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                 // Initialize empty data array
 124                 $data = [
 125                         'type'        => '',
 126                         'text'        => '',
 127                         'after'       => '',
 128                         'image'       => null,
 129                         'url'         => '',
 130                         'title'       => '',
 131                         'description' => '',
 132                 ];
 133
 134                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 135                         return self::getOldAttachmentData($body);
 136                 }
 137
 138                 $attributes = $match[2];
 139
 140                 $data["text"] = trim($match[1]);
 141
 142                 $type = "";
 143                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 144                 if (!empty($matches[1])) {
 145                         $type = strtolower($matches[1]);
 146                 }
 147
 148                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 149                 if (!empty($matches[1])) {
 150                         $type = strtolower($matches[1]);
 151                 }
 152
 153                 if ($type == "") {
 154                         return [];
 155                 }
 156
 157                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 158                         return [];
 159                 }
 160
 161                 if ($type != "") {
 162                         $data["type"] = $type;
 163                 }
 164
 165                 $url = "";
 166                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 167                 if (!empty($matches[1])) {
 168                         $url = $matches[1];
 169                 }
 170
 171                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 172                 if (!empty($matches[1])) {
 173                         $url = $matches[1];
 174                 }
 175
 176                 if ($url != "") {
 177                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 178                 }
 179
 180                 $title = "";
 181                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 182                 if (!empty($matches[1])) {
 183                         $title = $matches[1];
 184                 }
 185
 186                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 187                 if (!empty($matches[1])) {
 188                         $title = $matches[1];
 189                 }
 190
 191                 if ($title != "") {
 192                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 193                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 194                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 195                         $data["title"] = $title;
 196                 }
 197
 198                 $image = "";
 199                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 200                 if (!empty($matches[1])) {
 201                         $image = $matches[1];
 202                 }
 203
 204                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 205                 if (!empty($matches[1])) {
 206                         $image = $matches[1];
 207                 }
 208
 209                 if ($image != "") {
 210                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 211                 }
 212
 213                 $preview = "";
 214                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 215                 if (!empty($matches[1])) {
 216                         $preview = $matches[1];
 217                 }
 218
 219                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 220                 if (!empty($matches[1])) {
 221                         $preview = $matches[1];
 222                 }
 223
 224                 if ($preview != "") {
 225                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 226                 }
 227
 228                 $data["description"] = trim($match[3]);
 229
 230                 $data["after"] = trim($match[4]);
 231
 232                 return $data;
 233         }
 234
 235         public static function getAttachedData($body, $item = [])
 236         {
 237                 /*
 238                 - text:
 239                 - type: link, video, photo
 240                 - title:
 241                 - url:
 242                 - image:
 243                 - description:
 244                 - (thumbnail)
 245                 */
 246
 247                 $has_title = !empty($item['title']);
 248                 $plink = defaults($item, 'plink', '');
 249                 $post = self::getAttachmentData($body);
 250
 251                 // Get all linked images with alternative image description
 252                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 253                         foreach ($pictures as $picture) {
 254                                 if (Photo::isLocal($picture[1])) {
 255                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 256                                 }
 257                         }
 258                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 259                                 $post['image_description'] = $post['images'][0]['description'];
 260                         }
 261                 }
 262
 263                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 264                         foreach ($pictures as $picture) {
 265                                 if (Photo::isLocal($picture[1])) {
 266                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 267                                 }
 268                         }
 269                 }
 270
 271                 // if nothing is found, it maybe having an image.
 272                 if (!isset($post["type"])) {
 273                         // Simplify image codes
 274                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 275                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 276                         $post["text"] = $body;
 277
 278                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 279                                 if ((count($pictures) == 1) && !$has_title) {
 280                                         if (!empty($item['object-type']) && ($item['object-type'] == ACTIVITY_OBJ_IMAGE)) {
 281                                                 // Replace the preview picture with the real picture
 282                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 283                                                 $data = ['url' => $url, 'type' => 'photo'];
 284                                         } else {
 285                                                 // Checking, if the link goes to a picture
 286                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 287                                         }
 288
 289                                         // Workaround:
 290                                         // Sometimes photo posts to the own album are not detected at the start.
 291                                         // So we seem to cannot use the cache for these cases. That's strange.
 292                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 293                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 294                                         }
 295
 296                                         if ($data["type"] == "photo") {
 297                                                 $post["type"] = "photo";
 298                                                 if (isset($data["images"][0])) {
 299                                                         $post["image"] = $data["images"][0]["src"];
 300                                                         $post["url"] = $data["url"];
 301                                                 } else {
 302                                                         $post["image"] = $data["url"];
 303                                                 }
 304
 305                                                 $post["preview"] = $pictures[0][2];
 306                                                 $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 307                                         } else {
 308                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 309                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 310                                                         $post["type"] = "photo";
 311                                                         $post["image"] = $pictures[0][1];
 312                                                         $post["preview"] = $pictures[0][2];
 313                                                         $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 314                                                 }
 315                                         }
 316                                 } elseif (count($pictures) > 0) {
 317                                         $post["type"] = "link";
 318                                         $post["url"] = $plink;
 319                                         $post["image"] = $pictures[0][2];
 320                                         $post["text"] = $body;
 321
 322                                         foreach ($pictures as $picture) {
 323                                                 $post["text"] = trim(str_replace($picture[0], "", $post["text"]));
 324                                         }
 325                                 }
 326                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 327                                 if ((count($pictures) == 1) && !$has_title) {
 328                                         $post["type"] = "photo";
 329                                         $post["image"] = $pictures[0][1];
 330                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 331                                 } elseif (count($pictures) > 0) {
 332                                         $post["type"] = "link";
 333                                         $post["url"] = $plink;
 334                                         $post["image"] = $pictures[0][1];
 335                                         $post["text"] = $body;
 336
 337                                         foreach ($pictures as $picture) {
 338                                                 $post["text"] = trim(str_replace($picture[0], "", $post["text"]));
 339                                         }
 340                                 }
 341                         }
 342
 343                         // Test for the external links
 344                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post["text"], $links1, PREG_SET_ORDER);
 345                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post["text"], $links2, PREG_SET_ORDER);
 346
 347                         $links = array_merge($links1, $links2);
 348
 349                         // If there is only a single one, then use it.
 350                         // This should cover link posts via API.
