saved_image];
 683         }
 684
 685         private static function interpolateSavedImagesIntoItemBody(int $uriid, string $body, array $images): string
 686         {
 687                 $newbody = $body;
 688
 689                 $cnt = 0;
 690                 foreach ($images as $image) {
 691                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 692                         // it loops over the array starting from the first element and going sequentially
 693                         // to the last element
 694                         $newbody = str_replace(
 695                                 '[$#saved_image' . $cnt . '#$]',
 696                                 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />',
 697                                 $newbody
 698                         );
 699                         $cnt++;
 700                 }
 701
 702                 return $newbody;
 703         }
 704
 705         /**
 706          * @param string $text A BBCode string
 707          * @return array Empty array if no share tag is present or the following array, missing attributes end up empty strings:
 708          *               - comment   : Text before the opening share tag
 709          *               - shared    : Text inside the share tags
 710          *               - author    : (Optional) Display name of the shared author
 711          *               - profile   : (Optional) Profile page URL of the shared author
 712          *               - avatar    : (Optional) Profile picture URL of the shared author
 713          *               - link      : (Optional) Canonical URL of the shared post
 714          *               - posted    : (Optional) Date the shared post was initially posted ("Y-m-d H:i:s" in GMT)
 715          *               - message_id: (Optional) Shared post URI if any
 716          *               - guid      : (Optional) Shared post GUID if any
 717          */
 718         public static function fetchShareAttributes(string $text): array
 719         {
 720                 DI::profiler()->startRecording('rendering');
 721                 if (preg_match('~(.*?)\[share](.*)\[/share]~ism', $text, $matches)) {
 722                         DI::profiler()->stopRecording();
 723                         return [
 724                                 'author'     => '',
 725                                 'profile'    => '',
 726                                 'avatar'     => '',
 727                                 'link'       => '',
 728                                 'posted'     => '',
 729                                 'guid'       => '',
 730                                 'message_id' => trim($matches[2]),
 731                                 'comment'    => trim($matches[1]),
 732                                 'shared'     => '',
 733                         ];
 734                 }
 735                 // See Issue https://github.com/friendica/friendica/issues/10454
 736                 // Hashtags in usernames are expanded to links. This here is a quick fix.
 737                 $text = preg_replace('~([@!#])\[url=.*?](.*?)\[/url]~ism', '$1$2', $text);
 738
 739                 if (!preg_match('~(.*?)\[share(.*?)](.*)\[/share]~ism', $text, $matches)) {
 740                         DI::profiler()->stopRecording();
 741                         return [];
 742                 }
 743
 744                 $attributes = self::extractShareAttributes($matches[2]);
 745
 746                 $attributes['comment'] = trim($matches[1]);
 747                 $attributes['shared'] = trim($matches[3]);
 748
 749                 DI::profiler()->stopRecording();
 750                 return $attributes;
 751         }
 752
 753         /**
 754          * @see BBCode::fetchShareAttributes()
 755          * @param string $shareString Internal opening share tag string matched by the regular expression
 756          * @return array A fixed attribute array where missing attribute are represented by empty strings
 757          */
 758         private static function extractShareAttributes(string $shareString): array
 759         {
 760                 $attributes = [];
 761                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid', 'message_id'] as $field) {
 762                         preg_match("/$field=(['\"])(.+?)\\1/ism", $shareString, $matches);
 763                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 764                 }
 765
 766                 return $attributes;
 767         }
 768
 769         /**
 770          * Remove the share block
 771          *
 772          * @param string $body
 773          * @return string
 774          */
 775         public static function removeSharedData(string $body): string
 776         {
 777                 return trim(preg_replace("/\s*\[share.*?\].*?\[\/share\]\s*/ism", '', $body));
 778         }
 779
 780         /**
 781          * This function converts a [share] block to text according to a provided callback function whose signature is:
 782          *
 783          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 784          *
 785          * Where:
 786          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 787          * data lookup
 788          * - $author_contact is a contact record array
 789          * - $content is the inner content of the [share] block
 790          * - $is_quote_share indicates whether there's any content before the [share] block
 791          * - Return value is the string that should replace the [share] block in the provided text
 792          *
 793          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 794          *
 795          * @param  string   $text     A BBCode string
 796          * @param  callable $callback
 797          * @return string The BBCode string with all [share] blocks replaced
 798          */
 799         public static function convertShare(string $text, callable $callback, int $uriid = 0): string
 800         {
 801                 DI::profiler()->startRecording('rendering');
 802                 $return = preg_replace_callback(
 803                         '~(.*?)\[share(.*?)](.*)\[/share]~ism',
 804                         function ($match) use ($callback, $uriid) {
 805                                 $attributes = self::extractShareAttributes($match[2]);
 806
 807                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
 808                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
 809                                 $author_contact['addr'] = ($author_contact['addr'] ?? '');
 810
 811                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 812                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 813                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 814
 815                                 if (!empty($author_contact['id'])) {
 816                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
 817                                 } elseif ($attributes['avatar']) {
 818                                         $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
 819                                 }
 820
 821                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
 822
 823                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
 824                         },
 825                         $text
 826                 );
 827
 828                 DI::profiler()->stopRecording();
 829                 return trim($return);
 830         }
 831
 832         /**
 833          * Convert complex IMG and ZMG elements
 834          *
 835          * @param [type] $text
 836          * @param integer $simplehtml
 837          * @param integer $uriid
 838          * @return string
 839          */
 840         private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
 841         {
 842                 DI::profiler()->startRecording('rendering');
 843                 $return = preg_replace_callback(
 844                         "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
 845                         function ($match) use ($simplehtml, $uriid) {
 846                                 $attribute_string = $match[1];
 847                                 $attributes = [];
 848                                 foreach (['alt', 'width', 'height'] as $field) {
 849                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 850                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 851                                 }
 852
 853                                 $img_str = '<img src="' . self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
 854                                 foreach ($attributes as $key => $value) {
 855                                         if (!empty($value)) {
 856                                                 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
 857                                         }
 858                                 }
 859                                 return $img_str . '>';
 860                         },
 861                         $text
 862                 );
 863
 864                 DI::profiler()->stopRecording();
 865                 return $return;
 866         }
 867
 868         /**
 869          * Default [share] tag conversion callback
 870          *
 871          * Note: Can produce a [bookmark] tag in the output
 872          *
 873          * @see BBCode::convertShare()
 874          * @param array   $attributes     [share] block attribute values
 875          * @param array   $author_contact Contact row of the shared author
 876          * @param string  $content        Inner content of the [share] block
 877          * @param boolean $is_quote_share Whether there is content before the [share] block
 878          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 879          * @return string
 880          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 881          */
 882         private static function convertShareCallback(array $attributes, array $author_contact, string $content, bool $is_quote_share, int $simplehtml): string
 883         {
 884                 DI::profiler()->startRecording('rendering');
 885                 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
 886
 887                 switch ($simplehtml) {
 888                         case self::MASTODON_API:
 889                         case self::TWITTER_API:
 890                                 $text = ($is_quote_share ? '<br>' : '') .
