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\Content\OEmbed;
  12 use Friendica\Content\Smilies;
  13 use Friendica\Core\Config;
  14 use Friendica\Core\Hook;
  15 use Friendica\Core\L10n;
  16 use Friendica\Core\Logger;
  17 use Friendica\Core\Protocol;
  18 use Friendica\Core\Renderer;
  19 use Friendica\Core\System;
  20 use Friendica\DI;
  21 use Friendica\Model\Contact;
  22 use Friendica\Model\Event;
  23 use Friendica\Model\Photo;
  24 use Friendica\Network\Probe;
  25 use Friendica\Object\Image;
  26 use Friendica\Protocol\Activity;
  27 use Friendica\Util\Images;
  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
  36 {
  37         /**
  38          * 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                                         $picturedata = Images::getInfoFromURLCached($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          * 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                                                 $imgdata = Images::getInfoFromURLCached($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                         // Simplify "video" element
 357                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 358
 359                         // Now count the number of external media links
 360                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 361                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 362                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 363                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 364
 365                         // Add them to the other external links
 366                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 367
 368                         // Are there more than one?
 369                         if (count($links) > 1) {
 370                                 // The post will be the type "text", which means a blog post
 371                                 unset($post['type']);
 372                                 $post['url'] = $plink;
 373                         }
 374
 375                         if (!isset($post['type'])) {
 376                                 $post['type'] = "text";
 377                                 $post['text'] = trim($body);
 378                         }
 379                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 380                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 381
 382                         if (isset($data['images'][0])) {
 383                                 $post['image'] = $data['images'][0]['src'];
 384                         }
 385                 }
 386
 387                 return $post;
 388         }
 389
 390         /**
 391          * Remove [attachment] BBCode and replaces it with a regular [url]
 392          *
 393          * @param string  $body
 394          * @param boolean $no_link_desc No link description
 395          *
 396          * @return string with replaced body
 397          */
 398         public static function removeAttachment($body, $no_link_desc = false)
 399         {
 400                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 401                         function ($match) use ($no_link_desc) {
 402                                 $attach_data = self::getAttachmentData($match[0]);
 403                                 if (empty($attach_data['url'])) {
 404                                         return $match[0];
 405                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 406                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 407                                 } else {
 408                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 409                                 }
 410                 }, $body);
 411         }
 412
 413         /**
 414          * Converts a BBCode text into plaintext
 415          *
 416          * @param      $text
 417          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 418          *
 419          * @return string
 420          */
 421         public static function toPlaintext($text, $keep_urls = true)
 422         {
 423                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 424
 425                 return $naked_text;
 426         }
 427
 428         private static function proxyUrl($image, $simplehtml = false)
 429         {
 430                 // Only send proxied pictures to API and for internal display
 431                 if (in_array($simplehtml, [false, 2])) {
 432                         return ProxyUtils::proxifyUrl($image);
 433                 } else {
 434                         return $image;
 435                 }
 436         }
 437
 438         public static function scaleExternalImages($srctext)
 439         {
 440                 $s = $srctext;
 441
 442                 $matches = null;
 443                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 444                 if ($c) {
 445                         foreach ($matches as $mtch) {
 446                                 Logger::log('scale_external_image: ' . $mtch[1]);
 447
 448                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 449                                 if (stristr($mtch[1], $hostname)) {
 450                                         continue;
 451                                 }
 452
 453                                 $i = Network::fetchUrl($mtch[1]);
 454                                 if (!$i) {
 455                                         return $srctext;
 456                                 }
 457
 458                                 // guess mimetype from headers or filename
 459                                 $type = Images::guessType($mtch[1], true);
 460
 461                                 if ($i) {
 462                                         $Image = new Image($i, $type);
 463                                         if ($Image->isValid()) {
 464                                                 $orig_width = $Image->getWidth();
 465                                                 $orig_height = $Image->getHeight();
 466
 467                                                 if ($orig_width > 640 || $orig_height > 640) {
 468                                                         $Image->scaleDown(640);
 469                                                         $new_width = $Image->getWidth();
 470                                                         $new_height = $Image->getHeight();
 471                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 472                                                         $s = str_replace(
 473                                                                 $mtch[0],
 474                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 475                                                                 . "\n",
 476                                                                 $s
 477                                                         );
 478                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 479                                                 }
 480                                         }
 481                                 }
 482                         }
 483                 }
 484
 485                 return $s;
 486         }
 487
 488         /**
 489          * Truncates imported message body string length to max_import_size
 490          *
 491          * The purpose of this function is to apply system message length limits to
 492          * imported messages without including any embedded photos in the length
 493          *
 494          * @param string $body
 495          * @return string
 496          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 497          */
 498         public static function limitBodySize($body)
 499         {
 500                 $maxlen = Config::get('config', 'max_import_size', 0);
 501
 502                 // If the length of the body, including the embedded images, is smaller
 503                 // than the maximum, then don't waste time looking for the images
 504                 if ($maxlen && (strlen($body) > $maxlen)) {
 505
 506                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 507
 508                         $orig_body = $body;
 509                         $new_body = '';
 510                         $textlen = 0;
 511
 512                         $img_start = strpos($orig_body, '[img');
 513                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 514                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 515                         while (($img_st_close !== false) && ($img_end !== false)) {
 516
 517                                 $img_st_close++; // make it point to AFTER the closing bracket
 518                                 $img_end += $img_start;
 519                                 $img_end += strlen('[/img]');
 520
 521                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 522                                         // This is an embedded image
 523
 524                                         if (($textlen + $img_start) > $maxlen) {
 525                                                 if ($textlen < $maxlen) {
 526                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 527                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 528                                                         $textlen = $maxlen;
 529                                                 }
 530                                         } else {
 531                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 532                                                 $textlen += $img_start;