 351                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 352                                 $post["type"] = "link";
 353                                 $post["url"] = $links[0][1];
 354                         }
 355
 356                         // Now count the number of external media links
 357                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post["text"], $links1, PREG_SET_ORDER);
 358                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post["text"], $links2, PREG_SET_ORDER);
 359                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post["text"], $links3, PREG_SET_ORDER);
 360                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post["text"], $links4, PREG_SET_ORDER);
 361
 362                         // Add them to the other external links
 363                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 364
 365                         // Are there more than one?
 366                         if (count($links) > 1) {
 367                                 // The post will be the type "text", which means a blog post
 368                                 unset($post["type"]);
 369                                 $post["url"] = $plink;
 370                         }
 371
 372                         if (!isset($post["type"])) {
 373                                 $post["type"] = "text";
 374                                 $post["text"] = trim($body);
 375                         }
 376                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 377                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 378
 379                         if (isset($data["images"][0])) {
 380                                 $post["image"] = $data["images"][0]["src"];
 381                         }
 382                 }
 383
 384                 return $post;
 385         }
 386
 387         /**
 388          * @brief Converts a BBCode text into plaintext
 389          *
 390          * @param      $text
 391          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 392          *
 393          * @return string
 394          */
 395         public static function toPlaintext($text, $keep_urls = true)
 396         {
 397                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 398
 399                 return $naked_text;
 400         }
 401
 402         private static function proxyUrl($image, $simplehtml = false)
 403         {
 404                 // Only send proxied pictures to API and for internal display
 405                 if (in_array($simplehtml, [false, 2])) {
 406                         return ProxyUtils::proxifyUrl($image);
 407                 } else {
 408                         return $image;
 409                 }
 410         }
 411
 412         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 413         {
 414                 // Suppress "view full size"
 415                 if (intval(Config::get('system', 'no_view_full_size'))) {
 416                         $include_link = false;
 417                 }
 418
 419                 // Picture addresses can contain special characters
 420                 $s = htmlspecialchars_decode($srctext);
 421
 422                 $matches = null;
 423                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 424                 if ($c) {
 425                         foreach ($matches as $mtch) {
 426                                 Logger::log('scale_external_image: ' . $mtch[1]);
 427
 428                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 429                                 if (stristr($mtch[1], $hostname)) {
 430                                         continue;
 431                                 }
 432
 433                                 // $scale_replace, if passed, is an array of two elements. The
 434                                 // first is the name of the full-size image. The second is the
 435                                 // name of a remote, scaled-down version of the full size image.
 436                                 // This allows Friendica to display the smaller remote image if
 437                                 // one exists, while still linking to the full-size image
 438                                 if ($scale_replace) {
 439                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 440                                 } else {
 441                                         $scaled = $mtch[1];
 442                                 }
 443                                 $i = Network::fetchUrl($scaled);
 444                                 if (!$i) {
 445                                         return $srctext;
 446                                 }
 447
 448                                 // guess mimetype from headers or filename
 449                                 $type = Image::guessType($mtch[1], true);
 450
 451                                 if ($i) {
 452                                         $Image = new Image($i, $type);
 453                                         if ($Image->isValid()) {
 454                                                 $orig_width = $Image->getWidth();
 455                                                 $orig_height = $Image->getHeight();
 456
 457                                                 if ($orig_width > 640 || $orig_height > 640) {
 458                                                         $Image->scaleDown(640);
 459                                                         $new_width = $Image->getWidth();
 460                                                         $new_height = $Image->getHeight();
 461                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 462                                                         $s = str_replace(
 463                                                                 $mtch[0],
 464                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 465                                                                 . "\n" . (($include_link)
 466                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 467                                                                         : ''),
 468                                                                 $s
 469                                                         );
 470                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 471                                                 }
 472                                         }
 473                                 }
 474                         }
 475                 }
 476
 477                 // replace the special char encoding
 478                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 479                 return $s;
 480         }
 481
 482         /**
 483          * The purpose of this function is to apply system message length limits to
 484          * imported messages without including any embedded photos in the length
 485          *
 486          * @brief Truncates imported message body string length to max_import_size
 487          * @param string $body
 488          * @return string
 489          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 490          */
 491         public static function limitBodySize($body)
 492         {
 493                 $maxlen = Config::get('config', 'max_import_size', 0);
 494
 495                 // If the length of the body, including the embedded images, is smaller
 496                 // than the maximum, then don't waste time looking for the images
 497                 if ($maxlen && (strlen($body) > $maxlen)) {
 498
 499                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 500
 501                         $orig_body = $body;
 502                         $new_body = '';
 503                         $textlen = 0;
 504
 505                         $img_start = strpos($orig_body, '[img');
 506                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 507                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 508                         while (($img_st_close !== false) && ($img_end !== false)) {
 509
 510                                 $img_st_close++; // make it point to AFTER the closing bracket
 511                                 $img_end += $img_start;
 512                                 $img_end += strlen('[/img]');
 513
 514                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 515                                         // This is an embedded image
 516
 517                                         if (($textlen + $img_start) > $maxlen) {
 518                                                 if ($textlen < $maxlen) {
 519                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 520                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 521                                                         $textlen = $maxlen;
 522                                                 }
 523                                         } else {
 524                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 525                                                 $textlen += $img_start;
 526                                         }
 527
 528                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 529                                 } else {
 530
 531                                         if (($textlen + $img_end) > $maxlen) {
 532                                                 if ($textlen < $maxlen) {
