src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\Hook;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Logger;
  19 use Friendica\Core\Protocol;
  20 use Friendica\Core\Renderer;
  21 use Friendica\Core\System;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Model\Photo;
  25 use Friendica\Network\Probe;
  26 use Friendica\Object\Image;
  27 use Friendica\Protocol\Activity;
  28 use Friendica\Util\Map;
  29 use Friendica\Util\Network;
  30 use Friendica\Util\ParseUrl;
  31 use Friendica\Util\Proxy as ProxyUtils;
  32 use Friendica\Util\Strings;
  33 use Friendica\Util\XML;
  34
  35 class BBCode extends BaseObject
  36 {
  37         /**
  38          * @brief Fetches attachment data that were generated the old way
  39          *
  40          * @param string $body Message body
  41          * @return array
  42          *                     'type' -> Message type ('link', 'video', 'photo')
  43          *                     'text' -> Text before the shared message
  44          *                     'after' -> Text after the shared message
  45          *                     'image' -> Preview image of the message
  46          *                     'url' -> Url to the attached message
  47          *                     'title' -> Title of the attachment
  48          *                     'description' -> Description of the attachment
  49          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  50          */
  51         private static function getOldAttachmentData($body)
  52         {
  53                 $post = [];
  54
  55                 // Simplify image codes
  56                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  57
  58                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  59                         foreach ($attached as $data) {
  60                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  61                                         continue;
  62                                 }
  63
  64                                 $post['type'] = substr($data[1], 5);
  65
  66                                 $pos = strpos($body, $data[0]);
  67                                 if ($pos > 0) {
  68                                         $post['text'] = trim(substr($body, 0, $pos));
  69                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  70                                 } else {
  71                                         $post['text'] = trim(str_replace($data[0], '', $body));
  72                                         $post['after'] = '';
  73                                 }
  74
  75                                 $attacheddata = $data[2];
  76
  77                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
  78
  79                                         $picturedata = Image::getInfoFromURL($matches[1]);
  80
  81                                         if ($picturedata) {
  82                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  83                                                         $post['image'] = $matches[1];
  84                                                 } else {
  85                                                         $post['preview'] = $matches[1];
  86                                                 }
  87                                         }
  88                                 }
  89
  90                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  91                                         $post['url'] = $matches[1];
  92                                         $post['title'] = $matches[2];
  93                                 }
  94                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
  95                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  96                                         $post['url'] = $matches[1];
  97                                 }
  98
  99                                 // Search for description
 100                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 101                                         $post['description'] = $matches[1];
 102                                 }
 103                         }
 104                 }
 105                 return $post;
 106         }
 107
 108         /**
 109          * @brief Fetches attachment data that were generated with the "attachment" element
 110          *
 111          * @param string $body Message body
 112          * @return array
 113          *                     'type' -> Message type ('link', 'video', 'photo')
 114          *                     'text' -> Text before the shared message
 115          *                     'after' -> Text after the shared message
 116          *                     'image' -> Preview image of the message
 117          *                     'url' -> Url to the attached message
 118          *                     'title' -> Title of the attachment
 119          *                     'description' -> Description of the attachment
 120          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 121          */
 122         public static function getAttachmentData($body)
 123         {
 124                 $data = [
 125                         'type'        => '',
 126                         'text'        => '',
 127                         'after'       => '',
 128                         'image'       => null,
 129                         'url'         => '',
 130                         'title'       => '',
 131                         'description' => '',
 132                 ];
 133
 134                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 135                         return self::getOldAttachmentData($body);
 136                 }
 137
 138                 $attributes = $match[2];
 139
 140                 $data['text'] = trim($match[1]);
 141
 142                 $type = '';
 143                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 144                 if (!empty($matches[1])) {
 145                         $type = strtolower($matches[1]);
 146                 }
 147
 148                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 149                 if (!empty($matches[1])) {
 150                         $type = strtolower($matches[1]);
 151                 }
 152
 153                 if ($type == '') {
 154                         return [];
 155                 }
 156
 157                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 158                         return [];
 159                 }
 160
 161                 if ($type != '') {
 162                         $data['type'] = $type;
 163                 }
 164
 165                 $url = '';
 166                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 167                 if (!empty($matches[1])) {
 168                         $url = $matches[1];
 169                 }
 170
 171                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 172                 if (!empty($matches[1])) {
 173                         $url = $matches[1];
 174                 }
 175
 176                 if ($url != '') {
 177                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 178                 }
 179
 180                 $title = '';
 181                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 182                 if (!empty($matches[1])) {
 183                         $title = $matches[1];
 184                 }
 185
 186                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 187                 if (!empty($matches[1])) {
 188                         $title = $matches[1];
 189                 }
 190
 191                 if ($title != '') {
 192                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 193                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 194                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 195                         $data['title'] = $title;
 196                 }
 197
 198                 $image = '';
 199                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 200                 if (!empty($matches[1])) {
 201                         $image = $matches[1];
 202                 }
 203
 204                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 205                 if (!empty($matches[1])) {
 206                         $image = $matches[1];
 207                 }
 208
 209                 if ($image != '') {
 210                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 211                 }
 212
 213                 $preview = '';
 214                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 215                 if (!empty($matches[1])) {
 216                         $preview = $matches[1];
 217                 }
 218
 219                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 220                 if (!empty($matches[1])) {
 221                         $preview = $matches[1];
 222                 }
 223
 224                 if ($preview != '') {
 225                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 226                 }
 227
 228                 $data['description'] = trim($match[3]);
 229
 230                 $data['after'] = trim($match[4]);
 231
 232                 return $data;
 233         }
 234
 235         public static function getAttachedData($body, $item = [])
 236         {
 237                 /*
 238                 - text:
 239                 - type: link, video, photo
 240                 - title:
 241                 - url:
 242                 - image:
 243                 - description:
 244                 - (thumbnail)
 245                 */
 246
 247                 $has_title = !empty($item['title']);
 248                 $plink = $item['plink'] ?? '';
 249                 $post = self::getAttachmentData($body);
 250
 251                 // Get all linked images with alternative image description
 252                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 253                         foreach ($pictures as $picture) {
 254                                 if (Photo::isLocal($picture[1])) {
 255                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 256                                 }
 257                         }
 258                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 259                                 $post['image_description'] = $post['images'][0]['description'];
 260                         }
 261                 }
 262
 263                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 264                         foreach ($pictures as $picture) {
 265                                 if (Photo::isLocal($picture[1])) {
 266                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 267                                 }
 268                         }
 269                 }
 270
 271                 // if nothing is found, it maybe having an image.