 533                                         }
 534
 535                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 536                                 } else {
 537
 538                                         if (($textlen + $img_end) > $maxlen) {
 539                                                 if ($textlen < $maxlen) {
 540                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 541                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 542                                                         $textlen = $maxlen;
 543                                                 }
 544                                         } else {
 545                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 546                                                 $textlen += $img_end;
 547                                         }
 548                                 }
 549                                 $orig_body = substr($orig_body, $img_end);
 550
 551                                 if ($orig_body === false) {
 552                                         // in case the body ends on a closing image tag
 553                                         $orig_body = '';
 554                                 }
 555
 556                                 $img_start = strpos($orig_body, '[img');
 557                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 558                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 559                         }
 560
 561                         if (($textlen + strlen($orig_body)) > $maxlen) {
 562                                 if ($textlen < $maxlen) {
 563                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 564                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 565                                 }
 566                         } else {
 567                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 568                                 $new_body = $new_body . $orig_body;
 569                         }
 570
 571                         return $new_body;
 572                 } else {
 573                         return $body;
 574                 }
 575         }
 576
 577         /**
 578          * Processes [attachment] tags
 579          *
 580          * Note: Can produce a [bookmark] tag in the returned string
 581          *
 582          * @param string   $text
 583          * @param bool|int $simplehtml
 584          * @param bool     $tryoembed
 585          * @return string
 586          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 587          */
 588         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 589         {
 590                 $data = self::getAttachmentData($text);
 591                 if (empty($data) || empty($data['url'])) {
 592                         return $text;
 593                 }
 594
 595                 if (isset($data['title'])) {
 596                         $data['title'] = strip_tags($data['title']);
 597                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 598                 } else {
 599                         $data['title'] = null;
 600                 }
 601
 602                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data['image'])) {
 603                         $data['preview'] = $data['image'];
 604                         $data['image'] = '';
 605                 }
 606
 607                 $return = '';
 608                 try {
 609                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 610                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 611                         } else {
 612                                 throw new Exception('OEmbed is disabled for this attachment.');
 613                         }
 614                 } catch (Exception $e) {
 615                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 616
 617                         if ($simplehtml != 4) {
 618                                 $return = sprintf('<div class="type-%s">', $data['type']);
 619                         }
 620
 621                         if (!empty($data['title']) && !empty($data['url'])) {
 622                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 623                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 624                                 } else {
 625                                         if (!empty($data['image'])) {
 626                                                 $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), $data['title']);
 627                                         } elseif (!empty($data['preview'])) {
 628                                                 $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']);
 629                                         }
 630                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 631                                 }
 632                         }
 633
 634                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 635                                 // Sanitize the HTML by converting it to BBCode
 636                                 $bbcode = HTML::toBBCode($data['description']);
 637                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 638                         }
 639
 640                         if (!empty($data['url'])) {
 641                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 642                         }
 643
 644                         if ($simplehtml != 4) {
 645                                 $return .= '</div>';
 646                         }
 647                 }
 648
 649                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 650         }
 651
 652         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 653         {
 654                 $data = self::getAttachmentData($Text);
 655
 656                 if (!$data) {
 657                         return $Text;
 658                 } elseif ($nolink) {
 659                         return $data['text'] . ($data['after'] ?? '');
 660                 }
 661
 662                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 663                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 664                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 665                         $data['title'] = $data['url'];
 666                 } elseif (($text != '') && strstr($title, $text)) {
 667                         $data['text'] = $data['title'];
 668                         $data['title'] = $data['url'];
 669                 }
 670
 671                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 672                         return $data['title'] . $data['after'];
 673                 }
 674
 675                 // If the link already is included in the post, don't add it again
 676                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 677                         return $data['text'] . $data['after'];
 678                 }
 679
 680                 $text = $data['text'];
 681
 682                 if (!empty($data['url']) && !empty($data['title'])) {
 683                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 684                 } elseif (!empty($data['url'])) {
 685                         $text .= "\n[url]" . $data['url'] . '[/url]';
 686                 }
 687
 688                 return $text . "\n" . $data['after'];
 689         }
 690
 691         /**
 692          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 693          *
 694          * @param array $match Array with the matching values
 695          * @return string reformatted link including HTML codes
 696          */
 697         private static function convertUrlForActivityPubCallback($match)
 698         {
 699                 $url = $match[1];
 700
 701                 if (isset($match[2]) && ($match[1] != $match[2])) {
 702                         return $match[0];
 703                 }
 704
 705                 $parts = parse_url($url);
 706                 if (!isset($parts['scheme'])) {
 707                         return $match[0];
 708                 }
 709
 710                 return self::convertUrlForActivityPub($url);
 711         }
 712
 713         /**
 714          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 715          * @param string $url URL that is about to be reformatted
 716          * @return string reformatted link including HTML codes
 717          */
 718         private static function convertUrlForActivityPub($url)
 719         {
 720                 $html = '<a href="%s" target="_blank">%s</a>';
 721                 return sprintf($html, $url, self::getStyledURL($url));
 722         }
 723
 724         /**
 725          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 726          * @param string $url URL that is about to be reformatted
 727          * @return string reformatted link
 728          */
 729         private static function getStyledURL($url)
 730         {
 731                 $parts = parse_url($url);
 732                 $scheme = $parts['scheme'] . '://';
 733                 $styled_url = str_replace($scheme, '', $url);
 734
 735                 if (strlen($styled_url) > 30) {
 736                         $styled_url = substr($styled_url, 0, 30) . "…";
 737                 }
 738
 739                 return $styled_url;
 740         }
 741
 742         /*
 743          * [noparse][i]italic[/i][/noparse] turns into
 744          * [noparse][ i ]italic[ /i ][/noparse],
 745          * to hide them from parser.
