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                         'type'        => '',
 125                         'text'        => '',
 126                         'after'       => '',
 127                         'image'       => null,
 128                         'url'         => '',
 129                         'title'       => '',
 130                         'description' => '',
 131                 ];
 132
 133                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 134                         return self::getOldAttachmentData($body);
 135                 }
 136
 137                 $attributes = $match[2];
 138
 139                 $data["text"] = trim($match[1]);
 140
 141                 $type = "";
 142                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 143                 if (!empty($matches[1])) {
 144                         $type = strtolower($matches[1]);
 145                 }
 146
 147                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 148                 if (!empty($matches[1])) {
 149                         $type = strtolower($matches[1]);
 150                 }
 151
 152                 if ($type == "") {
 153                         return [];
 154                 }
 155
 156                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 157                         return [];
 158                 }
 159
 160                 if ($type != "") {
 161                         $data["type"] = $type;
 162                 }
 163
 164                 $url = "";
 165                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 166                 if (!empty($matches[1])) {
 167                         $url = $matches[1];
 168                 }
 169
 170                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 171                 if (!empty($matches[1])) {
 172                         $url = $matches[1];
 173                 }
 174
 175                 if ($url != "") {
 176                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 177                 }
 178
 179                 $title = "";
 180                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 181                 if (!empty($matches[1])) {
 182                         $title = $matches[1];
 183                 }
 184
 185                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 186                 if (!empty($matches[1])) {
 187                         $title = $matches[1];
 188                 }
 189
 190                 if ($title != "") {
 191                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 192                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 193                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 194                         $data["title"] = $title;
 195                 }
 196
 197                 $image = "";
 198                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 199                 if (!empty($matches[1])) {
 200                         $image = $matches[1];
 201                 }
 202
 203                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 204                 if (!empty($matches[1])) {
 205                         $image = $matches[1];
 206                 }
 207
 208                 if ($image != "") {
 209                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 210                 }
 211
 212                 $preview = "";
 213                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 214                 if (!empty($matches[1])) {
 215                         $preview = $matches[1];
 216                 }
 217
 218                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 219                 if (!empty($matches[1])) {
 220                         $preview = $matches[1];
 221                 }
 222
 223                 if ($preview != "") {
 224                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 225                 }
 226
 227                 $data["description"] = trim($match[3]);
 228
 229                 $data["after"] = trim($match[4]);
 230
 231                 return $data;
 232         }
 233
 234         public static function getAttachedData($body, $item = [])
 235         {
 236                 /*
 237                 - text:
 238                 - type: link, video, photo
 239                 - title:
 240                 - url:
 241                 - image:
 242                 - description:
 243                 - (thumbnail)
 244                 */
 245
 246                 $has_title = !empty($item['title']);
 247                 $plink = defaults($item, 'plink', '');
 248                 $post = self::getAttachmentData($body);
 249
 250                 // Get all linked images with alternative image description
 251                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 252                         foreach ($pictures as $picture) {
 253                                 if (Photo::isLocal($picture[1])) {
 254                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 255                                 }
 256                         }
 257                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 258                                 $post['image_description'] = $post['images'][0]['description'];
 259                         }
 260                 }
 261
 262                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 263                         foreach ($pictures as $picture) {
 264                                 if (Photo::isLocal($picture[1])) {
 265                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 266                                 }
 267                         }
 268                 }
 269
 270                 // if nothing is found, it maybe having an image.
 271                 if (!isset($post["type"])) {
 272                         // Simplify image codes
 273                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 274                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 275                         $post["text"] = $body;
 276
 277                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 278                                 if ((count($pictures) == 1) && !$has_title) {
 279                                         if (!empty($item['object-type']) && ($item['object-type'] == ACTIVITY_OBJ_IMAGE)) {
 280                                                 // Replace the preview picture with the real picture
 281                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 282                                                 $data = ['url' => $url, 'type' => 'photo'];
 283                                         } else {
 284                                                 // Checking, if the link goes to a picture
 285                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 286                                         }
 287
 288                                         // Workaround:
 289                                         // Sometimes photo posts to the own album are not detected at the start.
 290                                         // So we seem to cannot use the cache for these cases. That's strange.
 291                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 292                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 293                                         }
 294
 295                                         if ($data["type"] == "photo") {
 296                                                 $post["type"] = "photo";
 297                                                 if (isset($data["images"][0])) {
 298                                                         $post["image"] = $data["images"][0]["src"];
 299                                                         $post["url"] = $data["url"];
 300                                                 } else {
 301                                                         $post["image"] = $data["url"];
 302                                                 }
 303
 304                                                 $post["preview"] = $pictures[0][2];
 305                                                 $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 306                                         } else {
 307                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 308                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 309                                                         $post["type"] = "photo";
 310                                                         $post["image"] = $pictures[0][1];
 311                                                         $post["preview"] = $pictures[0][2];
 312                                                         $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 313                                                 }
 314                                         }
 315                                 } elseif (count($pictures) > 0) {
 316                                         $post["type"] = "link";
 317                                         $post["url"] = $plink;
 318                                         $post["image"] = $pictures[0][2];
 319                                         $post["text"] = $body;
 320
 321                                         foreach ($pictures as $picture) {
 322                                                 $post["text"] = trim(str_replace($picture[0], "", $post["text"]));
 323                                         }
 324                                 }
 325                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 326                                 if ((count($pictures) == 1) && !$has_title) {
 327                                         $post["type"] = "photo";
 328                                         $post["image"] = $pictures[0][1];
 329                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 330                                 } elseif (count($pictures) > 0) {
 331                                         $post["type"] = "link";
 332                                         $post["url"] = $plink;
 333                                         $post["image"] = $pictures[0][1];
 334                                         $post["text"] = $body;
 335
 336                                         foreach ($pictures as $picture) {
 337                                                 $post["text"] = trim(str_replace($picture[0], "", $post["text"]));
 338                                         }
 339                                 }
 340                         }
 341
 342                         // Test for the external links
 343                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post["text"], $links1, PREG_SET_ORDER);
 344                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post["text"], $links2, PREG_SET_ORDER);
 345
 346                         $links = array_merge($links1, $links2);
 347
 348                         // If there is only a single one, then use it.
 349                         // This should cover link posts via API.
