src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\Hook;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Logger;
  19 use Friendica\Core\Protocol;
  20 use Friendica\Core\Renderer;
  21 use Friendica\Core\System;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Model\Photo;
  25 use Friendica\Network\Probe;
  26 use Friendica\Object\Image;
  27 use Friendica\Protocol\Activity;
  28 use Friendica\Util\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 extends BaseObject
  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                         // Now count the number of external media links
 358                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 359                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 360                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 361                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 362
 363                         // Add them to the other external links
 364                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 365
 366                         // Are there more than one?
 367                         if (count($links) > 1) {
 368                                 // The post will be the type "text", which means a blog post
 369                                 unset($post['type']);
 370                                 $post['url'] = $plink;
 371                         }
 372
 373                         if (!isset($post['type'])) {
 374                                 $post['type'] = "text";
 375                                 $post['text'] = trim($body);
 376                         }
 377                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 378                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 379
 380                         if (isset($data['images'][0])) {
 381                                 $post['image'] = $data['images'][0]['src'];
 382                         }
 383                 }
 384
 385                 return $post;
 386         }
 387
 388         /**
 389          * Remove [attachment] BBCode and replaces it with a regular [url]
 390          *
 391          * @param string  $body
 392          * @param boolean $no_link_desc No link description
 393          *
 394          * @return string with replaced body
 395          */
 396         public static function removeAttachment($body, $no_link_desc = false)
 397         {
 398                 return preg_replace_callback("/\[attachment (.*)\](.*?)\[\/attachment\]/ism",
 399                         function ($match) use ($no_link_desc) {
 400                                 $attach_data = self::getAttachmentData($match[0]);
 401                                 if (empty($attach_data['url'])) {
 402                                         return $match[0];
 403                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 404                                         return '[url]' . $attach_data['url'] . "[/url]\n";
 405                                 } else {
 406                                         return '[url=' . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 407                                 }
 408                 }, $body);
 409         }
 410
 411         /**
 412          * @brief Converts a BBCode text into plaintext
 413          *
 414          * @param      $text
 415          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 416          *
 417          * @return string
 418          */
 419         public static function toPlaintext($text, $keep_urls = true)
 420         {
 421                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 422
 423                 return $naked_text;
 424         }
 425
 426         private static function proxyUrl($image, $simplehtml = false)
 427         {
 428                 // Only send proxied pictures to API and for internal display
 429                 if (in_array($simplehtml, [false, 2])) {
 430                         return ProxyUtils::proxifyUrl($image);
 431                 } else {
 432                         return $image;
 433                 }
 434         }
 435
 436         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 437         {
 438                 // Suppress "view full size"
 439                 if (intval(Config::get('system', 'no_view_full_size'))) {
 440                         $include_link = false;
 441                 }
 442
 443                 // Picture addresses can contain special characters
 444                 $s = htmlspecialchars_decode($srctext);
 445
 446                 $matches = null;
 447                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 448                 if ($c) {
 449                         foreach ($matches as $mtch) {
 450                                 Logger::log('scale_external_image: ' . $mtch[1]);
 451
 452                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 453                                 if (stristr($mtch[1], $hostname)) {
 454                                         continue;
 455                                 }
 456
 457                                 // $scale_replace, if passed, is an array of two elements. The
 458                                 // first is the name of the full-size image. The second is the
 459                                 // name of a remote, scaled-down version of the full size image.
 460                                 // This allows Friendica to display the smaller remote image if
 461                                 // one exists, while still linking to the full-size image
 462                                 if ($scale_replace) {
 463                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 464                                 } else {
 465                                         $scaled = $mtch[1];
 466                                 }
 467                                 $i = Network::fetchUrl($scaled);
 468                                 if (!$i) {
 469                                         return $srctext;
 470                                 }
 471
 472                                 // guess mimetype from headers or filename
 473                                 $type = Images::guessType($mtch[1], true);
 474
 475                                 if ($i) {
 476                                         $Image = new Image($i, $type);
 477                                         if ($Image->isValid()) {
 478                                                 $orig_width = $Image->getWidth();
 479                                                 $orig_height = $Image->getHeight();
 480
 481                                                 if ($orig_width > 640 || $orig_height > 640) {
 482                                                         $Image->scaleDown(640);
 483                                                         $new_width = $Image->getWidth();
 484                                                         $new_height = $Image->getHeight();
 485                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 486                                                         $s = str_replace(
 487                                                                 $mtch[0],
 488                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 489                                                                 . "\n" . (($include_link)
 490                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 491                                                                         : ''),
 492                                                                 $s
 493                                                         );
 494                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 495                                                 }
 496                                         }
 497                                 }
 498                         }
 499                 }
 500
 501                 // replace the special char encoding
 502                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 503                 return $s;
 504         }
 505
 506         /**
 507          * The purpose of this function is to apply system message length limits to
 508          * imported messages without including any embedded photos in the length
 509          *
 510          * @brief Truncates imported message body string length to max_import_size
 511          * @param string $body
 512          * @return string
 513          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 514          */
 515         public static function limitBodySize($body)
 516         {
 517                 $maxlen = Config::get('config', 'max_import_size', 0);
 518
 519                 // If the length of the body, including the embedded images, is smaller
 520                 // than the maximum, then don't waste time looking for the images
 521                 if ($maxlen && (strlen($body) > $maxlen)) {
 522
 523                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 524
 525                         $orig_body = $body;
 526                         $new_body = '';
 527                         $textlen = 0;
 528
 529                         $img_start = strpos($orig_body, '[img');
 530                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 531                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 532                         while (($img_st_close !== false) && ($img_end !== false)) {
 533
 534                                 $img_st_close++; // make it point to AFTER the closing bracket
 535                                 $img_end += $img_start;
 536                                 $img_end += strlen('[/img]');
 537
 538                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 539                                         // This is an embedded image
 540
 541                                         if (($textlen + $img_start) > $maxlen) {
 542                                                 if ($textlen < $maxlen) {