 272                 if (!isset($post['type'])) {
 273                         // Simplify image codes
 274                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 275                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 276                         $post['text'] = $body;
 277
 278                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 279                                 if ((count($pictures) == 1) && !$has_title) {
 280                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 281                                                 // Replace the preview picture with the real picture
 282                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 283                                                 $data = ['url' => $url, 'type' => 'photo'];
 284                                         } else {
 285                                                 // Checking, if the link goes to a picture
 286                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 287                                         }
 288
 289                                         // Workaround:
 290                                         // Sometimes photo posts to the own album are not detected at the start.
 291                                         // So we seem to cannot use the cache for these cases. That's strange.
 292                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 293                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 294                                         }
 295
 296                                         if ($data['type'] == 'photo') {
 297                                                 $post['type'] = 'photo';
 298                                                 if (isset($data['images'][0])) {
 299                                                         $post['image'] = $data['images'][0]['src'];
 300                                                         $post['url'] = $data['url'];
 301                                                 } else {
 302                                                         $post['image'] = $data['url'];
 303                                                 }
 304
 305                                                 $post['preview'] = $pictures[0][2];
 306                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 307                                         } else {
 308                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 309                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 310                                                         $post['type'] = 'photo';
 311                                                         $post['image'] = $pictures[0][1];
 312                                                         $post['preview'] = $pictures[0][2];
 313                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 314                                                 }
 315                                         }
 316                                 } elseif (count($pictures) > 0) {
 317                                         $post['type'] = 'link';
 318                                         $post['url'] = $plink;
 319                                         $post['image'] = $pictures[0][2];
 320                                         $post['text'] = $body;
 321
 322                                         foreach ($pictures as $picture) {
 323                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 324                                         }
 325                                 }
 326                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 327                                 if ((count($pictures) == 1) && !$has_title) {
 328                                         $post['type'] = 'photo';
 329                                         $post['image'] = $pictures[0][1];
 330                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 331                                 } elseif (count($pictures) > 0) {
 332                                         $post['type'] = 'link';
 333                                         $post['url'] = $plink;
 334                                         $post['image'] = $pictures[0][1];
 335                                         $post['text'] = $body;
 336
 337                                         foreach ($pictures as $picture) {
 338                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 339                                         }
 340                                 }
 341                         }
 342
 343                         // Test for the external links
 344                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 345                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 346
 347                         $links = array_merge($links1, $links2);
 348
 349                         // If there is only a single one, then use it.
 350                         // This should cover link posts via API.
 351                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 352                                 $post['type'] = 'link';
 353                                 $post['url'] = $links[0][1];
 354                         }
 355
 356                         // Now count the number of external media links
 357                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 358                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 359                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 360                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 361
 362                         // Add them to the other external links
 363                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 364
 365                         // Are there more than one?
 366                         if (count($links) > 1) {
 367                                 // The post will be the type "text", which means a blog post
 368                                 unset($post['type']);
 369                                 $post['url'] = $plink;
 370                         }
 371
 372                         if (!isset($post['type'])) {
 373                                 $post['type'] = "text";
 374                                 $post['text'] = trim($body);
 375                         }
 376                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 377                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 378
 379                         if (isset($data['images'][0])) {
 380                                 $post['image'] = $data['images'][0]['src'];
 381                         }
 382                 }
 383
 384                 return $post;
 385         }
 386
 387         /**
 388          * @brief Converts a BBCode text into plaintext
 389          *
 390          * @param      $text
 391          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 392          *
 393          * @return string
 394          */
 395         public static function toPlaintext($text, $keep_urls = true)
 396         {
 397                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 398
 399                 return $naked_text;
 400         }
 401
 402         private static function proxyUrl($image, $simplehtml = false)
 403         {
 404                 // Only send proxied pictures to API and for internal display
 405                 if (in_array($simplehtml, [false, 2])) {
 406                         return ProxyUtils::proxifyUrl($image);
 407                 } else {
 408                         return $image;
 409                 }
 410         }
 411
 412         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 413         {
 414                 // Suppress "view full size"
 415                 if (intval(Config::get('system', 'no_view_full_size'))) {
 416                         $include_link = false;
 417                 }
 418
 419                 // Picture addresses can contain special characters
 420                 $s = htmlspecialchars_decode($srctext);
 421
 422                 $matches = null;
 423                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 424                 if ($c) {
 425                         foreach ($matches as $mtch) {
 426                                 Logger::log('scale_external_image: ' . $mtch[1]);
 427
 428                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 429                                 if (stristr($mtch[1], $hostname)) {
 430                                         continue;
 431                                 }
 432
 433                                 // $scale_replace, if passed, is an array of two elements. The
 434                                 // first is the name of the full-size image. The second is the
 435                                 // name of a remote, scaled-down version of the full size image.
 436                                 // This allows Friendica to display the smaller remote image if
 437                                 // one exists, while still linking to the full-size image
 438                                 if ($scale_replace) {
 439                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 440                                 } else {
 441                                         $scaled = $mtch[1];
 442                                 }
 443                                 $i = Network::fetchUrl($scaled);
 444                                 if (!$i) {
 445                                         return $srctext;
 446                                 }
 447
 448                                 // guess mimetype from headers or filename
 449                                 $type = Image::guessType($mtch[1], true);
 450
 451                                 if ($i) {
 452                                         $Image = new Image($i, $type);
 453                                         if ($Image->isValid()) {
 454                                                 $orig_width = $Image->getWidth();
 455                                                 $orig_height = $Image->getHeight();
 456
 457                                                 if ($orig_width > 640 || $orig_height > 640) {
 458                                                         $Image->scaleDown(640);
 459                                                         $new_width = $Image->getWidth();
 460                                                         $new_height = $Image->getHeight();
 461                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 462                                                         $s = str_replace(
 463                                                                 $mtch[0],
 464                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 465                                                                 . "\n" . (($include_link)
 466                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 467                                                                         : ''),
 468                                                                 $s
 469                                                         );
 470                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 471                                                 }
 472                                         }
 473                                 }
 474                         }
 475                 }
 476
 477                 // replace the special char encoding
 478                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 479                 return $s;
 480         }
 481
 482         /**
 483          * The purpose of this function is to apply system message length limits to
 484          * imported messages without including any embedded photos in the length
 485          *
 486          * @brief Truncates imported message body string length to max_import_size
 487          * @param string $body
 488          * @return string
 489          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 490          */
 491         public static function limitBodySize($body)
 492         {
 493                 $maxlen = Config::get('config', 'max_import_size', 0);
 494