 350                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 351                                 $post["type"] = "link";
 352                                 $post["url"] = $links[0][1];
 353                         }
 354
 355                         // Now count the number of external media links
 356                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post["text"], $links1, PREG_SET_ORDER);
 357                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post["text"], $links2, PREG_SET_ORDER);
 358                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post["text"], $links3, PREG_SET_ORDER);
 359                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post["text"], $links4, PREG_SET_ORDER);
 360
 361                         // Add them to the other external links
 362                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 363
 364                         // Are there more than one?
 365                         if (count($links) > 1) {
 366                                 // The post will be the type "text", which means a blog post
 367                                 unset($post["type"]);
 368                                 $post["url"] = $plink;
 369                         }
 370
 371                         if (!isset($post["type"])) {
 372                                 $post["type"] = "text";
 373                                 $post["text"] = trim($body);
 374                         }
 375                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 376                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 377
 378                         if (isset($data["images"][0])) {
 379                                 $post["image"] = $data["images"][0]["src"];
 380                         }
 381                 }
 382
 383                 return $post;
 384         }
 385
 386         /**
 387          * @brief Converts a BBCode text into plaintext
 388          *
 389          * @param      $text
 390          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 391          *
 392          * @return string
 393          */
 394         public static function toPlaintext($text, $keep_urls = true)
 395         {
 396                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 397
 398                 return $naked_text;
 399         }
 400
 401         private static function proxyUrl($image, $simplehtml = false)
 402         {
 403                 // Only send proxied pictures to API and for internal display
 404                 if (in_array($simplehtml, [false, 2])) {
 405                         return ProxyUtils::proxifyUrl($image);
 406                 } else {
 407                         return $image;
 408                 }
 409         }
 410
 411         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 412         {
 413                 // Suppress "view full size"
 414                 if (intval(Config::get('system', 'no_view_full_size'))) {
 415                         $include_link = false;
 416                 }
 417
 418                 // Picture addresses can contain special characters
 419                 $s = htmlspecialchars_decode($srctext);
 420
 421                 $matches = null;
 422                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 423                 if ($c) {
 424                         foreach ($matches as $mtch) {
 425                                 Logger::log('scale_external_image: ' . $mtch[1]);
 426
 427                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 428                                 if (stristr($mtch[1], $hostname)) {
 429                                         continue;
 430                                 }
 431
 432                                 // $scale_replace, if passed, is an array of two elements. The
 433                                 // first is the name of the full-size image. The second is the
 434                                 // name of a remote, scaled-down version of the full size image.
 435                                 // This allows Friendica to display the smaller remote image if
 436                                 // one exists, while still linking to the full-size image
 437                                 if ($scale_replace) {
 438                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 439                                 } else {
 440                                         $scaled = $mtch[1];
 441                                 }
 442                                 $i = Network::fetchUrl($scaled);
 443                                 if (!$i) {
 444                                         return $srctext;
 445                                 }
 446
 447                                 // guess mimetype from headers or filename
 448                                 $type = Image::guessType($mtch[1], true);
 449
 450                                 if ($i) {
 451                                         $Image = new Image($i, $type);
 452                                         if ($Image->isValid()) {
 453                                                 $orig_width = $Image->getWidth();
 454                                                 $orig_height = $Image->getHeight();
 455
 456                                                 if ($orig_width > 640 || $orig_height > 640) {
 457                                                         $Image->scaleDown(640);
 458                                                         $new_width = $Image->getWidth();
 459                                                         $new_height = $Image->getHeight();
 460                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 461                                                         $s = str_replace(
 462                                                                 $mtch[0],
 463                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 464                                                                 . "\n" . (($include_link)
 465                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 466                                                                         : ''),
 467                                                                 $s
 468                                                         );
 469                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 470                                                 }
 471                                         }
 472                                 }
 473                         }
 474                 }
 475
 476                 // replace the special char encoding
 477                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 478                 return $s;
 479         }
 480
 481         /**
 482          * The purpose of this function is to apply system message length limits to
 483          * imported messages without including any embedded photos in the length
 484          *
 485          * @brief Truncates imported message body string length to max_import_size
 486          * @param string $body
 487          * @return string
 488          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 489          */
 490         public static function limitBodySize($body)
 491         {
 492                 $maxlen = Config::get('config', 'max_import_size', 0);
 493
 494                 // If the length of the body, including the embedded images, is smaller
 495                 // than the maximum, then don't waste time looking for the images
 496                 if ($maxlen && (strlen($body) > $maxlen)) {
 497
 498                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 499
 500                         $orig_body = $body;
 501                         $new_body = '';
 502                         $textlen = 0;
 503
 504                         $img_start = strpos($orig_body, '[img');
 505                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 506                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 507                         while (($img_st_close !== false) && ($img_end !== false)) {
 508
 509                                 $img_st_close++; // make it point to AFTER the closing bracket
 510                                 $img_end += $img_start;
 511                                 $img_end += strlen('[/img]');
 512
 513                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 514                                         // This is an embedded image
 515
 516                                         if (($textlen + $img_start) > $maxlen) {
 517                                                 if ($textlen < $maxlen) {
 518                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 519                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 520                                                         $textlen = $maxlen;
 521                                                 }
 522                                         } else {
 523                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 524                                                 $textlen += $img_start;
 525                                         }
 526
 527                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 528                                 } else {
 529
 530                                         if (($textlen + $img_end) > $maxlen) {
 531                                                 if ($textlen < $maxlen) {