 746          */
 747         private static function escapeNoparseCallback($match)
 748         {
 749                 $whole_match = $match[0];
 750                 $captured = $match[1];
 751                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 752                 $new_str = str_replace($captured, $spacefied, $whole_match);
 753                 return $new_str;
 754         }
 755
 756         /*
 757          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 758          * now turns back and the [noparse] tags are trimed
 759          * returning [i]italic[/i]
 760          */
 761         private static function unescapeNoparseCallback($match)
 762         {
 763                 $captured = $match[1];
 764                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 765                 return $unspacefied;
 766         }
 767
 768         /**
 769          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 770          * occurrences
 771          *
 772          * @param string $text        Text to search
 773          * @param string $name        Tag name
 774          * @param int    $occurrences Number of first occurrences to skip
 775          * @return boolean|array
 776          */
 777         public static function getTagPosition($text, $name, $occurrences = 0)
 778         {
 779                 if ($occurrences < 0) {
 780                         $occurrences = 0;
 781                 }
 782
 783                 $start_open = -1;
 784                 for ($i = 0; $i <= $occurrences; $i++) {
 785                         if ($start_open !== false) {
 786                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 787                         }
 788                 }
 789
 790                 if ($start_open === false) {
 791                         return false;
 792                 }
 793
 794                 $start_equal = strpos($text, '=', $start_open);
 795                 $start_close = strpos($text, ']', $start_open);
 796
 797                 if ($start_close === false) {
 798                         return false;
 799                 }
 800
 801                 $start_close++;
 802
 803                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 804
 805                 if ($end_open === false) {
 806                         return false;
 807                 }
 808
 809                 $res = [
 810                         'start' => [
 811                                 'open' => $start_open,
 812                                 'close' => $start_close
 813                         ],
 814                         'end' => [
 815                                 'open' => $end_open,
 816                                 'close' => $end_open + strlen('[/' . $name . ']')
 817                         ],
 818                 ];
 819
 820                 if ($start_equal !== false) {
 821                         $res['start']['equal'] = $start_equal + 1;
 822                 }
 823
 824                 return $res;
 825         }
 826
 827         /**
 828          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 829          *
 830          * @param string $pattern Preg pattern string
 831          * @param string $replace Preg replace string
 832          * @param string $name    BBCode tag name
 833          * @param string $text    Text to search
 834          * @return string
 835          */
 836         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 837         {
 838                 $occurrences = 0;
 839                 $pos = self::getTagPosition($text, $name, $occurrences);
 840                 while ($pos !== false && $occurrences++ < 1000) {
 841                         $start = substr($text, 0, $pos['start']['open']);
 842                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 843                         $end = substr($text, $pos['end']['close']);
 844                         if ($end === false) {
 845                                 $end = '';
 846                         }
 847
 848                         $subject = preg_replace($pattern, $replace, $subject);
 849                         $text = $start . $subject . $end;
 850
 851                         $pos = self::getTagPosition($text, $name, $occurrences);
 852                 }
 853
 854                 return $text;
 855         }
 856
 857         private static function extractImagesFromItemBody($body)
 858         {
 859                 $saved_image = [];
 860                 $orig_body = $body;
 861                 $new_body = '';
 862
 863                 $cnt = 0;
 864                 $img_start = strpos($orig_body, '[img');
 865                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 866                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 867                 while (($img_st_close !== false) && ($img_end !== false)) {
 868                         $img_st_close++; // make it point to AFTER the closing bracket
 869                         $img_end += $img_start;
 870
 871                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 872                                 // This is an embedded image
 873                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 874                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 875
 876                                 $cnt++;
 877                         } else {
 878                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 879                         }
 880
 881                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 882
 883                         if ($orig_body === false) {
 884                                 // in case the body ends on a closing image tag
 885                                 $orig_body = '';
 886                         }
 887
 888                         $img_start = strpos($orig_body, '[img');
 889                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 890                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 891                 }
 892
 893                 $new_body = $new_body . $orig_body;
 894
 895                 return ['body' => $new_body, 'images' => $saved_image];
 896         }
 897
 898         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 899         {
 900                 $newbody = $body;
 901
 902                 $cnt = 0;
 903                 foreach ($images as $image) {
 904                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 905                         // it loops over the array starting from the first element and going sequentially
 906                         // to the last element
 907                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 908                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 909                         $cnt++;
 910                 }
 911
 912                 return $newbody;
 913         }
 914
 915         /**
 916          * This function converts a [share] block to text according to a provided callback function whose signature is:
 917          *
 918          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 919          *
 920          * Where:
 921          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 922          * data lookup
 923          * - $author_contact is a contact record array
 924          * - $content is the inner content of the [share] block
 925          * - $is_quote_share indicates whether there's any content before the [share] block
 926          * - Return value is the string that should replace the [share] block in the provided text
 927          *
 928          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 929          *
 930          * @param  string   $text     A BBCode string
 931          * @param  callable $callback
 932          * @return string The BBCode string with all [share] blocks replaced
 933          */
 934         public static function convertShare($text, callable $callback)
 935         {
 936                 $return = preg_replace_callback(
 937                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 938                         function ($match) use ($callback) {
 939                                 $attribute_string = $match[2];
 940                                 $attributes = [];
 941                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 942                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 943                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 944                                 }
 945
 946                                 // We only call this so that a previously unknown contact can be added.
 947                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 948                                 // This function then can fetch an entry from the contact table.
