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\Cache;
  14 use Friendica\Core\Config;
  15 use Friendica\Core\Hook;
  16 use Friendica\Core\L10n;
  17 use Friendica\Core\Logger;
  18 use Friendica\Core\Protocol;
  19 use Friendica\Core\Renderer;
  20 use Friendica\Core\System;
  21 use Friendica\DI;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Model\Photo;
  25 use Friendica\Network\Probe;
  26 use Friendica\Object\Image;
  27 use Friendica\Protocol\Activity;
  28 use Friendica\Util\Images;
  29 use Friendica\Util\Map;
  30 use Friendica\Util\Network;
  31 use Friendica\Util\ParseUrl;
  32 use Friendica\Util\Proxy as ProxyUtils;
  33 use Friendica\Util\Strings;
  34 use Friendica\Util\XML;
  35
  36 class BBCode
  37 {
  38         /**
  39          * @brief Fetches attachment data that were generated the old way
  40          *
  41          * @param string $body Message body
  42          * @return array
  43          *                     'type' -> Message type ('link', 'video', 'photo')
  44          *                     'text' -> Text before the shared message
  45          *                     'after' -> Text after the shared message
  46          *                     'image' -> Preview image of the message
  47          *                     'url' -> Url to the attached message
  48          *                     'title' -> Title of the attachment
  49          *                     'description' -> Description of the attachment
  50          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  51          */
  52         private static function getOldAttachmentData($body)
  53         {
  54                 $post = [];
  55
  56                 // Simplify image codes
  57                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  58
  59                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  60                         foreach ($attached as $data) {
  61                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  62                                         continue;
  63                                 }
  64
  65                                 $post['type'] = substr($data[1], 5);
  66
  67                                 $pos = strpos($body, $data[0]);
  68                                 if ($pos > 0) {
  69                                         $post['text'] = trim(substr($body, 0, $pos));
  70                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  71                                 } else {
  72                                         $post['text'] = trim(str_replace($data[0], '', $body));
  73                                         $post['after'] = '';
  74                                 }
  75
  76                                 $attacheddata = $data[2];
  77
  78                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
  79
  80                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
  81
  82                                         if ($picturedata) {
  83                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  84                                                         $post['image'] = $matches[1];
  85                                                 } else {
  86                                                         $post['preview'] = $matches[1];
  87                                                 }
  88                                         }
  89                                 }
  90
  91                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  92                                         $post['url'] = $matches[1];
  93                                         $post['title'] = $matches[2];
  94                                 }
  95                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
  96                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  97                                         $post['url'] = $matches[1];
  98                                 }
  99
 100                                 // Search for description
 101                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 102                                         $post['description'] = $matches[1];
 103                                 }
 104                         }
 105                 }
 106                 return $post;
 107         }
 108
 109         /**
 110          * @brief Fetches attachment data that were generated with the "attachment" element
 111          *
 112          * @param string $body Message body
 113          * @return array
 114          *                     'type' -> Message type ('link', 'video', 'photo')
 115          *                     'text' -> Text before the shared message
 116          *                     'after' -> Text after the shared message
 117          *                     'image' -> Preview image of the message
 118          *                     'url' -> Url to the attached message
 119          *                     'title' -> Title of the attachment
 120          *                     'description' -> Description of the attachment
 121          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 122          */
 123         public static function getAttachmentData($body)
 124         {
 125                 $data = [
 126                         'type'        => '',
 127                         'text'        => '',
 128                         'after'       => '',
 129                         'image'       => null,
 130                         'url'         => '',
 131                         'title'       => '',
 132                         'description' => '',
 133                 ];
 134
 135                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 136                         return self::getOldAttachmentData($body);
 137                 }
 138
 139                 $attributes = $match[2];
 140
 141                 $data['text'] = trim($match[1]);
 142
 143                 $type = '';
 144                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 145                 if (!empty($matches[1])) {
 146                         $type = strtolower($matches[1]);
 147                 }
 148
 149                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 150                 if (!empty($matches[1])) {
 151                         $type = strtolower($matches[1]);
 152                 }
 153
 154                 if ($type == '') {
 155                         return [];
 156                 }
 157
 158                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 159                         return [];
 160                 }
 161
 162                 if ($type != '') {
 163                         $data['type'] = $type;
 164                 }
 165
 166                 $url = '';
 167                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 168                 if (!empty($matches[1])) {
 169                         $url = $matches[1];
 170                 }
 171
 172                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 173                 if (!empty($matches[1])) {
 174                         $url = $matches[1];
 175                 }
 176
 177                 if ($url != '') {
 178                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 179                 }
 180
 181                 $title = '';
 182                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 183                 if (!empty($matches[1])) {
 184                         $title = $matches[1];
 185                 }
 186
 187                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 188                 if (!empty($matches[1])) {
 189                         $title = $matches[1];
 190                 }
 191
 192                 if ($title != '') {
 193                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 194                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 195                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 196                         $data['title'] = $title;
 197                 }
 198
 199                 $image = '';
 200                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 201                 if (!empty($matches[1])) {
 202                         $image = $matches[1];
 203                 }
 204
 205                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 206                 if (!empty($matches[1])) {
 207                         $image = $matches[1];
 208                 }
 209
 210                 if ($image != '') {
 211                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 212                 }
 213
 214                 $preview = '';
 215                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 216                 if (!empty($matches[1])) {
 217                         $preview = $matches[1];
 218                 }
 219
 220                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 221                 if (!empty($matches[1])) {
 222                         $preview = $matches[1];
 223                 }
 224
 225                 if ($preview != '') {
 226                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 227                 }
 228
 229                 $data['description'] = trim($match[3]);
 230
 231                 $data['after'] = trim($match[4]);
 232
 233                 return $data;
 234         }
 235
 236         public static function getAttachedData($body, $item = [])
 237         {
 238                 /*
 239                 - text:
 240                 - type: link, video, photo
 241                 - title:
 242                 - url:
 243                 - image:
 244                 - description:
 245                 - (thumbnail)
 246                 */
 247
 248                 $has_title = !empty($item['title']);
 249                 $plink = $item['plink'] ?? '';
 250                 $post = self::getAttachmentData($body);
 251
 252                 // Get all linked images with alternative image description
 253                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 254                         foreach ($pictures as $picture) {
 255                                 if (Photo::isLocal($picture[1])) {
 256                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 257                                 }
 258                         }
 259                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 260                                 $post['image_description'] = $post['images'][0]['description'];
 261                         }
 262                 }
 263
 264                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 265                         foreach ($pictures as $picture) {
 266                                 if (Photo::isLocal($picture[1])) {
 267                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 268                                 }
 269                         }
 270                 }
 271
 272                 // if nothing is found, it maybe having an image.