 543                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 544                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 545                                                         $textlen = $maxlen;
 546                                                 }
 547                                         } else {
 548                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 549                                                 $textlen += $img_start;
 550                                         }
 551
 552                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 553                                 } else {
 554
 555                                         if (($textlen + $img_end) > $maxlen) {
 556                                                 if ($textlen < $maxlen) {
 557                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 558                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 559                                                         $textlen = $maxlen;
 560                                                 }
 561                                         } else {
 562                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 563                                                 $textlen += $img_end;
 564                                         }
 565                                 }
 566                                 $orig_body = substr($orig_body, $img_end);
 567
 568                                 if ($orig_body === false) {
 569                                         // in case the body ends on a closing image tag
 570                                         $orig_body = '';
 571                                 }
 572
 573                                 $img_start = strpos($orig_body, '[img');
 574                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 575                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 576                         }
 577
 578                         if (($textlen + strlen($orig_body)) > $maxlen) {
 579                                 if ($textlen < $maxlen) {
 580                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 581                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 582                                 }
 583                         } else {
 584                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 585                                 $new_body = $new_body . $orig_body;
 586                         }
 587
 588                         return $new_body;
 589                 } else {
 590                         return $body;
 591                 }
 592         }
 593
 594         /**
 595          * Processes [attachment] tags
 596          *
 597          * Note: Can produce a [bookmark] tag in the returned string
 598          *
 599          * @brief Processes [attachment] tags
 600          * @param string   $text
 601          * @param bool|int $simplehtml
 602          * @param bool     $tryoembed
 603          * @return string
 604          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 605          */
 606         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 607         {
 608                 $data = self::getAttachmentData($text);
 609                 if (empty($data) || empty($data['url'])) {
 610                         return $text;
 611                 }
 612
 613                 if (isset($data['title'])) {
 614                         $data['title'] = strip_tags($data['title']);
 615                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 616                 } else {
 617                         $data['title'] = null;
 618                 }
 619
 620                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data['image'])) {
 621                         $data['preview'] = $data['image'];
 622                         $data['image'] = '';
 623                 }
 624
 625                 $return = '';
 626                 try {
 627                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 628                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 629                         } else {
 630                                 throw new Exception('OEmbed is disabled for this attachment.');
 631                         }
 632                 } catch (Exception $e) {
 633                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 634
 635                         if ($simplehtml != 4) {
 636                                 $return = sprintf('<div class="type-%s">', $data['type']);
 637                         }
 638
 639                         if (!empty($data['title']) && !empty($data['url'])) {
 640                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 641                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 642                                 } else {
 643                                         if (!empty($data['image'])) {
 644                                                 $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), $data['title']);
 645                                         } elseif (!empty($data['preview'])) {
 646                                                 $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']);
 647                                         }
 648                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 649                                 }
 650                         }
 651
 652                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 653                                 // Sanitize the HTML by converting it to BBCode
 654                                 $bbcode = HTML::toBBCode($data['description']);
 655                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 656                         }
 657
 658                         if (!empty($data['url'])) {
 659                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 660                         }
 661
 662                         if ($simplehtml != 4) {
 663                                 $return .= '</div>';
 664                         }
 665                 }
 666
 667                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 668         }
 669
 670         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 671         {
 672                 $data = self::getAttachmentData($Text);
 673
 674                 if (!$data) {
 675                         return $Text;
 676                 } elseif ($nolink) {
 677                         return $data['text'] . ($data['after'] ?? '');
 678                 }
 679
 680                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 681                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 682                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 683                         $data['title'] = $data['url'];
 684                 } elseif (($text != '') && strstr($title, $text)) {
 685                         $data['text'] = $data['title'];
 686                         $data['title'] = $data['url'];
 687                 }
 688
 689                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 690                         return $data['title'] . $data['after'];
 691                 }
 692
 693                 // If the link already is included in the post, don't add it again
 694                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 695                         return $data['text'] . $data['after'];
 696                 }
 697
 698                 $text = $data['text'];
 699
 700                 if (!empty($data['url']) && !empty($data['title'])) {
 701                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 702                 } elseif (!empty($data['url'])) {
 703                         $text .= "\n[url]" . $data['url'] . '[/url]';
 704                 }
 705
 706                 return $text . "\n" . $data['after'];
 707         }
 708
 709         /**
 710          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 711          *
 712          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 713          * @param array $match Array with the matching values
 714          * @return string reformatted link including HTML codes
 715          */
 716         private static function convertUrlForActivityPubCallback($match)
 717         {
 718                 $url = $match[1];
 719
 720                 if (isset($match[2]) && ($match[1] != $match[2])) {
 721                         return $match[0];
 722                 }
 723
 724                 $parts = parse_url($url);
 725                 if (!isset($parts['scheme'])) {
 726                         return $match[0];
 727                 }
 728
 729                 return self::convertUrlForActivityPub($url);
 730         }
 731
 732         /**
 733          * @brief Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 734          * @param string $url URL that is about to be reformatted
 735          * @return string reformatted link including HTML codes
 736          */
 737         private static function convertUrlForActivityPub($url)
 738         {
 739                 $html = '<a href="%s" target="_blank">%s</a>';
 740                 return sprintf($html, $url, self::getStyledURL($url));
 741         }
 742
 743         /**
 744          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 745          * @param string $url URL that is about to be reformatted
 746          * @return string reformatted link
 747          */
 748         private static function getStyledURL($url)
 749         {
 750                 $parts = parse_url($url);
 751                 $scheme = $parts['scheme'] . '://';
 752                 $styled_url = str_replace($scheme, '', $url);
 753
 754                 if (strlen($styled_url) > 30) {
 755                         $styled_url = substr($styled_url, 0, 30) . "…";
 756                 }
 757
 758                 return $styled_url;
 759         }
 760
 761         /*
 762          * [noparse][i]italic[/i][/noparse] turns into
 763          * [noparse][ i ]italic[ /i ][/noparse],
 764          * to hide them from parser.