 532                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 533                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 534                                                         $textlen = $maxlen;
 535                                                 }
 536                                         } else {
 537                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 538                                                 $textlen += $img_end;
 539                                         }
 540                                 }
 541                                 $orig_body = substr($orig_body, $img_end);
 542
 543                                 if ($orig_body === false) {
 544                                         // in case the body ends on a closing image tag
 545                                         $orig_body = '';
 546                                 }
 547
 548                                 $img_start = strpos($orig_body, '[img');
 549                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 550                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 551                         }
 552
 553                         if (($textlen + strlen($orig_body)) > $maxlen) {
 554                                 if ($textlen < $maxlen) {
 555                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 556                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 557                                 }
 558                         } else {
 559                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 560                                 $new_body = $new_body . $orig_body;
 561                         }
 562
 563                         return $new_body;
 564                 } else {
 565                         return $body;
 566                 }
 567         }
 568
 569         /**
 570          * Processes [attachment] tags
 571          *
 572          * Note: Can produce a [bookmark] tag in the returned string
 573          *
 574          * @brief Processes [attachment] tags
 575          * @param string   $return
 576          * @param bool|int $simplehtml
 577          * @param bool     $tryoembed
 578          * @return string
 579          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 580          */
 581         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 582         {
 583                 $data = self::getAttachmentData($return);
 584                 if (empty($data) || empty($data["url"])) {
 585                         return $return;
 586                 }
 587
 588                 if (isset($data["title"])) {
 589                         $data["title"] = strip_tags($data["title"]);
 590                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 591                 } else {
 592                         $data["title"] = null;
 593                 }
 594
 595                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 596                         $data["preview"] = $data["image"];
 597                         $data["image"] = "";
 598                 }
 599
 600                 $return = '';
 601                 if (in_array($simplehtml, [7, 9])) {
 602                         $return = self::convertUrlForOStatus($data["url"]);
 603                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 604                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 605                 } else {
 606                         try {
 607                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 608                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 609                                 } else {
 610                                         throw new Exception('OEmbed is disabled for this attachment.');
 611                                 }
 612                         } catch (Exception $e) {
 613                                 $data["title"] = defaults($data, 'title', $data['url']);
 614
 615                                 if ($simplehtml != 4) {
 616                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 617                                 }
 618
 619                                 if (!empty($data['title']) && !empty($data['url'])) {
 620                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 621                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 622                                         } else {
 623                                                 if (!empty($data["image"])) {
 624                                                         $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), $data["title"]);
 625                                                 } elseif (!empty($data["preview"])) {
 626                                                         $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"]);
 627                                                 }
 628                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 629                                         }
 630                                 }
 631
 632                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 633                                         // Sanitize the HTML by converting it to BBCode
 634                                         $bbcode = HTML::toBBCode($data["description"]);
 635                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 636                                 }
 637
 638                                 if (!empty($data['url'])) {
 639                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 640                                 }
 641
 642                                 if ($simplehtml != 4) {
 643                                         $return .= '</div>';
 644                                 }
 645                         }
 646                 }
 647
 648                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 649         }
 650
 651         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 652         {
 653                 $data = self::getAttachmentData($Text);
 654
 655                 if (!$data) {
 656                         return $Text;
 657                 } elseif ($nolink) {
 658                         return $data["text"] . defaults($data, 'after', '');
 659                 }
 660
 661                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 662                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 663                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 664                         $data["title"] = $data["url"];
 665                 } elseif (($text != "") && strstr($title, $text)) {
 666                         $data["text"] = $data["title"];
 667                         $data["title"] = $data["url"];
 668                 }
 669
 670                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 671                         return $data["title"] . $data["after"];
 672                 }
 673
 674                 // If the link already is included in the post, don't add it again
 675                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 676                         return $data["text"] . $data["after"];
 677                 }
 678
 679                 $text = $data["text"];
 680
 681                 if (!empty($data["url"]) && !empty($data["title"])) {
 682                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 683                 } elseif (!empty($data["url"])) {
 684                         $text .= "\n[url]" . $data["url"] . "[/url]";
 685                 }
 686
 687                 return $text . "\n" . $data["after"];
 688         }
 689
 690         /**
 691          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 692          *
 693          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 694          * @param array $match Array with the matching values
 695          * @return string reformatted link including HTML codes
 696          */
 697         private static function convertUrlForOStatusCallback($match)
 698         {
 699                 $url = $match[1];
 700
 701                 if (isset($match[2]) && ($match[1] != $match[2])) {
 702                         return $match[0];
 703                 }
 704
 705                 $parts = parse_url($url);
 706                 if (!isset($parts['scheme'])) {
 707                         return $match[0];
 708                 }
 709
 710                 return self::convertUrlForOStatus($url);
 711         }
 712
 713         /**
 714          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 715          * @param string $url URL that is about to be reformatted
 716          * @return string reformatted link including HTML codes
 717          */
 718         private static function convertUrlForOStatus($url)
 719         {
 720                 $parts = parse_url($url);
 721                 $scheme = $parts['scheme'] . '://';
 722                 $styled_url = str_replace($scheme, '', $url);
 723
 724                 if (strlen($styled_url) > 30) {
 725                         $styled_url = substr($styled_url, 0, 30) . "…";
 726                 }
 727
 728                 $html = '<a href="%s" target="_blank">%s</a>';
 729
 730                 return sprintf($html, $url, $styled_url);
 731         }
 732
 733         /*
 734          * [noparse][i]italic[/i][/noparse] turns into
 735          * [noparse][ i ]italic[ /i ][/noparse],
 736          * to hide them from parser.