 949                                 $default['url'] = $attributes['profile'];
 950
 951                                 if (!empty($attributes['author'])) {
 952                                         $default['name'] = $attributes['author'];
 953                                 }
 954
 955                                 if (!empty($attributes['avatar'])) {
 956                                         $default['photo'] = $attributes['avatar'];
 957                                 }
 958
 959                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 960
 961                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 962                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 963
 964                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 965                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 966                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 967
 968                                 if ($attributes['avatar']) {
 969                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 970                                 }
 971
 972                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 973                         },
 974                         $text
 975                 );
 976
 977                 return $return;
 978         }
 979
 980         /**
 981          * Default [share] tag conversion callback
 982          *
 983          * Note: Can produce a [bookmark] tag in the output
 984          *
 985          * @see BBCode::convertShare()
 986          * @param array   $attributes     [share] block attribute values
 987          * @param array   $author_contact Contact row of the shared author
 988          * @param string  $content        Inner content of the [share] block
 989          * @param boolean $is_quote_share Whether there is content before the [share] block
 990          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 991          * @return string
 992          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 993          */
 994         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 995         {
 996                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 997
 998                 switch ($simplehtml) {
 999                         case 1:
1000                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1001                                 break;
1002                         case 2:
1003                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1004                                 break;
1005                         case 3: // Diaspora
1006                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1007                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1008                                 } else {
1009                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1010
1011                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1012                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1013                                         }
1014
1015                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1016
1017                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1018                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1019                                         }
1020                                 }
1021
1022                                 break;
1023                         case 4:
1024                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1025                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1026                                 $headline .= ':</b></p>' . "\n";
1027
1028                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1029
1030                                 break;
1031                         case 5:
1032                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1033                                 break;
1034                         case 7: // statusnet/GNU Social
1035                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1036                                 break;
1037                         case 9: // ActivityPub
1038                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1039                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1040                                 break;
1041                         default:
1042                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1043                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1044                                         try {
1045                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1046                                         } catch (Exception $e) {
1047                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1048                                         }
1049                                 } else {
1050                                         $text = ($is_quote_share? "\n" : '');
1051
1052                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1053                                         $text .= Renderer::replaceMacros($tpl, [
1054                                                 '$profile' => $attributes['profile'],
1055                                                 '$avatar'  => $attributes['avatar'],
1056                                                 '$author'  => $attributes['author'],
1057                                                 '$link'    => $attributes['link'],
1058                                                 '$posted'  => $attributes['posted'],
1059                                                 '$content' => trim($content)
1060                                         ]);
1061                                 }
1062                                 break;
1063                 }
1064
1065                 return $text;
1066         }
1067
1068         private static function removePictureLinksCallback($match)
1069         {
1070                 $cache_key = 'remove:' . $match[1];
1071                 $text = DI::cache()->get($cache_key);
1072
1073                 if (is_null($text)) {
1074                         $a = DI::app();
1075
1076                         $stamp1 = microtime(true);
1077
1078                         $ch = @curl_init($match[1]);
1079                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1080                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1081                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1082                         @curl_exec($ch);
1083                         $curl_info = @curl_getinfo($ch);
1084
1085                         DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1086
1087                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1088                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1089                         } else {
1090                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1091
1092                                 // if its not a picture then look if its a page that contains a picture link
1093                                 $body = Network::fetchUrl($match[1]);
1094
1095                                 $doc = new DOMDocument();
1096                                 @$doc->loadHTML($body);
1097                                 $xpath = new DOMXPath($doc);
1098                                 $list = $xpath->query("//meta[@name]");
1099                                 foreach ($list as $node) {
1100                                         $attr = [];
1101
1102                                         if ($node->attributes->length) {
1103                                                 foreach ($node->attributes as $attribute) {
1104                                                         $attr[$attribute->name] = $attribute->value;
1105                                                 }
1106                                         }
1107
1108                                         if (strtolower($attr['name']) == 'twitter:image') {
1109                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1110                                         }
1111                                 }
1112                         }
1113                         DI::cache()->set($cache_key, $text);
1114                 }
1115
1116                 return $text;
1117         }
1118
1119         private static function expandLinksCallback($match)
1120         {
1121                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1122                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1123                 } else {
1124                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1125                 }
1126         }
1127
1128         private static function cleanPictureLinksCallback($match)
1129         {
1130                 $a = DI::app();
1131
1132                 // When the picture link is the own photo path then we can avoid fetching the link
1133                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1134                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1135                         if (!empty($match[3])) {
1136                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1137                         } else {
1138                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1139                         }
1140                         return $text;
1141                 }
1142
1143                 $cache_key = 'clean:' . $match[1];
1144                 $text = DI::cache()->get($cache_key);
1145                 if (!is_null($text)) {
1146                         return $text;
1147                 }
1148
1149                 // Only fetch the header, not the content
1150                 $stamp1 = microtime(true);
1151
1152                 $ch = @curl_init($match[1]);
1153                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1154                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1155                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1156                 @curl_exec($ch);
1157                 $curl_info = @curl_getinfo($ch);
1158
1159                 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1160
1161                 // if its a link to a picture then embed this picture
1162                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1163                         $text = '[img]' . $match[1] . '[/img]';
1164                 } else {
1165                         if (!empty($match[3])) {
1166                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1167                         } else {
1168                                 $text = '[img]' . $match[2] . '[/img]';
1169                         }
1170
1171                         // if its not a picture then look if its a page that contains a picture link
1172                         $body = Network::fetchUrl($match[1]);
1173
1174                         $doc = new DOMDocument();
1175                         @$doc->loadHTML($body);
1176                         $xpath = new DOMXPath($doc);
1177                         $list = $xpath->query("//meta[@name]");
1178                         foreach ($list as $node) {
1179                                 $attr = [];
1180                                 if ($node->attributes->length) {
1181                                         foreach ($node->attributes as $attribute) {