 495                 // If the length of the body, including the embedded images, is smaller
 496                 // than the maximum, then don't waste time looking for the images
 497                 if ($maxlen && (strlen($body) > $maxlen)) {
 498
 499                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 500
 501                         $orig_body = $body;
 502                         $new_body = '';
 503                         $textlen = 0;
 504
 505                         $img_start = strpos($orig_body, '[img');
 506                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 507                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 508                         while (($img_st_close !== false) && ($img_end !== false)) {
 509
 510                                 $img_st_close++; // make it point to AFTER the closing bracket
 511                                 $img_end += $img_start;
 512                                 $img_end += strlen('[/img]');
 513
 514                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 515                                         // This is an embedded image
 516
 517                                         if (($textlen + $img_start) > $maxlen) {
 518                                                 if ($textlen < $maxlen) {
 519                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 520                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 521                                                         $textlen = $maxlen;
 522                                                 }
 523                                         } else {
 524                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 525                                                 $textlen += $img_start;
 526                                         }
 527
 528                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 529                                 } else {
 530
 531                                         if (($textlen + $img_end) > $maxlen) {
 532                                                 if ($textlen < $maxlen) {
 533                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 534                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 535                                                         $textlen = $maxlen;
 536                                                 }
 537                                         } else {
 538                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 539                                                 $textlen += $img_end;
 540                                         }
 541                                 }
 542                                 $orig_body = substr($orig_body, $img_end);
 543
 544                                 if ($orig_body === false) {
 545                                         // in case the body ends on a closing image tag
 546                                         $orig_body = '';
 547                                 }
 548
 549                                 $img_start = strpos($orig_body, '[img');
 550                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 551                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 552                         }
 553
 554                         if (($textlen + strlen($orig_body)) > $maxlen) {
 555                                 if ($textlen < $maxlen) {
 556                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 557                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 558                                 }
 559                         } else {
 560                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 561                                 $new_body = $new_body . $orig_body;
 562                         }
 563
 564                         return $new_body;
 565                 } else {
 566                         return $body;
 567                 }
 568         }
 569
 570         /**
 571          * Processes [attachment] tags
 572          *
 573          * Note: Can produce a [bookmark] tag in the returned string
 574          *
 575          * @brief Processes [attachment] tags
 576          * @param string   $text
 577          * @param bool|int $simplehtml
 578          * @param bool     $tryoembed
 579          * @return string
 580          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 581          */
 582         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 583         {
 584                 $data = self::getAttachmentData($text);
 585                 if (empty($data) || empty($data['url'])) {
 586                         return $text;
 587                 }
 588
 589                 if (isset($data['title'])) {
 590                         $data['title'] = strip_tags($data['title']);
 591                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 592                 } else {
 593                         $data['title'] = null;
 594                 }
 595
 596                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data['image'])) {
 597                         $data['preview'] = $data['image'];
 598                         $data['image'] = '';
 599                 }
 600
 601                 $return = '';
 602                 if (in_array($simplehtml, [7, 9])) {
 603                         // Only add the link when it isn't already part of the body
 604                         if (substr_count($text, $data['url']) == 1) {
 605                                 $return = self::convertUrlForActivityPub($data['url']);
 606                         }
 607                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 608                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data['url'], $data['title']);
 609                 } else {
 610                         try {
 611                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 612                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 613                                 } else {
 614                                         throw new Exception('OEmbed is disabled for this attachment.');
 615                                 }
 616                         } catch (Exception $e) {
 617                                 $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 618
 619                                 if ($simplehtml != 4) {
 620                                         $return = sprintf('<div class="type-%s">', $data['type']);
 621                                 }
 622
 623                                 if (!empty($data['title']) && !empty($data['url'])) {
 624                                         if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 625                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 626                                         } else {
 627                                                 if (!empty($data['image'])) {
 628                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 629                                                 } elseif (!empty($data['preview'])) {
 630                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']);
 631                                                 }
 632                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 633                                         }
 634                                 }
 635
 636                                 if (!empty($data['description']) && $data['description'] != $data['title']) {
 637                                         // Sanitize the HTML by converting it to BBCode
 638                                         $bbcode = HTML::toBBCode($data['description']);
 639                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 640                                 }
 641
 642                                 if (!empty($data['url'])) {
 643                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 644                                 }
 645
 646                                 if ($simplehtml != 4) {
 647                                         $return .= '</div>';
 648                                 }
 649                         }
 650                 }
 651
 652                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 653         }
 654
 655         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 656         {
 657                 $data = self::getAttachmentData($Text);
 658
 659                 if (!$data) {
 660                         return $Text;
 661                 } elseif ($nolink) {
 662                         return $data['text'] . ($data['after'] ?? '');
 663                 }
 664
 665                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 666                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 667                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 668                         $data['title'] = $data['url'];
 669                 } elseif (($text != '') && strstr($title, $text)) {
 670                         $data['text'] = $data['title'];
 671                         $data['title'] = $data['url'];
 672                 }
 673
 674                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 675                         return $data['title'] . $data['after'];
 676                 }
 677
 678                 // If the link already is included in the post, don't add it again
 679                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 680                         return $data['text'] . $data['after'];
 681                 }
 682
 683                 $text = $data['text'];
 684
 685                 if (!empty($data['url']) && !empty($data['title'])) {
 686                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 687                 } elseif (!empty($data['url'])) {
 688                         $text .= "\n[url]" . $data['url'] . '[/url]';
 689                 }
 690
 691                 return $text . "\n" . $data['after'];
 692         }
 693
 694         /**
 695          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 696          *
 697          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 698          * @param array $match Array with the matching values
 699          * @return string reformatted link including HTML codes
 700          */
 701         private static function convertUrlForActivityPubCallback($match)
 702         {
 703                 $url = $match[1];
 704
 705                 if (isset($match[2]) && ($match[1] != $match[2])) {
 706                         return $match[0];
 707                 }
 708
 709                 $parts = parse_url($url);
 710                 if (!isset($parts['scheme'])) {
 711                         return $match[0];
 712                 }
 713
 714                 return self::convertUrlForActivityPub($url);
 715         }
 716
 717         /**
 718          * @brief Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 719          * @param string $url URL that is about to be reformatted
 720          * @return string reformatted link including HTML codes
 721          */
 722         private static function convertUrlForActivityPub($url)
 723         {
 724                 $html = '<a href="%s" target="_blank">%s</a>';
 725                 return sprintf($html, $url, self::getStyledURL($url));
 726         }
 727
 728         /**
 729          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 730          * @param string $url URL that is about to be reformatted
 731          * @return string reformatted link
 732          */
 733         private static function getStyledURL($url)
 734         {
 735                 $parts = parse_url($url);
 736                 $scheme = $parts['scheme'] . '://';
 737                 $styled_url = str_replace($scheme, '', $url);
 738
 739                 if (strlen($styled_url) > 30) {
 740                         $styled_url = substr($styled_url, 0, 30) . "…";
 741                 }
 742
 743                 return $styled_url;
 744         }
 745
 746         /*
 747          * [noparse][i]italic[/i][/noparse] turns into
 748          * [noparse][ i ]italic[ /i ][/noparse],
 749          * to hide them from parser.