 737          */
 738         private static function escapeNoparseCallback($match)
 739         {
 740                 $whole_match = $match[0];
 741                 $captured = $match[1];
 742                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 743                 $new_str = str_replace($captured, $spacefied, $whole_match);
 744                 return $new_str;
 745         }
 746
 747         /*
 748          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 749          * now turns back and the [noparse] tags are trimed
 750          * returning [i]italic[/i]
 751          */
 752         private static function unescapeNoparseCallback($match)
 753         {
 754                 $captured = $match[1];
 755                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 756                 return $unspacefied;
 757         }
 758
 759         /**
 760          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 761          * occurrences
 762          *
 763          * @param string $text        Text to search
 764          * @param string $name        Tag name
 765          * @param int    $occurrences Number of first occurrences to skip
 766          * @return boolean|array
 767          */
 768         public static function getTagPosition($text, $name, $occurrences = 0)
 769         {
 770                 if ($occurrences < 0) {
 771                         $occurrences = 0;
 772                 }
 773
 774                 $start_open = -1;
 775                 for ($i = 0; $i <= $occurrences; $i++) {
 776                         if ($start_open !== false) {
 777                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 778                         }
 779                 }
 780
 781                 if ($start_open === false) {
 782                         return false;
 783                 }
 784
 785                 $start_equal = strpos($text, '=', $start_open);
 786                 $start_close = strpos($text, ']', $start_open);
 787
 788                 if ($start_close === false) {
 789                         return false;
 790                 }
 791
 792                 $start_close++;
 793
 794                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 795
 796                 if ($end_open === false) {
 797                         return false;
 798                 }
 799
 800                 $res = [
 801                         'start' => [
 802                                 'open' => $start_open,
 803                                 'close' => $start_close
 804                         ],
 805                         'end' => [
 806                                 'open' => $end_open,
 807                                 'close' => $end_open + strlen('[/' . $name . ']')
 808                         ],
 809                 ];
 810
 811                 if ($start_equal !== false) {
 812                         $res['start']['equal'] = $start_equal + 1;
 813                 }
 814
 815                 return $res;
 816         }
 817
 818         /**
 819          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 820          *
 821          * @param string $pattern Preg pattern string
 822          * @param string $replace Preg replace string
 823          * @param string $name    BBCode tag name
 824          * @param string $text    Text to search
 825          * @return string
 826          */
 827         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 828         {
 829                 $occurrences = 0;
 830                 $pos = self::getTagPosition($text, $name, $occurrences);
 831                 while ($pos !== false && $occurrences++ < 1000) {
 832                         $start = substr($text, 0, $pos['start']['open']);
 833                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 834                         $end = substr($text, $pos['end']['close']);
 835                         if ($end === false) {
 836                                 $end = '';
 837                         }
 838
 839                         $subject = preg_replace($pattern, $replace, $subject);
 840                         $text = $start . $subject . $end;
 841
 842                         $pos = self::getTagPosition($text, $name, $occurrences);
 843                 }
 844
 845                 return $text;
 846         }
 847
 848         private static function extractImagesFromItemBody($body)
 849         {
 850                 $saved_image = [];
 851                 $orig_body = $body;
 852                 $new_body = '';
 853
 854                 $cnt = 0;
 855                 $img_start = strpos($orig_body, '[img');
 856                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 857                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 858                 while (($img_st_close !== false) && ($img_end !== false)) {
 859                         $img_st_close++; // make it point to AFTER the closing bracket
 860                         $img_end += $img_start;
 861
 862                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 863                                 // This is an embedded image
 864                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 865                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 866
 867                                 $cnt++;
 868                         } else {
 869                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 870                         }
 871
 872                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 873
 874                         if ($orig_body === false) {
 875                                 // in case the body ends on a closing image tag
 876                                 $orig_body = '';
 877                         }
 878
 879                         $img_start = strpos($orig_body, '[img');
 880                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 881                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 882                 }
 883
 884                 $new_body = $new_body . $orig_body;
 885
 886                 return ['body' => $new_body, 'images' => $saved_image];
 887         }
 888
 889         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 890         {
 891                 $newbody = $body;
 892
 893                 $cnt = 0;
 894                 foreach ($images as $image) {
 895                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 896                         // it loops over the array starting from the first element and going sequentially
 897                         // to the last element
 898                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 899                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 900                         $cnt++;
 901                 }
 902
 903                 return $newbody;
 904         }
 905
 906         /**
 907          * This function converts a [share] block to text according to a provided callback function whose signature is:
 908          *
 909          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 910          *
 911          * Where:
 912          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 913          * data lookup
 914          * - $author_contact is a contact record array
 915          * - $content is the inner content of the [share] block
 916          * - $is_quote_share indicates whether there's any content before the [share] block
 917          * - Return value is the string that should replace the [share] block in the provided text
 918          *
 919          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 920          *
 921          * @param  string   $text     A BBCode string
 922          * @param  callable $callback
 923          * @return string The BBCode string with all [share] blocks replaced
 924          */
 925         public static function convertShare($text, callable $callback)
 926         {
 927                 $return = preg_replace_callback(
 928                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 929                         function ($match) use ($callback) {
 930                                 $attribute_string = $match[2];
 931
 932                                 $attributes = [];
 933                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 934                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 935                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 936                                 }
 937
 938                                 // We only call this so that a previously unknown contact can be added.
 939                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 940                                 // This function then can fetch an entry from the contact table.
 941                                 $default['url'] = $attributes['profile'];
 942
 943                                 if (!empty($attributes['author'])) {
 944                                         $default['name'] = $attributes['author'];
 945                                 }
 946
 947                                 if (!empty($attributes['avatar'])) {
 948                                         $default['photo'] = $attributes['avatar'];
 949                                 }
 950
 951                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 952
 953                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 954                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 955
 956                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 957                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 958                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 959
 960                                 if ($attributes['avatar']) {
 961                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 962                                 }
 963
 964                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 965                         },
 966                         $text
 967                 );
 968
 969                 return $return;
 970         }
 971
 972         /**
 973          * Default [share] tag conversion callback
 974          *
 975          * Note: Can produce a [bookmark] tag in the output
 976          *
 977          * @see BBCode::convertShare()
 978          * @param array   $attributes     [share] block attribute values
 979          * @param array   $author_contact Contact row of the shared author
 980          * @param string  $content        Inner content of the [share] block
 981          * @param boolean $is_quote_share Whether there is content before the [share] block
 982          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 983          * @return string
 984          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 985          */
 986         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 987         {
 988                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 989
 990                 switch ($simplehtml) {
 991                         case 1:
 992                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 993                                 break;
 994                         case 2:
 995                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 996                                 break;
 997                         case 3: // Diaspora
 998                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 999                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1000                                 } else {
1001                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1002
1003                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1004                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1005                                         }
1006
1007                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1008
1009                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1010                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1011                                         }
1012                                 }
1013
1014                                 break;
1015                         case 4:
1016                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1017                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1018                                 $headline .= ':</b></p>' . "\n";
1019
1020                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1021
1022                                 break;
1023                         case 5:
1024                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1025                                 break;
1026                         case 7: // statusnet/GNU Social
1027                         case 9: // ActivityPub
1028                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1029                                 break;