 273                 if (!isset($post['type'])) {
 274                         // Simplify image codes
 275                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 276                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 277                         $post['text'] = $body;
 278
 279                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 280                                 if ((count($pictures) == 1) && !$has_title) {
 281                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 282                                                 // Replace the preview picture with the real picture
 283                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 284                                                 $data = ['url' => $url, 'type' => 'photo'];
 285                                         } else {
 286                                                 // Checking, if the link goes to a picture
 287                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 288                                         }
 289
 290                                         // Workaround:
 291                                         // Sometimes photo posts to the own album are not detected at the start.
 292                                         // So we seem to cannot use the cache for these cases. That's strange.
 293                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 294                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 295                                         }
 296
 297                                         if ($data['type'] == 'photo') {
 298                                                 $post['type'] = 'photo';
 299                                                 if (isset($data['images'][0])) {
 300                                                         $post['image'] = $data['images'][0]['src'];
 301                                                         $post['url'] = $data['url'];
 302                                                 } else {
 303                                                         $post['image'] = $data['url'];
 304                                                 }
 305
 306                                                 $post['preview'] = $pictures[0][2];
 307                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 308                                         } else {
 309                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 310                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 311                                                         $post['type'] = 'photo';
 312                                                         $post['image'] = $pictures[0][1];
 313                                                         $post['preview'] = $pictures[0][2];
 314                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 315                                                 }
 316                                         }
 317                                 } elseif (count($pictures) > 0) {
 318                                         $post['type'] = 'link';
 319                                         $post['url'] = $plink;
 320                                         $post['image'] = $pictures[0][2];
 321                                         $post['text'] = $body;
 322
 323                                         foreach ($pictures as $picture) {
 324                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 325                                         }
 326                                 }
 327                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 328                                 if ((count($pictures) == 1) && !$has_title) {
 329                                         $post['type'] = 'photo';
 330                                         $post['image'] = $pictures[0][1];
 331                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 332                                 } elseif (count($pictures) > 0) {
 333                                         $post['type'] = 'link';
 334                                         $post['url'] = $plink;
 335                                         $post['image'] = $pictures[0][1];
 336                                         $post['text'] = $body;
 337
 338                                         foreach ($pictures as $picture) {
 339                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 340                                         }
 341                                 }
 342                         }
 343
 344                         // Test for the external links
 345                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 346                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 347
 348                         $links = array_merge($links1, $links2);
 349
 350                         // If there is only a single one, then use it.
 351                         // This should cover link posts via API.
 352                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 353                                 $post['type'] = 'link';
 354                                 $post['url'] = $links[0][1];
 355                         }
 356
 357                         // Simplify "video" element
 358                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 359
 360                         // Now count the number of external media links
 361                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 362                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 363                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 364                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 365
 366                         // Add them to the other external links
 367                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 368
 369                         // Are there more than one?
 370                         if (count($links) > 1) {
 371                                 // The post will be the type "text", which means a blog post
 372                                 unset($post['type']);
 373                                 $post['url'] = $plink;
 374                         }
 375
 376                         if (!isset($post['type'])) {
 377                                 $post['type'] = "text";
 378                                 $post['text'] = trim($body);
 379                         }
 380                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 381                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 382
 383                         if (isset($data['images'][0])) {
 384                                 $post['image'] = $data['images'][0]['src'];
 385                         }
 386                 }
 387
 388                 return $post;
 389         }
 390
 391         /**
 392          * Remove [attachment] BBCode and replaces it with a regular [url]
 393          *
 394          * @param string  $body
 395          * @param boolean $no_link_desc No link description
 396          *
 397          * @return string with replaced body
 398          */
 399         public static function removeAttachment($body, $no_link_desc = false)
 400         {
 401                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 402                         function ($match) use ($no_link_desc) {
 403                                 $attach_data = self::getAttachmentData($match[0]);
 404                                 if (empty($attach_data['url'])) {
 405                                         return $match[0];
 406                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 407                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 408                                 } else {
 409                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 410                                 }
 411                 }, $body);
 412         }
 413
 414         /**
 415          * @brief Converts a BBCode text into plaintext
 416          *
 417          * @param      $text
 418          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 419          *
 420          * @return string
 421          */
 422         public static function toPlaintext($text, $keep_urls = true)
 423         {
 424                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 425
 426                 return $naked_text;
 427         }
 428
 429         private static function proxyUrl($image, $simplehtml = false)
 430         {
 431                 // Only send proxied pictures to API and for internal display
 432                 if (in_array($simplehtml, [false, 2])) {
 433                         return ProxyUtils::proxifyUrl($image);
 434                 } else {
 435                         return $image;
 436                 }
 437         }
 438
 439         public static function scaleExternalImages($srctext)
 440         {
 441                 // Picture addresses can contain special characters
 442                 $s = $srctext;
 443
 444                 $matches = null;
 445                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 446                 if ($c) {
 447                         foreach ($matches as $mtch) {
 448                                 Logger::log('scale_external_image: ' . $mtch[1]);
 449
 450                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 451                                 if (stristr($mtch[1], $hostname)) {
 452                                         continue;
 453                                 }
 454
 455                                 $i = Network::fetchUrl($mtch[1]);
 456                                 if (!$i) {
 457                                         return $srctext;
 458                                 }
 459
 460                                 // guess mimetype from headers or filename
 461                                 $type = Images::guessType($mtch[1], true);
 462
 463                                 if ($i) {
 464                                         $Image = new Image($i, $type);
 465                                         if ($Image->isValid()) {
 466                                                 $orig_width = $Image->getWidth();
 467                                                 $orig_height = $Image->getHeight();
 468
 469                                                 if ($orig_width > 640 || $orig_height > 640) {
 470                                                         $Image->scaleDown(640);
 471                                                         $new_width = $Image->getWidth();
 472                                                         $new_height = $Image->getHeight();
 473                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 474                                                         $s = str_replace(
 475                                                                 $mtch[0],
 476                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 477                                                                 . "\n",
 478                                                                 $s
 479                                                         );
 480                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 481                                                 }
 482                                         }
 483                                 }
 484                         }
 485                 }
 486
 487                 return $s;
 488         }
 489
 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          * @brief Truncates imported message body string length to max_import_size