 750          */
 751         private static function escapeNoparseCallback($match)
 752         {
 753                 $whole_match = $match[0];
 754                 $captured = $match[1];
 755                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 756                 $new_str = str_replace($captured, $spacefied, $whole_match);
 757                 return $new_str;
 758         }
 759
 760         /*
 761          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 762          * now turns back and the [noparse] tags are trimed
 763          * returning [i]italic[/i]
 764          */
 765         private static function unescapeNoparseCallback($match)
 766         {
 767                 $captured = $match[1];
 768                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 769                 return $unspacefied;
 770         }
 771
 772         /**
 773          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 774          * occurrences
 775          *
 776          * @param string $text        Text to search
 777          * @param string $name        Tag name
 778          * @param int    $occurrences Number of first occurrences to skip
 779          * @return boolean|array
 780          */
 781         public static function getTagPosition($text, $name, $occurrences = 0)
 782         {
 783                 if ($occurrences < 0) {
 784                         $occurrences = 0;
 785                 }
 786
 787                 $start_open = -1;
 788                 for ($i = 0; $i <= $occurrences; $i++) {
 789                         if ($start_open !== false) {
 790                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 791                         }
 792                 }
 793
 794                 if ($start_open === false) {
 795                         return false;
 796                 }
 797
 798                 $start_equal = strpos($text, '=', $start_open);
 799                 $start_close = strpos($text, ']', $start_open);
 800
 801                 if ($start_close === false) {
 802                         return false;
 803                 }
 804
 805                 $start_close++;
 806
 807                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 808
 809                 if ($end_open === false) {
 810                         return false;
 811                 }
 812
 813                 $res = [
 814                         'start' => [
 815                                 'open' => $start_open,
 816                                 'close' => $start_close
 817                         ],
 818                         'end' => [
 819                                 'open' => $end_open,
 820                                 'close' => $end_open + strlen('[/' . $name . ']')
 821                         ],
 822                 ];
 823
 824                 if ($start_equal !== false) {
 825                         $res['start']['equal'] = $start_equal + 1;
 826                 }
 827
 828                 return $res;
 829         }
 830
 831         /**
 832          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 833          *
 834          * @param string $pattern Preg pattern string
 835          * @param string $replace Preg replace string
 836          * @param string $name    BBCode tag name
 837          * @param string $text    Text to search
 838          * @return string
 839          */
 840         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 841         {
 842                 $occurrences = 0;
 843                 $pos = self::getTagPosition($text, $name, $occurrences);
 844                 while ($pos !== false && $occurrences++ < 1000) {
 845                         $start = substr($text, 0, $pos['start']['open']);
 846                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 847                         $end = substr($text, $pos['end']['close']);
 848                         if ($end === false) {
 849                                 $end = '';
 850                         }
 851
 852                         $subject = preg_replace($pattern, $replace, $subject);
 853                         $text = $start . $subject . $end;
 854
 855                         $pos = self::getTagPosition($text, $name, $occurrences);
 856                 }
 857
 858                 return $text;
 859         }
 860
 861         private static function extractImagesFromItemBody($body)
 862         {
 863                 $saved_image = [];
 864                 $orig_body = $body;
 865                 $new_body = '';
 866
 867                 $cnt = 0;
 868                 $img_start = strpos($orig_body, '[img');
 869                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 870                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 871                 while (($img_st_close !== false) && ($img_end !== false)) {
 872                         $img_st_close++; // make it point to AFTER the closing bracket
 873                         $img_end += $img_start;
 874
 875                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 876                                 // This is an embedded image
 877                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 878                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 879
 880                                 $cnt++;
 881                         } else {
 882                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 883                         }
 884
 885                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 886
 887                         if ($orig_body === false) {
 888                                 // in case the body ends on a closing image tag
 889                                 $orig_body = '';
 890                         }
 891
 892                         $img_start = strpos($orig_body, '[img');
 893                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 894                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 895                 }
 896
 897                 $new_body = $new_body . $orig_body;
 898
 899                 return ['body' => $new_body, 'images' => $saved_image];
 900         }
 901
 902         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 903         {
 904                 $newbody = $body;
 905
 906                 $cnt = 0;
 907                 foreach ($images as $image) {
 908                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 909                         // it loops over the array starting from the first element and going sequentially
 910                         // to the last element
 911                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 912                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 913                         $cnt++;
 914                 }
 915
 916                 return $newbody;
 917         }
 918
 919         /**
 920          * This function converts a [share] block to text according to a provided callback function whose signature is:
 921          *
 922          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 923          *
 924          * Where:
 925          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 926          * data lookup
 927          * - $author_contact is a contact record array
 928          * - $content is the inner content of the [share] block
 929          * - $is_quote_share indicates whether there's any content before the [share] block
 930          * - Return value is the string that should replace the [share] block in the provided text
 931          *
 932          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 933          *
 934          * @param  string   $text     A BBCode string
 935          * @param  callable $callback
 936          * @return string The BBCode string with all [share] blocks replaced
 937          */
 938         public static function convertShare($text, callable $callback)
 939         {
 940                 $return = preg_replace_callback(
 941                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 942                         function ($match) use ($callback) {
 943                                 $attribute_string = $match[2];
 944
 945                                 $attributes = [];
 946                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 947                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 948                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 949                                 }
 950
 951                                 // We only call this so that a previously unknown contact can be added.
 952                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 953                                 // This function then can fetch an entry from the contact table.