1030                         default:
1031                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1032                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1033                                         try {
1034                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1035                                         } catch (Exception $e) {
1036                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1037                                         }
1038                                 } else {
1039                                         $text = ($is_quote_share? "\n" : '');
1040
1041                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1042                                         $text .= Renderer::replaceMacros($tpl, [
1043                                                 '$profile' => $attributes['profile'],
1044                                                 '$avatar'  => $attributes['avatar'],
1045                                                 '$author'  => $attributes['author'],
1046                                                 '$link'    => $attributes['link'],
1047                                                 '$posted'  => $attributes['posted'],
1048                                                 '$content' => trim($content)
1049                                         ]);
1050                                 }
1051                                 break;
1052                 }
1053
1054                 return $text;
1055         }
1056
1057         private static function removePictureLinksCallback($match)
1058         {
1059                 $text = Cache::get($match[1]);
1060
1061                 if (is_null($text)) {
1062                         $a = self::getApp();
1063
1064                         $stamp1 = microtime(true);
1065
1066                         $ch = @curl_init($match[1]);
1067                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1068                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1069                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1070                         @curl_exec($ch);
1071                         $curl_info = @curl_getinfo($ch);
1072
1073                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1074
1075                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1076                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1077                         } else {
1078                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1079
1080                                 // if its not a picture then look if its a page that contains a picture link
1081                                 $body = Network::fetchUrl($match[1]);
1082
1083                                 $doc = new DOMDocument();
1084                                 @$doc->loadHTML($body);
1085                                 $xpath = new DOMXPath($doc);
1086                                 $list = $xpath->query("//meta[@name]");
1087                                 foreach ($list as $node) {
1088                                         $attr = [];
1089
1090                                         if ($node->attributes->length) {
1091                                                 foreach ($node->attributes as $attribute) {
1092                                                         $attr[$attribute->name] = $attribute->value;
1093                                                 }
1094                                         }
1095
1096                                         if (strtolower($attr["name"]) == "twitter:image") {
1097                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1098                                         }
1099                                 }
1100                         }
1101                         Cache::set($match[1], $text);
1102                 }
1103
1104                 return $text;
1105         }
1106
1107         private static function expandLinksCallback($match)
1108         {
1109                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1110                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1111                 } else {
1112                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1113                 }
1114         }
1115
1116         private static function cleanPictureLinksCallback($match)
1117         {
1118                 $text = Cache::get($match[1]);
1119
1120                 if (is_null($text)) {
1121                         $a = self::getApp();
1122
1123                         $stamp1 = microtime(true);
1124
1125                         $ch = @curl_init($match[1]);
1126                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1127                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1128                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1129                         @curl_exec($ch);
1130                         $curl_info = @curl_getinfo($ch);
1131
1132                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1133
1134                         // if its a link to a picture then embed this picture
1135                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1136                                 $text = "[img]" . $match[1] . "[/img]";
1137                         } else {
1138                                 if (!empty($match[3])) {
1139                                         $text = "[img=" . $match[2] . "]" . $match[3] . "[/img]";
1140                                 } else {
1141                                         $text = "[img]" . $match[2] . "[/img]";
1142                                 }
1143
1144                                 // if its not a picture then look if its a page that contains a picture link
1145                                 $body = Network::fetchUrl($match[1]);
1146
1147                                 $doc = new DOMDocument();
1148                                 @$doc->loadHTML($body);
1149                                 $xpath = new DOMXPath($doc);
1150                                 $list = $xpath->query("//meta[@name]");
1151                                 foreach ($list as $node) {
1152                                         $attr = [];
1153                                         if ($node->attributes->length) {
1154                                                 foreach ($node->attributes as $attribute) {
1155                                                         $attr[$attribute->name] = $attribute->value;
1156                                                 }
1157                                         }
1158
1159                                         if (strtolower($attr["name"]) == "twitter:image") {
1160                                                 if (!empty($match[3])) {
1161                                                         $text = "[img=" . $attr["content"] . "]" . $match[3] . "[/img]";
1162                                                 } else {
1163                                                         $text = "[img]" . $attr["content"] . "[/img]";
1164                                                 }
1165                                         }
1166                                 }
1167                         }
1168                         Cache::set($match[1], $text);
1169                 }
1170
1171                 return $text;
1172         }
1173
1174         public static function cleanPictureLinks($text)
1175         {
1176                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1177                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1178                 return $return;
1179         }
1180
1181         /**
1182          * @brief Converts a BBCode message to HTML message
1183          *
1184          * BBcode 2 HTML was written by WAY2WEB.net
1185          * extended to work with Mistpark/Friendica - Mike Macgirvin
1186          *
1187          * Simple HTML values meaning:
1188          * - 0: Friendica display
1189          * - 1: Unused
1190          * - 2: Used for Windows Phone push, Friendica API
1191          * - 3: Used before converting to Markdown in bb2diaspora.php
1192          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1193          * - 5: Unused
1194          * - 6: Unused
1195          * - 7: Used for dfrn, OStatus
1196          * - 8: Used for WP backlink text setting
1197          * - 9: ActivityPub
1198          *
1199          * @param string $text
1200          * @param bool   $try_oembed
1201          * @param int    $simple_html
1202          * @param bool   $for_plaintext
1203          * @return string
1204          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1205          */
1206         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1207         {
1208                 $a = self::getApp();
1209
1210                 /*
1211                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1212                  *
1213                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1214                  * $match[1] = $url
1215                  * $match[2] = $title or absent
1216                  */
1217                 $try_oembed_callback = function ($match)
1218                 {
1219                         $url = $match[1];
1220                         $title = defaults($match, 2, null);
1221
1222                         try {
1223                                 $return = OEmbed::getHTML($url, $title);
1224                         } catch (Exception $ex) {
1225                                 $return = $match[0];
1226                         }
1227
1228                         return $return;
1229                 };
1230
1231                 // Extracting code blocks before the whitespace processing and the autolinker
1232                 $codeblocks = [];
1233
1234                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1235                         function ($matches) use (&$codeblocks) {
1236                                 $return = '#codeblock-' . count($codeblocks) . '#';
1237                                 if (strpos($matches[2], "\n") !== false) {
1238                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1239                                 } else {
1240                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1241                                 }
1242
1243                                 return $return;
1244                         },
1245                         $text
1246                 );
1247
1248                 // Hide all [noparse] contained bbtags by spacefying them
1249                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1250
1251                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1252                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1253                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1254
1255                 // Remove the abstract element. It is a non visible element.
1256                 $text = self::stripAbstract($text);
1257
1258                 // Move all spaces out of the tags
1259                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1260                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1261
1262                 // Extract the private images which use data urls since preg has issues with
1263                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1264                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1265
1266                 $extracted = self::extractImagesFromItemBody($text);
1267                 $text = $extracted['body'];
1268                 $saved_image = $extracted['images'];
1269
1270                 // If we find any event code, turn it into an event.
1271                 // After we're finished processing the bbcode we'll
1272                 // replace all of the event code with a reformatted version.
1273
1274                 $ev = Event::fromBBCode($text);
1275
1276                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1277                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1278
1279                 $text = str_replace("<", "&lt;", $text);
1280                 $text = str_replace(">", "&gt;", $text);
1281
1282                 // remove some newlines before the general conversion
1283                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1284                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1285
1286                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1287                 if (!$try_oembed) {
1288                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1289                 }
1290
1291                 // Convert new line chars to html <br /> tags
1292
1293                 // nlbr seems to be hopelessly messed up
1294                 //      $Text = nl2br($Text);
1295
1296                 // We'll emulate it.
