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