1182                                                 $attr[$attribute->name] = $attribute->value;
1183                                         }
1184                                 }
1185
1186                                 if (strtolower($attr['name']) == "twitter:image") {
1187                                         if (!empty($match[3])) {
1188                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1189                                         } else {
1190                                                 $text = "[img]" . $attr['content'] . "[/img]";
1191                                         }
1192                                 }
1193                         }
1194                 }
1195                 DI::cache()->set($cache_key, $text);
1196
1197                 return $text;
1198         }
1199
1200         public static function cleanPictureLinks($text)
1201         {
1202                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1203                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1204                 return $return;
1205         }
1206
1207         /**
1208          * Converts a BBCode message to HTML message
1209          *
1210          * BBcode 2 HTML was written by WAY2WEB.net
1211          * extended to work with Mistpark/Friendica - Mike Macgirvin
1212          *
1213          * Simple HTML values meaning:
1214          * - 0: Friendica display
1215          * - 1: Unused
1216          * - 2: Used for Windows Phone push, Friendica API
1217          * - 3: Used before converting to Markdown in bb2diaspora.php
1218          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1219          * - 5: Unused
1220          * - 6: Unused
1221          * - 7: Used for dfrn, OStatus
1222          * - 8: Used for Twitter, WP backlink text setting
1223          * - 9: ActivityPub
1224          *
1225          * @param string $text
1226          * @param bool   $try_oembed
1227          * @param int    $simple_html
1228          * @param bool   $for_plaintext
1229          * @return string
1230          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1231          */
1232         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1233         {
1234                 $a = DI::app();
1235
1236                 /*
1237                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1238                  *
1239                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1240                  * $match[1] = $url
1241                  * $match[2] = $title or absent
1242                  */
1243                 $try_oembed_callback = function ($match)
1244                 {
1245                         $url = $match[1];
1246                         $title = $match[2] ?? null;
1247
1248                         try {
1249                                 $return = OEmbed::getHTML($url, $title);
1250                         } catch (Exception $ex) {
1251                                 $return = $match[0];
1252                         }
1253
1254                         return $return;
1255                 };
1256
1257                 // Extracting code blocks before the whitespace processing and the autolinker
1258                 $codeblocks = [];
1259
1260                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1261                         function ($matches) use (&$codeblocks) {
1262                                 $return = '#codeblock-' . count($codeblocks) . '#';
1263                                 if (strpos($matches[2], "\n") !== false) {
1264                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1265                                 } else {
1266                                         $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1267                                 }
1268
1269                                 return $return;
1270                         },
1271                         $text
1272                 );
1273
1274                 // Hide all [noparse] contained bbtags by spacefying them
1275                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1276
1277                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1278                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1279                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1280
1281                 // Remove the abstract element. It is a non visible element.
1282                 $text = self::stripAbstract($text);
1283
1284                 // Move all spaces out of the tags
1285                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1286                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1287
1288                 // Extract the private images which use data urls since preg has issues with
1289                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1290                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1291
1292                 $extracted = self::extractImagesFromItemBody($text);
1293                 $text = $extracted['body'];
1294                 $saved_image = $extracted['images'];
1295
1296                 // If we find any event code, turn it into an event.
1297                 // After we're finished processing the bbcode we'll
1298                 // replace all of the event code with a reformatted version.
1299
1300                 $ev = Event::fromBBCode($text);
1301
1302                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1303                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1304
1305                 $text = str_replace("<", "&lt;", $text);
1306                 $text = str_replace(">", "&gt;", $text);
1307
1308                 // remove some newlines before the general conversion
1309                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1310                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1311
1312                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1313                 if (!$try_oembed) {
1314                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1315                 }
1316
1317                 // Convert new line chars to html <br /> tags
1318
1319                 // nlbr seems to be hopelessly messed up
1320                 //      $Text = nl2br($Text);
1321
1322                 // We'll emulate it.
1323
1324                 $text = trim($text);
1325                 $text = str_replace("\r\n", "\n", $text);
1326
1327                 // Remove linefeeds inside of the table elements. See issue #6799
1328                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1329                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1330                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1331                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1332                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1333                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1334                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1335                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1336                 do {
1337                         $oldtext = $text;
1338                         $text = str_replace($search, $replace, $text);
1339                 } while ($oldtext != $text);
1340
1341                 // Replace these here only once
1342                 $search = ["\n[table]", "[/table]\n"];
1343                 $replace = ["[table]", "[/table]"];
1344                 $text = str_replace($search, $replace, $text);
1345
1346                 // removing multiplicated newlines
1347                 if (Config::get('system', 'remove_multiplicated_lines')) {
1348                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1349                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1350                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1351                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1352                         do {
1353                                 $oldtext = $text;
1354                                 $text = str_replace($search, $replace, $text);
1355                         } while ($oldtext != $text);
1356                 }
1357
1358                 /// @todo Have a closer look at the different html modes
1359                 // Handle attached links or videos
1360                 if (in_array($simple_html, [9])) {
1361                         $text = self::removeAttachment($text);
1362                 } elseif (!in_array($simple_html, [0, 4])) {
1363                         $text = self::removeAttachment($text, true);
1364                 } else {
1365                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1366                 }
1367
1368                 // leave open the posibility of [map=something]
1369                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1370                 if (strpos($text, '[/map]') !== false) {
1371                         $text = preg_replace_callback(
1372                                 "/\[map\](.*?)\[\/map\]/ism",
1373                                 function ($match) use ($simple_html) {
1374                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1375                                 },
1376                                 $text
1377                         );
1378                 }
1379
1380                 if (strpos($text, '[map=') !== false) {
1381                         $text = preg_replace_callback(
1382                                 "/\[map=(.*?)\]/ism",
1383                                 function ($match) use ($simple_html) {
1384                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1385                                 },
1386                                 $text
1387                         );
1388                 }
1389
1390                 if (strpos($text, '[map]') !== false) {
1391                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1392                 }
1393
1394                 // Check for headers
1395                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1396                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1397                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1398                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1399                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1400                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1401
1402                 // Check for paragraph
1403                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1404
1405                 // Check for bold text
1406                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1407
1408                 // Check for Italics text
1409                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1410
1411                 // Check for Underline text
1412                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1413
1414                 // Check for strike-through text
1415                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1416
1417                 // Check for over-line text
1418                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1419
1420                 // Check for colored text
1421                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1422
1423                 // Check for sized text
1424                 // [size=50] --> font-size: 50px (with the unit).