1297
1298                 $text = trim($text);
1299                 $text = str_replace("\r\n", "\n", $text);
1300
1301                 // Remove linefeeds inside of the table elements. See issue #6799
1302                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1303                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1304                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1305                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1306                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1307                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1308                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1309                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1310                 do {
1311                         $oldtext = $text;
1312                         $text = str_replace($search, $replace, $text);
1313                 } while ($oldtext != $text);
1314
1315                 // Replace these here only once
1316                 $search = ["\n[table]", "[/table]\n"];
1317                 $replace = ["[table]", "[/table]"];
1318                 $text = str_replace($search, $replace, $text);
1319
1320                 // removing multiplicated newlines
1321                 if (Config::get("system", "remove_multiplicated_lines")) {
1322                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1323                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1324                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1325                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1326                         do {
1327                                 $oldtext = $text;
1328                                 $text = str_replace($search, $replace, $text);
1329                         } while ($oldtext != $text);
1330                 }
1331
1332                 // Handle attached links or videos
1333                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1334
1335                 // leave open the posibility of [map=something]
1336                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1337                 if (strpos($text, '[/map]') !== false) {
1338                         $text = preg_replace_callback(
1339                                 "/\[map\](.*?)\[\/map\]/ism",
1340                                 function ($match) use ($simple_html) {
1341                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1342                                 },
1343                                 $text
1344                         );
1345                 }
1346
1347                 if (strpos($text, '[map=') !== false) {
1348                         $text = preg_replace_callback(
1349                                 "/\[map=(.*?)\]/ism",
1350                                 function ($match) use ($simple_html) {
1351                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1352                                 },
1353                                 $text
1354                         );
1355                 }
1356
1357                 if (strpos($text, '[map]') !== false) {
1358                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1359                 }
1360
1361                 // Check for headers
1362                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1363                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1364                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1365                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1366                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1367                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1368
1369                 // Check for paragraph
1370                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1371
1372                 // Check for bold text
1373                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1374
1375                 // Check for Italics text
1376                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1377
1378                 // Check for Underline text
1379                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1380
1381                 // Check for strike-through text
1382                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1383
1384                 // Check for over-line text
1385                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1386
1387                 // Check for colored text
1388                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1389
1390                 // Check for sized text
1391                 // [size=50] --> font-size: 50px (with the unit).
1392                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1393                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1394
1395                 // Check for centered text
1396                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1397
1398                 // Check for list text
1399                 $text = str_replace("[*]", "<li>", $text);
1400
1401                 // Check for style sheet commands
1402                 $text = preg_replace_callback(
1403                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1404                         function ($match) {
1405                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1406                         },
1407                         $text
1408                 );
1409
1410                 // Check for CSS classes
1411                 $text = preg_replace_callback(
1412                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1413                         function ($match) {
1414                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1415                         },
1416                         $text
1417                 );
1418
1419                 // handle nested lists
1420                 $endlessloop = 0;
1421
1422                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1423                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1424                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1425                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1426                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1427                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1428                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1429                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1430                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1431                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1432                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1433                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1434                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1435                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1436                 }
1437
1438                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1439                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1440                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1441                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1442
1443                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1444                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1445
1446                 $text = str_replace('[hr]', '<hr />', $text);
1447
1448                 // This is actually executed in Item::prepareBody()
1449
1450                 $nosmile = strpos($text, '[nosmile]') !== false;
1451                 $text = str_replace('[nosmile]', '', $text);
1452
1453                 // Check for font change text
1454                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1455
1456                 // Declare the format for [spoiler] layout
1457                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1458
1459                 // Check for [spoiler] text
1460                 // handle nested quotes
1461                 $endlessloop = 0;
1462                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1463                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1464                 }
1465
1466                 // Check for [spoiler=Author] text
1467
1468                 $t_wrote = L10n::t('$1 wrote:');
1469
1470                 // handle nested quotes
1471                 $endlessloop = 0;
1472                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1473                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1474                                 "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1475                                 $text);
1476                 }
1477
1478                 // Declare the format for [quote] layout
1479                 $QuoteLayout = '<blockquote>$1</blockquote>';
1480
1481                 // Check for [quote] text
1482                 // handle nested quotes
1483                 $endlessloop = 0;
1484                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1485                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1486                 }
1487
1488                 // Check for [quote=Author] text
1489
1490                 $t_wrote = L10n::t('$1 wrote:');
1491
1492                 // handle nested quotes
1493                 $endlessloop = 0;
1494                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1495                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1496                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1497                                 $text);
1498                 }
1499
1500
1501                 // [img=widthxheight]image source[/img]
1502                 $text = preg_replace_callback(
1503                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1504                         function ($matches) use ($simple_html) {
1505                                 if (strpos($matches[3], "data:image/") === 0) {
1506                                         return $matches[0];
1507                                 }
1508
1509                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1510                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1511                         },
1512                         $text
1513                 );
1514
1515                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1516                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1517
1518                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1519                         function ($matches) use ($simple_html) {
1520                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1521                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1522                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1523                         },
1524                         $text);
1525
1526                 // Images
1527                 // [img]pathtoimage[/img]
1528                 $text = preg_replace_callback(
1529                         "/\[img\](.*?)\[\/img\]/ism",
1530                         function ($matches) use ($simple_html) {
1531                                 if (strpos($matches[1], "data:image/") === 0) {
1532                                         return $matches[0];
1533                                 }
1534
1535                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1536                                 return "[img]" . $matches[1] . "[/img]";
1537                         },
1538                         $text
1539                 );
1540
1541                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1542                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1543
1544                 $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);
1545                 $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 />', $text);
1546                 //$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 />', $Text);
1547
1548                 // Try to Oembed
1549                 if ($try_oembed) {
1550                         $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);
1551                         $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);
1552
1553                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1554                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1555                 } else {
1556                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1557                                 '<a href="$1" target="_blank">$1</a>', $text);
1558                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1559                                 '<a href="$1" target="_blank">$1</a>', $text);
1560                 }
1561
1562                 // html5 video and audio
1563
1564
1565                 if ($try_oembed) {
1566                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1567                 } else {
1568                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1569                 }
1570
1571                 // Youtube extensions
1572                 if ($try_oembed) {
1573                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1574                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1575                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1576                 }
1577
1578                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1579                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1580                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1581
1582                 if ($try_oembed) {
1583                         $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);
1584                 } else {
1585                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1586                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1587                 }
1588
1589                 if ($try_oembed) {
1590                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1591                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1592                 }
1593
1594                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1595                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1596
1597                 if ($try_oembed) {
1598                         $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);
1599                 } else {
1600                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1601                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1602                 }
1603
1604                 // oembed tag
1605                 $text = OEmbed::BBCode2HTML($text);
1606
1607                 // Avoid triple linefeeds through oembed
1608                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1609
1610                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1611                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1612                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1613                 // start which is always required). Allow desc with a missing summary for compatibility.