 954                                 $default['url'] = $attributes['profile'];
 955
 956                                 if (!empty($attributes['author'])) {
 957                                         $default['name'] = $attributes['author'];
 958                                 }
 959
 960                                 if (!empty($attributes['avatar'])) {
 961                                         $default['photo'] = $attributes['avatar'];
 962                                 }
 963
 964                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 965
 966                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 967                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 968
 969                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 970                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 971                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 972
 973                                 if ($attributes['avatar']) {
 974                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 975                                 }
 976
 977                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 978                         },
 979                         $text
 980                 );
 981
 982                 return $return;
 983         }
 984
 985         /**
 986          * Default [share] tag conversion callback
 987          *
 988          * Note: Can produce a [bookmark] tag in the output
 989          *
 990          * @see BBCode::convertShare()
 991          * @param array   $attributes     [share] block attribute values
 992          * @param array   $author_contact Contact row of the shared author
 993          * @param string  $content        Inner content of the [share] block
 994          * @param boolean $is_quote_share Whether there is content before the [share] block
 995          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 996          * @return string
 997          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 998          */
 999         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1000         {
1001                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1002
1003                 switch ($simplehtml) {
1004                         case 1:
1005                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1006                                 break;
1007                         case 2:
1008                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1009                                 break;
1010                         case 3: // Diaspora
1011                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1012                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1013                                 } else {
1014                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1015
1016                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1017                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1018                                         }
1019
1020                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1021
1022                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1023                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1024                                         }
1025                                 }
1026
1027                                 break;
1028                         case 4:
1029                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1030                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1031                                 $headline .= ':</b></p>' . "\n";
1032
1033                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1034
1035                                 break;
1036                         case 5:
1037                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1038                                 break;
1039                         case 7: // statusnet/GNU Social
1040                         case 9: // ActivityPub
1041                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1042                                 break;
1043                         default:
1044                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1045                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1046                                         try {
1047                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1048                                         } catch (Exception $e) {
1049                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1050                                         }
1051                                 } else {
1052                                         $text = ($is_quote_share? "\n" : '');
1053
1054                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1055                                         $text .= Renderer::replaceMacros($tpl, [
1056                                                 '$profile' => $attributes['profile'],
1057                                                 '$avatar'  => $attributes['avatar'],
1058                                                 '$author'  => $attributes['author'],
1059                                                 '$link'    => $attributes['link'],
1060                                                 '$posted'  => $attributes['posted'],
1061                                                 '$content' => trim($content)
1062                                         ]);
1063                                 }
1064                                 break;
1065                 }
1066
1067                 return $text;
1068         }
1069
1070         private static function removePictureLinksCallback($match)
1071         {
1072                 $cache_key = 'remove:' . $match[1];
1073                 $text = Cache::get($cache_key);
1074
1075                 if (is_null($text)) {
1076                         $a = self::getApp();
1077
1078                         $stamp1 = microtime(true);
1079
1080                         $ch = @curl_init($match[1]);
1081                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1082                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1083                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1084                         @curl_exec($ch);
1085                         $curl_info = @curl_getinfo($ch);
1086
1087                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1088
1089                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1090                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1091                         } else {
1092                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1093
1094                                 // if its not a picture then look if its a page that contains a picture link
1095                                 $body = Network::fetchUrl($match[1]);
1096
1097                                 $doc = new DOMDocument();
1098                                 @$doc->loadHTML($body);
1099                                 $xpath = new DOMXPath($doc);
1100                                 $list = $xpath->query("//meta[@name]");
1101                                 foreach ($list as $node) {
1102                                         $attr = [];
1103
1104                                         if ($node->attributes->length) {
1105                                                 foreach ($node->attributes as $attribute) {
1106                                                         $attr[$attribute->name] = $attribute->value;
1107                                                 }
1108                                         }
1109
1110                                         if (strtolower($attr['name']) == 'twitter:image') {
1111                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1112                                         }
1113                                 }
1114                         }
1115                         Cache::set($cache_key, $text);
1116                 }
1117
1118                 return $text;
1119         }
1120
1121         private static function expandLinksCallback($match)
1122         {
1123                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1124                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1125                 } else {
1126                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1127                 }
1128         }
1129
1130         private static function cleanPictureLinksCallback($match)
1131         {
1132                 $a = self::getApp();
1133
1134                 // When the picture link is the own photo path then we can avoid fetching the link
1135                 $own_photo_url = preg_quote(Strings::normaliseLink($a->getBaseURL()) . '/photos/');
1136                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1137                         if (!empty($match[3])) {
1138                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1139                         } else {
1140                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1141                         }
1142                         return $text;
1143                 }
1144
1145                 $cache_key = 'clean:' . $match[1];
1146                 $text = Cache::get($cache_key);
1147                 if (!is_null($text)) {
1148                         return $text;
1149                 }
1150
1151                 // Only fetch the header, not the content
1152                 $stamp1 = microtime(true);
1153
1154                 $ch = @curl_init($match[1]);
1155                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1156                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1157                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1158                 @curl_exec($ch);
1159                 $curl_info = @curl_getinfo($ch);
1160
1161                 $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1162
1163                 // if its a link to a picture then embed this picture
1164                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1165                         $text = '[img]' . $match[1] . '[/img]';
1166                 } else {
1167                         if (!empty($match[3])) {
1168                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1169                         } else {
1170                                 $text = '[img]' . $match[2] . '[/img]';
1171                         }
1172
1173                         // if its not a picture then look if its a page that contains a picture link
1174                         $body = Network::fetchUrl($match[1]);
1175
1176                         $doc = new DOMDocument();
1177                         @$doc->loadHTML($body);
1178                         $xpath = new DOMXPath($doc);
1179                         $list = $xpath->query("//meta[@name]");
1180                         foreach ($list as $node) {
1181                                 $attr = [];
1182                                 if ($node->attributes->length) {
1183                                         foreach ($node->attributes as $attribute) {
1184                                                 $attr[$attribute->name] = $attribute->value;
1185                                         }
1186                                 }
1187
1188                                 if (strtolower($attr['name']) == "twitter:image") {
1189                                         if (!empty($match[3])) {
1190                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1191                                         } else {
1192                                                 $text = "[img]" . $attr['content'] . "[/img]";
1193                                         }
1194                                 }
1195                         }
1196                 }
1197                 Cache::set($cache_key, $text);
1198
1199                 return $text;
1200         }
1201
1202         public static function cleanPictureLinks($text)
1203         {
1204                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1205                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1206                 return $return;
1207         }
1208
1209         /**
1210          * @brief Converts a BBCode message to HTML message
1211          *
1212          * BBcode 2 HTML was written by WAY2WEB.net
1213          * extended to work with Mistpark/Friendica - Mike Macgirvin
1214          *
1215          * Simple HTML values meaning:
1216          * - 0: Friendica display
1217          * - 1: Unused
1218          * - 2: Used for Windows Phone push, Friendica API
1219          * - 3: Used before converting to Markdown in bb2diaspora.php
1220          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1221          * - 5: Unused
1222          * - 6: Unused
1223          * - 7: Used for dfrn, OStatus
1224          * - 8: Used for WP backlink text setting
1225          * - 9: ActivityPub
1226          *
1227          * @param string $text
1228          * @param bool   $try_oembed
1229          * @param int    $simple_html
1230          * @param bool   $for_plaintext
1231          * @return string
1232          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1233          */
1234         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1235         {
1236                 $a = self::getApp();
1237
1238                 /*
1239                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1240                  *
1241                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1242                  * $match[1] = $url
1243                  * $match[2] = $title or absent
1244                  */
1245                 $try_oembed_callback = function ($match)
1246                 {
1247                         $url = $match[1];
1248                         $title = $match[2] ?? null;
1249
1250                         try {
1251                                 $return = OEmbed::getHTML($url, $title);
1252                         } catch (Exception $ex) {
1253                                 $return = $match[0];
1254                         }
1255
1256                         return $return;
1257                 };
1258
1259                 // Extracting code blocks before the whitespace processing and the autolinker
1260                 $codeblocks = [];
1261
1262                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1263                         function ($matches) use (&$codeblocks) {
1264                                 $return = '#codeblock-' . count($codeblocks) . '#';
1265                                 if (strpos($matches[2], "\n") !== false) {
1266                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1267                                 } else {
1268                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1269                                 }
1270
1271                                 return $return;
1272                         },
1273                         $text
1274                 );
1275
1276                 // Hide all [noparse] contained bbtags by spacefying them
1277                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1278
1279                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1280                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1281                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1282
1283                 // Remove the abstract element. It is a non visible element.