1425                 if ($simple_html != 3) {
1426                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1427                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1428                 } else {
1429                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1430                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1431                 }
1432
1433
1434                 // Check for centered text
1435                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1436
1437                 // Check for list text
1438                 $text = str_replace("[*]", "<li>", $text);
1439
1440                 // Check for style sheet commands
1441                 $text = preg_replace_callback(
1442                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1443                         function ($match) {
1444                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1445                         },
1446                         $text
1447                 );
1448
1449                 // Check for CSS classes
1450                 $text = preg_replace_callback(
1451                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1452                         function ($match) {
1453                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1454                         },
1455                         $text
1456                 );
1457
1458                 // handle nested lists
1459                 $endlessloop = 0;
1460
1461                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1462                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1463                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1464                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1465                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1466                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1467                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1468                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1469                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1470                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1471                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1472                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1473                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1474                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1475                 }
1476
1477                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1478                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1479                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1480                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1481
1482                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1483                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1484
1485                 $text = str_replace('[hr]', '<hr />', $text);
1486
1487                 if (!$for_plaintext) {
1488                         $escaped = [];
1489
1490                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1491                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1492                                 function ($matches) use (&$escaped) {
1493                                         $return = '{escaped-' . count($escaped) . '}';
1494                                         $escaped[] = $matches[0];
1495
1496                                         return $return;
1497                                 },
1498                                 $text
1499                         );
1500
1501                         // Autolinker for isolated URLs
1502                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1503
1504                         // Restoring escaped blocks
1505                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1506                                 function ($matches) use ($escaped) {
1507                                         return $escaped[intval($matches[1])] ?? $matches[0];
1508                                 },
1509                                 $text
1510                         );
1511                 }
1512
1513                 // This is actually executed in Item::prepareBody()
1514
1515                 $nosmile = strpos($text, '[nosmile]') !== false;
1516                 $text = str_replace('[nosmile]', '', $text);
1517
1518                 // Check for font change text
1519                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1520
1521                 // Declare the format for [spoiler] layout
1522                 $SpoilerLayout = '<details class="spoiler"><summary>' . L10n::t('Click to open/close') . '</summary>$1</details>';
1523
1524                 // Check for [spoiler] text
1525                 // handle nested quotes
1526                 $endlessloop = 0;
1527                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1528                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1529                 }
1530
1531                 // Check for [spoiler=Title] text
1532
1533                 // handle nested quotes
1534                 $endlessloop = 0;
1535                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1536                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1537                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1538                                 $text);
1539                 }
1540
1541                 // Declare the format for [quote] layout
1542                 $QuoteLayout = '<blockquote>$1</blockquote>';
1543
1544                 // Check for [quote] text
1545                 // handle nested quotes
1546                 $endlessloop = 0;
1547                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1548                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1549                 }
1550
1551                 // Check for [quote=Author] text
1552
1553                 $t_wrote = L10n::t('$1 wrote:');
1554
1555                 // handle nested quotes
1556                 $endlessloop = 0;
1557                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1558                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1559                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1560                                 $text);
1561                 }
1562
1563
1564                 // [img=widthxheight]image source[/img]
1565                 $text = preg_replace_callback(
1566                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1567                         function ($matches) use ($simple_html) {
1568                                 if (strpos($matches[3], "data:image/") === 0) {
1569                                         return $matches[0];
1570                                 }
1571
1572                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1573                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1574                         },
1575                         $text
1576                 );
1577
1578                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1579                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1580
1581                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1582                         function ($matches) use ($simple_html) {
1583                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1584                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1585                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1586                         },
1587                         $text);
1588
1589                 // Images
1590                 // [img]pathtoimage[/img]
1591                 $text = preg_replace_callback(
1592                         "/\[img\](.*?)\[\/img\]/ism",
1593                         function ($matches) use ($simple_html) {
1594                                 if (strpos($matches[1], "data:image/") === 0) {
1595                                         return $matches[0];
1596                                 }
1597
1598                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1599                                 return "[img]" . $matches[1] . "[/img]";
1600                         },
1601                         $text
1602                 );
1603
1604                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1605                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1606
1607                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1608                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1609                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1610
1611                 // Simplify "video" element
1612                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1613
1614                 // Try to Oembed
1615                 if ($try_oembed) {
1616                         $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);
1617                         $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);
1618
1619                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1620                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1621                 } else {
1622                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1623                                 '<a href="$1" target="_blank">$1</a>', $text);
1624                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1625                                 '<a href="$1" target="_blank">$1</a>', $text);
1626                 }
1627
1628                 // html5 video and audio
1629
1630
1631                 if ($try_oembed) {
1632                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1633                 } else {
1634                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1635                 }
1636
1637                 // Youtube extensions
1638                 if ($try_oembed) {
1639                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1640                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1641                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1642                 }
1643
1644                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1645                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1646                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1647
1648                 if ($try_oembed) {
1649                         $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);
1650                 } else {
1651                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1652                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1653                 }
1654
1655                 if ($try_oembed) {
1656                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1657                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1658                 }
1659
1660                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1661                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1662
1663                 if ($try_oembed) {
1664                         $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);
1665                 } else {
1666                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1667                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1668                 }
1669
1670                 // oembed tag
1671                 $text = OEmbed::BBCode2HTML($text);
1672
1673                 // Avoid triple linefeeds through oembed
1674                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1675
1676                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1677                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1678                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1679                 // start which is always required). Allow desc with a missing summary for compatibility.