 765          */
 766         private static function escapeNoparseCallback($match)
 767         {
 768                 $whole_match = $match[0];
 769                 $captured = $match[1];
 770                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 771                 $new_str = str_replace($captured, $spacefied, $whole_match);
 772                 return $new_str;
 773         }
 774
 775         /*
 776          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 777          * now turns back and the [noparse] tags are trimed
 778          * returning [i]italic[/i]
 779          */
 780         private static function unescapeNoparseCallback($match)
 781         {
 782                 $captured = $match[1];
 783                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 784                 return $unspacefied;
 785         }
 786
 787         /**
 788          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 789          * occurrences
 790          *
 791          * @param string $text        Text to search
 792          * @param string $name        Tag name
 793          * @param int    $occurrences Number of first occurrences to skip
 794          * @return boolean|array
 795          */
 796         public static function getTagPosition($text, $name, $occurrences = 0)
 797         {
 798                 if ($occurrences < 0) {
 799                         $occurrences = 0;
 800                 }
 801
 802                 $start_open = -1;
 803                 for ($i = 0; $i <= $occurrences; $i++) {
 804                         if ($start_open !== false) {
 805                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 806                         }
 807                 }
 808
 809                 if ($start_open === false) {
 810                         return false;
 811                 }
 812
 813                 $start_equal = strpos($text, '=', $start_open);
 814                 $start_close = strpos($text, ']', $start_open);
 815
 816                 if ($start_close === false) {
 817                         return false;
 818                 }
 819
 820                 $start_close++;
 821
 822                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 823
 824                 if ($end_open === false) {
 825                         return false;
 826                 }
 827
 828                 $res = [
 829                         'start' => [
 830                                 'open' => $start_open,
 831                                 'close' => $start_close
 832                         ],
 833                         'end' => [
 834                                 'open' => $end_open,
 835                                 'close' => $end_open + strlen('[/' . $name . ']')
 836                         ],
 837                 ];
 838
 839                 if ($start_equal !== false) {
 840                         $res['start']['equal'] = $start_equal + 1;
 841                 }
 842
 843                 return $res;
 844         }
 845
 846         /**
 847          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 848          *
 849          * @param string $pattern Preg pattern string
 850          * @param string $replace Preg replace string
 851          * @param string $name    BBCode tag name
 852          * @param string $text    Text to search
 853          * @return string
 854          */
 855         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 856         {
 857                 $occurrences = 0;
 858                 $pos = self::getTagPosition($text, $name, $occurrences);
 859                 while ($pos !== false && $occurrences++ < 1000) {
 860                         $start = substr($text, 0, $pos['start']['open']);
 861                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 862                         $end = substr($text, $pos['end']['close']);
 863                         if ($end === false) {
 864                                 $end = '';
 865                         }
 866
 867                         $subject = preg_replace($pattern, $replace, $subject);
 868                         $text = $start . $subject . $end;
 869
 870                         $pos = self::getTagPosition($text, $name, $occurrences);
 871                 }
 872
 873                 return $text;
 874         }
 875
 876         private static function extractImagesFromItemBody($body)
 877         {
 878                 $saved_image = [];
 879                 $orig_body = $body;
 880                 $new_body = '';
 881
 882                 $cnt = 0;
 883                 $img_start = strpos($orig_body, '[img');
 884                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 885                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 886                 while (($img_st_close !== false) && ($img_end !== false)) {
 887                         $img_st_close++; // make it point to AFTER the closing bracket
 888                         $img_end += $img_start;
 889
 890                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 891                                 // This is an embedded image
 892                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 893                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 894
 895                                 $cnt++;
 896                         } else {
 897                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 898                         }
 899
 900                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 901
 902                         if ($orig_body === false) {
 903                                 // in case the body ends on a closing image tag
 904                                 $orig_body = '';
 905                         }
 906
 907                         $img_start = strpos($orig_body, '[img');
 908                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 909                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 910                 }
 911
 912                 $new_body = $new_body . $orig_body;
 913
 914                 return ['body' => $new_body, 'images' => $saved_image];
 915         }
 916
 917         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 918         {
 919                 $newbody = $body;
 920
 921                 $cnt = 0;
 922                 foreach ($images as $image) {
 923                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 924                         // it loops over the array starting from the first element and going sequentially
 925                         // to the last element
 926                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 927                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 928                         $cnt++;
 929                 }
 930
 931                 return $newbody;
 932         }
 933
 934         /**
 935          * This function converts a [share] block to text according to a provided callback function whose signature is:
 936          *
 937          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 938          *
 939          * Where:
 940          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 941          * data lookup
 942          * - $author_contact is a contact record array
 943          * - $content is the inner content of the [share] block
 944          * - $is_quote_share indicates whether there's any content before the [share] block
 945          * - Return value is the string that should replace the [share] block in the provided text
 946          *
 947          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 948          *
 949          * @param  string   $text     A BBCode string
 950          * @param  callable $callback
 951          * @return string The BBCode string with all [share] blocks replaced
 952          */
 953         public static function convertShare($text, callable $callback)
 954         {
 955                 $return = preg_replace_callback(
 956                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 957                         function ($match) use ($callback) {
 958                                 $attribute_string = $match[2];
 959
 960                                 $attributes = [];
 961                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 962                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 963                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 964                                 }
 965
 966                                 // We only call this so that a previously unknown contact can be added.
 967                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 968                                 // This function then can fetch an entry from the contact table.