 495          * @param string $body
 496          * @return string
 497          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 498          */
 499         public static function limitBodySize($body)
 500         {
 501                 $maxlen = Config::get('config', 'max_import_size', 0);
 502
 503                 // If the length of the body, including the embedded images, is smaller
 504                 // than the maximum, then don't waste time looking for the images
 505                 if ($maxlen && (strlen($body) > $maxlen)) {
 506
 507                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 508
 509                         $orig_body = $body;
 510                         $new_body = '';
 511                         $textlen = 0;
 512
 513                         $img_start = strpos($orig_body, '[img');
 514                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 515                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 516                         while (($img_st_close !== false) && ($img_end !== false)) {
 517
 518                                 $img_st_close++; // make it point to AFTER the closing bracket
 519                                 $img_end += $img_start;
 520                                 $img_end += strlen('[/img]');
 521
 522                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 523                                         // This is an embedded image
 524
 525                                         if (($textlen + $img_start) > $maxlen) {
 526                                                 if ($textlen < $maxlen) {
 527                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 528                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 529                                                         $textlen = $maxlen;
 530                                                 }
 531                                         } else {
 532                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 533                                                 $textlen += $img_start;
 534                                         }
 535
 536                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 537                                 } else {
 538
 539                                         if (($textlen + $img_end) > $maxlen) {
 540                                                 if ($textlen < $maxlen) {
 541                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 542                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 543                                                         $textlen = $maxlen;
 544                                                 }
 545                                         } else {
 546                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 547                                                 $textlen += $img_end;
 548                                         }
 549                                 }
 550                                 $orig_body = substr($orig_body, $img_end);
 551
 552                                 if ($orig_body === false) {
 553                                         // in case the body ends on a closing image tag
 554                                         $orig_body = '';
 555                                 }
 556
 557                                 $img_start = strpos($orig_body, '[img');
 558                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 559                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 560                         }
 561
 562                         if (($textlen + strlen($orig_body)) > $maxlen) {
 563                                 if ($textlen < $maxlen) {
 564                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 565                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 566                                 }
 567                         } else {
 568                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 569                                 $new_body = $new_body . $orig_body;
 570                         }
 571
 572                         return $new_body;
 573                 } else {
 574                         return $body;
 575                 }
 576         }
 577
 578         /**
 579          * Processes [attachment] tags
 580          *
 581          * Note: Can produce a [bookmark] tag in the returned string
 582          *
 583          * @brief Processes [attachment] tags
 584          * @param string   $text
 585          * @param bool|int $simplehtml
 586          * @param bool     $tryoembed
 587          * @return string
 588          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 589          */
 590         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 591         {
 592                 $data = self::getAttachmentData($text);
 593                 if (empty($data) || empty($data['url'])) {
 594                         return $text;
 595                 }
 596
 597                 if (isset($data['title'])) {
 598                         $data['title'] = strip_tags($data['title']);
 599                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 600                 } else {
 601                         $data['title'] = null;
 602                 }
 603
 604                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data['image'])) {
 605                         $data['preview'] = $data['image'];
 606                         $data['image'] = '';
 607                 }
 608
 609                 $return = '';
 610                 try {
 611                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 612                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 613                         } else {
 614                                 throw new Exception('OEmbed is disabled for this attachment.');
 615                         }
 616                 } catch (Exception $e) {
 617                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 618
 619                         if ($simplehtml != 4) {
 620                                 $return = sprintf('<div class="type-%s">', $data['type']);
 621                         }
 622
 623                         if (!empty($data['title']) && !empty($data['url'])) {
 624                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 625                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 626                                 } else {
 627                                         if (!empty($data['image'])) {
 628                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 629                                         } elseif (!empty($data['preview'])) {
 630                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']);
 631                                         }
 632                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 633                                 }
 634                         }
 635
 636                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 637                                 // Sanitize the HTML by converting it to BBCode
 638                                 $bbcode = HTML::toBBCode($data['description']);
 639                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 640                         }
 641
 642                         if (!empty($data['url'])) {
 643                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 644                         }
 645
 646                         if ($simplehtml != 4) {
 647                                 $return .= '</div>';
 648                         }
 649                 }
 650
 651                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 652         }
 653
 654         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 655         {
 656                 $data = self::getAttachmentData($Text);
 657
 658                 if (!$data) {
 659                         return $Text;
 660                 } elseif ($nolink) {
 661                         return $data['text'] . ($data['after'] ?? '');
 662                 }
 663
 664                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 665                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 666                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 667                         $data['title'] = $data['url'];
 668                 } elseif (($text != '') && strstr($title, $text)) {
 669                         $data['text'] = $data['title'];
 670                         $data['title'] = $data['url'];
 671                 }
 672
 673                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 674                         return $data['title'] . $data['after'];
 675                 }
 676
 677                 // If the link already is included in the post, don't add it again
 678                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 679                         return $data['text'] . $data['after'];
 680                 }
 681
 682                 $text = $data['text'];
 683
 684                 if (!empty($data['url']) && !empty($data['title'])) {
 685                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 686                 } elseif (!empty($data['url'])) {
 687                         $text .= "\n[url]" . $data['url'] . '[/url]';
 688                 }
 689
 690                 return $text . "\n" . $data['after'];
 691         }
 692
 693         /**
 694          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 695          *
 696          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 697          * @param array $match Array with the matching values
 698          * @return string reformatted link including HTML codes
 699          */
 700         private static function convertUrlForActivityPubCallback($match)
 701         {
 702                 $url = $match[1];
 703
 704                 if (isset($match[2]) && ($match[1] != $match[2])) {
 705                         return $match[0];
 706                 }
 707
 708                 $parts = parse_url($url);
 709                 if (!isset($parts['scheme'])) {
 710                         return $match[0];
 711                 }
 712
 713                 return self::convertUrlForActivityPub($url);
 714         }
 715
 716         /**
 717          * @brief Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 718          * @param string $url URL that is about to be reformatted
 719          * @return string reformatted link including HTML codes
 720          */
 721         private static function convertUrlForActivityPub($url)
 722         {
 723                 $html = '<a href="%s" target="_blank">%s</a>';
 724                 return sprintf($html, $url, self::getStyledURL($url));
 725         }
 726
 727         /**
 728          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 729          * @param string $url URL that is about to be reformatted
 730          * @return string reformatted link
 731          */
 732         private static function getStyledURL($url)
 733         {
 734                 $parts = parse_url($url);
 735                 $scheme = $parts['scheme'] . '://';
 736                 $styled_url = str_replace($scheme, '', $url);
 737
 738                 if (strlen($styled_url) > 30) {
 739                         $styled_url = substr($styled_url, 0, 30) . "…";
 740                 }
 741
 742                 return $styled_url;
 743         }
 744
 745         /*
 746          * [noparse][i]italic[/i][/noparse] turns into
 747          * [noparse][ i ]italic[ /i ][/noparse],
 748          * to hide them from parser.