1614
1615                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1616                         $sub = Event::getHTML($ev, $simple_html);
1617
1618                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1619                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1620                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1621                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1622                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1623                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1624                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1625                 }
1626
1627                 // Replace non graphical smilies for external posts
1628                 if (!$nosmile && !$for_plaintext) {
1629                         $text = Smilies::replace($text);
1630                 }
1631
1632                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1633                 if (!$for_plaintext) {
1634                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1635                         if (in_array($simple_html, [7, 9])) {
1636                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1637                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1638                         }
1639                 } else {
1640                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1641                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1642                 }
1643
1644                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1645
1646                 // Remove all hashtag addresses
1647                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7, 9])) {
1648                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1649                 } elseif ($simple_html == 3) {
1650                         // The ! is converted to @ since Diaspora only understands the @
1651                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1652                                 '@<a href="$2">$3</a>',
1653                                 $text);
1654                 } elseif (in_array($simple_html, [7, 9])) {
1655                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1656                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1657                                 $text);
1658                 } elseif (!$simple_html) {
1659                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1660                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1661                                 $text);
1662                 }
1663
1664                 // Bookmarks in red - will be converted to bookmarks in friendica
1665                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1666                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1667                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1668                                         "[bookmark=$1]$2[/bookmark]", $text);
1669
1670                 if (in_array($simple_html, [2, 6, 7, 8])) {
1671                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1672                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1673                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1674                 }
1675
1676                 if ($simple_html == 5) {
1677                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1678                 }
1679
1680                 // Perform URL Search
1681                 if ($try_oembed) {
1682                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1683                 }
1684
1685                 if ($simple_html == 5) {
1686                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1687                 } else {
1688                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1689                 }
1690
1691                 // Handle Diaspora posts
1692                 $text = preg_replace_callback(
1693                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1694                         function ($match) {
1695                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1696                         }, $text
1697                 );
1698
1699                 $text = preg_replace_callback(
1700                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1701                         function ($match) {
1702                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1703                         }, $text
1704                 );
1705
1706                 // Server independent link to posts and comments
1707                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1708                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1709                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1710
1711                 /* Tag conversion
1712                  * Supports:
1713                  * - #[url=<anything>]<term>[/url]
1714                  * - [url=<anything>]#<term>[/url]
1715                  */
1716                 $text = preg_replace_callback("/(?:#\[url\=.*?\]|\[url\=.*?\]#)(.*?)\[\/url\]/ism", function($matches) {
1717                         return '#<a href="'
1718                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1719                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1720                                 . XML::escape($matches[1])
1721                                 . '</a>';
1722                 }, $text);
1723
1724                 // We need no target="_blank" for local links
1725                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1726                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1727                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1728                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1729
1730                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1731                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1732
1733                 // Red compatibility, though the link can't be authenticated on Friendica
1734                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1735
1736
1737                 // we may need to restrict this further if it picks up too many strays
1738                 // link acct:user@host to a webfinger profile redirector
1739
1740                 $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);
1741
1742                 // Perform MAIL Search
1743                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1744                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1745
1746                 // Unhide all [noparse] contained bbtags unspacefying them
1747                 // and triming the [noparse] tag.
1748
1749                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1750                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1751                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1752
1753                 /// @todo What is the meaning of these lines?
1754                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1755                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1756
1757                 // Currently deactivated, it made problems with " inside of alt texts.
1758                 //$text = preg_replace('/\&quot\;/', '"', $text);
1759
1760                 // fix any escaped ampersands that may have been converted into links
1761                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1762
1763                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1764                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1765
1766                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1767
1768                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1769                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1770
1771                 // sanitize href attributes (only whitelisted protocols URLs)
1772                 // default value for backward compatibility
1773                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1774
1775                 // Always allowed protocol even if config isn't set or not including it
1776                 $allowed_link_protocols[] = '//';
1777                 $allowed_link_protocols[] = 'http://';
1778                 $allowed_link_protocols[] = 'https://';
1779                 $allowed_link_protocols[] = 'redir/';
1780
1781                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1782
1783                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1784                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1785
1786                 // Shared content
1787                 $text = self::convertShare(
1788                         $text,
1789                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1790                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1791                         }
1792                 );
1793
1794                 if ($saved_image) {
1795                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1796                 }
1797
1798                 // Restore code blocks
1799                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1800                         function ($matches) use ($codeblocks) {
1801                                 $return = $matches[0];
1802                                 if (isset($codeblocks[intval($matches[1])])) {
1803                                         $return = $codeblocks[$matches[1]];
1804                                 }
1805                                 return $return;
1806                         },
1807                         $text
1808                 );
1809
1810                 // Clean up the HTML by loading and saving the HTML with the DOM.
1811                 // Bad structured html can break a whole page.
1812                 // For performance reasons do it only with activated item cache or at export.