 891                                         '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
 892                                         '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
 893                                 break;
 894                         case self::DIASPORA:
 895                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 896                                         $text = ($is_quote_share ? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 897                                 } else {
 898                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
 899
 900                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
 901                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
 902                                         }
 903
 904                                         $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 905
 906                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
 907                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
 908                                         }
 909                                 }
 910
 911                                 break;
 912                         case self::CONNECTORS:
 913                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 914                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 915                                 $headline .= ':</b></p>' . "\n";
 916
 917                                 $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
 918
 919                                 break;
 920                         case self::OSTATUS:
 921                                 $text = ($is_quote_share ? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 922                                 break;
 923                         case self::ACTIVITYPUB:
 924                                 $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>';
 925                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
 926                                 break;
 927                         default:
 928                                 $text = ($is_quote_share ? "\n" : '');
 929
 930                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
 931                                 $network = $contact['network'] ?? Protocol::PHANTOM;
 932
 933                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
 934                                 $text .= BBCode::SHARED_ANCHOR . Renderer::replaceMacros($tpl, [
 935                                         '$profile'      => $attributes['profile'],
 936                                         '$avatar'       => $attributes['avatar'],
 937                                         '$author'       => $attributes['author'],
 938                                         '$link'         => $attributes['link'],
 939                                         '$link_title'   => DI::l10n()->t('Link to source'),
 940                                         '$posted'       => $attributes['posted'],
 941                                         '$guid'         => $attributes['guid'],
 942                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
 943                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
 944                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
 945                                 ]);
 946                                 break;
 947                 }
 948
 949                 return $text;
 950         }
 951
 952         private static function removePictureLinksCallback(array $match): string
 953         {
 954                 $cache_key = 'remove:' . $match[1];
 955                 $text = DI::cache()->get($cache_key);
 956
 957                 if (is_null($text)) {
 958                         $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
 959                         if ($curlResult->isSuccess()) {
 960                                 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
 961                         } else {
 962                                 $mimetype = '';
 963                         }
 964
 965                         if (substr($mimetype, 0, 6) == 'image/') {
 966                                 $text = '[url=' . $match[1] . ']' . $match[1] . '[/url]';
 967                         } else {
 968                                 $text = '[url=' . $match[2] . ']' . $match[2] . '[/url]';
 969
 970                                 // if its not a picture then look if its a page that contains a picture link
 971                                 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
 972                                 if (empty($body)) {
 973                                         DI::cache()->set($cache_key, $text);
 974                                         return $text;
 975                                 }
 976
 977                                 $doc = new DOMDocument();
 978                                 @$doc->loadHTML($body);
 979                                 $xpath = new DOMXPath($doc);
 980                                 $list = $xpath->query('//meta[@name]');
 981                                 foreach ($list as $node) {
 982                                         $attr = [];
 983
 984                                         if ($node->attributes->length) {
 985                                                 foreach ($node->attributes as $attribute) {
 986                                                         $attr[$attribute->name] = $attribute->value;
 987                                                 }
 988                                         }
 989
 990                                         if (strtolower($attr['name']) == 'twitter:image') {
 991                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
 992                                         }
 993                                 }
 994                         }
 995                         DI::cache()->set($cache_key, $text);
 996                 }
 997
 998                 return $text;
 999         }
1000
1001         /**
1002          * Callback: Expands links from given $match array
1003          *
1004          * @param array $match Array with link match
1005          * @return string BBCode
1006          */
1007         private static function expandLinksCallback(array $match): string
1008         {
1009                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1010                         return ($match[1] . '[url]' . $match[2] . '[/url]');
1011                 } else {
1012                         return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1013                 }
1014         }
1015
1016         /**
1017          * Callback: Cleans picture links
1018          *
1019          * @param array $match Array with link match
1020          * @return string BBCode
1021          */
1022         private static function cleanPictureLinksCallback(array $match): string
1023         {
1024                 // When the picture link is the own photo path then we can avoid fetching the link
1025                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()) . '/photos/');
1026                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1027                         if (!empty($match[3])) {
1028                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1029                         } else {
1030                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1031                         }
1032                         return $text;
1033                 }
1034
1035                 $cache_key = 'clean:' . $match[1];
1036                 $text = DI::cache()->get($cache_key);
1037                 if (!is_null($text)) {
1038                         return $text;
1039                 }
1040
1041                 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1042                 if ($curlResult->isSuccess()) {
1043                         $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1044                 } else {
1045                         $mimetype = '';
1046                 }
1047
1048                 // if its a link to a picture then embed this picture
1049                 if (substr($mimetype, 0, 6) == 'image/') {
1050                         $text = '[img]' . $match[1] . '[/img]';
1051                 } else {
1052                         if (!empty($match[3])) {
1053                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1054                         } else {
1055                                 $text = '[img]' . $match[2] . '[/img]';
1056                         }
1057
1058                         // if its not a picture then look if its a page that contains a picture link
1059                         $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1060                         if (empty($body)) {
1061                                 DI::cache()->set($cache_key, $text);
1062                                 return $text;
1063                         }
1064
1065                         $doc = new DOMDocument();
1066                         @$doc->loadHTML($body);
1067                         $xpath = new DOMXPath($doc);
1068                         $list = $xpath->query('//meta[@name]');