1680
1681                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1682                         $sub = Event::getHTML($ev, $simple_html);
1683
1684                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1685                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1686                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1687                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1688                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1689                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1690                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1691                 }
1692
1693                 // Replace non graphical smilies for external posts
1694                 if (!$nosmile && !$for_plaintext) {
1695                         $text = Smilies::replace($text);
1696                 }
1697
1698                 if (!$for_plaintext) {
1699                         if (in_array($simple_html, [7, 9])) {
1700                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1701                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1702                         }
1703                 } else {
1704                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1705                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1706                 }
1707
1708                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1709
1710                 // Remove all hashtag addresses
1711                 if ($simple_html && !in_array($simple_html, [3, 7, 9])) {
1712                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1713                 } elseif ($simple_html == 3) {
1714                         // The ! is converted to @ since Diaspora only understands the @
1715                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1716                                 '@<a href="$2">$3</a>',
1717                                 $text);
1718                 } elseif (in_array($simple_html, [7, 9])) {
1719                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1720                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1721                                 $text);
1722                 } elseif (!$simple_html) {
1723                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1724                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1725                                 $text);
1726                 }
1727
1728                 // Bookmarks in red - will be converted to bookmarks in friendica
1729                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1730                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1731                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1732                                         "[bookmark=$1]$2[/bookmark]", $text);
1733
1734                 if (in_array($simple_html, [2, 6, 7, 8])) {
1735                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1736                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1737                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1738                 }
1739
1740                 if ($simple_html == 5) {
1741                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1742                 }
1743
1744                 // Perform URL Search
1745                 if ($try_oembed) {
1746                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1747                 }
1748
1749                 if ($simple_html == 5) {
1750                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1751                 } else {
1752                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1753                 }
1754
1755                 // Handle Diaspora posts
1756                 $text = preg_replace_callback(
1757                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1758                         function ($match) {
1759                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1760                         }, $text
1761                 );
1762
1763                 $text = preg_replace_callback(
1764                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1765                         function ($match) {
1766                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1767                         }, $text
1768                 );
1769
1770                 // Server independent link to posts and comments
1771                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1772                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1773                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1774
1775                 /* Tag conversion
1776                  * Supports:
1777                  * - #[url=<anything>]<term>[/url]
1778                  * - [url=<anything>]#<term>[/url]
1779                  */
1780                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1781                         return '#<a href="'
1782                                 . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1783                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1784                                 . XML::escape($matches[1])
1785                                 . '</a>';
1786                 }, $text);
1787
1788                 // We need no target="_blank" for local links
1789                 // convert links start with DI::baseUrl() as local link without the target="_blank" attribute
1790                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1791                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1792                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1793
1794                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1795                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1796
1797                 // Red compatibility, though the link can't be authenticated on Friendica
1798                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1799
1800
1801                 // we may need to restrict this further if it picks up too many strays
1802                 // link acct:user@host to a webfinger profile redirector
1803
1804                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1805
1806                 // Perform MAIL Search
1807                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1808                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1809
1810                 // Unhide all [noparse] contained bbtags unspacefying them
1811                 // and triming the [noparse] tag.
1812
1813                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1814                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1815                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1816
1817                 /// @todo What is the meaning of these lines?
1818                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1819                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1820
1821                 // Currently deactivated, it made problems with " inside of alt texts.
1822                 //$text = preg_replace('/\&quot\;/', '"', $text);
1823
1824                 // fix any escaped ampersands that may have been converted into links
1825                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1826
1827                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1828                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1829
1830                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1831
1832                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1833                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1834
1835                 // sanitize href attributes (only whitelisted protocols URLs)
1836                 // default value for backward compatibility
1837                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1838
1839                 // Always allowed protocol even if config isn't set or not including it
1840                 $allowed_link_protocols[] = '//';
1841                 $allowed_link_protocols[] = 'http://';
1842                 $allowed_link_protocols[] = 'https://';
1843                 $allowed_link_protocols[] = 'redir/';
1844
1845                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1846
1847                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1848                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1849
1850                 // Shared content
1851                 $text = self::convertShare(
1852                         $text,
1853                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1854                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1855                         }
1856                 );
1857
1858                 if ($saved_image) {
1859                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1860                 }
1861
1862                 // Restore code blocks
1863                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1864                         function ($matches) use ($codeblocks) {
1865                                 $return = $matches[0];
1866                                 if (isset($codeblocks[intval($matches[1])])) {
1867                                         $return = $codeblocks[$matches[1]];
1868                                 }
1869                                 return $return;
1870                         },
1871                         $text
1872                 );
1873
1874                 // Clean up the HTML by loading and saving the HTML with the DOM.
1875                 // Bad structured html can break a whole page.
1876                 // For performance reasons do it only with activated item cache or at export.