 969                                 $default['url'] = $attributes['profile'];
 970
 971                                 if (!empty($attributes['author'])) {
 972                                         $default['name'] = $attributes['author'];
 973                                 }
 974
 975                                 if (!empty($attributes['avatar'])) {
 976                                         $default['photo'] = $attributes['avatar'];
 977                                 }
 978
 979                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 980
 981                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 982                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 983
 984                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 985                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 986                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 987
 988                                 if ($attributes['avatar']) {
 989                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 990                                 }
 991
 992                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 993                         },
 994                         $text
 995                 );
 996
 997                 return $return;
 998         }
 999
1000         /**
1001          * Default [share] tag conversion callback
1002          *
1003          * Note: Can produce a [bookmark] tag in the output
1004          *
1005          * @see BBCode::convertShare()
1006          * @param array   $attributes     [share] block attribute values
1007          * @param array   $author_contact Contact row of the shared author
1008          * @param string  $content        Inner content of the [share] block
1009          * @param boolean $is_quote_share Whether there is content before the [share] block
1010          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1011          * @return string
1012          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1013          */
1014         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1015         {
1016                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1017
1018                 switch ($simplehtml) {
1019                         case 1:
1020                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1021                                 break;
1022                         case 2:
1023                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1024                                 break;
1025                         case 3: // Diaspora
1026                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1027                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1028                                 } else {
1029                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1030
1031                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1032                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1033                                         }
1034
1035                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1036
1037                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1038                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1039                                         }
1040                                 }
1041
1042                                 break;
1043                         case 4:
1044                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1045                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1046                                 $headline .= ':</b></p>' . "\n";
1047
1048                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1049
1050                                 break;
1051                         case 5:
1052                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1053                                 break;
1054                         case 7: // statusnet/GNU Social
1055                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1056                                 break;
1057                         case 9: // ActivityPub
1058                                 $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>';
1059                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1060                                 break;
1061                         default:
1062                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1063                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1064                                         try {
1065                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1066                                         } catch (Exception $e) {
1067                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1068                                         }
1069                                 } else {
1070                                         $text = ($is_quote_share? "\n" : '');
1071
1072                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1073                                         $text .= Renderer::replaceMacros($tpl, [
1074                                                 '$profile' => $attributes['profile'],
1075                                                 '$avatar'  => $attributes['avatar'],
1076                                                 '$author'  => $attributes['author'],
1077                                                 '$link'    => $attributes['link'],
1078                                                 '$posted'  => $attributes['posted'],
1079                                                 '$content' => trim($content)
1080                                         ]);
1081                                 }
1082                                 break;
1083                 }
1084
1085                 return $text;
1086         }
1087
1088         private static function removePictureLinksCallback($match)
1089         {
1090                 $cache_key = 'remove:' . $match[1];
1091                 $text = Cache::get($cache_key);
1092
1093                 if (is_null($text)) {
1094                         $a = self::getApp();
1095
1096                         $stamp1 = microtime(true);
1097
1098                         $ch = @curl_init($match[1]);
1099                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1100                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1101                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1102                         @curl_exec($ch);
1103                         $curl_info = @curl_getinfo($ch);
1104
1105                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1106
1107                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1108                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1109                         } else {
1110                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1111
1112                                 // if its not a picture then look if its a page that contains a picture link
1113                                 $body = Network::fetchUrl($match[1]);
1114
1115                                 $doc = new DOMDocument();
1116                                 @$doc->loadHTML($body);
1117                                 $xpath = new DOMXPath($doc);
1118                                 $list = $xpath->query("//meta[@name]");
1119                                 foreach ($list as $node) {
1120                                         $attr = [];
1121
1122                                         if ($node->attributes->length) {
1123                                                 foreach ($node->attributes as $attribute) {
1124                                                         $attr[$attribute->name] = $attribute->value;
1125                                                 }
1126                                         }
1127
1128                                         if (strtolower($attr['name']) == 'twitter:image') {
1129                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1130                                         }
1131                                 }
1132                         }
1133                         Cache::set($cache_key, $text);
1134                 }
1135
1136                 return $text;
1137         }
1138
1139         private static function expandLinksCallback($match)
1140         {
1141                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1142                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1143                 } else {
1144                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1145                 }
1146         }
1147
1148         private static function cleanPictureLinksCallback($match)
1149         {
1150                 $a = self::getApp();
1151
1152                 // When the picture link is the own photo path then we can avoid fetching the link
1153                 $own_photo_url = preg_quote(Strings::normaliseLink($a->getBaseURL()) . '/photos/');
1154                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1155                         if (!empty($match[3])) {
1156                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1157                         } else {
1158                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1159                         }
1160                         return $text;
1161                 }
1162
1163                 $cache_key = 'clean:' . $match[1];
1164                 $text = Cache::get($cache_key);
1165                 if (!is_null($text)) {
1166                         return $text;
1167                 }
1168
1169                 // Only fetch the header, not the content
1170                 $stamp1 = microtime(true);
1171
1172                 $ch = @curl_init($match[1]);
1173                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1174                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1175                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1176                 @curl_exec($ch);
1177                 $curl_info = @curl_getinfo($ch);
1178
1179                 $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1180
1181                 // if its a link to a picture then embed this picture
1182                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1183                         $text = '[img]' . $match[1] . '[/img]';
1184                 } else {
1185                         if (!empty($match[3])) {
1186                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1187                         } else {
1188                                 $text = '[img]' . $match[2] . '[/img]';
1189                         }
1190
1191                         // if its not a picture then look if its a page that contains a picture link
1192                         $body = Network::fetchUrl($match[1]);
1193
1194                         $doc = new DOMDocument();
1195                         @$doc->loadHTML($body);
1196                         $xpath = new DOMXPath($doc);
1197                         $list = $xpath->query("//meta[@name]");
1198                         foreach ($list as $node) {
1199                                 $attr = [];
1200                                 if ($node->attributes->length) {
1201                                         foreach ($node->attributes as $attribute) {