1284                 $text = self::stripAbstract($text);
1285
1286                 // Move all spaces out of the tags
1287                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1288                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1289
1290                 // Extract the private images which use data urls since preg has issues with
1291                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1292                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1293
1294                 $extracted = self::extractImagesFromItemBody($text);
1295                 $text = $extracted['body'];
1296                 $saved_image = $extracted['images'];
1297
1298                 // If we find any event code, turn it into an event.
1299                 // After we're finished processing the bbcode we'll
1300                 // replace all of the event code with a reformatted version.
1301
1302                 $ev = Event::fromBBCode($text);
1303
1304                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1305                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1306
1307                 $text = str_replace("<", "&lt;", $text);
1308                 $text = str_replace(">", "&gt;", $text);
1309
1310                 // remove some newlines before the general conversion
1311                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1312                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1313
1314                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1315                 if (!$try_oembed) {
1316                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1317                 }
1318
1319                 // Convert new line chars to html <br /> tags
1320
1321                 // nlbr seems to be hopelessly messed up
1322                 //      $Text = nl2br($Text);
1323
1324                 // We'll emulate it.
1325
1326                 $text = trim($text);
1327                 $text = str_replace("\r\n", "\n", $text);
1328
1329                 // Remove linefeeds inside of the table elements. See issue #6799
1330                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1331                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1332                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1333                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1334                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1335                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1336                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1337                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1338                 do {
1339                         $oldtext = $text;
1340                         $text = str_replace($search, $replace, $text);
1341                 } while ($oldtext != $text);
1342
1343                 // Replace these here only once
1344                 $search = ["\n[table]", "[/table]\n"];
1345                 $replace = ["[table]", "[/table]"];
1346                 $text = str_replace($search, $replace, $text);
1347
1348                 // removing multiplicated newlines
1349                 if (Config::get('system', 'remove_multiplicated_lines')) {
1350                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1351                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1352                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1353                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1354                         do {
1355                                 $oldtext = $text;
1356                                 $text = str_replace($search, $replace, $text);
1357                         } while ($oldtext != $text);
1358                 }
1359
1360                 // Handle attached links or videos
1361                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1362
1363                 // leave open the posibility of [map=something]
1364                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1365                 if (strpos($text, '[/map]') !== false) {
1366                         $text = preg_replace_callback(
1367                                 "/\[map\](.*?)\[\/map\]/ism",
1368                                 function ($match) use ($simple_html) {
1369                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1370                                 },
1371                                 $text
1372                         );
1373                 }
1374
1375                 if (strpos($text, '[map=') !== false) {
1376                         $text = preg_replace_callback(
1377                                 "/\[map=(.*?)\]/ism",
1378                                 function ($match) use ($simple_html) {
1379                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1380                                 },
1381                                 $text
1382                         );
1383                 }
1384
1385                 if (strpos($text, '[map]') !== false) {
1386                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1387                 }
1388
1389                 // Check for headers
1390                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1391                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1392                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1393                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1394                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1395                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1396
1397                 // Check for paragraph
1398                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1399
1400                 // Check for bold text
1401                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1402
1403                 // Check for Italics text
1404                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1405
1406                 // Check for Underline text
1407                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1408
1409                 // Check for strike-through text
1410                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1411
1412                 // Check for over-line text
1413                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1414
1415                 // Check for colored text
1416                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1417
1418                 // Check for sized text
1419                 // [size=50] --> font-size: 50px (with the unit).
1420                 if ($simple_html != 3) {
1421                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1422                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1423                 } else {
1424                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1425                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1426                 }
1427
1428
1429                 // Check for centered text
1430                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1431
1432                 // Check for list text
1433                 $text = str_replace("[*]", "<li>", $text);
1434
1435                 // Check for style sheet commands
1436                 $text = preg_replace_callback(
1437                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1438                         function ($match) {
1439                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1440                         },
1441                         $text
1442                 );
1443
1444                 // Check for CSS classes
1445                 $text = preg_replace_callback(
1446                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1447                         function ($match) {
1448                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1449                         },
1450                         $text
1451                 );
1452
1453                 // handle nested lists
1454                 $endlessloop = 0;
1455
1456                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1457                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1458                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1459                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1460                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1461                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1462                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1463                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1464                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1465                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1466                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1467                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1468                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1469                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1470                 }
1471
1472                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1473                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1474                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1475                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1476
1477                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1478                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1479
1480                 $text = str_replace('[hr]', '<hr />', $text);
1481
1482                 if (!$for_plaintext) {
1483                         // Autolinker for isolated URLs
1484                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1485                 }
1486
1487                 // This is actually executed in Item::prepareBody()
1488
1489                 $nosmile = strpos($text, '[nosmile]') !== false;
1490                 $text = str_replace('[nosmile]', '', $text);
1491
1492                 // Check for font change text
1493                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1494
1495                 // Declare the format for [spoiler] layout
1496                 $SpoilerLayout = '<details class="spoiler"><summary>' . L10n::t('Click to open/close') . '</summary>$1</details>';
1497
1498                 // Check for [spoiler] text
1499                 // handle nested quotes
1500                 $endlessloop = 0;
1501                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1502                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1503                 }
1504
1505                 // Check for [spoiler=Title] text
1506
1507                 // handle nested quotes
1508                 $endlessloop = 0;
1509                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1510                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1511                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1512                                 $text);
1513                 }
1514
1515                 // Declare the format for [quote] layout
1516                 $QuoteLayout = '<blockquote>$1</blockquote>';
1517
1518                 // Check for [quote] text
1519                 // handle nested quotes
1520                 $endlessloop = 0;
1521                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1522                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1523                 }
1524
1525                 // Check for [quote=Author] text
1526
1527                 $t_wrote = L10n::t('$1 wrote:');
1528
1529                 // handle nested quotes
1530                 $endlessloop = 0;
1531                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1532                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1533                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1534                                 $text);
1535                 }
1536
1537
1538                 // [img=widthxheight]image source[/img]
1539                 $text = preg_replace_callback(
1540                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1541                         function ($matches) use ($simple_html) {
1542                                 if (strpos($matches[3], "data:image/") === 0) {
1543                                         return $matches[0];
1544                                 }
1545
1546                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1547                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1548                         },
1549                         $text
1550                 );
1551
1552                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1553                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1554
1555                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1556                         function ($matches) use ($simple_html) {
1557                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1558                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1559                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1560                         },
1561                         $text);
1562
1563                 // Images
1564                 // [img]pathtoimage[/img]
1565                 $text = preg_replace_callback(
1566                         "/\[img\](.*?)\[\/img\]/ism",
1567                         function ($matches) use ($simple_html) {
1568                                 if (strpos($matches[1], "data:image/") === 0) {
1569                                         return $matches[0];
1570                                 }
1571
1572                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1573                                 return "[img]" . $matches[1] . "[/img]";
1574                         },
1575                         $text
1576                 );
1577
1578                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1579                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1580
1581                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1582                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1583                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1584
1585                 // Try to Oembed
1586                 if ($try_oembed) {
1587                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism", '<video src="$1" controls="controls" width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1588                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3).*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1589
1590                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1591                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1592                 } else {
1593                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1594                                 '<a href="$1" target="_blank">$1</a>', $text);
1595                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1596                                 '<a href="$1" target="_blank">$1</a>', $text);
1597                 }
1598
1599                 // html5 video and audio
1600
1601
1602                 if ($try_oembed) {
1603                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1604                 } else {
1605                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1606                 }
1607
1608                 // Youtube extensions
1609                 if ($try_oembed) {
1610                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1611                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1612                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1613                 }
1614
1615                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1616                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1617                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1618
1619                 if ($try_oembed) {
1620                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1621                 } else {
1622                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1623                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1624                 }
1625
1626                 if ($try_oembed) {
1627                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1628                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1629                 }
1630
1631                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1632                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1633
1634                 if ($try_oembed) {
1635                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1636                 } else {
1637                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1638                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1639                 }
1640
1641                 // oembed tag
1642                 $text = OEmbed::BBCode2HTML($text);
1643
1644                 // Avoid triple linefeeds through oembed
1645                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1646
1647                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1648                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1649                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1650                 // start which is always required). Allow desc with a missing summary for compatibility.