1813                 if (!$try_oembed || (get_itemcachepath() != "")) {
1814                         $doc = new DOMDocument();
1815                         $doc->preserveWhiteSpace = false;
1816
1817                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1818
1819                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1820                         $encoding = '<?xml encoding="UTF-8">';
1821                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1822                         $doc->encoding = 'UTF-8';
1823                         $text = $doc->saveHTML();
1824                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1825
1826                         $text = str_replace('<br></li>', '</li>', $text);
1827
1828                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1829                 }
1830
1831                 // Clean up some useless linebreaks in lists
1832                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1833                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1834                 //$Text = str_replace('</li><br />', '</li>', $Text);
1835                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1836                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1837
1838                 Hook::callAll('bbcode', $text);
1839
1840                 return trim($text);
1841         }
1842
1843         /**
1844          * @brief Strips the "abstract" tag from the provided text
1845          *
1846          * @param string $text The text with BBCode
1847          * @return string The same text - but without "abstract" element
1848          */
1849         public static function stripAbstract($text)
1850         {
1851                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1852                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1853
1854                 return $text;
1855         }
1856
1857         /**
1858          * @brief Returns the value of the "abstract" element
1859          *
1860          * @param string $text The text that maybe contains the element
1861          * @param string $addon The addon for which the abstract is meant for
1862          * @return string The abstract
1863          */
1864         public static function getAbstract($text, $addon = "")
1865         {
1866                 $abstract = "";
1867                 $abstracts = [];
1868                 $addon = strtolower($addon);
1869
1870                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1871                         foreach ($results AS $result) {
1872                                 $abstracts[strtolower($result[1])] = $result[2];
1873                         }
1874                 }
1875
1876                 if (isset($abstracts[$addon])) {
1877                         $abstract = $abstracts[$addon];
1878                 }
1879
1880                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1881                         $abstract = $result[1];
1882                 }
1883
1884                 return $abstract;
1885         }
1886
1887         /**
1888          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1889          *
1890          * @param array $match Matching values for the callback
1891          *                     [1] = Mention type (! or @)
1892          *                     [2] = Name
1893          *                     [3] = Address
1894          * @return string Replaced mention
1895          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1896          * @throws \ImagickException
1897          */
1898         private static function bbCodeMention2DiasporaCallback($match)
1899         {
1900                 $contact = Contact::getDetailsByURL($match[3]);
1901
1902                 if (empty($contact['addr'])) {
1903                         $contact = Probe::uri($match[3]);
1904                 }
1905
1906                 if (empty($contact['addr'])) {
1907                         return $match[0];
1908                 }
1909
1910                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1911                 return $mention;
1912         }
1913
1914         /**
1915          * @brief Converts a BBCode text into Markdown
1916          *
1917          * This function converts a BBCode item body to be sent to Markdown-enabled
1918          * systems like Diaspora and Libertree
1919          *
1920          * @param string $text
1921          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1922          * @return string
1923          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1924          */
1925         public static function toMarkdown($text, $for_diaspora = true)
1926         {
1927                 $a = self::getApp();
1928
1929                 $original_text = $text;
1930
1931                 // Since Diaspora is creating a summary for links, this function removes them before posting
1932                 if ($for_diaspora) {
1933                         $text = self::removeShareInformation($text);
1934                 }
1935
1936                 /**
1937                  * Transform #tags, strip off the [url] and replace spaces with underscore
1938                  */
1939                 $url_search_string = "^\[\]";
1940                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1941                         function ($matches) {
1942                                 return '#' . str_replace(' ', '_', $matches[2]);
1943                         },
1944                         $text
1945                 );
1946
1947                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1948                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1949
1950                 // Convert it to HTML - don't try oembed
1951                 if ($for_diaspora) {
1952                         $text = self::convert($text, false, 3);
1953
1954                         // Add all tags that maybe were removed
1955                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1956                                 $tagline = "";
1957                                 foreach ($tags[2] as $tag) {
1958                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1959                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1960                                                 $tagline .= '#' . $tag . ' ';
1961                                         }
1962                                 }
1963                                 $text = $text . " " . $tagline;
1964                         }
1965                 } else {
1966                         $text = self::convert($text, false, 4);
1967                 }
1968
1969                 // mask some special HTML chars from conversation to markdown
1970                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1971
1972                 // If a link is followed by a quote then there should be a newline before it
1973                 // Maybe we should make this newline at every time before a quote.
1974                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1975
1976                 $stamp1 = microtime(true);
1977
1978                 // Now convert HTML to Markdown
1979                 $text = HTML::toMarkdown($text);
1980
1981                 // unmask the special chars back to HTML
1982                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1983
1984                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
1985
1986                 // Libertree has a problem with escaped hashtags.
1987                 $text = str_replace(['\#'], ['#'], $text);
1988
1989                 // Remove any leading or trailing whitespace, as this will mess up
1990                 // the Diaspora signature verification and cause the item to disappear
1991                 $text = trim($text);
1992
1993                 if ($for_diaspora) {
1994                         $url_search_string = "^\[\]";
1995                         $text = preg_replace_callback(
1996                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1997                                 ['self', 'bbCodeMention2DiasporaCallback'],
1998                                 $text
1999                         );
2000                 }
2001
2002                 Hook::callAll('bb2diaspora', $text);
2003
2004                 return $text;
2005         }
2006
2007         /**
2008      * @brief Pull out all #hashtags and @person tags from $string.
2009      *
2010      * We also get @person@domain.com - which would make
2011      * the regex quite complicated as tags can also
2012      * end a sentence. So we'll run through our results
2013      * and strip the period from any tags which end with one.
2014      * Returns array of tags found, or empty array.
2015      *
2016      * @param string $string Post content
2017      *
2018      * @return array List of tag and person names
2019      */
2020     public static function getTags($string)
2021     {
2022         $ret = [];
2023
2024         // Convert hashtag links to hashtags
2025         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2026
2027         // ignore anything in a code block
2028         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2029
2030         // Force line feeds at bbtags
2031         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2032
2033         // ignore anything in a bbtag
2034         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2035
2036         // Match full names against @tags including the space between first and last
2037         // We will look these up afterward to see if they are full names or not recognisable.
2038
2039         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2040             foreach ($matches[1] as $match) {
2041                 if (strstr($match, ']')) {
2042                     // we might be inside a bbcode color tag - leave it alone
2043                     continue;
2044                 }
2045
2046                 if (substr($match, -1, 1) === '.') {
2047                     $ret[] = substr($match, 0, -1);
2048                 } else {
2049                     $ret[] = $match;
2050                 }
2051             }
2052         }
2053
2054         // Otherwise pull out single word tags. These can be @nickname, @first_last
2055         // and #hash tags.
2056
2057         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2058             foreach ($matches[1] as $match) {
2059                 if (strstr($match, ']')) {
2060                     // we might be inside a bbcode color tag - leave it alone
2061                     continue;
2062                 }
2063                 if (substr($match, -1, 1) === '.') {
2064                     $match = substr($match,0,-1);
2065                 }
2066                 // ignore strictly numeric tags like #1
2067                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2068                     continue;
2069                 }
2070                 // try not to catch url fragments
2071                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2072                     continue;
2073                 }
2074                 $ret[] = $match;
2075             }
2076         }
2077
2078         return $ret;
2079     }
2080 }