1202                                                 $attr[$attribute->name] = $attribute->value;
1203                                         }
1204                                 }
1205
1206                                 if (strtolower($attr['name']) == "twitter:image") {
1207                                         if (!empty($match[3])) {
1208                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1209                                         } else {
1210                                                 $text = "[img]" . $attr['content'] . "[/img]";
1211                                         }
1212                                 }
1213                         }
1214                 }
1215                 Cache::set($cache_key, $text);
1216
1217                 return $text;
1218         }
1219
1220         public static function cleanPictureLinks($text)
1221         {
1222                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1223                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1224                 return $return;
1225         }
1226
1227         /**
1228          * @brief Converts a BBCode message to HTML message
1229          *
1230          * BBcode 2 HTML was written by WAY2WEB.net
1231          * extended to work with Mistpark/Friendica - Mike Macgirvin
1232          *
1233          * Simple HTML values meaning:
1234          * - 0: Friendica display
1235          * - 1: Unused
1236          * - 2: Used for Windows Phone push, Friendica API
1237          * - 3: Used before converting to Markdown in bb2diaspora.php
1238          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1239          * - 5: Unused
1240          * - 6: Unused
1241          * - 7: Used for dfrn, OStatus
1242          * - 8: Used for Twitter, WP backlink text setting
1243          * - 9: ActivityPub
1244          *
1245          * @param string $text
1246          * @param bool   $try_oembed
1247          * @param int    $simple_html
1248          * @param bool   $for_plaintext
1249          * @return string
1250          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1251          */
1252         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1253         {
1254                 $a = self::getApp();
1255
1256                 /*
1257                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1258                  *
1259                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1260                  * $match[1] = $url
1261                  * $match[2] = $title or absent
1262                  */
1263                 $try_oembed_callback = function ($match)
1264                 {
1265                         $url = $match[1];
1266                         $title = $match[2] ?? null;
1267
1268                         try {
1269                                 $return = OEmbed::getHTML($url, $title);
1270                         } catch (Exception $ex) {
1271                                 $return = $match[0];
1272                         }
1273
1274                         return $return;
1275                 };
1276
1277                 // Extracting code blocks before the whitespace processing and the autolinker
1278                 $codeblocks = [];
1279
1280                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1281                         function ($matches) use (&$codeblocks) {
1282                                 $return = '#codeblock-' . count($codeblocks) . '#';
1283                                 if (strpos($matches[2], "\n") !== false) {
1284                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1285                                 } else {
1286                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1287                                 }
1288
1289                                 return $return;
1290                         },
1291                         $text
1292                 );
1293
1294                 // Hide all [noparse] contained bbtags by spacefying them
1295                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1296
1297                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1298                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1299                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1300
1301                 // Remove the abstract element. It is a non visible element.
1302                 $text = self::stripAbstract($text);
1303
1304                 // Move all spaces out of the tags
1305                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1306                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1307
1308                 // Extract the private images which use data urls since preg has issues with
1309                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1310                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1311
1312                 $extracted = self::extractImagesFromItemBody($text);
1313                 $text = $extracted['body'];
1314                 $saved_image = $extracted['images'];
1315
1316                 // If we find any event code, turn it into an event.
1317                 // After we're finished processing the bbcode we'll
1318                 // replace all of the event code with a reformatted version.
1319
1320                 $ev = Event::fromBBCode($text);
1321
1322                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1323                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1324
1325                 $text = str_replace("<", "&lt;", $text);
1326                 $text = str_replace(">", "&gt;", $text);
1327
1328                 // remove some newlines before the general conversion
1329                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1330                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1331
1332                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1333                 if (!$try_oembed) {
1334                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1335                 }
1336
1337                 // Convert new line chars to html <br /> tags
1338
1339                 // nlbr seems to be hopelessly messed up
1340                 //      $Text = nl2br($Text);
1341
1342                 // We'll emulate it.
1343
1344                 $text = trim($text);
1345                 $text = str_replace("\r\n", "\n", $text);
1346
1347                 // Remove linefeeds inside of the table elements. See issue #6799
1348                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1349                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1350                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1351                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1352                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1353                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1354                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1355                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1356                 do {
1357                         $oldtext = $text;
1358                         $text = str_replace($search, $replace, $text);
1359                 } while ($oldtext != $text);
1360
1361                 // Replace these here only once
1362                 $search = ["\n[table]", "[/table]\n"];
1363                 $replace = ["[table]", "[/table]"];
1364                 $text = str_replace($search, $replace, $text);
1365
1366                 // removing multiplicated newlines
1367                 if (Config::get('system', 'remove_multiplicated_lines')) {
1368                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1369                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1370                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1371                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1372                         do {
1373                                 $oldtext = $text;
1374                                 $text = str_replace($search, $replace, $text);
1375                         } while ($oldtext != $text);
1376                 }
1377
1378                 /// @todo Have a closer look at the different html modes
1379                 // Handle attached links or videos
1380                 if (in_array($simple_html, [9])) {
1381                         $text = self::removeAttachment($text);
1382                 } elseif (!in_array($simple_html, [0, 4])) {
1383                         $text = self::removeAttachment($text, true);
1384                 } else {
1385                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1386                 }
1387
1388                 // leave open the posibility of [map=something]
1389                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1390                 if (strpos($text, '[/map]') !== false) {
1391                         $text = preg_replace_callback(
1392                                 "/\[map\](.*?)\[\/map\]/ism",
1393                                 function ($match) use ($simple_html) {
1394                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1395                                 },
1396                                 $text
1397                         );
1398                 }
1399
1400                 if (strpos($text, '[map=') !== false) {
1401                         $text = preg_replace_callback(
1402                                 "/\[map=(.*?)\]/ism",
1403                                 function ($match) use ($simple_html) {
1404                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1405                                 },
1406                                 $text
1407                         );
1408                 }
1409
1410                 if (strpos($text, '[map]') !== false) {
1411                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1412                 }
1413
1414                 // Check for headers
1415                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1416                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1417                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1418                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1419                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1420                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1421
1422                 // Check for paragraph
1423                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1424
1425                 // Check for bold text
1426                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1427
1428                 // Check for Italics text
1429                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1430
1431                 // Check for Underline text
1432                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1433
1434                 // Check for strike-through text
1435                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1436
1437                 // Check for over-line text
1438                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1439
1440                 // Check for colored text
1441                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1442
1443                 // Check for sized text
1444                 // [size=50] --> font-size: 50px (with the unit).