1651
1652                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1653                         $sub = Event::getHTML($ev, $simple_html);
1654
1655                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1656                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1657                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1658                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1659                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1660                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1661                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1662                 }
1663
1664                 // Replace non graphical smilies for external posts
1665                 if (!$nosmile && !$for_plaintext) {
1666                         $text = Smilies::replace($text);
1667                 }
1668
1669                 if (!$for_plaintext) {
1670                         if (in_array($simple_html, [7, 9])) {
1671                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1672                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1673                         }
1674                 } else {
1675                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1676                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1677                 }
1678
1679                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1680
1681                 // Remove all hashtag addresses
1682                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7, 9])) {
1683                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1684                 } elseif ($simple_html == 3) {
1685                         // The ! is converted to @ since Diaspora only understands the @
1686                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1687                                 '@<a href="$2">$3</a>',
1688                                 $text);
1689                 } elseif (in_array($simple_html, [7, 9])) {
1690                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1691                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1692                                 $text);
1693                 } elseif (!$simple_html) {
1694                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1695                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1696                                 $text);
1697                 }
1698
1699                 // Bookmarks in red - will be converted to bookmarks in friendica
1700                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1701                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1702                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1703                                         "[bookmark=$1]$2[/bookmark]", $text);
1704
1705                 if (in_array($simple_html, [2, 6, 7, 8])) {
1706                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1707                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1708                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1709                 }
1710
1711                 if ($simple_html == 5) {
1712                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1713                 }
1714
1715                 // Perform URL Search
1716                 if ($try_oembed) {
1717                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1718                 }
1719
1720                 if ($simple_html == 5) {
1721                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1722                 } else {
1723                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1724                 }
1725
1726                 // Handle Diaspora posts
1727                 $text = preg_replace_callback(
1728                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1729                         function ($match) {
1730                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1731                         }, $text
1732                 );
1733
1734                 $text = preg_replace_callback(
1735                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1736                         function ($match) {
1737                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1738                         }, $text
1739                 );
1740
1741                 // Server independent link to posts and comments
1742                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1743                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1744                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1745
1746                 /* Tag conversion
1747                  * Supports:
1748                  * - #[url=<anything>]<term>[/url]
1749                  * - [url=<anything>]#<term>[/url]
1750                  */
1751                 $text = preg_replace_callback("/(?:#\[url\=.*?\]|\[url\=.*?\]#)(.*?)\[\/url\]/ism", function($matches) {
1752                         return '#<a href="'
1753                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1754                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1755                                 . XML::escape($matches[1])
1756                                 . '</a>';
1757                 }, $text);
1758
1759                 // We need no target="_blank" for local links
1760                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1761                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1762                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1763                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1764
1765                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1766                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1767
1768                 // Red compatibility, though the link can't be authenticated on Friendica
1769                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1770
1771
1772                 // we may need to restrict this further if it picks up too many strays
1773                 // link acct:user@host to a webfinger profile redirector
1774
1775                 $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="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1776
1777                 // Perform MAIL Search
1778                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1779                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1780
1781                 // Unhide all [noparse] contained bbtags unspacefying them
1782                 // and triming the [noparse] tag.
1783
1784                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1785                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1786                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1787
1788                 /// @todo What is the meaning of these lines?
1789                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1790                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1791
1792                 // Currently deactivated, it made problems with " inside of alt texts.
1793                 //$text = preg_replace('/\&quot\;/', '"', $text);
1794
1795                 // fix any escaped ampersands that may have been converted into links
1796                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1797
1798                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1799                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1800
1801                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1802
1803                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1804                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1805
1806                 // sanitize href attributes (only whitelisted protocols URLs)
1807                 // default value for backward compatibility
1808                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1809
1810                 // Always allowed protocol even if config isn't set or not including it
1811                 $allowed_link_protocols[] = '//';
1812                 $allowed_link_protocols[] = 'http://';
1813                 $allowed_link_protocols[] = 'https://';
1814                 $allowed_link_protocols[] = 'redir/';
1815
1816                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1817
1818                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1819                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1820
1821                 // Shared content
1822                 $text = self::convertShare(
1823                         $text,
1824                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1825                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1826                         }
1827                 );
1828
1829                 if ($saved_image) {
1830                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1831                 }
1832
1833                 // Restore code blocks
1834                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1835                         function ($matches) use ($codeblocks) {
1836                                 $return = $matches[0];
1837                                 if (isset($codeblocks[intval($matches[1])])) {
1838                                         $return = $codeblocks[$matches[1]];
1839                                 }
1840                                 return $return;
1841                         },
1842                         $text
1843                 );
1844
1845                 // Clean up the HTML by loading and saving the HTML with the DOM.
1846                 // Bad structured html can break a whole page.
1847                 // For performance reasons do it only with activated item cache or at export.