 533                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 534                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 535                                                         $textlen = $maxlen;
 536                                                 }
 537                                         } else {
 538                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 539                                                 $textlen += $img_end;
 540                                         }
 541                                 }
 542                                 $orig_body = substr($orig_body, $img_end);
 543
 544                                 if ($orig_body === false) {
 545                                         // in case the body ends on a closing image tag
 546                                         $orig_body = '';
 547                                 }
 548
 549                                 $img_start = strpos($orig_body, '[img');
 550                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 551                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 552                         }
 553
 554                         if (($textlen + strlen($orig_body)) > $maxlen) {
 555                                 if ($textlen < $maxlen) {
 556                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 557                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 558                                 }
 559                         } else {
 560                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 561                                 $new_body = $new_body . $orig_body;
 562                         }
 563
 564                         return $new_body;
 565                 } else {
 566                         return $body;
 567                 }
 568         }
 569
 570         /**
 571          * Processes [attachment] tags
 572          *
 573          * Note: Can produce a [bookmark] tag in the returned string
 574          *
 575          * @brief Processes [attachment] tags
 576          * @param string   $return
 577          * @param bool|int $simplehtml
 578          * @param bool     $tryoembed
 579          * @return string
 580          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 581          */
 582         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 583         {
 584                 $data = self::getAttachmentData($return);
 585                 if (empty($data) || empty($data["url"])) {
 586                         return $return;
 587                 }
 588
 589                 if (isset($data["title"])) {
 590                         $data["title"] = strip_tags($data["title"]);
 591                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 592                 } else {
 593                         $data["title"] = null;
 594                 }
 595
 596                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 597                         $data["preview"] = $data["image"];
 598                         $data["image"] = "";
 599                 }
 600
 601                 $return = '';
 602                 if (in_array($simplehtml, [7, 9])) {
 603                         $return = self::convertUrlForOStatus($data["url"]);
 604                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 605                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 606                 } else {
 607                         try {
 608                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 609                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 610                                 } else {
 611                                         throw new Exception('OEmbed is disabled for this attachment.');
 612                                 }
 613                         } catch (Exception $e) {
 614                                 $data["title"] = defaults($data, 'title', $data['url']);
 615
 616                                 if ($simplehtml != 4) {
 617                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 618                                 }
 619
 620                                 if (!empty($data['title']) && !empty($data['url'])) {
 621                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 622                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 623                                         } else {
 624                                                 if (!empty($data["image"])) {
 625                                                         $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), $data["title"]);
 626                                                 } elseif (!empty($data["preview"])) {
 627                                                         $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"]);
 628                                                 }
 629                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 630                                         }
 631                                 }
 632
 633                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 634                                         // Sanitize the HTML by converting it to BBCode
 635                                         $bbcode = HTML::toBBCode($data["description"]);
 636                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 637                                 }
 638
 639                                 if (!empty($data['url'])) {
 640                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 641                                 }
 642
 643                                 if ($simplehtml != 4) {
 644                                         $return .= '</div>';
 645                                 }
 646                         }
 647                 }
 648
 649                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 650         }
 651
 652         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 653         {
 654                 $data = self::getAttachmentData($Text);
 655
 656                 if (!$data) {
 657                         return $Text;
 658                 } elseif ($nolink) {
 659                         return $data["text"] . defaults($data, 'after', '');
 660                 }
 661
 662                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 663                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 664                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 665                         $data["title"] = $data["url"];
 666                 } elseif (($text != "") && strstr($title, $text)) {
 667                         $data["text"] = $data["title"];
 668                         $data["title"] = $data["url"];
 669                 }
 670
 671                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 672                         return $data["title"] . $data["after"];
 673                 }
 674
 675                 // If the link already is included in the post, don't add it again
 676                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 677                         return $data["text"] . $data["after"];
 678                 }
 679
 680                 $text = $data["text"];
 681
 682                 if (!empty($data["url"]) && !empty($data["title"])) {
 683                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 684                 } elseif (!empty($data["url"])) {
 685                         $text .= "\n[url]" . $data["url"] . "[/url]";
 686                 }
 687
 688                 return $text . "\n" . $data["after"];
 689         }
 690
 691         /**
 692          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 693          *
 694          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 695          * @param array $match Array with the matching values
 696          * @return string reformatted link including HTML codes
 697          */
 698         private static function convertUrlForOStatusCallback($match)
 699         {
 700                 $url = $match[1];
 701
 702                 if (isset($match[2]) && ($match[1] != $match[2])) {
 703                         return $match[0];
 704                 }
 705
 706                 $parts = parse_url($url);
 707                 if (!isset($parts['scheme'])) {
 708                         return $match[0];
 709                 }
 710
 711                 return self::convertUrlForOStatus($url);
 712         }
 713
 714         /**
 715          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 716          * @param string $url URL that is about to be reformatted
 717          * @return string reformatted link including HTML codes
 718          */
 719         private static function convertUrlForOStatus($url)
 720         {
 721                 $parts = parse_url($url);
 722                 $scheme = $parts['scheme'] . '://';
 723                 $styled_url = str_replace($scheme, '', $url);
 724
 725                 if (strlen($styled_url) > 30) {
 726                         $styled_url = substr($styled_url, 0, 30) . "…";
 727                 }
 728
 729                 $html = '<a href="%s" target="_blank">%s</a>';
 730
 731                 return sprintf($html, $url, $styled_url);
 732         }
 733
 734         /*
 735          * [noparse][i]italic[/i][/noparse] turns into
 736          * [noparse][ i ]italic[ /i ][/noparse],
 737          * to hide them from parser.
