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