1877                 if (!$try_oembed || (get_itemcachepath() != '')) {
1878                         $doc = new DOMDocument();
1879                         $doc->preserveWhiteSpace = false;
1880
1881                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1882
1883                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1884                         $encoding = '<?xml encoding="UTF-8">';
1885                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1886                         $doc->encoding = 'UTF-8';
1887                         $text = $doc->saveHTML();
1888                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1889
1890                         $text = str_replace('<br></li>', '</li>', $text);
1891
1892                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1893                 }
1894
1895                 // Clean up some useless linebreaks in lists
1896                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1897                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1898                 //$Text = str_replace('</li><br />', '</li>', $Text);
1899                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1900                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1901
1902                 Hook::callAll('bbcode', $text);
1903
1904                 return trim($text);
1905         }
1906
1907         /**
1908          * Strips the "abstract" tag from the provided text
1909          *
1910          * @param string $text The text with BBCode
1911          * @return string The same text - but without "abstract" element
1912          */
1913         public static function stripAbstract($text)
1914         {
1915                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1916                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1917
1918                 return $text;
1919         }
1920
1921         /**
1922          * Returns the value of the "abstract" element
1923          *
1924          * @param string $text The text that maybe contains the element
1925          * @param string $addon The addon for which the abstract is meant for
1926          * @return string The abstract
1927          */
1928         public static function getAbstract($text, $addon = '')
1929         {
1930                 $abstract = '';
1931                 $abstracts = [];
1932                 $addon = strtolower($addon);
1933
1934                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1935                         foreach ($results AS $result) {
1936                                 $abstracts[strtolower($result[1])] = $result[2];
1937                         }
1938                 }
1939
1940                 if (isset($abstracts[$addon])) {
1941                         $abstract = $abstracts[$addon];
1942                 }
1943
1944                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1945                         $abstract = $result[1];
1946                 }
1947
1948                 return $abstract;
1949         }
1950
1951         /**
1952          * Callback function to replace a Friendica style mention in a mention for Diaspora
1953          *
1954          * @param array $match Matching values for the callback
1955          *                     [1] = Mention type (! or @)
1956          *                     [2] = Name
1957          *                     [3] = Address
1958          * @return string Replaced mention
1959          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1960          * @throws \ImagickException
1961          */
1962         private static function bbCodeMention2DiasporaCallback($match)
1963         {
1964                 $contact = Contact::getDetailsByURL($match[3]);
1965
1966                 if (empty($contact['addr'])) {
1967                         $contact = Probe::uri($match[3]);
1968                 }
1969
1970                 if (empty($contact['addr'])) {
1971                         return $match[0];
1972                 }
1973
1974                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1975                 return $mention;
1976         }
1977
1978         /**
1979          * Converts a BBCode text into Markdown
1980          *
1981          * This function converts a BBCode item body to be sent to Markdown-enabled
1982          * systems like Diaspora and Libertree
1983          *
1984          * @param string $text
1985          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1986          * @return string
1987          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1988          */
1989         public static function toMarkdown($text, $for_diaspora = true)
1990         {
1991                 $a = DI::app();
1992
1993                 $original_text = $text;
1994
1995                 // Since Diaspora is creating a summary for links, this function removes them before posting
1996                 if ($for_diaspora) {
1997                         $text = self::removeShareInformation($text);
1998                 }
1999
2000                 /**
2001                  * Transform #tags, strip off the [url] and replace spaces with underscore
2002                  */
2003                 $url_search_string = "^\[\]";
2004                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2005                         function ($matches) {
2006                                 return '#' . str_replace(' ', '_', $matches[2]);
2007                         },
2008                         $text
2009                 );
2010
2011                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2012                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2013
2014                 // Convert it to HTML - don't try oembed
2015                 if ($for_diaspora) {
2016                         $text = self::convert($text, false, 3);
2017
2018                         // Add all tags that maybe were removed
2019                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2020                                 $tagline = '';
2021                                 foreach ($tags[2] as $tag) {
2022                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2023                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2024                                                 $tagline .= '#' . $tag . ' ';
2025                                         }
2026                                 }
2027                                 $text = $text . " " . $tagline;
2028                         }
2029                 } else {
2030                         $text = self::convert($text, false, 4);
2031                 }
2032
2033                 // If a link is followed by a quote then there should be a newline before it
2034                 // Maybe we should make this newline at every time before a quote.
2035                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2036
2037                 $stamp1 = microtime(true);
2038
2039                 // Now convert HTML to Markdown
2040                 $text = HTML::toMarkdown($text);
2041
2042                 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2043
2044                 // Libertree has a problem with escaped hashtags.
2045                 $text = str_replace(['\#'], ['#'], $text);
2046
2047                 // Remove any leading or trailing whitespace, as this will mess up
2048                 // the Diaspora signature verification and cause the item to disappear
2049                 $text = trim($text);
2050
2051                 if ($for_diaspora) {
2052                         $url_search_string = "^\[\]";
2053                         $text = preg_replace_callback(
2054                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2055                                 ['self', 'bbCodeMention2DiasporaCallback'],
2056                                 $text
2057                         );
2058                 }
2059
2060                 Hook::callAll('bb2diaspora', $text);
2061
2062                 return $text;
2063         }
2064
2065         /**
2066      * Pull out all #hashtags and @person tags from $string.
2067      *
2068      * We also get @person@domain.com - which would make
2069      * the regex quite complicated as tags can also
2070      * end a sentence. So we'll run through our results
2071      * and strip the period from any tags which end with one.
2072      * Returns array of tags found, or empty array.
2073      *
2074      * @param string $string Post content
2075      *
2076      * @return array List of tag and person names
2077      */
2078     public static function getTags($string)
2079     {
2080         $ret = [];
2081
2082         // Convert hashtag links to hashtags
2083         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2084
2085         // ignore anything in a code block
2086         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2087
2088         // Force line feeds at bbtags
2089         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2090
2091         // ignore anything in a bbtag
2092         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2093
2094         // Match full names against @tags including the space between first and last
2095         // We will look these up afterward to see if they are full names or not recognisable.
2096
2097         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2098             foreach ($matches[1] as $match) {
2099                 if (strstr($match, ']')) {
2100                     // we might be inside a bbcode color tag - leave it alone
2101                     continue;
2102                 }
2103
2104                 if (substr($match, -1, 1) === '.') {
2105                     $ret[] = substr($match, 0, -1);
2106                 } else {
2107                     $ret[] = $match;
2108                 }
2109             }
2110         }
2111
2112         // Otherwise pull out single word tags. These can be @nickname, @first_last
2113         // and #hash tags.
2114
2115         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2116             foreach ($matches[1] as $match) {
2117                 if (strstr($match, ']')) {
2118                     // we might be inside a bbcode color tag - leave it alone
2119                     continue;
2120                 }
2121                 if (substr($match, -1, 1) === '.') {
2122                     $match = substr($match,0,-1);
2123                 }
2124                 // ignore strictly numeric tags like #1
2125                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2126                     continue;
2127                 }
2128                 // try not to catch url fragments
2129                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2130                     continue;
2131                 }
2132                 $ret[] = $match;
2133             }
2134         }
2135
2136         return $ret;
2137     }
2138 }