 749          */
 750         private static function escapeNoparseCallback($match)
 751         {
 752                 $whole_match = $match[0];
 753                 $captured = $match[1];
 754                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 755                 $new_str = str_replace($captured, $spacefied, $whole_match);
 756                 return $new_str;
 757         }
 758
 759         /*
 760          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 761          * now turns back and the [noparse] tags are trimed
 762          * returning [i]italic[/i]
 763          */
 764         private static function unescapeNoparseCallback($match)
 765         {
 766                 $captured = $match[1];
 767                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 768                 return $unspacefied;
 769         }
 770
 771         /**
 772          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 773          * occurrences
 774          *
 775          * @param string $text        Text to search
 776          * @param string $name        Tag name
 777          * @param int    $occurrences Number of first occurrences to skip
 778          * @return boolean|array
 779          */
 780         public static function getTagPosition($text, $name, $occurrences = 0)
 781         {
 782                 if ($occurrences < 0) {
 783                         $occurrences = 0;
 784                 }
 785
 786                 $start_open = -1;
 787                 for ($i = 0; $i <= $occurrences; $i++) {
 788                         if ($start_open !== false) {
 789                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 790                         }
 791                 }
 792
 793                 if ($start_open === false) {
 794                         return false;
 795                 }
 796
 797                 $start_equal = strpos($text, '=', $start_open);
 798                 $start_close = strpos($text, ']', $start_open);
 799
 800                 if ($start_close === false) {
 801                         return false;
 802                 }
 803
 804                 $start_close++;
 805
 806                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 807
 808                 if ($end_open === false) {
 809                         return false;
 810                 }
 811
 812                 $res = [
 813                         'start' => [
 814                                 'open' => $start_open,
 815                                 'close' => $start_close
 816                         ],
 817                         'end' => [
 818                                 'open' => $end_open,
 819                                 'close' => $end_open + strlen('[/' . $name . ']')
 820                         ],
 821                 ];
 822
 823                 if ($start_equal !== false) {
 824                         $res['start']['equal'] = $start_equal + 1;
 825                 }
 826
 827                 return $res;
 828         }
 829
 830         /**
 831          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 832          *
 833          * @param string $pattern Preg pattern string
 834          * @param string $replace Preg replace string
 835          * @param string $name    BBCode tag name
 836          * @param string $text    Text to search
 837          * @return string
 838          */
 839         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 840         {
 841                 $occurrences = 0;
 842                 $pos = self::getTagPosition($text, $name, $occurrences);
 843                 while ($pos !== false && $occurrences++ < 1000) {
 844                         $start = substr($text, 0, $pos['start']['open']);
 845                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 846                         $end = substr($text, $pos['end']['close']);
 847                         if ($end === false) {
 848                                 $end = '';
 849                         }
 850
 851                         $subject = preg_replace($pattern, $replace, $subject);
 852                         $text = $start . $subject . $end;
 853
 854                         $pos = self::getTagPosition($text, $name, $occurrences);
 855                 }
 856
 857                 return $text;
 858         }
 859
 860         private static function extractImagesFromItemBody($body)
 861         {
 862                 $saved_image = [];
 863                 $orig_body = $body;
 864                 $new_body = '';
 865
 866                 $cnt = 0;
 867                 $img_start = strpos($orig_body, '[img');
 868                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 869                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 870                 while (($img_st_close !== false) && ($img_end !== false)) {
 871                         $img_st_close++; // make it point to AFTER the closing bracket
 872                         $img_end += $img_start;
 873
 874                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 875                                 // This is an embedded image
 876                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 877                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 878
 879                                 $cnt++;
 880                         } else {
 881                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 882                         }
 883
 884                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 885
 886                         if ($orig_body === false) {
 887                                 // in case the body ends on a closing image tag
 888                                 $orig_body = '';
 889                         }
 890
 891                         $img_start = strpos($orig_body, '[img');
 892                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 893                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 894                 }
 895
 896                 $new_body = $new_body . $orig_body;
 897
 898                 return ['body' => $new_body, 'images' => $saved_image];
 899         }
 900
 901         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 902         {
 903                 $newbody = $body;
 904
 905                 $cnt = 0;
 906                 foreach ($images as $image) {
 907                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 908                         // it loops over the array starting from the first element and going sequentially
 909                         // to the last element
 910                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 911                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 912                         $cnt++;
 913                 }
 914
 915                 return $newbody;
 916         }
 917
 918         /**
 919          * This function converts a [share] block to text according to a provided callback function whose signature is:
 920          *
 921          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 922          *
 923          * Where:
 924          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 925          * data lookup
 926          * - $author_contact is a contact record array
 927          * - $content is the inner content of the [share] block
 928          * - $is_quote_share indicates whether there's any content before the [share] block
 929          * - Return value is the string that should replace the [share] block in the provided text
 930          *
 931          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 932          *
 933          * @param  string   $text     A BBCode string
 934          * @param  callable $callback
 935          * @return string The BBCode string with all [share] blocks replaced
 936          */
 937         public static function convertShare($text, callable $callback)
 938         {
 939                 $return = preg_replace_callback(
 940                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 941                         function ($match) use ($callback) {
 942                                 $attribute_string = $match[2];
 943                                 $attributes = [];
 944                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 945                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 946                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 947                                 }
 948
 949                                 // We only call this so that a previously unknown contact can be added.
 950                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 951                                 // This function then can fetch an entry from the contact table.
 952                                 $default['url'] = $attributes['profile'];
 953
 954                                 if (!empty($attributes['author'])) {
 955                                         $default['name'] = $attributes['author'];
 956                                 }
 957
 958                                 if (!empty($attributes['avatar'])) {
 959                                         $default['photo'] = $attributes['avatar'];
 960                                 }
 961
 962                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 963
 964                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 965                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 966
 967                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 968                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 969                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 970
 971                                 if ($attributes['avatar']) {
 972                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 973                                 }
 974
 975                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 976                         },
 977                         $text
 978                 );
 979
 980                 return $return;
 981         }
 982
 983         /**
 984          * Default [share] tag conversion callback
 985          *
 986          * Note: Can produce a [bookmark] tag in the output
 987          *
 988          * @see BBCode::convertShare()
 989          * @param array   $attributes     [share] block attribute values
 990          * @param array   $author_contact Contact row of the shared author
 991          * @param string  $content        Inner content of the [share] block
 992          * @param boolean $is_quote_share Whether there is content before the [share] block
 993          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 994          * @return string
 995          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 996          */
 997         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 998         {
 999                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1000
1001                 switch ($simplehtml) {
1002                         case 1:
1003                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1004                                 break;
1005                         case 2:
1006                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1007                                 break;
1008                         case 3: // Diaspora
1009                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1010                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1011                                 } else {
1012                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1013
1014                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1015                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1016                                         }
1017
1018                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1019
1020                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1021                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1022                                         }
1023                                 }
1024
1025                                 break;
1026                         case 4:
1027                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1028                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1029                                 $headline .= ':</b></p>' . "\n";
1030
1031                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1032
1033                                 break;
1034                         case 5:
1035                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1036                                 break;