 738          */
 739         private static function escapeNoparseCallback($match)
 740         {
 741                 $whole_match = $match[0];
 742                 $captured = $match[1];
 743                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 744                 $new_str = str_replace($captured, $spacefied, $whole_match);
 745                 return $new_str;
 746         }
 747
 748         /*
 749          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 750          * now turns back and the [noparse] tags are trimed
 751          * returning [i]italic[/i]
 752          */
 753         private static function unescapeNoparseCallback($match)
 754         {
 755                 $captured = $match[1];
 756                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 757                 return $unspacefied;
 758         }
 759
 760         /**
 761          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 762          * occurrences
 763          *
 764          * @param string $text        Text to search
 765          * @param string $name        Tag name
 766          * @param int    $occurrences Number of first occurrences to skip
 767          * @return boolean|array
 768          */
 769         public static function getTagPosition($text, $name, $occurrences = 0)
 770         {
 771                 if ($occurrences < 0) {
 772                         $occurrences = 0;
 773                 }
 774
 775                 $start_open = -1;
 776                 for ($i = 0; $i <= $occurrences; $i++) {
 777                         if ($start_open !== false) {
 778                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 779                         }
 780                 }
 781
 782                 if ($start_open === false) {
 783                         return false;
 784                 }
 785
 786                 $start_equal = strpos($text, '=', $start_open);
 787                 $start_close = strpos($text, ']', $start_open);
 788
 789                 if ($start_close === false) {
 790                         return false;
 791                 }
 792
 793                 $start_close++;
 794
 795                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 796
 797                 if ($end_open === false) {
 798                         return false;
 799                 }
 800
 801                 $res = [
 802                         'start' => [
 803                                 'open' => $start_open,
 804                                 'close' => $start_close
 805                         ],
 806                         'end' => [
 807                                 'open' => $end_open,
 808                                 'close' => $end_open + strlen('[/' . $name . ']')
 809                         ],
 810                 ];
 811
 812                 if ($start_equal !== false) {
 813                         $res['start']['equal'] = $start_equal + 1;
 814                 }
 815
 816                 return $res;
 817         }
 818
 819         /**
 820          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 821          *
 822          * @param string $pattern Preg pattern string
 823          * @param string $replace Preg replace string
 824          * @param string $name    BBCode tag name
 825          * @param string $text    Text to search
 826          * @return string
 827          */
 828         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 829         {
 830                 $occurrences = 0;
 831                 $pos = self::getTagPosition($text, $name, $occurrences);
 832                 while ($pos !== false && $occurrences++ < 1000) {
 833                         $start = substr($text, 0, $pos['start']['open']);
 834                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 835                         $end = substr($text, $pos['end']['close']);
 836                         if ($end === false) {
 837                                 $end = '';
 838                         }
 839
 840                         $subject = preg_replace($pattern, $replace, $subject);
 841                         $text = $start . $subject . $end;
 842
 843                         $pos = self::getTagPosition($text, $name, $occurrences);
 844                 }
 845
 846                 return $text;
 847         }
 848
 849         private static function extractImagesFromItemBody($body)
 850         {
 851                 $saved_image = [];
 852                 $orig_body = $body;
 853                 $new_body = '';
 854
 855                 $cnt = 0;
 856                 $img_start = strpos($orig_body, '[img');
 857                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 858                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 859                 while (($img_st_close !== false) && ($img_end !== false)) {
 860                         $img_st_close++; // make it point to AFTER the closing bracket
 861                         $img_end += $img_start;
 862
 863                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 864                                 // This is an embedded image
 865                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 866                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 867
 868                                 $cnt++;
 869                         } else {
 870                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 871                         }
 872
 873                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 874
 875                         if ($orig_body === false) {
 876                                 // in case the body ends on a closing image tag
 877                                 $orig_body = '';
 878                         }
 879
 880                         $img_start = strpos($orig_body, '[img');
 881                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 882                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 883                 }
 884
 885                 $new_body = $new_body . $orig_body;
 886
 887                 return ['body' => $new_body, 'images' => $saved_image];
 888         }
 889
 890         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 891         {
 892                 $newbody = $body;
 893
 894                 $cnt = 0;
 895                 foreach ($images as $image) {
 896                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 897                         // it loops over the array starting from the first element and going sequentially
 898                         // to the last element
 899                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 900                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 901                         $cnt++;
 902                 }
 903
 904                 return $newbody;
 905         }
 906
 907         /**
 908          * This function converts a [share] block to text according to a provided callback function whose signature is:
 909          *
 910          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 911          *
 912          * Where:
 913          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 914          * data lookup
 915          * - $author_contact is a contact record array
 916          * - $content is the inner content of the [share] block
 917          * - $is_quote_share indicates whether there's any content before the [share] block
 918          * - Return value is the string that should replace the [share] block in the provided text
 919          *
 920          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 921          *
 922          * @param  string   $text     A BBCode string
 923          * @param  callable $callback
 924          * @return string The BBCode string with all [share] blocks replaced
 925          */
 926         public static function convertShare($text, callable $callback)
 927         {
 928                 $return = preg_replace_callback(
 929                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 930                         function ($match) use ($callback) {
 931                                 $attribute_string = $match[2];
 932
 933                                 $attributes = [];
 934                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 935                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 936                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 937                                 }
 938
 939                                 // We only call this so that a previously unknown contact can be added.
 940                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 941                                 // This function then can fetch an entry from the contact table.
 942                                 $default['url'] = $attributes['profile'];
 943
 944                                 if (!empty($attributes['author'])) {
 945                                         $default['name'] = $attributes['author'];
 946                                 }
 947
 948                                 if (!empty($attributes['avatar'])) {
 949                                         $default['photo'] = $attributes['avatar'];
 950                                 }
 951
 952                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 953
 954                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 955                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 956
 957                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 958                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 959                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 960
 961                                 if ($attributes['avatar']) {
 962                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 963                                 }
 964
 965                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 966                         },
 967                         $text
 968                 );
 969
 970                 return $return;
 971         }
 972
 973         /**
 974          * Default [share] tag conversion callback
 975          *
 976          * Note: Can produce a [bookmark] tag in the output
 977          *
 978          * @see BBCode::convertShare()
 979          * @param array   $attributes     [share] block attribute values
 980          * @param array   $author_contact Contact row of the shared author
 981          * @param string  $content        Inner content of the [share] block
 982          * @param boolean $is_quote_share Whether there is content before the [share] block
 983          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 984          * @return string
 985          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 986          */
 987         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 988         {
 989                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 990
 991                 switch ($simplehtml) {
 992                         case 1:
 993                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 994                                 break;
 995                         case 2:
 996                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 997                                 break;
 998                         case 3: // Diaspora
 999                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1000                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1001                                 } else {
1002                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1003
1004                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1005                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1006                                         }
1007
1008                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1009
1010                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1011                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1012                                         }
1013                                 }
1014
1015                                 break;
1016                         case 4:
1017                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1018                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1019                                 $headline .= ':</b></p>' . "\n";
1020
1021                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1022
1023                                 break;
1024                         case 5:
1025                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1026                                 break;
1027                         case 7: // statusnet/GNU Social
1028                         case 9: // ActivityPub
1029                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1030                                 break;