1445                 if ($simple_html != 3) {
1446                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1447                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1448                 } else {
1449                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1450                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1451                 }
1452
1453
1454                 // Check for centered text
1455                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1456
1457                 // Check for list text
1458                 $text = str_replace("[*]", "<li>", $text);
1459
1460                 // Check for style sheet commands
1461                 $text = preg_replace_callback(
1462                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1463                         function ($match) {
1464                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1465                         },
1466                         $text
1467                 );
1468
1469                 // Check for CSS classes
1470                 $text = preg_replace_callback(
1471                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1472                         function ($match) {
1473                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1474                         },
1475                         $text
1476                 );
1477
1478                 // handle nested lists
1479                 $endlessloop = 0;
1480
1481                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1482                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1483                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1484                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1485                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1486                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1487                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1488                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1489                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1490                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1491                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1492                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1493                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1494                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1495                 }
1496
1497                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1498                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1499                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1500                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1501
1502                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1503                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1504
1505                 $text = str_replace('[hr]', '<hr />', $text);
1506
1507                 if (!$for_plaintext) {
1508                         // Autolinker for isolated URLs
1509                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1510                 }
1511
1512                 // This is actually executed in Item::prepareBody()
1513
1514                 $nosmile = strpos($text, '[nosmile]') !== false;
1515                 $text = str_replace('[nosmile]', '', $text);
1516
1517                 // Check for font change text
1518                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1519
1520                 // Declare the format for [spoiler] layout
1521                 $SpoilerLayout = '<details class="spoiler"><summary>' . L10n::t('Click to open/close') . '</summary>$1</details>';
1522
1523                 // Check for [spoiler] text
1524                 // handle nested quotes
1525                 $endlessloop = 0;
1526                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1527                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1528                 }
1529
1530                 // Check for [spoiler=Title] text
1531
1532                 // handle nested quotes
1533                 $endlessloop = 0;
1534                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1535                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1536                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1537                                 $text);
1538                 }
1539
1540                 // Declare the format for [quote] layout
1541                 $QuoteLayout = '<blockquote>$1</blockquote>';
1542
1543                 // Check for [quote] text
1544                 // handle nested quotes
1545                 $endlessloop = 0;
1546                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1547                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1548                 }
1549
1550                 // Check for [quote=Author] text
1551
1552                 $t_wrote = L10n::t('$1 wrote:');
1553
1554                 // handle nested quotes
1555                 $endlessloop = 0;
1556                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1557                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1558                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1559                                 $text);
1560                 }
1561
1562
1563                 // [img=widthxheight]image source[/img]
1564                 $text = preg_replace_callback(
1565                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1566                         function ($matches) use ($simple_html) {
1567                                 if (strpos($matches[3], "data:image/") === 0) {
1568                                         return $matches[0];
1569                                 }
1570
1571                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1572                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1573                         },
1574                         $text
1575                 );
1576
1577                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1578                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1579
1580                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1581                         function ($matches) use ($simple_html) {
1582                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1583                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1584                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1585                         },
1586                         $text);
1587
1588                 // Images
1589                 // [img]pathtoimage[/img]
1590                 $text = preg_replace_callback(
1591                         "/\[img\](.*?)\[\/img\]/ism",
1592                         function ($matches) use ($simple_html) {
1593                                 if (strpos($matches[1], "data:image/") === 0) {
1594                                         return $matches[0];
1595                                 }
1596
1597                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1598                                 return "[img]" . $matches[1] . "[/img]";
1599                         },
1600                         $text
1601                 );
1602
1603                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1604                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1605
1606                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1607                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1608                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1609
1610                 // Try to Oembed
1611                 if ($try_oembed) {
1612                         $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);
1613                         $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);
1614
1615                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1616                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1617                 } else {
1618                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1619                                 '<a href="$1" target="_blank">$1</a>', $text);
1620                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1621                                 '<a href="$1" target="_blank">$1</a>', $text);
1622                 }
1623
1624                 // html5 video and audio
1625
1626
1627                 if ($try_oembed) {
1628                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1629                 } else {
1630                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1631                 }
1632
1633                 // Youtube extensions
1634                 if ($try_oembed) {
1635                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1636                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1637                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1638                 }
1639
1640                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1641                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1642                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1643
1644                 if ($try_oembed) {
1645                         $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);
1646                 } else {
1647                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1648                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1649                 }
1650
1651                 if ($try_oembed) {
1652                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1653                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1654                 }
1655
1656                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1657                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1658
1659                 if ($try_oembed) {
1660                         $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);
1661                 } else {
1662                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1663                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1664                 }
1665
1666                 // oembed tag
1667                 $text = OEmbed::BBCode2HTML($text);
1668
1669                 // Avoid triple linefeeds through oembed
1670                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1671
1672                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1673                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1674                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1675                 // start which is always required). Allow desc with a missing summary for compatibility.
