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