1037                         case 7: // statusnet/GNU Social
1038                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1039                                 break;
1040                         case 9: // ActivityPub
1041                                 $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>';
1042                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1043                                 break;
1044                         default:
1045                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1046                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1047                                         try {
1048                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1049                                         } catch (Exception $e) {
1050                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1051                                         }
1052                                 } else {
1053                                         $text = ($is_quote_share? "\n" : '');
1054
1055                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1056                                         $text .= Renderer::replaceMacros($tpl, [
1057                                                 '$profile' => $attributes['profile'],
1058                                                 '$avatar'  => $attributes['avatar'],
1059                                                 '$author'  => $attributes['author'],
1060                                                 '$link'    => $attributes['link'],
1061                                                 '$posted'  => $attributes['posted'],
1062                                                 '$content' => trim($content)
1063                                         ]);
1064                                 }
1065                                 break;
1066                 }
1067
1068                 return $text;
1069         }
1070
1071         private static function removePictureLinksCallback($match)
1072         {
1073                 $cache_key = 'remove:' . $match[1];
1074                 $text = Cache::get($cache_key);
1075
1076                 if (is_null($text)) {
1077                         $a = DI::app();
1078
1079                         $stamp1 = microtime(true);
1080
1081                         $ch = @curl_init($match[1]);
1082                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1083                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1084                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1085                         @curl_exec($ch);
1086                         $curl_info = @curl_getinfo($ch);
1087
1088                         DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1089
1090                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1091                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1092                         } else {
1093                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1094
1095                                 // if its not a picture then look if its a page that contains a picture link
1096                                 $body = Network::fetchUrl($match[1]);
1097
1098                                 $doc = new DOMDocument();
1099                                 @$doc->loadHTML($body);
1100                                 $xpath = new DOMXPath($doc);
1101                                 $list = $xpath->query("//meta[@name]");
1102                                 foreach ($list as $node) {
1103                                         $attr = [];
1104
1105                                         if ($node->attributes->length) {
1106                                                 foreach ($node->attributes as $attribute) {
1107                                                         $attr[$attribute->name] = $attribute->value;
1108                                                 }
1109                                         }
1110
1111                                         if (strtolower($attr['name']) == 'twitter:image') {
1112                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1113                                         }
1114                                 }
1115                         }
1116                         Cache::set($cache_key, $text);
1117                 }
1118
1119                 return $text;
1120         }
1121
1122         private static function expandLinksCallback($match)
1123         {
1124                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1125                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1126                 } else {
1127                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1128                 }
1129         }
1130
1131         private static function cleanPictureLinksCallback($match)
1132         {
1133                 $a = DI::app();
1134
1135                 // When the picture link is the own photo path then we can avoid fetching the link
1136                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1137                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1138                         if (!empty($match[3])) {
1139                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1140                         } else {
1141                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1142                         }
1143                         return $text;
1144                 }
1145
1146                 $cache_key = 'clean:' . $match[1];
1147                 $text = Cache::get($cache_key);
1148                 if (!is_null($text)) {
1149                         return $text;
1150                 }
1151
1152                 // Only fetch the header, not the content
1153                 $stamp1 = microtime(true);
1154
1155                 $ch = @curl_init($match[1]);
1156                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1157                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1158                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1159                 @curl_exec($ch);
1160                 $curl_info = @curl_getinfo($ch);
1161
1162                 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1163
1164                 // if its a link to a picture then embed this picture
1165                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1166                         $text = '[img]' . $match[1] . '[/img]';
1167                 } else {
1168                         if (!empty($match[3])) {
1169                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1170                         } else {
1171                                 $text = '[img]' . $match[2] . '[/img]';
1172                         }
1173
1174                         // if its not a picture then look if its a page that contains a picture link
1175                         $body = Network::fetchUrl($match[1]);
1176
1177                         $doc = new DOMDocument();
1178                         @$doc->loadHTML($body);
1179                         $xpath = new DOMXPath($doc);
1180                         $list = $xpath->query("//meta[@name]");
1181                         foreach ($list as $node) {
1182                                 $attr = [];
1183                                 if ($node->attributes->length) {
1184                                         foreach ($node->attributes as $attribute) {
1185                                                 $attr[$attribute->name] = $attribute->value;
1186                                         }
1187                                 }
1188
1189                                 if (strtolower($attr['name']) == "twitter:image") {
1190                                         if (!empty($match[3])) {
1191                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1192                                         } else {
1193                                                 $text = "[img]" . $attr['content'] . "[/img]";
1194                                         }
1195                                 }
1196                         }
1197                 }
1198                 Cache::set($cache_key, $text);
1199
1200                 return $text;
1201         }
1202
1203         public static function cleanPictureLinks($text)
1204         {
1205                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1206                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1207                 return $return;
1208         }
1209
1210         /**
1211          * @brief Converts a BBCode message to HTML message
1212          *
1213          * BBcode 2 HTML was written by WAY2WEB.net
1214          * extended to work with Mistpark/Friendica - Mike Macgirvin
1215          *
1216          * Simple HTML values meaning:
1217          * - 0: Friendica display
1218          * - 1: Unused
1219          * - 2: Used for Windows Phone push, Friendica API
1220          * - 3: Used before converting to Markdown in bb2diaspora.php
1221          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1222          * - 5: Unused
1223          * - 6: Unused
1224          * - 7: Used for dfrn, OStatus
1225          * - 8: Used for Twitter, WP backlink text setting
1226          * - 9: ActivityPub
1227          *
1228          * @param string $text
1229          * @param bool   $try_oembed
1230          * @param int    $simple_html
1231          * @param bool   $for_plaintext
1232          * @return string
1233          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1234          */
1235         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1236         {
1237                 $a = DI::app();
1238
1239                 /*
1240                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1241                  *
1242                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1243                  * $match[1] = $url
1244                  * $match[2] = $title or absent
1245                  */
1246                 $try_oembed_callback = function ($match)
1247                 {
1248                         $url = $match[1];
1249                         $title = $match[2] ?? null;
1250
1251                         try {
1252                                 $return = OEmbed::getHTML($url, $title);
1253                         } catch (Exception $ex) {
1254                                 $return = $match[0];
1255                         }
1256
1257                         return $return;
1258                 };
1259
1260                 // Extracting code blocks before the whitespace processing and the autolinker
1261                 $codeblocks = [];
1262
1263                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1264                         function ($matches) use (&$codeblocks) {
1265                                 $return = '#codeblock-' . count($codeblocks) . '#';
1266                                 if (strpos($matches[2], "\n") !== false) {
1267                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1268                                 } else {
1269                                         $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1270                                 }
1271
1272                                 return $return;
1273                         },
1274                         $text
1275                 );
1276
1277                 // Hide all [noparse] contained bbtags by spacefying them
1278                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1279
1280                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1281                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1282                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1283
1284                 // Remove the abstract element. It is a non visible element.
1285                 $text = self::stripAbstract($text);
1286
1287                 // Move all spaces out of the tags
1288                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1289                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1290
1291                 // Extract the private images which use data urls since preg has issues with
1292                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1293                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1294
1295                 $extracted = self::extractImagesFromItemBody($text);
1296                 $text = $extracted['body'];
1297                 $saved_image = $extracted['images'];
1298
1299                 // If we find any event code, turn it into an event.
1300                 // After we're finished processing the bbcode we'll
1301                 // replace all of the event code with a reformatted version.