1031                         default:
1032                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1033                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1034                                         try {
1035                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1036                                         } catch (Exception $e) {
1037                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1038                                         }
1039                                 } else {
1040                                         $text = ($is_quote_share? "\n" : '');
1041
1042                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1043                                         $text .= Renderer::replaceMacros($tpl, [
1044                                                 '$profile' => $attributes['profile'],
1045                                                 '$avatar'  => $attributes['avatar'],
1046                                                 '$author'  => $attributes['author'],
1047                                                 '$link'    => $attributes['link'],
1048                                                 '$posted'  => $attributes['posted'],
1049                                                 '$content' => trim($content)
1050                                         ]);
1051                                 }
1052                                 break;
1053                 }
1054
1055                 return $text;
1056         }
1057
1058         private static function removePictureLinksCallback($match)
1059         {
1060                 $text = Cache::get($match[1]);
1061
1062                 if (is_null($text)) {
1063                         $a = self::getApp();
1064
1065                         $stamp1 = microtime(true);
1066
1067                         $ch = @curl_init($match[1]);
1068                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1069                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1070                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1071                         @curl_exec($ch);
1072                         $curl_info = @curl_getinfo($ch);
1073
1074                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1075
1076                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1077                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1078                         } else {
1079                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1080
1081                                 // if its not a picture then look if its a page that contains a picture link
1082                                 $body = Network::fetchUrl($match[1]);
1083
1084                                 $doc = new DOMDocument();
1085                                 @$doc->loadHTML($body);
1086                                 $xpath = new DOMXPath($doc);
1087                                 $list = $xpath->query("//meta[@name]");
1088                                 foreach ($list as $node) {
1089                                         $attr = [];
1090
1091                                         if ($node->attributes->length) {
1092                                                 foreach ($node->attributes as $attribute) {
1093                                                         $attr[$attribute->name] = $attribute->value;
1094                                                 }
1095                                         }
1096
1097                                         if (strtolower($attr["name"]) == "twitter:image") {
1098                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1099                                         }
1100                                 }
1101                         }
1102                         Cache::set($match[1], $text);
1103                 }
1104
1105                 return $text;
1106         }
1107
1108         private static function expandLinksCallback($match)
1109         {
1110                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1111                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1112                 } else {
1113                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1114                 }
1115         }
1116
1117         private static function cleanPictureLinksCallback($match)
1118         {
1119                 $text = Cache::get($match[1]);
1120
1121                 if (is_null($text)) {
1122                         $a = self::getApp();
1123
1124                         $stamp1 = microtime(true);
1125
1126                         $ch = @curl_init($match[1]);
1127                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1128                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1129                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1130                         @curl_exec($ch);
1131                         $curl_info = @curl_getinfo($ch);
1132
1133                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1134
1135                         // if its a link to a picture then embed this picture
1136                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1137                                 $text = "[img]" . $match[1] . "[/img]";
1138                         } else {
1139                                 if (!empty($match[3])) {
1140                                         $text = "[img=" . $match[2] . "]" . $match[3] . "[/img]";
1141                                 } else {
1142                                         $text = "[img]" . $match[2] . "[/img]";
1143                                 }
1144
1145                                 // if its not a picture then look if its a page that contains a picture link
1146                                 $body = Network::fetchUrl($match[1]);
1147
1148                                 $doc = new DOMDocument();
1149                                 @$doc->loadHTML($body);
1150                                 $xpath = new DOMXPath($doc);
1151                                 $list = $xpath->query("//meta[@name]");
1152                                 foreach ($list as $node) {
1153                                         $attr = [];
1154                                         if ($node->attributes->length) {
1155                                                 foreach ($node->attributes as $attribute) {
1156                                                         $attr[$attribute->name] = $attribute->value;
1157                                                 }
1158                                         }
1159
1160                                         if (strtolower($attr["name"]) == "twitter:image") {
1161                                                 if (!empty($match[3])) {
1162                                                         $text = "[img=" . $attr["content"] . "]" . $match[3] . "[/img]";
1163                                                 } else {
1164                                                         $text = "[img]" . $attr["content"] . "[/img]";
1165                                                 }
1166                                         }
1167                                 }
1168                         }
1169                         Cache::set($match[1], $text);
1170                 }
1171
1172                 return $text;
1173         }
1174
1175         public static function cleanPictureLinks($text)
1176         {
1177                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1178                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1179                 return $return;
1180         }
1181
1182         /**
1183          * @brief Converts a BBCode message to HTML message
1184          *
1185          * BBcode 2 HTML was written by WAY2WEB.net
1186          * extended to work with Mistpark/Friendica - Mike Macgirvin
1187          *
1188          * Simple HTML values meaning:
1189          * - 0: Friendica display
1190          * - 1: Unused
1191          * - 2: Used for Windows Phone push, Friendica API
1192          * - 3: Used before converting to Markdown in bb2diaspora.php
1193          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1194          * - 5: Unused
1195          * - 6: Unused
1196          * - 7: Used for dfrn, OStatus
1197          * - 8: Used for WP backlink text setting
1198          * - 9: ActivityPub
1199          *
1200          * @param string $text
1201          * @param bool   $try_oembed
1202          * @param int    $simple_html
1203          * @param bool   $for_plaintext
1204          * @return string
1205          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1206          */
1207         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1208         {
1209                 $a = self::getApp();
1210
1211                 /*
1212                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1213                  *
1214                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1215                  * $match[1] = $url
1216                  * $match[2] = $title or absent
1217                  */
1218                 $try_oembed_callback = function ($match)
1219                 {
1220                         $url = $match[1];
1221                         $title = defaults($match, 2, null);
1222
1223                         try {
1224                                 $return = OEmbed::getHTML($url, $title);
1225                         } catch (Exception $ex) {
1226                                 $return = $match[0];
1227                         }
1228
1229                         return $return;
1230                 };
1231
1232                 // Extracting code blocks before the whitespace processing and the autolinker
1233                 $codeblocks = [];
1234
1235                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1236                         function ($matches) use (&$codeblocks) {
1237                                 $return = '#codeblock-' . count($codeblocks) . '#';
1238                                 if (strpos($matches[2], "\n") !== false) {
1239                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1240                                 } else {
1241                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1242                                 }
1243
1244                                 return $return;
1245                         },
1246                         $text
1247                 );
1248
1249                 // Hide all [noparse] contained bbtags by spacefying them
1250                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1251
1252                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1253                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1254                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1255
1256                 // Remove the abstract element. It is a non visible element.
1257                 $text = self::stripAbstract($text);
1258
1259                 // Move all spaces out of the tags
1260                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1261                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1262
1263                 // Extract the private images which use data urls since preg has issues with
1264                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1265                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1266
1267                 $extracted = self::extractImagesFromItemBody($text);
1268                 $text = $extracted['body'];
1269                 $saved_image = $extracted['images'];
1270
1271                 // If we find any event code, turn it into an event.
1272                 // After we're finished processing the bbcode we'll
1273                 // replace all of the event code with a reformatted version.
1274
1275                 $ev = Event::fromBBCode($text);
1276
1277                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1278                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1279
1280                 $text = str_replace("<", "&lt;", $text);
1281                 $text = str_replace(">", "&gt;", $text);
1282
1283                 // remove some newlines before the general conversion
1284                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1285                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1286
1287                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1288                 if (!$try_oembed) {
1289                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1290                 }
1291
1292                 // Convert new line chars to html <br /> tags
1293
1294                 // nlbr seems to be hopelessly messed up
1295                 //      $Text = nl2br($Text);
1296
1297                 // We'll emulate it.
