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\Addon;
  15 use Friendica\Core\Cache;
  16 use Friendica\Core\Config;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Protocol;
  19 use Friendica\Core\System;
  20 use Friendica\Model\Contact;
  21 use Friendica\Model\Event;
  22 use Friendica\Network\Probe;
  23 use Friendica\Object\Image;
  24 use Friendica\Util\Map;
  25 use Friendica\Util\Network;
  26 use Friendica\Util\ParseUrl;
  27 use Friendica\Util\Proxy as ProxyUtils;
  28 use League\HTMLToMarkdown\HtmlConverter;
  29
  30 class BBCode extends BaseObject
  31 {
  32         /**
  33          * @brief Fetches attachment data that were generated the old way
  34          *
  35          * @param string $body Message body
  36          * @return array
  37          * 'type' -> Message type ("link", "video", "photo")
  38          * 'text' -> Text before the shared message
  39          * 'after' -> Text after the shared message
  40          * 'image' -> Preview image of the message
  41          * 'url' -> Url to the attached message
  42          * 'title' -> Title of the attachment
  43          * 'description' -> Description of the attachment
  44          */
  45         private static function getOldAttachmentData($body)
  46         {
  47                 $post = [];
  48
  49                 // Simplify image codes
  50                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  51
  52                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  53                         foreach ($attached as $data) {
  54                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  55                                         continue;
  56                                 }
  57
  58                                 $post["type"] = substr($data[1], 5);
  59
  60                                 $pos = strpos($body, $data[0]);
  61                                 if ($pos > 0) {
  62                                         $post["text"] = trim(substr($body, 0, $pos));
  63                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  64                                 } else {
  65                                         $post["text"] = trim(str_replace($data[0], "", $body));
  66                                 }
  67
  68                                 $attacheddata = $data[2];
  69
  70                                 $URLSearchString = "^\[\]";
  71
  72                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  73
  74                                         $picturedata = Image::getInfoFromURL($matches[1]);
  75
  76                                         if ($picturedata) {
  77                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  78                                                         $post["image"] = $matches[1];
  79                                                 } else {
  80                                                         $post["preview"] = $matches[1];
  81                                                 }
  82                                         }
  83                                 }
  84
  85                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  86                                         $post["url"] = $matches[1];
  87                                         $post["title"] = $matches[2];
  88                                 }
  89                                 if (!empty($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  90                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  91                                         $post["url"] = $matches[1];
  92                                 }
  93
  94                                 // Search for description
  95                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  96                                         $post["description"] = $matches[1];
  97                                 }
  98                         }
  99                 }
 100                 return $post;
 101         }
 102
 103         /**
 104          * @brief Fetches attachment data that were generated with the "attachment" element
 105          *
 106          * @param string $body Message body
 107          * @return array
 108          * 'type' -> Message type ("link", "video", "photo")
 109          * 'text' -> Text before the shared message
 110          * 'after' -> Text after the shared message
 111          * 'image' -> Preview image of the message
 112          * 'url' -> Url to the attached message
 113          * 'title' -> Title of the attachment
 114          * 'description' -> Description of the attachment
 115          */
 116         public static function getAttachmentData($body)
 117         {
 118                 $data = [];
 119
 120                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 121                         return self::getOldAttachmentData($body);
 122                 }
 123
 124                 $attributes = $match[2];
 125
 126                 $data["text"] = trim($match[1]);
 127
 128                 $type = "";
 129                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 130                 if (x($matches, 1)) {
 131                         $type = strtolower($matches[1]);
 132                 }
 133
 134                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 135                 if (x($matches, 1)) {
 136                         $type = strtolower($matches[1]);
 137                 }
 138
 139                 if ($type == "") {
 140                         return [];
 141                 }
 142
 143                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 144                         return [];
 145                 }
 146
 147                 if ($type != "") {
 148                         $data["type"] = $type;
 149                 }
 150
 151                 $url = "";
 152                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 153                 if (x($matches, 1)) {
 154                         $url = $matches[1];
 155                 }
 156
 157                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 158                 if (x($matches, 1)) {
 159                         $url = $matches[1];
 160                 }
 161
 162                 if ($url != "") {
 163                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 164                 }
 165
 166                 $title = "";
 167                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 168                 if (x($matches, 1)) {
 169                         $title = $matches[1];
 170                 }
 171
 172                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 173                 if (x($matches, 1)) {
 174                         $title = $matches[1];
 175                 }
 176
 177                 if ($title != "") {
 178                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 179                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 180                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 181                         $data["title"] = $title;
 182                 }
 183
 184                 $image = "";
 185                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 186                 if (x($matches, 1)) {
 187                         $image = $matches[1];
 188                 }
 189
 190                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 191                 if (x($matches, 1)) {
 192                         $image = $matches[1];
 193                 }
 194
 195                 if ($image != "") {
 196                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 197                 }
 198
 199                 $preview = "";
 200                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 201                 if (x($matches, 1)) {
 202                         $preview = $matches[1];
 203                 }
 204
 205                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 206                 if (x($matches, 1)) {
 207                         $preview = $matches[1];
 208                 }
 209
 210                 if ($preview != "") {
 211                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 212                 }
 213
 214                 $data["description"] = trim($match[3]);
 215
 216                 $data["after"] = trim($match[4]);
 217
 218                 return $data;
 219         }
 220
 221         public static function getAttachedData($body, $item = [])
 222         {
 223                 /*
 224                 - text:
 225                 - type: link, video, photo
 226                 - title:
 227                 - url:
 228                 - image:
 229                 - description:
 230                 - (thumbnail)
 231                 */
 232
 233                 $has_title = !empty($item['title']);
 234                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 235                 $post = self::getAttachmentData($body);
 236
 237                 // if nothing is found, it maybe having an image.