1302
1303                 $ev = Event::fromBBCode($text);
1304
1305                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1306                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1307
1308                 $text = str_replace("<", "&lt;", $text);
1309                 $text = str_replace(">", "&gt;", $text);
1310
1311                 // remove some newlines before the general conversion
1312                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1313                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1314
1315                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1316                 if (!$try_oembed) {
1317                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1318                 }
1319
1320                 // Convert new line chars to html <br /> tags
1321
1322                 // nlbr seems to be hopelessly messed up
1323                 //      $Text = nl2br($Text);
1324
1325                 // We'll emulate it.
1326
1327                 $text = trim($text);
1328                 $text = str_replace("\r\n", "\n", $text);
1329
1330                 // Remove linefeeds inside of the table elements. See issue #6799
1331                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1332                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1333                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1334                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1335                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1336                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1337                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1338                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1339                 do {
1340                         $oldtext = $text;
1341                         $text = str_replace($search, $replace, $text);
1342                 } while ($oldtext != $text);
1343
1344                 // Replace these here only once
1345                 $search = ["\n[table]", "[/table]\n"];
1346                 $replace = ["[table]", "[/table]"];
1347                 $text = str_replace($search, $replace, $text);
1348
1349                 // removing multiplicated newlines
1350                 if (Config::get('system', 'remove_multiplicated_lines')) {
1351                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1352                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1353                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1354                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1355                         do {
1356                                 $oldtext = $text;
1357                                 $text = str_replace($search, $replace, $text);
1358                         } while ($oldtext != $text);
1359                 }
1360
1361                 /// @todo Have a closer look at the different html modes
1362                 // Handle attached links or videos
1363                 if (in_array($simple_html, [9])) {
1364                         $text = self::removeAttachment($text);
1365                 } elseif (!in_array($simple_html, [0, 4])) {
1366                         $text = self::removeAttachment($text, true);
1367                 } else {
1368                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1369                 }
1370
1371                 // leave open the posibility of [map=something]
1372                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1373                 if (strpos($text, '[/map]') !== false) {
1374                         $text = preg_replace_callback(
1375                                 "/\[map\](.*?)\[\/map\]/ism",
1376                                 function ($match) use ($simple_html) {
1377                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1378                                 },
1379                                 $text
1380                         );
1381                 }
1382
1383                 if (strpos($text, '[map=') !== false) {
1384                         $text = preg_replace_callback(
1385                                 "/\[map=(.*?)\]/ism",
1386                                 function ($match) use ($simple_html) {
1387                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1388                                 },
1389                                 $text
1390                         );
1391                 }
1392
1393                 if (strpos($text, '[map]') !== false) {
1394                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1395                 }
1396
1397                 // Check for headers
1398                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1399                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1400                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1401                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1402                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1403                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1404
1405                 // Check for paragraph
1406                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1407
1408                 // Check for bold text
1409                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1410
1411                 // Check for Italics text
1412                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1413
1414                 // Check for Underline text
1415                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1416
1417                 // Check for strike-through text
1418                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1419
1420                 // Check for over-line text
1421                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1422
1423                 // Check for colored text
1424                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1425
1426                 // Check for sized text
1427                 // [size=50] --> font-size: 50px (with the unit).
1428                 if ($simple_html != 3) {
1429                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1430                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1431                 } else {
1432                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1433                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1434                 }
1435
1436
1437                 // Check for centered text
1438                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1439
1440                 // Check for list text
1441                 $text = str_replace("[*]", "<li>", $text);
1442
1443                 // Check for style sheet commands
1444                 $text = preg_replace_callback(
1445                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1446                         function ($match) {
1447                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1448                         },
1449                         $text
1450                 );
1451
1452                 // Check for CSS classes
1453                 $text = preg_replace_callback(
1454                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1455                         function ($match) {
1456                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1457                         },
1458                         $text
1459                 );
1460
1461                 // handle nested lists
1462                 $endlessloop = 0;
1463
1464                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1465                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1466                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1467                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1468                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1469                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1470                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1471                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1472                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1473                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1474                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1475                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1476                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1477                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1478                 }
1479
1480                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1481                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1482                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1483                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1484
1485                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1486                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1487
1488                 $text = str_replace('[hr]', '<hr />', $text);
1489
1490                 if (!$for_plaintext) {
1491                         $escaped = [];
1492
1493                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1494                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1495                                 function ($matches) use (&$escaped) {
1496                                         $return = '{escaped-' . count($escaped) . '}';
1497                                         $escaped[] = $matches[0];
1498
1499                                         return $return;
1500                                 },
1501                                 $text
1502                         );
1503
1504                         // Autolinker for isolated URLs
1505                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1506
1507                         // Restoring escaped blocks
1508                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1509                                 function ($matches) use ($escaped) {
1510                                         return $escaped[intval($matches[1])] ?? $matches[0];
1511                                 },
1512                                 $text
1513                         );
1514                 }
1515
1516                 // This is actually executed in Item::prepareBody()
1517
1518                 $nosmile = strpos($text, '[nosmile]') !== false;
1519                 $text = str_replace('[nosmile]', '', $text);
1520
1521                 // Check for font change text
1522                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1523
1524                 // Declare the format for [spoiler] layout
1525                 $SpoilerLayout = '<details class="spoiler"><summary>' . L10n::t('Click to open/close') . '</summary>$1</details>';
1526
1527                 // Check for [spoiler] text
1528                 // handle nested quotes
1529                 $endlessloop = 0;
1530                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1531                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1532                 }
1533
1534                 // Check for [spoiler=Title] text
1535
1536                 // handle nested quotes
1537                 $endlessloop = 0;
1538                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1539                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1540                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1541                                 $text);
1542                 }
1543
1544                 // Declare the format for [quote] layout
1545                 $QuoteLayout = '<blockquote>$1</blockquote>';
1546
1547                 // Check for [quote] text
1548                 // handle nested quotes
1549                 $endlessloop = 0;
1550                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1551                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1552                 }
1553
1554                 // Check for [quote=Author] text
1555
1556                 $t_wrote = L10n::t('$1 wrote:');
1557
1558                 // handle nested quotes
1559                 $endlessloop = 0;
1560                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1561                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1562                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1563                                 $text);
1564                 }
1565
1566
1567                 // [img=widthxheight]image source[/img]
1568                 $text = preg_replace_callback(
1569                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1570                         function ($matches) use ($simple_html) {
1571                                 if (strpos($matches[3], "data:image/") === 0) {
1572                                         return $matches[0];
1573                                 }
1574
1575                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1576                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1577                         },
1578                         $text
1579                 );
1580
1581                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1582                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1583
1584                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1585                         function ($matches) use ($simple_html) {
1586                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1587                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1588                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1589                         },
1590                         $text);
1591
1592                 // Images
1593                 // [img]pathtoimage[/img]
1594                 $text = preg_replace_callback(
1595                         "/\[img\](.*?)\[\/img\]/ism",
1596                         function ($matches) use ($simple_html) {
1597                                 if (strpos($matches[1], "data:image/") === 0) {
1598                                         return $matches[0];
1599                                 }
1600
1601                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1602                                 return "[img]" . $matches[1] . "[/img]";
1603                         },
1604                         $text
1605                 );
1606
1607                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1608                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1609
1610                 $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);
1611                 $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 />', $text);
1612                 //$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);
1613
1614                 // Simplify "video" element
1615                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1616
1617                 // Try to Oembed
1618                 if ($try_oembed) {
1619                         $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);
1620                         $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);
1621
1622                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1623                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1624                 } else {
1625                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1626                                 '<a href="$1" target="_blank">$1</a>', $text);
1627                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1628                                 '<a href="$1" target="_blank">$1</a>', $text);
1629                 }
1630
1631                 // html5 video and audio
1632
1633
1634                 if ($try_oembed) {
1635                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1636                 } else {
1637                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1638                 }
1639
1640                 // Youtube extensions
1641                 if ($try_oembed) {
1642                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1643                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1644                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1645                 }
1646
1647                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1648                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1649                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1650
1651                 if ($try_oembed) {
1652                         $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);
1653                 } else {
1654                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1655                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1656                 }
1657
1658                 if ($try_oembed) {
1659                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1660                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1661                 }
1662
1663                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1664                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1665
1666                 if ($try_oembed) {
1667                         $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);
1668                 } else {
1669                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1670                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1671                 }
1672
1673                 // oembed tag
1674                 $text = OEmbed::BBCode2HTML($text);
1675
1676                 // Avoid triple linefeeds through oembed
1677                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1678
1679                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1680                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1681                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1682                 // start which is always required). Allow desc with a missing summary for compatibility.