1298
1299                 $text = trim($text);
1300                 $text = str_replace("\r\n", "\n", $text);
1301
1302                 // Remove linefeeds inside of the table elements. See issue #6799
1303                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1304                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1305                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1306                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1307                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1308                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1309                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1310                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1311                 do {
1312                         $oldtext = $text;
1313                         $text = str_replace($search, $replace, $text);
1314                 } while ($oldtext != $text);
1315
1316                 // Replace these here only once
1317                 $search = ["\n[table]", "[/table]\n"];
1318                 $replace = ["[table]", "[/table]"];
1319                 $text = str_replace($search, $replace, $text);
1320
1321                 // removing multiplicated newlines
1322                 if (Config::get("system", "remove_multiplicated_lines")) {
1323                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1324                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1325                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1326                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1327                         do {
1328                                 $oldtext = $text;
1329                                 $text = str_replace($search, $replace, $text);
1330                         } while ($oldtext != $text);
1331                 }
1332
1333                 // Handle attached links or videos
1334                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1335
1336                 // leave open the posibility of [map=something]
1337                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1338                 if (strpos($text, '[/map]') !== false) {
1339                         $text = preg_replace_callback(
1340                                 "/\[map\](.*?)\[\/map\]/ism",
1341                                 function ($match) use ($simple_html) {
1342                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1343                                 },
1344                                 $text
1345                         );
1346                 }
1347
1348                 if (strpos($text, '[map=') !== false) {
1349                         $text = preg_replace_callback(
1350                                 "/\[map=(.*?)\]/ism",
1351                                 function ($match) use ($simple_html) {
1352                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1353                                 },
1354                                 $text
1355                         );
1356                 }
1357
1358                 if (strpos($text, '[map]') !== false) {
1359                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1360                 }
1361
1362                 // Check for headers
1363                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1364                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1365                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1366                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1367                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1368                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1369
1370                 // Check for paragraph
1371                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1372
1373                 // Check for bold text
1374                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1375
1376                 // Check for Italics text
1377                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1378
1379                 // Check for Underline text
1380                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1381
1382                 // Check for strike-through text
1383                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1384
1385                 // Check for over-line text
1386                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1387
1388                 // Check for colored text
1389                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1390
1391                 // Check for sized text
1392                 // [size=50] --> font-size: 50px (with the unit).
1393                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1394                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1395
1396                 // Check for centered text
1397                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1398
1399                 // Check for list text
1400                 $text = str_replace("[*]", "<li>", $text);
1401
1402                 // Check for style sheet commands
1403                 $text = preg_replace_callback(
1404                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1405                         function ($match) {
1406                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1407                         },
1408                         $text
1409                 );
1410
1411                 // Check for CSS classes
1412                 $text = preg_replace_callback(
1413                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1414                         function ($match) {
1415                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1416                         },
1417                         $text
1418                 );
1419
1420                 // handle nested lists
1421                 $endlessloop = 0;
1422
1423                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1424                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1425                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1426                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1427                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1428                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1429                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1430                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1431                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1432                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1433                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1434                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1435                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1436                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1437                 }
1438
1439                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1440                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1441                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1442                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1443
1444                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1445                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1446
1447                 $text = str_replace('[hr]', '<hr />', $text);
1448
1449                 // This is actually executed in Item::prepareBody()
1450
1451                 $nosmile = strpos($text, '[nosmile]') !== false;
1452                 $text = str_replace('[nosmile]', '', $text);
1453
1454                 // Check for font change text
1455                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1456
1457                 // Declare the format for [spoiler] layout
1458                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1459
1460                 // Check for [spoiler] text
1461                 // handle nested quotes
1462                 $endlessloop = 0;
1463                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1464                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1465                 }
1466
1467                 // Check for [spoiler=Author] text
1468
1469                 $t_wrote = L10n::t('$1 wrote:');
1470
1471                 // handle nested quotes
1472                 $endlessloop = 0;
1473                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1474                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1475                                 "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1476                                 $text);
1477                 }
1478
1479                 // Declare the format for [quote] layout
1480                 $QuoteLayout = '<blockquote>$1</blockquote>';
1481
1482                 // Check for [quote] text
1483                 // handle nested quotes
1484                 $endlessloop = 0;
1485                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1486                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1487                 }
1488
1489                 // Check for [quote=Author] text
1490
1491                 $t_wrote = L10n::t('$1 wrote:');
1492
1493                 // handle nested quotes
1494                 $endlessloop = 0;
1495                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1496                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1497                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1498                                 $text);
1499                 }
1500
1501
1502                 // [img=widthxheight]image source[/img]
1503                 $text = preg_replace_callback(
1504                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1505                         function ($matches) use ($simple_html) {
1506                                 if (strpos($matches[3], "data:image/") === 0) {
1507                                         return $matches[0];
1508                                 }
1509
1510                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1511                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1512                         },
1513                         $text
1514                 );
1515
1516                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1517                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1518
1519                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1520                         function ($matches) use ($simple_html) {
1521                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1522                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1523                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1524                         },
1525                         $text);
1526
1527                 // Images
1528                 // [img]pathtoimage[/img]
1529                 $text = preg_replace_callback(
1530                         "/\[img\](.*?)\[\/img\]/ism",
1531                         function ($matches) use ($simple_html) {
1532                                 if (strpos($matches[1], "data:image/") === 0) {
1533                                         return $matches[0];
1534                                 }
1535
1536                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1537                                 return "[img]" . $matches[1] . "[/img]";
1538                         },
1539                         $text
1540                 );
1541
1542                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1543                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1544
1545                 $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);
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 />', $text);
1547                 //$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 />', $Text);
1548
1549                 // Try to Oembed
1550                 if ($try_oembed) {
1551                         $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);
1552                         $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);
1553
1554                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1555                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1556                 } else {
1557                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1558                                 '<a href="$1" target="_blank">$1</a>', $text);
1559                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1560                                 '<a href="$1" target="_blank">$1</a>', $text);
1561                 }
1562
1563                 // html5 video and audio
1564
1565
1566                 if ($try_oembed) {
1567                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1568                 } else {
1569                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1570                 }
1571
1572                 // Youtube extensions
1573                 if ($try_oembed) {
1574                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1575                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1576                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1577                 }
1578
1579                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1580                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1581                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1582
1583                 if ($try_oembed) {
1584                         $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);
1585                 } else {
1586                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1587                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1588                 }
1589
1590                 if ($try_oembed) {
1591                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1592                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1593                 }
1594
1595                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1596                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1597
1598                 if ($try_oembed) {
1599                         $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);
1600                 } else {
1601                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1602                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1603                 }
1604
1605                 // oembed tag
1606                 $text = OEmbed::BBCode2HTML($text);
1607
1608                 // Avoid triple linefeeds through oembed
1609                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1610
1611                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1612                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1613                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1614                 // start which is always required). Allow desc with a missing summary for compatibility.