1676
1677                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1678                         $sub = Event::getHTML($ev, $simple_html);
1679
1680                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1681                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1682                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1683                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1684                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1685                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1686                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1687                 }
1688
1689                 // Replace non graphical smilies for external posts
1690                 if (!$nosmile && !$for_plaintext) {
1691                         $text = Smilies::replace($text);
1692                 }
1693
1694                 if (!$for_plaintext) {
1695                         if (in_array($simple_html, [7, 9])) {
1696                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1697                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1698                         }
1699                 } else {
1700                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1701                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1702                 }
1703
1704                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1705
1706                 // Remove all hashtag addresses
1707                 if ($simple_html && !in_array($simple_html, [3, 7, 9])) {
1708                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1709                 } elseif ($simple_html == 3) {
1710                         // The ! is converted to @ since Diaspora only understands the @
1711                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1712                                 '@<a href="$2">$3</a>',
1713                                 $text);
1714                 } elseif (in_array($simple_html, [7, 9])) {
1715                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1716                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1717                                 $text);
1718                 } elseif (!$simple_html) {
1719                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1720                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1721                                 $text);
1722                 }
1723
1724                 // Bookmarks in red - will be converted to bookmarks in friendica
1725                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1726                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1727                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1728                                         "[bookmark=$1]$2[/bookmark]", $text);
1729
1730                 if (in_array($simple_html, [2, 6, 7, 8])) {
1731                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1732                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1733                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1734                 }
1735
1736                 if ($simple_html == 5) {
1737                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1738                 }
1739
1740                 // Perform URL Search
1741                 if ($try_oembed) {
1742                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1743                 }
1744
1745                 if ($simple_html == 5) {
1746                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1747                 } else {
1748                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1749                 }
1750
1751                 // Handle Diaspora posts
1752                 $text = preg_replace_callback(
1753                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1754                         function ($match) {
1755                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1756                         }, $text
1757                 );
1758
1759                 $text = preg_replace_callback(
1760                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1761                         function ($match) {
1762                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1763                         }, $text
1764                 );
1765
1766                 // Server independent link to posts and comments
1767                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1768                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1769                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1770
1771                 /* Tag conversion
1772                  * Supports:
1773                  * - #[url=<anything>]<term>[/url]
1774                  * - [url=<anything>]#<term>[/url]
1775                  */
1776                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1777                         return '#<a href="'
1778                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1779                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1780                                 . XML::escape($matches[1])
1781                                 . '</a>';
1782                 }, $text);
1783
1784                 // We need no target="_blank" for local links
1785                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1786                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1787                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1788                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1789
1790                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1791                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1792
1793                 // Red compatibility, though the link can't be authenticated on Friendica
1794                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1795
1796
1797                 // we may need to restrict this further if it picks up too many strays
1798                 // link acct:user@host to a webfinger profile redirector
1799
1800                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1801
1802                 // Perform MAIL Search
1803                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1804                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1805
1806                 // Unhide all [noparse] contained bbtags unspacefying them
1807                 // and triming the [noparse] tag.
1808
1809                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1810                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1811                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1812
1813                 /// @todo What is the meaning of these lines?
1814                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1815                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1816
1817                 // Currently deactivated, it made problems with " inside of alt texts.
1818                 //$text = preg_replace('/\&quot\;/', '"', $text);
1819
1820                 // fix any escaped ampersands that may have been converted into links
1821                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1822
1823                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1824                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1825
1826                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1827
1828                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1829                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1830
1831                 // sanitize href attributes (only whitelisted protocols URLs)
1832                 // default value for backward compatibility
1833                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1834
1835                 // Always allowed protocol even if config isn't set or not including it
1836                 $allowed_link_protocols[] = '//';
1837                 $allowed_link_protocols[] = 'http://';
1838                 $allowed_link_protocols[] = 'https://';
1839                 $allowed_link_protocols[] = 'redir/';
1840
1841                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1842
1843                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1844                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1845
1846                 // Shared content
1847                 $text = self::convertShare(
1848                         $text,
1849                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1850                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1851                         }
1852                 );
1853
1854                 if ($saved_image) {
1855                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1856                 }
1857
1858                 // Restore code blocks
1859                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1860                         function ($matches) use ($codeblocks) {
1861                                 $return = $matches[0];
1862                                 if (isset($codeblocks[intval($matches[1])])) {
1863                                         $return = $codeblocks[$matches[1]];
1864                                 }
1865                                 return $return;
1866                         },
1867                         $text
1868                 );
1869
1870                 // Clean up the HTML by loading and saving the HTML with the DOM.
1871                 // Bad structured html can break a whole page.
1872                 // For performance reasons do it only with activated item cache or at export.