1683
1684                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1685                         $sub = Event::getHTML($ev, $simple_html);
1686
1687                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1688                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1689                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1690                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1691                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1692                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1693                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1694                 }
1695
1696                 // Replace non graphical smilies for external posts
1697                 if (!$nosmile && !$for_plaintext) {
1698                         $text = Smilies::replace($text);
1699                 }
1700
1701                 if (!$for_plaintext) {
1702                         if (in_array($simple_html, [7, 9])) {
1703                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1704                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1705                         }
1706                 } else {
1707                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1708                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1709                 }
1710
1711                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1712
1713                 // Remove all hashtag addresses
1714                 if ($simple_html && !in_array($simple_html, [3, 7, 9])) {
1715                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1716                 } elseif ($simple_html == 3) {
1717                         // The ! is converted to @ since Diaspora only understands the @
1718                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1719                                 '@<a href="$2">$3</a>',
1720                                 $text);
1721                 } elseif (in_array($simple_html, [7, 9])) {
1722                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1723                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1724                                 $text);
1725                 } elseif (!$simple_html) {
1726                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1727                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1728                                 $text);
1729                 }
1730
1731                 // Bookmarks in red - will be converted to bookmarks in friendica
1732                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1733                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1734                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1735                                         "[bookmark=$1]$2[/bookmark]", $text);
1736
1737                 if (in_array($simple_html, [2, 6, 7, 8])) {
1738                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1739                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1740                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1741                 }
1742
1743                 if ($simple_html == 5) {
1744                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1745                 }
1746
1747                 // Perform URL Search
1748                 if ($try_oembed) {
1749                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1750                 }
1751
1752                 if ($simple_html == 5) {
1753                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1754                 } else {
1755                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1756                 }
1757
1758                 // Handle Diaspora posts
1759                 $text = preg_replace_callback(
1760                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1761                         function ($match) {
1762                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1763                         }, $text
1764                 );
1765
1766                 $text = preg_replace_callback(
1767                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1768                         function ($match) {
1769                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1770                         }, $text
1771                 );
1772
1773                 // Server independent link to posts and comments
1774                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1775                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1776                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1777
1778                 /* Tag conversion
1779                  * Supports:
1780                  * - #[url=<anything>]<term>[/url]
1781                  * - [url=<anything>]#<term>[/url]
1782                  */
1783                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1784                         return '#<a href="'
1785                                 . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1786                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1787                                 . XML::escape($matches[1])
1788                                 . '</a>';
1789                 }, $text);
1790
1791                 // We need no target="_blank" for local links
1792                 // convert links start with DI::baseUrl() as local link without the target="_blank" attribute
1793                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1794                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1795                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1796
1797                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1798                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1799
1800                 // Red compatibility, though the link can't be authenticated on Friendica
1801                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1802
1803
1804                 // we may need to restrict this further if it picks up too many strays
1805                 // link acct:user@host to a webfinger profile redirector
1806
1807                 $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);
1808
1809                 // Perform MAIL Search
1810                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1811                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1812
1813                 // Unhide all [noparse] contained bbtags unspacefying them
1814                 // and triming the [noparse] tag.
1815
1816                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1817                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1818                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1819
1820                 /// @todo What is the meaning of these lines?
1821                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1822                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1823
1824                 // Currently deactivated, it made problems with " inside of alt texts.
1825                 //$text = preg_replace('/\&quot\;/', '"', $text);
1826
1827                 // fix any escaped ampersands that may have been converted into links
1828                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1829
1830                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1831                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1832
1833                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1834
1835                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1836                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1837
1838                 // sanitize href attributes (only whitelisted protocols URLs)
1839                 // default value for backward compatibility
1840                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1841
1842                 // Always allowed protocol even if config isn't set or not including it
1843                 $allowed_link_protocols[] = '//';
1844                 $allowed_link_protocols[] = 'http://';
1845                 $allowed_link_protocols[] = 'https://';
1846                 $allowed_link_protocols[] = 'redir/';
1847
1848                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1849
1850                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1851                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1852
1853                 // Shared content
1854                 $text = self::convertShare(
1855                         $text,
1856                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1857                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1858                         }
1859                 );
1860
1861                 if ($saved_image) {
1862                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1863                 }
1864
1865                 // Restore code blocks
1866                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1867                         function ($matches) use ($codeblocks) {
1868                                 $return = $matches[0];
1869                                 if (isset($codeblocks[intval($matches[1])])) {
1870                                         $return = $codeblocks[$matches[1]];
1871                                 }
1872                                 return $return;
1873                         },
1874                         $text
1875                 );
1876
1877                 // Clean up the HTML by loading and saving the HTML with the DOM.
1878                 // Bad structured html can break a whole page.
1879                 // For performance reasons do it only with activated item cache or at export.