1615
1616                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1617                         $sub = Event::getHTML($ev, $simple_html);
1618
1619                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1620                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1621                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1622                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1623                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1624                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1625                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1626                 }
1627
1628                 // Replace non graphical smilies for external posts
1629                 if (!$nosmile && !$for_plaintext) {
1630                         $text = Smilies::replace($text);
1631                 }
1632
1633                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1634                 if (!$for_plaintext) {
1635                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1636                         if (in_array($simple_html, [7, 9])) {
1637                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1638                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1639                         }
1640                 } else {
1641                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1642                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1643                 }
1644
1645                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1646
1647                 // Remove all hashtag addresses
1648                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7, 9])) {
1649                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1650                 } elseif ($simple_html == 3) {
1651                         // The ! is converted to @ since Diaspora only understands the @
1652                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1653                                 '@<a href="$2">$3</a>',
1654                                 $text);
1655                 } elseif (in_array($simple_html, [7, 9])) {
1656                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1657                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1658                                 $text);
1659                 } elseif (!$simple_html) {
1660                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1661                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1662                                 $text);
1663                 }
1664
1665                 // Bookmarks in red - will be converted to bookmarks in friendica
1666                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1667                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1668                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1669                                         "[bookmark=$1]$2[/bookmark]", $text);
1670
1671                 if (in_array($simple_html, [2, 6, 7, 8])) {
1672                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1673                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1674                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1675                 }
1676
1677                 if ($simple_html == 5) {
1678                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1679                 }
1680
1681                 // Perform URL Search
1682                 if ($try_oembed) {
1683                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1684                 }
1685
1686                 if ($simple_html == 5) {
1687                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1688                 } else {
1689                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1690                 }
1691
1692                 // Handle Diaspora posts
1693                 $text = preg_replace_callback(
1694                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1695                         function ($match) {
1696                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1697                         }, $text
1698                 );
1699
1700                 $text = preg_replace_callback(
1701                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1702                         function ($match) {
1703                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1704                         }, $text
1705                 );
1706
1707                 // Server independent link to posts and comments
1708                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1709                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1710                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1711
1712                 /* Tag conversion
1713                  * Supports:
1714                  * - #[url=<anything>]<term>[/url]
1715                  * - [url=<anything>]#<term>[/url]
1716                  */
1717                 $text = preg_replace_callback("/(?:#\[url\=.*?\]|\[url\=.*?\]#)(.*?)\[\/url\]/ism", function($matches) {
1718                         return '#<a href="'
1719                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1720                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1721                                 . XML::escape($matches[1])
1722                                 . '</a>';
1723                 }, $text);
1724
1725                 // We need no target="_blank" for local links
1726                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1727                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1728                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1729                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1730
1731                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1732                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1733
1734                 // Red compatibility, though the link can't be authenticated on Friendica
1735                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1736
1737
1738                 // we may need to restrict this further if it picks up too many strays
1739                 // link acct:user@host to a webfinger profile redirector
1740
1741                 $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);
1742
1743                 // Perform MAIL Search
1744                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1745                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1746
1747                 // Unhide all [noparse] contained bbtags unspacefying them
1748                 // and triming the [noparse] tag.
1749
1750                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1751                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1752                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1753
1754                 /// @todo What is the meaning of these lines?
1755                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1756                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1757
1758                 // Currently deactivated, it made problems with " inside of alt texts.
1759                 //$text = preg_replace('/\&quot\;/', '"', $text);
1760
1761                 // fix any escaped ampersands that may have been converted into links
1762                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1763
1764                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1765                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1766
1767                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1768
1769                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1770                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1771
1772                 // sanitize href attributes (only whitelisted protocols URLs)
1773                 // default value for backward compatibility
1774                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1775
1776                 // Always allowed protocol even if config isn't set or not including it
1777                 $allowed_link_protocols[] = '//';
1778                 $allowed_link_protocols[] = 'http://';
1779                 $allowed_link_protocols[] = 'https://';
1780                 $allowed_link_protocols[] = 'redir/';
1781
1782                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1783
1784                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1785                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1786
1787                 // Shared content
1788                 $text = self::convertShare(
1789                         $text,
1790                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1791                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1792                         }
1793                 );
1794
1795                 if ($saved_image) {
1796                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1797                 }
1798
1799                 // Restore code blocks
1800                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1801                         function ($matches) use ($codeblocks) {
1802                                 $return = $matches[0];
1803                                 if (isset($codeblocks[intval($matches[1])])) {
1804                                         $return = $codeblocks[$matches[1]];
1805                                 }
1806                                 return $return;
1807                         },
1808                         $text
1809                 );
1810
1811                 // Clean up the HTML by loading and saving the HTML with the DOM.
1812                 // Bad structured html can break a whole page.
1813                 // For performance reasons do it only with activated item cache or at export.