1069                         foreach ($list as $node) {
1070                                 $attr = [];
1071                                 if ($node->attributes->length) {
1072                                         foreach ($node->attributes as $attribute) {
1073                                                 $attr[$attribute->name] = $attribute->value;
1074                                         }
1075                                 }
1076
1077                                 if (strtolower($attr['name']) == "twitter:image") {
1078                                         if (!empty($match[3])) {
1079                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1080                                         } else {
1081                                                 $text = "[img]" . $attr['content'] . "[/img]";
1082                                         }
1083                                 }
1084                         }
1085                 }
1086                 DI::cache()->set($cache_key, $text);
1087
1088                 return $text;
1089         }
1090
1091         /**
1092          * Cleans picture links
1093          *
1094          * @param string $text HTML/BBCode string
1095          * @return string Cleaned HTML/BBCode
1096          */
1097         public static function cleanPictureLinks(string $text): string
1098         {
1099                 DI::profiler()->startRecording('rendering');
1100                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $text);
1101                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $return);
1102                 DI::profiler()->stopRecording();
1103                 return $return;
1104         }
1105
1106         /**
1107          * Removes links
1108          *
1109          * @param string $text HTML/BBCode string
1110          * @return string Cleaned HTML/BBCode
1111          */
1112         public static function removeLinks(string $bbcode): string
1113         {
1114                 DI::profiler()->startRecording('rendering');
1115                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1116                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1117
1118                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1119                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1120                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1121                 DI::profiler()->stopRecording();
1122                 return $bbcode;
1123         }
1124
1125         /**
1126          * Replace names in mentions with nicknames
1127          *
1128          * @param string $body HTML/BBCode
1129          * @return string Body with replaced mentions
1130          */
1131         public static function setMentionsToNicknames(string $body): string
1132         {
1133                 DI::profiler()->startRecording('rendering');
1134                 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1135                 $body = preg_replace_callback($regexp, [self::class, 'mentionCallback'], $body);
1136                 DI::profiler()->stopRecording();
1137                 return $body;
1138         }
1139
1140         /**
1141          * Callback function to replace a Friendica style mention in a mention with the nickname
1142          *
1143          * @param array $match Matching values for the callback
1144          * @return string Replaced mention or empty string
1145          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1146          */
1147         private static function mentionCallback(array $match): string
1148         {
1149                 if (empty($match[2])) {
1150                         return '';
1151                 }
1152
1153                 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1154                 if (empty($data['nick'])) {
1155                         return $match[0];
1156                 }
1157
1158                 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1159         }
1160
1161         /**
1162          * Converts a BBCode message for a given URI-ID to a HTML message
1163          *
1164          * BBcode 2 HTML was written by WAY2WEB.net
1165          * extended to work with Mistpark/Friendica - Mike Macgirvin
1166          *
1167          * Simple HTML values meaning:
1168          * - 0: Friendica display
1169          * - 1: Unused
1170          * - 2: Used for Windows Phone push, Friendica API
1171          * - 3: Used before converting to Markdown in bb2diaspora.php
1172          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1173          * - 5: Unused
1174          * - 6: Unused
1175          * - 7: Used for dfrn, OStatus
1176          * - 8: Used for Twitter, WP backlink text setting
1177          * - 9: ActivityPub
1178          *
1179          * @param int    $uriid
1180          * @param string $text
1181          * @param int    $simple_html
1182          * @return string
1183          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1184          */
1185         public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1186         {
1187                 $try_oembed = ($simple_html == self::INTERNAL);
1188
1189                 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1190         }
1191
1192         /**
1193          * Converts a BBCode message to HTML message
1194          *
1195          * BBcode 2 HTML was written by WAY2WEB.net
1196          * extended to work with Mistpark/Friendica - Mike Macgirvin
1197          *
1198          * Simple HTML values meaning:
1199          * - 0: Friendica display
1200          * - 1: Unused
1201          * - 2: Used for Windows Phone push, Friendica API
1202          * - 3: Used before converting to Markdown in bb2diaspora.php
1203          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1204          * - 5: Unused
1205          * - 6: Unused
1206          * - 7: Used for dfrn, OStatus
1207          * - 8: Used for Twitter, WP backlink text setting
1208          * - 9: ActivityPub
1209          *
1210          * @param string $text
1211          * @param bool   $try_oembed
1212          * @param int    $simple_html
1213          * @param bool   $for_plaintext
1214          * @param int    $uriid
1215          * @return string Converted code or empty string
1216          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1217          */
1218         public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1219         {
1220                 // Accounting for null default column values
1221                 if (is_null($text) || $text === '') {
1222                         return '';
1223                 }
1224
1225                 DI::profiler()->startRecording('rendering');
1226
1227                 Hook::callAll('bbcode', $text);
1228
1229                 $a = DI::app();
1230
1231                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1232                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1233                                 /*
1234                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1235                                  *
1236                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1237                                  * $match[1] = $url
1238                                  * $match[2] = $title or absent
1239                                  */
1240                                 $try_oembed_callback = function (array $match) {
1241                                         $url = $match[1];
1242                                         $title = $match[2] ?? '';
1243
1244                                         try {
1245                                                 $return = OEmbed::getHTML($url, $title);
1246                                         } catch (Exception $ex) {
1247                                                 $return = $match[0];
1248                                         }
1249
1250                                         return $return;
1251                                 };
1252
1253                                 // Remove the abstract element. It is a non visible element.
1254                                 $text = self::stripAbstract($text);
1255
1256                                 // Line ending normalisation
1257                                 $text = str_replace("\r\n", "\n", $text);
1258
1259                                 // Move new lines outside of tags
1260                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1261                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1262
1263                                 // Extract the private images which use data urls since preg has issues with
1264                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1265                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1266
1267                                 $extracted = self::extractImagesFromItemBody($text);
1268                                 $text = $extracted['body'];
1269                                 $saved_image = $extracted['images'];
1270
1271                                 // If we find any event code, turn it into an event.
1272                                 // After we're finished processing the bbcode we'll
1273                                 // replace all of the event code with a reformatted version.