1873                 if (!$try_oembed || (get_itemcachepath() != '')) {
1874                         $doc = new DOMDocument();
1875                         $doc->preserveWhiteSpace = false;
1876
1877                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1878
1879                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1880                         $encoding = '<?xml encoding="UTF-8">';
1881                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1882                         $doc->encoding = 'UTF-8';
1883                         $text = $doc->saveHTML();
1884                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1885
1886                         $text = str_replace('<br></li>', '</li>', $text);
1887
1888                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1889                 }
1890
1891                 // Clean up some useless linebreaks in lists
1892                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1893                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1894                 //$Text = str_replace('</li><br />', '</li>', $Text);
1895                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1896                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1897
1898                 Hook::callAll('bbcode', $text);
1899
1900                 return trim($text);
1901         }
1902
1903         /**
1904          * @brief Strips the "abstract" tag from the provided text
1905          *
1906          * @param string $text The text with BBCode
1907          * @return string The same text - but without "abstract" element
1908          */
1909         public static function stripAbstract($text)
1910         {
1911                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1912                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1913
1914                 return $text;
1915         }
1916
1917         /**
1918          * @brief Returns the value of the "abstract" element
1919          *
1920          * @param string $text The text that maybe contains the element
1921          * @param string $addon The addon for which the abstract is meant for
1922          * @return string The abstract
1923          */
1924         public static function getAbstract($text, $addon = '')
1925         {
1926                 $abstract = '';
1927                 $abstracts = [];
1928                 $addon = strtolower($addon);
1929
1930                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1931                         foreach ($results AS $result) {
1932                                 $abstracts[strtolower($result[1])] = $result[2];
1933                         }
1934                 }
1935
1936                 if (isset($abstracts[$addon])) {
1937                         $abstract = $abstracts[$addon];
1938                 }
1939
1940                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1941                         $abstract = $result[1];
1942                 }
1943
1944                 return $abstract;
1945         }
1946
1947         /**
1948          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1949          *
1950          * @param array $match Matching values for the callback
1951          *                     [1] = Mention type (! or @)
1952          *                     [2] = Name
1953          *                     [3] = Address
1954          * @return string Replaced mention
1955          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1956          * @throws \ImagickException
1957          */
1958         private static function bbCodeMention2DiasporaCallback($match)
1959         {
1960                 $contact = Contact::getDetailsByURL($match[3]);
1961
1962                 if (empty($contact['addr'])) {
1963                         $contact = Probe::uri($match[3]);
1964                 }
1965
1966                 if (empty($contact['addr'])) {
1967                         return $match[0];
1968                 }
1969
1970                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1971                 return $mention;
1972         }
1973
1974         /**
1975          * @brief Converts a BBCode text into Markdown
1976          *
1977          * This function converts a BBCode item body to be sent to Markdown-enabled
1978          * systems like Diaspora and Libertree
1979          *
1980          * @param string $text
1981          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1982          * @return string
1983          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1984          */
1985         public static function toMarkdown($text, $for_diaspora = true)
1986         {
1987                 $a = self::getApp();
1988
1989                 $original_text = $text;
1990
1991                 // Since Diaspora is creating a summary for links, this function removes them before posting
1992                 if ($for_diaspora) {
1993                         $text = self::removeShareInformation($text);
1994                 }
1995
1996                 /**
1997                  * Transform #tags, strip off the [url] and replace spaces with underscore
1998                  */
1999                 $url_search_string = "^\[\]";
2000                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2001                         function ($matches) {
2002                                 return '#' . str_replace(' ', '_', $matches[2]);
2003                         },
2004                         $text
2005                 );
2006
2007                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2008                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2009
2010                 // Convert it to HTML - don't try oembed
2011                 if ($for_diaspora) {
2012                         $text = self::convert($text, false, 3);
2013
2014                         // Add all tags that maybe were removed
2015                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2016                                 $tagline = '';
2017                                 foreach ($tags[2] as $tag) {
2018                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2019                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2020                                                 $tagline .= '#' . $tag . ' ';
2021                                         }
2022                                 }
2023                                 $text = $text . " " . $tagline;
2024                         }
2025                 } else {
2026                         $text = self::convert($text, false, 4);
2027                 }
2028
2029                 // mask some special HTML chars from conversation to markdown
2030                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2031
2032                 // If a link is followed by a quote then there should be a newline before it
2033                 // Maybe we should make this newline at every time before a quote.
2034                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2035
2036                 $stamp1 = microtime(true);
2037
2038                 // Now convert HTML to Markdown
2039                 $text = HTML::toMarkdown($text);
2040
2041                 // unmask the special chars back to HTML
2042                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2043
2044                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
2045
2046                 // Libertree has a problem with escaped hashtags.
2047                 $text = str_replace(['\#'], ['#'], $text);
2048
2049                 // Remove any leading or trailing whitespace, as this will mess up
2050                 // the Diaspora signature verification and cause the item to disappear
2051                 $text = trim($text);
2052
2053                 if ($for_diaspora) {
2054                         $url_search_string = "^\[\]";
2055                         $text = preg_replace_callback(
2056                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2057                                 ['self', 'bbCodeMention2DiasporaCallback'],
2058                                 $text
2059                         );
2060                 }
2061
2062                 Hook::callAll('bb2diaspora', $text);
2063
2064                 return $text;
2065         }
2066
2067         /**
2068      * @brief Pull out all #hashtags and @person tags from $string.
2069      *
2070      * We also get @person@domain.com - which would make
2071      * the regex quite complicated as tags can also
2072      * end a sentence. So we'll run through our results
2073      * and strip the period from any tags which end with one.
2074      * Returns array of tags found, or empty array.
2075      *
2076      * @param string $string Post content
2077      *
2078      * @return array List of tag and person names
2079      */
2080     public static function getTags($string)
2081     {
2082         $ret = [];
2083
2084         // Convert hashtag links to hashtags
2085         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2086
2087         // ignore anything in a code block
2088         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2089
2090         // Force line feeds at bbtags
2091         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2092
2093         // ignore anything in a bbtag
2094         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2095
2096         // Match full names against @tags including the space between first and last
2097         // We will look these up afterward to see if they are full names or not recognisable.
2098
2099         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2100             foreach ($matches[1] as $match) {
2101                 if (strstr($match, ']')) {
2102                     // we might be inside a bbcode color tag - leave it alone
2103                     continue;
2104                 }
2105
2106                 if (substr($match, -1, 1) === '.') {
2107                     $ret[] = substr($match, 0, -1);
2108                 } else {
2109                     $ret[] = $match;
2110                 }
2111             }
2112         }
2113
2114         // Otherwise pull out single word tags. These can be @nickname, @first_last
2115         // and #hash tags.
2116
2117         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2118             foreach ($matches[1] as $match) {
2119                 if (strstr($match, ']')) {
2120                     // we might be inside a bbcode color tag - leave it alone
2121                     continue;
2122                 }
2123                 if (substr($match, -1, 1) === '.') {
2124                     $match = substr($match,0,-1);
2125                 }
2126                 // ignore strictly numeric tags like #1
2127                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2128                     continue;
2129                 }
2130                 // try not to catch url fragments
2131                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2132                     continue;
2133                 }
2134                 $ret[] = $match;
2135             }
2136         }
2137
2138         return $ret;
2139     }
2140 }