1814                 if (!$try_oembed || (get_itemcachepath() != "")) {
1815                         $doc = new DOMDocument();
1816                         $doc->preserveWhiteSpace = false;
1817
1818                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1819
1820                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1821                         $encoding = '<?xml encoding="UTF-8">';
1822                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1823                         $doc->encoding = 'UTF-8';
1824                         $text = $doc->saveHTML();
1825                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1826
1827                         $text = str_replace('<br></li>', '</li>', $text);
1828
1829                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1830                 }
1831
1832                 // Clean up some useless linebreaks in lists
1833                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1834                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1835                 //$Text = str_replace('</li><br />', '</li>', $Text);
1836                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1837                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1838
1839                 Hook::callAll('bbcode', $text);
1840
1841                 return trim($text);
1842         }
1843
1844         /**
1845          * @brief Strips the "abstract" tag from the provided text
1846          *
1847          * @param string $text The text with BBCode
1848          * @return string The same text - but without "abstract" element
1849          */
1850         public static function stripAbstract($text)
1851         {
1852                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1853                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1854
1855                 return $text;
1856         }
1857
1858         /**
1859          * @brief Returns the value of the "abstract" element
1860          *
1861          * @param string $text The text that maybe contains the element
1862          * @param string $addon The addon for which the abstract is meant for
1863          * @return string The abstract
1864          */
1865         public static function getAbstract($text, $addon = "")
1866         {
1867                 $abstract = "";
1868                 $abstracts = [];
1869                 $addon = strtolower($addon);
1870
1871                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1872                         foreach ($results AS $result) {
1873                                 $abstracts[strtolower($result[1])] = $result[2];
1874                         }
1875                 }
1876
1877                 if (isset($abstracts[$addon])) {
1878                         $abstract = $abstracts[$addon];
1879                 }
1880
1881                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1882                         $abstract = $result[1];
1883                 }
1884
1885                 return $abstract;
1886         }
1887
1888         /**
1889          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1890          *
1891          * @param array $match Matching values for the callback
1892          *                     [1] = Mention type (! or @)
1893          *                     [2] = Name
1894          *                     [3] = Address
1895          * @return string Replaced mention
1896          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1897          * @throws \ImagickException
1898          */
1899         private static function bbCodeMention2DiasporaCallback($match)
1900         {
1901                 $contact = Contact::getDetailsByURL($match[3]);
1902
1903                 if (empty($contact['addr'])) {
1904                         $contact = Probe::uri($match[3]);
1905                 }
1906
1907                 if (empty($contact['addr'])) {
1908                         return $match[0];
1909                 }
1910
1911                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1912                 return $mention;
1913         }
1914
1915         /**
1916          * @brief Converts a BBCode text into Markdown
1917          *
1918          * This function converts a BBCode item body to be sent to Markdown-enabled
1919          * systems like Diaspora and Libertree
1920          *
1921          * @param string $text
1922          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1923          * @return string
1924          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1925          */
1926         public static function toMarkdown($text, $for_diaspora = true)
1927         {
1928                 $a = self::getApp();
1929
1930                 $original_text = $text;
1931
1932                 // Since Diaspora is creating a summary for links, this function removes them before posting
1933                 if ($for_diaspora) {
1934                         $text = self::removeShareInformation($text);
1935                 }
1936
1937                 /**
1938                  * Transform #tags, strip off the [url] and replace spaces with underscore
1939                  */
1940                 $url_search_string = "^\[\]";
1941                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1942                         function ($matches) {
1943                                 return '#' . str_replace(' ', '_', $matches[2]);
1944                         },
1945                         $text
1946                 );
1947
1948                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1949                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1950
1951                 // Convert it to HTML - don't try oembed
1952                 if ($for_diaspora) {
1953                         $text = self::convert($text, false, 3);
1954
1955                         // Add all tags that maybe were removed
1956                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1957                                 $tagline = "";
1958                                 foreach ($tags[2] as $tag) {
1959                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1960                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1961                                                 $tagline .= '#' . $tag . ' ';
1962                                         }
1963                                 }
1964                                 $text = $text . " " . $tagline;
1965                         }
1966                 } else {
1967                         $text = self::convert($text, false, 4);
1968                 }
1969
1970                 // mask some special HTML chars from conversation to markdown
1971                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1972
1973                 // If a link is followed by a quote then there should be a newline before it
1974                 // Maybe we should make this newline at every time before a quote.
1975                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1976
1977                 $stamp1 = microtime(true);
1978
1979                 // Now convert HTML to Markdown
1980                 $text = HTML::toMarkdown($text);
1981
1982                 // unmask the special chars back to HTML
1983                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1984
1985                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
1986
1987                 // Libertree has a problem with escaped hashtags.
1988                 $text = str_replace(['\#'], ['#'], $text);
1989
1990                 // Remove any leading or trailing whitespace, as this will mess up
1991                 // the Diaspora signature verification and cause the item to disappear
1992                 $text = trim($text);
1993
1994                 if ($for_diaspora) {
1995                         $url_search_string = "^\[\]";
1996                         $text = preg_replace_callback(
1997                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1998                                 ['self', 'bbCodeMention2DiasporaCallback'],
1999                                 $text
2000                         );
2001                 }
2002
2003                 Hook::callAll('bb2diaspora', $text);
2004
2005                 return $text;
2006         }
2007
2008         /**
2009      * @brief Pull out all #hashtags and @person tags from $string.
2010      *
2011      * We also get @person@domain.com - which would make
2012      * the regex quite complicated as tags can also
2013      * end a sentence. So we'll run through our results
2014      * and strip the period from any tags which end with one.
2015      * Returns array of tags found, or empty array.
2016      *
2017      * @param string $string Post content
2018      *
2019      * @return array List of tag and person names
2020      */
2021     public static function getTags($string)
2022     {
2023         $ret = [];
2024
2025         // Convert hashtag links to hashtags
2026         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2027
2028         // ignore anything in a code block
2029         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2030
2031         // Force line feeds at bbtags
2032         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2033
2034         // ignore anything in a bbtag
2035         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2036
2037         // Match full names against @tags including the space between first and last
2038         // We will look these up afterward to see if they are full names or not recognisable.
2039
2040         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2041             foreach ($matches[1] as $match) {
2042                 if (strstr($match, ']')) {
2043                     // we might be inside a bbcode color tag - leave it alone
2044                     continue;
2045                 }
2046
2047                 if (substr($match, -1, 1) === '.') {
2048                     $ret[] = substr($match, 0, -1);
2049                 } else {
2050                     $ret[] = $match;
2051                 }
2052             }
2053         }
2054
2055         // Otherwise pull out single word tags. These can be @nickname, @first_last
2056         // and #hash tags.
2057
2058         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2059             foreach ($matches[1] as $match) {
2060                 if (strstr($match, ']')) {
2061                     // we might be inside a bbcode color tag - leave it alone
2062                     continue;
2063                 }
2064                 if (substr($match, -1, 1) === '.') {
2065                     $match = substr($match,0,-1);
2066                 }
2067                 // ignore strictly numeric tags like #1
2068                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2069                     continue;
2070                 }
2071                 // try not to catch url fragments
2072                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2073                     continue;
2074                 }
2075                 $ret[] = $match;
2076             }
2077         }
2078
2079         return $ret;
2080     }
2081 }