 238                 if (!isset($post["type"])) {
 239                         // Simplify image codes
 240                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 241
 242                         $URLSearchString = "^\[\]";
 243
 244                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 245
 246                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 247                                 if ((count($pictures) == 1) && !$has_title) {
 248                                         // Checking, if the link goes to a picture
 249                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 250
 251                                         // Workaround:
 252                                         // Sometimes photo posts to the own album are not detected at the start.
 253                                         // So we seem to cannot use the cache for these cases. That's strange.
 254                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 255                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 256                                         }
 257
 258                                         if ($data["type"] == "photo") {
 259                                                 $post["type"] = "photo";
 260                                                 if (isset($data["images"][0])) {
 261                                                         $post["image"] = $data["images"][0]["src"];
 262                                                         $post["url"] = $data["url"];
 263                                                 } else {
 264                                                         $post["image"] = $data["url"];
 265                                                 }
 266
 267                                                 $post["preview"] = $pictures[0][2];
 268                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 269                                         } else {
 270                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 271                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 272                                                         $post["type"] = "photo";
 273                                                         $post["image"] = $pictures[0][1];
 274                                                         $post["preview"] = $pictures[0][2];
 275                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 276                                                 }
 277                                         }
 278                                 } elseif (count($pictures) > 0) {
 279                                         $post["type"] = "link";
 280                                         $post["url"] = $plink;
 281                                         $post["image"] = $pictures[0][2];
 282                                         $post["text"] = $body;
 283                                 }
 284                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 285                                 if ((count($pictures) == 1) && !$has_title) {
 286                                         $post["type"] = "photo";
 287                                         $post["image"] = $pictures[0][1];
 288                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 289                                 } elseif (count($pictures) > 0) {
 290                                         $post["type"] = "link";
 291                                         $post["url"] = $plink;
 292                                         $post["image"] = $pictures[0][1];
 293                                         $post["text"] = $body;
 294                                 }
 295                         }
 296
 297                         // Test for the external links
 298                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 299                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 300
 301                         $links = array_merge($links1, $links2);
 302
 303                         // If there is only a single one, then use it.
 304                         // This should cover link posts via API.
 305                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 306                                 $post["type"] = "link";
 307                                 $post["text"] = trim($body);
 308                                 $post["url"] = $links[0][1];
 309                         }
 310
 311                         // Now count the number of external media links
 312                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 313                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 314                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 315                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 316
 317                         // Add them to the other external links
 318                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 319
 320                         // Are there more than one?
 321                         if (count($links) > 1) {
 322                                 // The post will be the type "text", which means a blog post
 323                                 unset($post["type"]);
 324                                 $post["url"] = $plink;
 325                         }
 326
 327                         if (!isset($post["type"])) {
 328                                 $post["type"] = "text";
 329                                 $post["text"] = trim($body);
 330                         }
 331                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 332                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 333
 334                         if (isset($data["images"][0])) {
 335                                 $post["image"] = $data["images"][0]["src"];
 336                         }
 337                 }
 338
 339                 return $post;
 340         }
 341
 342         /**
 343          * @brief Converts a BBCode text into plaintext
 344          *
 345          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 346          *
 347          * @return string
 348          */
 349         public static function toPlaintext($text, $keep_urls = true)
 350         {
 351                 $naked_text = preg_replace('/\[(.+?)\]/','', $text);
 352                 if (!$keep_urls) {
 353                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 354                 }
 355
 356                 return $naked_text;
 357         }
 358
 359         private static function proxyUrl($image, $simplehtml = false)
 360         {
 361                 // Only send proxied pictures to API and for internal display
 362                 if (in_array($simplehtml, [false, 2])) {
 363                         return ProxyUtils::proxifyUrl($image);
 364                 } else {
 365                         return $image;
 366                 }
 367         }
 368
 369         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 370         {
 371                 // Suppress "view full size"
 372                 if (intval(Config::get('system', 'no_view_full_size'))) {
 373                         $include_link = false;
 374                 }
 375
 376                 // Picture addresses can contain special characters
 377                 $s = htmlspecialchars_decode($srctext);
 378
 379                 $matches = null;
 380                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 381                 if ($c) {
 382                         foreach ($matches as $mtch) {
 383                                 logger('scale_external_image: ' . $mtch[1]);
 384
 385                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 386                                 if (stristr($mtch[1], $hostname)) {
 387                                         continue;
 388                                 }
 389
 390                                 // $scale_replace, if passed, is an array of two elements. The
 391                                 // first is the name of the full-size image. The second is the
 392                                 // name of a remote, scaled-down version of the full size image.