1848                 if (!$try_oembed || (get_itemcachepath() != '')) {
1849                         $doc = new DOMDocument();
1850                         $doc->preserveWhiteSpace = false;
1851
1852                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1853
1854                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1855                         $encoding = '<?xml encoding="UTF-8">';
1856                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1857                         $doc->encoding = 'UTF-8';
1858                         $text = $doc->saveHTML();
1859                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1860
1861                         $text = str_replace('<br></li>', '</li>', $text);
1862
1863                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1864                 }
1865
1866                 // Clean up some useless linebreaks in lists
1867                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1868                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1869                 //$Text = str_replace('</li><br />', '</li>', $Text);
1870                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1871                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1872
1873                 Hook::callAll('bbcode', $text);
1874
1875                 return trim($text);
1876         }
1877
1878         /**
1879          * @brief Strips the "abstract" tag from the provided text
1880          *
1881          * @param string $text The text with BBCode
1882          * @return string The same text - but without "abstract" element
1883          */
1884         public static function stripAbstract($text)
1885         {
1886                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1887                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1888
1889                 return $text;
1890         }
1891
1892         /**
1893          * @brief Returns the value of the "abstract" element
1894          *
1895          * @param string $text The text that maybe contains the element
1896          * @param string $addon The addon for which the abstract is meant for
1897          * @return string The abstract
1898          */
1899         public static function getAbstract($text, $addon = '')
1900         {
1901                 $abstract = '';
1902                 $abstracts = [];
1903                 $addon = strtolower($addon);
1904
1905                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1906                         foreach ($results AS $result) {
1907                                 $abstracts[strtolower($result[1])] = $result[2];
1908                         }
1909                 }
1910
1911                 if (isset($abstracts[$addon])) {
1912                         $abstract = $abstracts[$addon];
1913                 }
1914
1915                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1916                         $abstract = $result[1];
1917                 }
1918
1919                 return $abstract;
1920         }
1921
1922         /**
1923          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1924          *
1925          * @param array $match Matching values for the callback
1926          *                     [1] = Mention type (! or @)
1927          *                     [2] = Name
1928          *                     [3] = Address
1929          * @return string Replaced mention
1930          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1931          * @throws \ImagickException
1932          */
1933         private static function bbCodeMention2DiasporaCallback($match)
1934         {
1935                 $contact = Contact::getDetailsByURL($match[3]);
1936
1937                 if (empty($contact['addr'])) {
1938                         $contact = Probe::uri($match[3]);
1939                 }
1940
1941                 if (empty($contact['addr'])) {
1942                         return $match[0];
1943                 }
1944
1945                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1946                 return $mention;
1947         }
1948
1949         /**
1950          * @brief Converts a BBCode text into Markdown
1951          *
1952          * This function converts a BBCode item body to be sent to Markdown-enabled
1953          * systems like Diaspora and Libertree
1954          *
1955          * @param string $text
1956          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1957          * @return string
1958          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1959          */
1960         public static function toMarkdown($text, $for_diaspora = true)
1961         {
1962                 $a = self::getApp();
1963
1964                 $original_text = $text;
1965
1966                 // Since Diaspora is creating a summary for links, this function removes them before posting
1967                 if ($for_diaspora) {
1968                         $text = self::removeShareInformation($text);
1969                 }
1970
1971                 /**
1972                  * Transform #tags, strip off the [url] and replace spaces with underscore
1973                  */
1974                 $url_search_string = "^\[\]";
1975                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1976                         function ($matches) {
1977                                 return '#' . str_replace(' ', '_', $matches[2]);
1978                         },
1979                         $text
1980                 );
1981
1982                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1983                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1984
1985                 // Convert it to HTML - don't try oembed
1986                 if ($for_diaspora) {
1987                         $text = self::convert($text, false, 3);
1988
1989                         // Add all tags that maybe were removed
1990                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1991                                 $tagline = '';
1992                                 foreach ($tags[2] as $tag) {
1993                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1994                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1995                                                 $tagline .= '#' . $tag . ' ';
1996                                         }
1997                                 }
1998                                 $text = $text . " " . $tagline;
1999                         }
2000                 } else {
2001                         $text = self::convert($text, false, 4);
2002                 }
2003
2004                 // mask some special HTML chars from conversation to markdown
2005                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2006
2007                 // If a link is followed by a quote then there should be a newline before it
2008                 // Maybe we should make this newline at every time before a quote.
2009                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2010
2011                 $stamp1 = microtime(true);
2012
2013                 // Now convert HTML to Markdown
2014                 $text = HTML::toMarkdown($text);
2015
2016                 // unmask the special chars back to HTML
2017                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2018
2019                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
2020
2021                 // Libertree has a problem with escaped hashtags.
2022                 $text = str_replace(['\#'], ['#'], $text);
2023
2024                 // Remove any leading or trailing whitespace, as this will mess up
2025                 // the Diaspora signature verification and cause the item to disappear
2026                 $text = trim($text);
2027
2028                 if ($for_diaspora) {
2029                         $url_search_string = "^\[\]";
2030                         $text = preg_replace_callback(
2031                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2032                                 ['self', 'bbCodeMention2DiasporaCallback'],
2033                                 $text
2034                         );
2035                 }
2036
2037                 Hook::callAll('bb2diaspora', $text);
2038
2039                 return $text;
2040         }
2041
2042         /**
2043      * @brief Pull out all #hashtags and @person tags from $string.
2044      *
2045      * We also get @person@domain.com - which would make
2046      * the regex quite complicated as tags can also
2047      * end a sentence. So we'll run through our results
2048      * and strip the period from any tags which end with one.
2049      * Returns array of tags found, or empty array.
2050      *
2051      * @param string $string Post content
2052      *
2053      * @return array List of tag and person names
2054      */
2055     public static function getTags($string)
2056     {
2057         $ret = [];
2058
2059         // Convert hashtag links to hashtags
2060         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2061
2062         // ignore anything in a code block
2063         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2064
2065         // Force line feeds at bbtags
2066         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2067
2068         // ignore anything in a bbtag
2069         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2070
2071         // Match full names against @tags including the space between first and last
2072         // We will look these up afterward to see if they are full names or not recognisable.
2073
2074         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2075             foreach ($matches[1] as $match) {
2076                 if (strstr($match, ']')) {
2077                     // we might be inside a bbcode color tag - leave it alone
2078                     continue;
2079                 }
2080
2081                 if (substr($match, -1, 1) === '.') {
2082                     $ret[] = substr($match, 0, -1);
2083                 } else {
2084                     $ret[] = $match;
2085                 }
2086             }
2087         }
2088
2089         // Otherwise pull out single word tags. These can be @nickname, @first_last
2090         // and #hash tags.
2091
2092         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2093             foreach ($matches[1] as $match) {
2094                 if (strstr($match, ']')) {
2095                     // we might be inside a bbcode color tag - leave it alone
2096                     continue;
2097                 }
2098                 if (substr($match, -1, 1) === '.') {
2099                     $match = substr($match,0,-1);
2100                 }
2101                 // ignore strictly numeric tags like #1
2102                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2103                     continue;
2104                 }
2105                 // try not to catch url fragments
2106                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2107                     continue;
2108                 }
2109                 $ret[] = $match;
2110             }
2111         }
2112
2113         return $ret;
2114     }
2115 }