1274
1275                                 $ev = Event::fromBBCode($text);
1276
1277                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1278                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1279
1280                                 $text = str_replace("<", "&lt;", $text);
1281                                 $text = str_replace(">", "&gt;", $text);
1282
1283                                 // remove some newlines before the general conversion
1284                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1]$2[/share]\n", $text);
1285                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "\n[quote$1]$2[/quote]\n", $text);
1286
1287                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1288                                 if (!$try_oembed) {
1289                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1290                                 }
1291
1292                                 // Remove linefeeds inside of the table elements. See issue #6799
1293                                 $search = [
1294                                         "\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1295                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1296                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1297                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1298                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1299                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] ",
1300                                         " \n", "\t\n", "[/li]\n", "\n[li]", "\n[*]",
1301                                 ];
1302                                 $replace = [
1303                                         "[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1304                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1305                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1306                                         "[hr]", "[hr]", "[hr]", "[hr]",
1307                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1308                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]",
1309                                         "\n", "\n", "[/li]", "[li]", "[*]",
1310                                 ];
1311                                 do {
1312                                         $oldtext = $text;
1313                                         $text = str_replace($search, $replace, $text);
1314                                 } while ($oldtext != $text);
1315
1316                                 // Replace these here only once
1317                                 $search = ["\n[table]", "[/table]\n"];
1318                                 $replace = ["[table]", "[/table]"];
1319                                 $text = str_replace($search, $replace, $text);
1320
1321                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1322                                 $text = trim($text, "\n");
1323
1324                                 // removing multiplicated newlines
1325                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1326                                         $search = [
1327                                                 "\n\n\n", "[/quote]\n\n", "\n[/quote]", "\n[ul]", "[/ul]\n", "\n[ol]", "[/ol]\n", "\n\n[share ", "[/attachment]\n",
1328                                                 "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"
1329                                         ];
1330                                         $replace = [
1331                                                 "\n\n", "[/quote]\n", "[/quote]", "[ul]", "[/ul]", "[ol]", "[/ol]", "\n[share ", "[/attachment]",
1332                                                 "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"
1333                                         ];
1334                                         do {
1335                                                 $oldtext = $text;
1336                                                 $text = str_replace($search, $replace, $text);
1337                                         } while ($oldtext != $text);
1338                                 }
1339
1340                                 /// @todo Have a closer look at the different html modes
1341                                 // Handle attached links or videos
1342                                 if ($simple_html == self::NPF) {
1343                                         $text = self::removeAttachment($text);
1344                                 } elseif (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1345                                         $text = self::replaceAttachment($text);
1346                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1347                                         $text = self::replaceAttachment($text, true);
1348                                 } else {
1349                                         $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1350                                 }
1351
1352                                 $nosmile = strpos($text, '[nosmile]') !== false;
1353                                 $text = str_replace('[nosmile]', '', $text);
1354
1355                                 // Replace non graphical smilies for external posts
1356                                 if (!$nosmile) {
1357                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1358                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1359                                         });
1360                                 }
1361
1362                                 // leave open the possibility of [map=something]
1363                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1364                                 if (strpos($text, '[/map]') !== false) {
1365                                         $text = preg_replace_callback(
1366                                                 "/\[map\](.*?)\[\/map\]/ism",
1367                                                 function ($match) use ($simple_html) {
1368                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1369                                                 },
1370                                                 $text
1371                                         );
1372                                 }
1373
1374                                 if (strpos($text, '[map=') !== false) {
1375                                         $text = preg_replace_callback(
1376                                                 "/\[map=(.*?)\]/ism",
1377                                                 function ($match) use ($simple_html) {
1378                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1379                                                 },
1380                                                 $text
1381                                         );
1382                                 }
1383
1384                                 if (strpos($text, '[map]') !== false) {
1385                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1386                                 }
1387
1388                                 // Check for headers
1389
1390                                 if ($simple_html == self::INTERNAL) {
1391                                         //Ensure to always start with <h4> if possible
1392                                         $heading_count = 0;
1393                                         for ($level = 6; $level > 0; $level--) {
1394                                                 if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1395                                                         $heading_count++;
1396                                                 }
1397                                         }
1398                                         if ($heading_count > 0) {
1399                                                 $heading = min($heading_count + 3, 6);
1400                                                 for ($level = 6; $level > 0; $level--) {
1401                                                         if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1402                                                                 $text = preg_replace("(\[h$level\](.*?)\[\/h$level\])ism", "</p><h$heading>$1</h$heading><p>", $text);
1403                                                                 $heading--;
1404                                                         }
1405                                                 }
1406                                         }
1407                                 } else {
1408                                         $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '</p><h1>$1</h1><p>', $text);
1409                                         $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '</p><h2>$1</h2><p>', $text);
1410                                         $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '</p><h3>$1</h3><p>', $text);
1411                                         $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '</p><h4>$1</h4><p>', $text);
1412                                         $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '</p><h5>$1</h5><p>', $text);
1413                                         $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '</p><h6>$1</h6><p>', $text);
1414                                 }
1415
1416                                 // Check for paragraph
1417                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1418
1419                                 // Check for bold text
1420                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1421
1422                                 // Check for Italics text
1423                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1424
1425                                 // Check for Underline text
1426                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1427
1428                                 // Check for strike-through text
1429                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1430
1431                                 // Check for over-line text
1432                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1433
1434                                 // Check for colored text
1435                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1436
1437                                 // Check for sized text
1438                                 // [size=50] --> font-size: 50px (with the unit).