1880                 if (!$try_oembed || (get_itemcachepath() != '')) {
1881                         $doc = new DOMDocument();
1882                         $doc->preserveWhiteSpace = false;
1883
1884                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1885
1886                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1887                         $encoding = '<?xml encoding="UTF-8">';
1888                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1889                         $doc->encoding = 'UTF-8';
1890                         $text = $doc->saveHTML();
1891                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1892
1893                         $text = str_replace('<br></li>', '</li>', $text);
1894
1895                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1896                 }
1897
1898                 // Clean up some useless linebreaks in lists
1899                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1900                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1901                 //$Text = str_replace('</li><br />', '</li>', $Text);
1902                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1903                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1904
1905                 Hook::callAll('bbcode', $text);
1906
1907                 return trim($text);
1908         }
1909
1910         /**
1911          * @brief Strips the "abstract" tag from the provided text
1912          *
1913          * @param string $text The text with BBCode
1914          * @return string The same text - but without "abstract" element
1915          */
1916         public static function stripAbstract($text)
1917         {
1918                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1919                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1920
1921                 return $text;
1922         }
1923
1924         /**
1925          * @brief Returns the value of the "abstract" element
1926          *
1927          * @param string $text The text that maybe contains the element
1928          * @param string $addon The addon for which the abstract is meant for
1929          * @return string The abstract
1930          */
1931         public static function getAbstract($text, $addon = '')
1932         {
1933                 $abstract = '';
1934                 $abstracts = [];
1935                 $addon = strtolower($addon);
1936
1937                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1938                         foreach ($results AS $result) {
1939                                 $abstracts[strtolower($result[1])] = $result[2];
1940                         }
1941                 }
1942
1943                 if (isset($abstracts[$addon])) {
1944                         $abstract = $abstracts[$addon];
1945                 }
1946
1947                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1948                         $abstract = $result[1];
1949                 }
1950
1951                 return $abstract;
1952         }
1953
1954         /**
1955          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1956          *
1957          * @param array $match Matching values for the callback
1958          *                     [1] = Mention type (! or @)
1959          *                     [2] = Name
1960          *                     [3] = Address
1961          * @return string Replaced mention
1962          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1963          * @throws \ImagickException
1964          */
1965         private static function bbCodeMention2DiasporaCallback($match)
1966         {
1967                 $contact = Contact::getDetailsByURL($match[3]);
1968
1969                 if (empty($contact['addr'])) {
1970                         $contact = Probe::uri($match[3]);
1971                 }
1972
1973                 if (empty($contact['addr'])) {
1974                         return $match[0];
1975                 }
1976
1977                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1978                 return $mention;
1979         }
1980
1981         /**
1982          * @brief Converts a BBCode text into Markdown
1983          *
1984          * This function converts a BBCode item body to be sent to Markdown-enabled
1985          * systems like Diaspora and Libertree
1986          *
1987          * @param string $text
1988          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1989          * @return string
1990          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1991          */
1992         public static function toMarkdown($text, $for_diaspora = true)
1993         {
1994                 $a = DI::app();
1995
1996                 $original_text = $text;
1997
1998                 // Since Diaspora is creating a summary for links, this function removes them before posting
1999                 if ($for_diaspora) {
2000                         $text = self::removeShareInformation($text);
2001                 }
2002
2003                 /**
2004                  * Transform #tags, strip off the [url] and replace spaces with underscore
2005                  */
2006                 $url_search_string = "^\[\]";
2007                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2008                         function ($matches) {
2009                                 return '#' . str_replace(' ', '_', $matches[2]);
2010                         },
2011                         $text
2012                 );
2013
2014                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2015                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2016
2017                 // Convert it to HTML - don't try oembed
2018                 if ($for_diaspora) {
2019                         $text = self::convert($text, false, 3);
2020
2021                         // Add all tags that maybe were removed
2022                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2023                                 $tagline = '';
2024                                 foreach ($tags[2] as $tag) {
2025                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2026                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2027                                                 $tagline .= '#' . $tag . ' ';
2028                                         }
2029                                 }
2030                                 $text = $text . " " . $tagline;
2031                         }
2032                 } else {
2033                         $text = self::convert($text, false, 4);
2034                 }
2035
2036                 // If a link is followed by a quote then there should be a newline before it
2037                 // Maybe we should make this newline at every time before a quote.
2038                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2039
2040                 $stamp1 = microtime(true);
2041
2042                 // Now convert HTML to Markdown
2043                 $text = HTML::toMarkdown($text);
2044
2045                 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2046
2047                 // Libertree has a problem with escaped hashtags.
2048                 $text = str_replace(['\#'], ['#'], $text);
2049
2050                 // Remove any leading or trailing whitespace, as this will mess up
2051                 // the Diaspora signature verification and cause the item to disappear
2052                 $text = trim($text);
2053
2054                 if ($for_diaspora) {
2055                         $url_search_string = "^\[\]";
2056                         $text = preg_replace_callback(
2057                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2058                                 ['self', 'bbCodeMention2DiasporaCallback'],
2059                                 $text
2060                         );
2061                 }
2062
2063                 Hook::callAll('bb2diaspora', $text);
2064
2065                 return $text;
2066         }
2067
2068         /**
2069      * @brief Pull out all #hashtags and @person tags from $string.
2070      *
2071      * We also get @person@domain.com - which would make
2072      * the regex quite complicated as tags can also
2073      * end a sentence. So we'll run through our results
2074      * and strip the period from any tags which end with one.
2075      * Returns array of tags found, or empty array.
2076      *
2077      * @param string $string Post content
2078      *
2079      * @return array List of tag and person names
2080      */
2081     public static function getTags($string)
2082     {
2083         $ret = [];
2084
2085         // Convert hashtag links to hashtags
2086         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2087
2088         // ignore anything in a code block
2089         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2090
2091         // Force line feeds at bbtags
2092         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2093
2094         // ignore anything in a bbtag
2095         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2096
2097         // Match full names against @tags including the space between first and last
2098         // We will look these up afterward to see if they are full names or not recognisable.
2099
2100         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2101             foreach ($matches[1] as $match) {
2102                 if (strstr($match, ']')) {
2103                     // we might be inside a bbcode color tag - leave it alone
2104                     continue;
2105                 }
2106
2107                 if (substr($match, -1, 1) === '.') {
2108                     $ret[] = substr($match, 0, -1);
2109                 } else {
2110                     $ret[] = $match;
2111                 }
2112             }
2113         }
2114
2115         // Otherwise pull out single word tags. These can be @nickname, @first_last
2116         // and #hash tags.
2117
2118         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2119             foreach ($matches[1] as $match) {
2120                 if (strstr($match, ']')) {
2121                     // we might be inside a bbcode color tag - leave it alone
2122                     continue;
2123                 }
2124                 if (substr($match, -1, 1) === '.') {
2125                     $match = substr($match,0,-1);
2126                 }
2127                 // ignore strictly numeric tags like #1
2128                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2129                     continue;
2130                 }
2131                 // try not to catch url fragments
2132                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2133                     continue;
2134                 }
2135                 $ret[] = $match;
2136             }
2137         }
2138
2139         return $ret;
2140     }
2141 }