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