1439                                 if ($simple_html != self::DIASPORA) {
1440                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1441                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1442                                 } else {
1443                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1444                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1445                                 }
1446
1447
1448                                 // Check for centered text
1449                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1450
1451                                 // Check for list text
1452                                 $text = str_replace("[*]", "<li>", $text);
1453
1454                                 // Check for block-level custom CSS
1455                                 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1456
1457                                 // Check for inline custom CSS
1458                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1459
1460                                 // Mastodon Emoji (internal tag, do not document for users)
1461                                 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1462
1463                                 // Check for CSS classes
1464                                 // @deprecated since 2021.12, left for backward-compatibility reasons
1465                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1466                                 // Add HTML new lines
1467                                 $text = str_replace("\n\n", '</p><p>', $text);
1468                                 $text = str_replace("\n", '<br>', $text);
1469
1470                                 // handle nested lists
1471                                 $endlessloop = 0;
1472
1473                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1474                                         ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1475                                         ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1476                                         ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1477                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '</p><ul class="listbullet" style="list-style-type: circle;">$1</ul><p>', $text);
1478                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '</p><ul class="listnone" style="list-style-type: none;">$1</ul><p>', $text);
1479                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '</p><ul class="listdecimal" style="list-style-type: decimal;">$1</ul><p>', $text);
1480                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '</p><ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul><p>', $text);
1481                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '</p><ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul><p>', $text);
1482                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '</p><ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul><p>', $text);
1483                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '</p><ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul><p>', $text);
1484                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '</p><ul>$1</ul><p>', $text);
1485                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '</p><ol>$1</ol><p>', $text);
1486                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1487                                 }
1488
1489                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1490                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1491                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1492                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '</p><table>$1</table><p>', $text);
1493
1494                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '</p><table border="1" >$1</table><p>', $text);
1495                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '</p><table border="0" >$1</table><p>', $text);
1496
1497                                 $text = str_replace('[hr]', '</p><hr /><p>', $text);
1498
1499                                 if (!$for_plaintext) {
1500                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1501                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1502                                         });
1503                                 }
1504
1505                                 // Check for font change text
1506                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1507
1508                                 // Declare the format for [spoiler] layout
1509                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1510
1511                                 // Check for [spoiler] text
1512                                 // handle nested quotes
1513                                 $endlessloop = 0;
1514                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1515                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1516                                 }
1517
1518                                 // Check for [spoiler=Title] text
1519
1520                                 // handle nested quotes
1521                                 $endlessloop = 0;
1522                                 while ((strpos($text, "[/spoiler]") !== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1523                                         $text = preg_replace(
1524                                                 "/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1525                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1526                                                 $text
1527                                         );
1528                                 }
1529
1530                                 // Declare the format for [quote] layout
1531                                 $QuoteLayout = '</p><blockquote>$1</blockquote><p>';
1532
1533                                 // Check for [quote] text
1534                                 // handle nested quotes
1535                                 $endlessloop = 0;
1536                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1537                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1538                                 }
1539
1540                                 // Check for [quote=Author] text
1541
1542                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1543
1544                                 // handle nested quotes
1545                                 $endlessloop = 0;
1546                                 while ((strpos($text, "[/quote]") !== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1547                                         $text = preg_replace(
1548                                                 "/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1549                                                 "<p><strong class=" . '"author"' . ">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1550                                                 $text
1551                                         );
1552                                 }
1553
1554
1555                                 // [img=widthxheight]image source[/img]
1556                                 $text = preg_replace_callback(
1557                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1558                                         function ($matches) use ($simple_html, $uriid) {
1559                                                 if (strpos($matches[3], "data:image/") === 0) {
1560                                                         return $matches[0];
1561                                                 }
1562
1563                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1564                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1565                                         },
1566                                         $text
1567                                 );
1568
1569                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1570                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1571
1572                                 $text = preg_replace_callback(
1573                                         "/\[[iz]mg\=(.*?)\](.*?)\[\/[iz]mg\]/ism",
1574                                         function ($matches) use ($simple_html, $uriid) {
1575                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1576                                                 $alt = htmlspecialchars($matches[2], ENT_COMPAT);
1577                                                 // Fix for Markdown problems with Diaspora, see issue #12701
1578                                                 if (($simple_html != self::DIASPORA) || strpos($matches[2], '"') === false) {
1579                                                         return '<img src="' . $matches[1] . '" alt="' . $alt . '" title="' . $alt . '">';
1580                                                 } else {
1581                                                         return '<img src="' . $matches[1] . '" alt="' . $alt . '">';
1582                                                 }
1583                                         },
1584                                         $text
1585                                 );
1586
1587                                 // Images
1588                                 // [img]pathtoimage[/img]
1589                                 $text = preg_replace_callback(
1590                                         "/\[[iz]mg\](.*?)\[\/[iz]mg\]/ism",
1591                                         function ($matches) use ($simple_html, $uriid) {
1592                                                 if (strpos($matches[1], "data:image/") === 0) {
1593                                                         return $matches[0];
1594                                                 }
1595
1596                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1597                                                 return "[img]" . $matches[1] . "[/img]";
1598                                         },
1599                                         $text
1600                                 );
1601
1602                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1603                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1604
1605                                 $text = self::convertImages($text, $simple_html, $uriid);
1606
1607                                 $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);
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>', $text);
1609                                 //$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);
1610
1611                                 // Simplify "video" element
1612                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1613
1614                                 if ($simple_html == self::NPF) {
1615                                         $text = preg_replace(
1616                                                 "/\[video\](.*?)\[\/video\]/ism",
1617                                                 '</p><video src="$1" controls width="100%" height="auto">$1</video><p>',
1618                                                 $text
1619                                         );
1620                                         $text = preg_replace(
1621                                                 "/\[audio\](.*?)\[\/audio\]/ism",
1622                                                 '</p><audio src="$1" controls>$1">$1</audio><p>',
1623                                                 $text
1624                                         );
1625                                 } elseif ($try_oembed) {
1626                                         // html5 video and audio
1627                                         $text = preg_replace(
1628                                                 "/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1629                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>',
1630                                                 $text
1631                                         );
1632
1633                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1634                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1635
1636                                         $text = preg_replace(
1637                                                 "/\[video\](.*?)\[\/video\]/ism",
1638                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1639                                                 $text
1640                                         );
1641                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1642                                 } else {
1643                                         $text = preg_replace(
1644                                                 "/\[video\](.*?)\[\/video\]/ism",
1645                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1646                                                 $text
1647                                         );
1648                                         $text = preg_replace(
1649                                                 "/\[audio\](.*?)\[\/audio\]/ism",
1650                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1651                                                 $text
1652                                         );
1653                                 }
1654
1655                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1656                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1657
1658                                 // Youtube extensions
1659                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1660                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1661                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1662                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1663
1664                                 if ($try_oembed) {
1665                                         $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);
1666                                 } else {
1667                                         $text = preg_replace(
1668                                                 "/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1669                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>',
1670                                                 $text
1671                                         );
1672                                 }
1673
1674                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1675                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1676
1677                                 if ($try_oembed) {
1678                                         $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);
1679                                 } else {
1680                                         $text = preg_replace(
1681                                                 "/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1682                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>',
1683                                                 $text
1684                                         );
1685                                 }
1686
1687                                 // oembed tag
1688                                 $text = OEmbed::BBCode2HTML($text);
1689
1690                                 // Avoid triple linefeeds through oembed
1691                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1692
1693                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1694                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1695                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1696                                 // start which is always required). Allow desc with a missing summary for compatibility.
1697
1698                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1699                                         $sub = Event::getHTML($ev, $simple_html, $uriid);
1700
1701                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1702                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1703                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1704                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1705                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1706                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1707                                 }
1708
1709                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1710                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1711                                         // Emojis are always 4 byte Unicode characters
1712                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1713                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1714                                         }
1715                                 }
1716
1717                                 // Handle mentions and hashtag links
1718                                 if ($simple_html == self::DIASPORA) {
1719                                         // The ! is converted to @ since Diaspora only understands the @
1720                                         $text = preg_replace(
1721                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1722                                                 '@<a href="$2">$3</a>',
1723                                                 $text
1724                                         );
1725                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1726                                         $text = preg_replace(
1727                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1728                                                 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1729                                                 $text
1730                                         );
1731                                         $text = preg_replace(
1732                                                 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1733                                                 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1734                                                 $text
1735                                         );
1736                                 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::TWITTER_API])) {
1737                                         $text = preg_replace(
1738                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1739                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1740                                                 $text
1741                                         );
1742                                 } elseif ($simple_html == self::MASTODON_API) {
1743                                         $text = preg_replace(
1744                                                 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1745                                                 '<a class="u-url mention status-link" href="$2" rel="nofollow noopener noreferrer" target="_blank" title="$3">$1<span>$3</span></a>',
1746                                                 $text
1747                                         );
1748                                         $text = preg_replace(
1749                                                 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1750                                                 '<a class="mention hashtag status-link" href="$2" rel="tag">$1<span>$3</span></a>',
1751                                                 $text
1752                                         );
1753                                 } else {
1754                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1755                                 }
1756
1757                                 if (!$for_plaintext) {
1758                                         if (in_array($simple_html, [self::OSTATUS, self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1759                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1760                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1761                                         }
1762                                 } else {
1763                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1764                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'removePictureLinksCallback'], $text);
1765                                 }
1766
1767                                 // Bookmarks in red - will be converted to bookmarks in friendica
1768                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1769                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1770                                 $text = preg_replace(
1771                                         "/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1772                                         "[bookmark=$1]$2[/bookmark]",
1773                                         $text
1774                                 );
1775
1776                                 if (in_array($simple_html, [self::OSTATUS, self::TWITTER, self::BLUESKY])) {
1777                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", [self::class, 'expandLinksCallback'], $text);
1778                                         //$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
1779                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
1780                                 }
1781
1782                                 // Perform URL Search
1783                                 if ($try_oembed) {
1784                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1785                                 }
1786
1787                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1788
1789                                 // Handle Diaspora posts
1790                                 $text = preg_replace_callback(
1791                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1792                                         function ($match) {
1793                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1794                                         },
1795                                         $text
1796                                 );
1797
1798                                 $text = preg_replace_callback(
1799                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1800                                         function ($match) {
1801                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1802                                         },
1803                                         $text
1804                                 );
1805
1806                                 // Server independent link to posts and comments
1807                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1808                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1809                                 $text = preg_replace($expression, DI::baseUrl() . "/display/$1", $text);
1810
1811                                 /* Tag conversion
1812                                  * Supports:
1813                                  * - #[url=<anything>]<term>[/url]
1814                                  * - [url=<anything>]#<term>[/url]
1815                                  */
1816                                 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
1817                                         $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function ($matches) use ($simple_html) {
1818                                                 if ($simple_html == self::ACTIVITYPUB) {
1819                                                         return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1820                                                                 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1821                                                                 . XML::escape($matches[1]) . '</a>';
1822                                                 } else {
1823                                                         return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1824                                                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1825                                                                 . XML::escape($matches[1]) . '</a>';
1826                                                 }
1827                                         }, $text);
1828                                         return $text;
1829                                 });
1830
1831                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1832                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1833                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1834                                 $text = preg_replace("/\[url\](" . $escapedBaseUrl . ".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1835                                 $text = preg_replace("/\[url\=(" . $escapedBaseUrl . ".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1836
1837                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1838                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1839
1840                                 // Red compatibility, though the link can't be authenticated on Friendica
1841                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1842
1843
1844                                 // we may need to restrict this further if it picks up too many strays
1845                                 // link acct:user@host to a webfinger profile redirector
1846
1847                                 $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);
1848
1849                                 // Perform MAIL Search
1850                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1851                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1852
1853                                 /// @todo What is the meaning of these lines?
1854                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1855                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1856
1857                                 // Currently deactivated, it made problems with " inside of alt texts.
1858                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1859
1860                                 // fix any escaped ampersands that may have been converted into links
1861                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1862
1863                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1864                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'contact/redir/', 'cid:'];
1865
1866                                 array_walk($allowed_src_protocols, function (&$value) {
1867                                         $value = preg_quote($value, '#');
1868                                 });
1869
1870                                 $text = preg_replace(
1871                                         '#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1872                                         '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">',
1873                                         $text
1874                                 );
1875
1876                                 // sanitize href attributes (only allowlisted protocols URLs)
1877                                 // default value for backward compatibility
1878                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1879
1880                                 // Always allowed protocol even if config isn't set or not including it
1881                                 $allowed_link_protocols[] = '//';
1882                                 $allowed_link_protocols[] = 'http://';
1883                                 $allowed_link_protocols[] = 'https://';
1884                                 $allowed_link_protocols[] = 'contact/redir/';
1885
1886                                 array_walk($allowed_link_protocols, function (&$value) {
1887                                         $value = preg_quote($value, '#');
1888                                 });
1889
1890                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1891                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1892
1893                                 // Shared content
1894                                 $text = self::convertShare(
1895                                         $text,
1896                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1897                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1898                                         },
1899                                         $uriid
1900                                 );
1901
1902                                 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
1903
1904                                 return $text;
1905                         }); // Escaped noparse, nobb, pre
1906
1907                         // Remove escaping tags and replace new lines that remain
1908                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1909                                 return str_replace("\n", "<br>", $match[2]);
1910                         }, $text);
1911
1912                         // Additionally, [pre] tags preserve spaces
1913                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1914                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES, 'UTF-8'));
1915                         }, $text);
1916
1917                         return $text;
1918                 }); // Escaped code
1919
1920                 $text = preg_replace_callback(
1921                         "#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1922                         function ($matches) {
1923                                 if (strpos($matches[2], "\n") !== false) {
1924                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1925                                 } else {
1926                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1927                                 }
1928
1929                                 return $return;
1930                         },
1931                         $text
1932                 );
1933
1934                 // Default iframe allowed domains/path
1935                 $allowedIframeDomains = [
1936                         DI::baseUrl()->getHost()
1937                                 . (DI::baseUrl()->getPath() ? '/' . DI::baseUrl()->getPath() : '')
1938                                 . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1939                         'www.youtube.com/embed/',
1940                         'player.vimeo.com/video/',
1941                 ];
1942
1943                 $allowedIframeDomains = array_merge(
1944                         $allowedIframeDomains,
1945                         DI::config()->get('system', 'allowed_oembed') ?
1946                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1947                                 : []
1948                 );
1949
1950                 if (strpos($text, '<p>') !== false || strpos($text, '</p>') !== false) {
1951                         $text = '<p>' . $text . '</p>';
1952                 }
1953
1954                 $text = HTML::purify($text, $allowedIframeDomains);
1955                 DI::profiler()->stopRecording();
1956
1957                 return trim($text);
1958         }
1959
1960         /**
1961          * Strips the "abstract" tag from the provided text
1962          *
1963          * @param string $text The text with BBCode
1964          * @return string The same text - but without "abstract" element
1965          */
1966         public static function stripAbstract(string $text): string
1967         {
1968                 DI::profiler()->startRecording('rendering');
1969
1970                 $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
1971                         $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
1972                         $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
1973                         return $text;
1974                 });
1975
1976                 DI::profiler()->stopRecording();
1977                 return $text;
1978         }
1979
1980         /**
1981          * Returns the value of the "abstract" element
1982          *
1983          * @param string $text  The text that maybe contains the element
1984          * @param string $addon The addon for which the abstract is meant for
1985          * @return string The abstract
1986          */
1987         public static function getAbstract(string $text, string $addon = ''): string
1988         {
1989                 DI::profiler()->startRecording('rendering');
1990                 $addon = strtolower($addon);
1991
1992                 $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
1993                         if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
1994                                 return $matches[1];
1995                         }
1996
1997                         if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
1998                                 return $matches[1];
1999                         }
2000
2001                         return '';
2002                 });
2003
2004                 DI::profiler()->stopRecording();
2005                 return $abstract;
2006         }
2007
2008         /**
2009          * Callback function to replace a Friendica style mention in a mention for Diaspora
2010          *
2011          * @param array $match Matching values for the callback
2012          *                     [1] = Mention type (! or @)
2013          *                     [2] = Name
2014          *                     [3] = Address
2015          * @return string Replaced mention
2016          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2017          * @throws \ImagickException
2018          */
2019         private static function bbCodeMention2DiasporaCallback(array $match): string
2020         {
2021                 $contact = Contact::getByURL($match[3], false, ['addr']);
2022                 if (empty($contact['addr'])) {
2023                         return $match[0];
2024                 }
2025
2026                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2027                 return $mention;
2028         }
2029
2030         /**
2031          * Converts a BBCode text into Markdown
2032          *
2033          * This function converts a BBCode item body to be sent to Markdown-enabled
2034          * systems like Diaspora and Libertree
2035          *
2036          * @param string $text
2037          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2038          * @return string
2039          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2040          */
2041         public static function toMarkdown(string $text, bool $for_diaspora = true): string
2042         {
2043                 DI::profiler()->startRecording('rendering');
2044                 $original_text = $text;
2045
2046                 // Since Diaspora is creating a summary for links, this function removes them before posting
2047                 if ($for_diaspora) {
2048                         $text = self::removeShareInformation($text);
2049                 }
2050
2051                 /**
2052                  * Transform #tags, strip off the [url] and replace spaces with underscore
2053                  */
2054                 $url_search_string = "^\[\]";
2055                 $text = preg_replace_callback(
2056                         "/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2057                         function ($matches) {
2058                                 return '#' . str_replace(' ', '_', $matches[2]);
2059                         },
2060                         $text
2061                 );
2062
2063                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2064                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2065
2066                 // Convert it to HTML - don't try oembed
2067                 if ($for_diaspora) {
2068                         $text = self::convertForUriId(0, $text, self::DIASPORA);
2069
2070                         // Add all tags that maybe were removed
2071                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2072                                 $tagline = '';
2073                                 foreach ($tags[2] as $tag) {
2074                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2075                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2076                                                 $tagline .= '#' . $tag . ' ';
2077                                         }
2078                                 }
2079                                 $text = $text . ' ' . $tagline;
2080                         }
2081                 } else {
2082                         $text = self::convertForUriId(0, $text, self::CONNECTORS);
2083                 }
2084
2085                 // If a link is followed by a quote then there should be a newline before it
2086                 // Maybe we should make this newline at every time before a quote.
2087                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2088
2089                 // Now convert HTML to Markdown
2090                 $text = HTML::toMarkdown($text);
2091
2092                 // Libertree has a problem with escaped hashtags.
2093                 $text = str_replace(['\#'], ['#'], $text);
2094
2095                 // Remove any leading or trailing whitespace, as this will mess up
2096                 // the Diaspora signature verification and cause the item to disappear
2097                 $text = trim($text);
2098
2099                 if ($for_diaspora) {
2100                         $url_search_string = "^\[\]";
2101                         $text = preg_replace_callback(
2102                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2103                                 [self::class, 'bbCodeMention2DiasporaCallback'],
2104                                 $text
2105                         );
2106                 }
2107
2108                 Hook::callAll('bb2diaspora', $text);
2109
2110                 DI::profiler()->stopRecording();
2111                 return $text;
2112         }
2113
2114         /**
2115          * Pull out all #hashtags and @person tags from $string.
2116          *
2117          * We also get @person@domain.com - which would make
2118          * the regex quite complicated as tags can also
2119          * end a sentence. So we'll run through our results
2120          * and strip the period from any tags which end with one.
2121          * Returns array of tags found, or empty array.
2122          *
2123          * @param string $string Post content
2124          * @return array List of tag and person names
2125          */
2126         public static function getTags(string $string): array
2127         {
2128                 DI::profiler()->startRecording('rendering');
2129                 $ret = [];
2130
2131                 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img', 'attachment'], function ($string) use (&$ret) {
2132                         // Convert hashtag links to hashtags
2133                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2134
2135                         // Force line feeds at bbtags
2136                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2137
2138                         // ignore anything in a bbtag
2139                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2140
2141                         // Match full names against @tags including the space between first and last
2142                         // We will look these up afterward to see if they are full names or not recognisable.
2143
2144                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2145                                 foreach ($matches[1] as $match) {
2146                                         if (strstr($match, ']')) {
2147                                                 // we might be inside a bbcode color tag - leave it alone
2148                                                 continue;
2149                                         }
2150
2151                                         if (substr($match, -1, 1) === '.') {
2152                                                 $ret[] = substr($match, 0, -1);
2153                                         } else {
2154                                                 $ret[] = $match;
2155                                         }
2156                                 }
2157                         }
2158
2159                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2160                         // and #hash tags.
2161
2162                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2163                                 foreach ($matches[1] as $match) {
2164                                         if (strstr($match, ']')) {
2165                                                 // we might be inside a bbcode color tag - leave it alone
2166                                                 continue;
2167                                         }
2168
2169                                         // try not to catch url fragments
2170                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2171                                                 continue;
2172                                         }
2173
2174                                         $ret[] = $match;
2175                                 }
2176                         }
2177                 });
2178
2179                 DI::profiler()->stopRecording();
2180                 return array_unique($ret);
2181         }
2182
2183         /**
2184          * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2185          *
2186          * @param string $body HTML/BBCode
2187          * @return string body with expanded tags
2188          */
2189         public static function expandTags(string $body): string
2190         {
2191                 return preg_replace_callback(
2192                         "/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2193                         function (array $match) {
2194                                 switch ($match[1]) {
2195                                         case '!':
2196                                         case '@':
2197                                                 $contact = Contact::getByURL($match[2]);
2198                                                 if (!empty($contact)) {
2199                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2200                                                 } else {
2201                                                         return $match[1] . $match[2];
2202                                                 }
2203                                                 break;
2204
2205                                         case '#':
2206                                         default:
2207                                                 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2208                                 }
2209                         },
2210                         $body
2211                 );
2212         }
2213
2214         /**
2215          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2216          *
2217          * @param string   $text HTML/BBCode
2218          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2219          * @param callable $callback
2220          * @return string
2221          * @see Strings::performWithEscapedBlocks
2222          */
2223         public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2224         {
2225                 $tagList = array_map('preg_quote', $tagList);
2226
2227                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2228         }
2229
2230         /**
2231          * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2232          *
2233          * @param string $body HTML/BBCode
2234          * @param int $profile_uid Profile user id
2235          * @param string $network Network name
2236          * @return string HTML/BBCode with inserted images
2237          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2238          * @throws \ImagickException
2239          */
2240         public static function setMentions(string $body, $profile_uid = 0, $network = '')
2241         {
2242                 DI::profiler()->startRecording('rendering');
2243                 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2244                         $tags = self::getTags($body);
2245
2246                         $tagged = [];
2247
2248                         foreach ($tags as $tag) {
2249                                 $tag_type = substr($tag, 0, 1);
2250
2251                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2252                                         continue;
2253                                 }
2254
2255                                 /*
2256                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2257                                  * Robert Johnson should be first in the $tags array
2258                                  */
2259                                 foreach ($tagged as $nextTag) {
2260                                         if (stristr($nextTag, $tag . ' ')) {
2261                                                 continue 2;
2262                                         }
2263                                 }
2264
2265                                 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2266                                         $tagged[] = $tag;
2267                                 }
2268                         }
2269
2270                         return $body;
2271                 });
2272
2273                 DI::profiler()->stopRecording();
2274                 return $body;
2275         }
2276
2277         /**
2278          * @param string      $author  Author display name
2279          * @param string      $profile Author profile URL
2280          * @param string      $avatar  Author profile picture URL
2281          * @param string      $link    Post source URL
2282          * @param string      $posted  Post created date
2283          * @param string|null $guid    Post guid (if any)
2284          * @param string|null $uri     Post uri (if any)
2285          * @return string
2286          * @TODO Rewrite to handle over whole record array
2287          */
2288         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null, string $uri = null): string
2289         {
2290                 DI::profiler()->startRecording('rendering');
2291                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2292                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2293                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2294                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2295                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2296
2297                 if ($guid) {
2298                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2299                 }
2300
2301                 if ($uri) {
2302                         $header .= "' message_id='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $uri);
2303                 }
2304
2305                 $header  .= "']";
2306
2307                 DI::profiler()->stopRecording();
2308                 return $header;
2309         }
2310
2311         /**
2312          * Returns the BBCode relevant to embed the provided URL in a post body.
2313          * For media type, it will return [img], [video] and [audio] tags.
2314          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2315          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2316          *
2317          * @param string      $url
2318          * @param bool        $tryAttachment
2319          * @param string|null $title
2320          * @param string|null $description
2321          * @param string|null $tags
2322          * @return string
2323          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2324          * @see ParseUrl::getSiteinfoCached
2325          */
2326         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2327         {
2328                 DI::profiler()->startRecording('rendering');
2329                 DI::logger()->info($url);
2330
2331                 // If there is already some content information submitted we don't
2332                 // need to parse the url for content.
2333                 if (!empty($title) && !empty($description)) {
2334                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2335
2336                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2337
2338                         $str_tags = '';
2339                         if (!empty($tags)) {
2340                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2341                                 if (count($arr_tags)) {
2342                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2343                                 }
2344                         }
2345
2346                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2347
2348                         DI::logger()->info('(unparsed): returns: ' . $result);
2349
2350                         DI::profiler()->stopRecording();
2351                         return $result;
2352                 }
2353
2354                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2355
2356                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2357                         switch ($siteinfo['type']) {
2358                                 case 'video':
2359                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2360                                         break;
2361                                 case 'audio':
2362                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2363                                         break;
2364                                 default:
2365                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2366                                         break;
2367                         }
2368
2369                         DI::profiler()->stopRecording();
2370                         return $bbcode;
2371                 }
2372
2373                 unset($siteinfo['keywords']);
2374
2375                 // Bypass attachment if parse url for a comment
2376                 if (!$tryAttachment) {
2377                         DI::profiler()->stopRecording();
2378                         return "\n" . '[url=' . $url . ']' . ($siteinfo['title'] ?? $url) . '[/url]';
2379                 }
2380
2381                 // Format it as BBCode attachment
2382                 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2383                 DI::profiler()->stopRecording();
2384                 return $bbcode;
2385         }
2386 }