 393                                 // This allows Friendica to display the smaller remote image if
 394                                 // one exists, while still linking to the full-size image
 395                                 if ($scale_replace) {
 396                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 397                                 } else {
 398                                         $scaled = $mtch[1];
 399                                 }
 400                                 $i = Network::fetchUrl($scaled);
 401                                 if (!$i) {
 402                                         return $srctext;
 403                                 }
 404
 405                                 // guess mimetype from headers or filename
 406                                 $type = Image::guessType($mtch[1], true);
 407
 408                                 if ($i) {
 409                                         $Image = new Image($i, $type);
 410                                         if ($Image->isValid()) {
 411                                                 $orig_width = $Image->getWidth();
 412                                                 $orig_height = $Image->getHeight();
 413
 414                                                 if ($orig_width > 640 || $orig_height > 640) {
 415                                                         $Image->scaleDown(640);
 416                                                         $new_width = $Image->getWidth();
 417                                                         $new_height = $Image->getHeight();
 418                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 419                                                         $s = str_replace(
 420                                                                 $mtch[0],
 421                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 422                                                                 . "\n" . (($include_link)
 423                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 424                                                                         : ''),
 425                                                                 $s
 426                                                         );
 427                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 428                                                 }
 429                                         }
 430                                 }
 431                         }
 432                 }
 433
 434                 // replace the special char encoding
 435                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 436                 return $s;
 437         }
 438
 439         /**
 440          * The purpose of this function is to apply system message length limits to
 441          * imported messages without including any embedded photos in the length
 442          *
 443          * @brief Truncates imported message body string length to max_import_size
 444          * @param string $body
 445          * @return string
 446          */
 447         public static function limitBodySize($body)
 448         {
 449                 $maxlen = Config::get('config', 'max_import_size', 0);
 450
 451                 // If the length of the body, including the embedded images, is smaller
 452                 // than the maximum, then don't waste time looking for the images
 453                 if ($maxlen && (strlen($body) > $maxlen)) {
 454
 455                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 456
 457                         $orig_body = $body;
 458                         $new_body = '';
 459                         $textlen = 0;
 460
 461                         $img_start = strpos($orig_body, '[img');
 462                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 463                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 464                         while (($img_st_close !== false) && ($img_end !== false)) {
 465
 466                                 $img_st_close++; // make it point to AFTER the closing bracket
 467                                 $img_end += $img_start;
 468                                 $img_end += strlen('[/img]');
 469
 470                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 471                                         // This is an embedded image
 472
 473                                         if (($textlen + $img_start) > $maxlen) {
 474                                                 if ($textlen < $maxlen) {
 475                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 476                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 477                                                         $textlen = $maxlen;
 478                                                 }
 479                                         } else {
 480                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 481                                                 $textlen += $img_start;
 482                                         }
 483
 484                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 485                                 } else {
 486
 487                                         if (($textlen + $img_end) > $maxlen) {
 488                                                 if ($textlen < $maxlen) {
 489                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 490                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 491                                                         $textlen = $maxlen;
 492                                                 }
 493                                         } else {
 494                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 495                                                 $textlen += $img_end;
 496                                         }
 497                                 }
 498                                 $orig_body = substr($orig_body, $img_end);
 499
 500                                 if ($orig_body === false) {
 501                                         // in case the body ends on a closing image tag
 502                                         $orig_body = '';
 503                                 }
 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, $img_start), '[/img]') : false);
 508                         }
 509
 510                         if (($textlen + strlen($orig_body)) > $maxlen) {
 511                                 if ($textlen < $maxlen) {
 512                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 513                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 514                                 }
 515                         } else {
 516                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 517                                 $new_body = $new_body . $orig_body;
 518                         }
 519
 520                         return $new_body;
 521                 } else {
 522                         return $body;
 523                 }
 524         }
 525
 526         /**
 527          * Processes [attachment] tags
 528          *
 529          * Note: Can produce a [bookmark] tag in the returned string
 530          *
 531          * @brief Processes [attachment] tags
 532          * @param string $return
 533          * @param bool|int $simplehtml
 534          * @param bool $tryoembed
 535          * @return string
 536          */
 537         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 538         {
 539                 $data = self::getAttachmentData($return);
 540                 if (empty($data) || empty($data["url"])) {
 541                         return $return;
 542                 }
 543
 544                 if (isset($data["title"])) {
 545                         $data["title"] = strip_tags($data["title"]);
 546                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 547                 } else {
 548                         $data["title"] = null;
 549                 }
 550
 551                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 552                         $data["preview"] = $data["image"];
 553                         $data["image"] = "";
 554                 }
 555
 556                 $return = '';
 557                 if ($simplehtml == 7) {
 558                         $return = self::convertUrlForOStatus($data["url"]);
 559                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 560                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 561                 } else {
 562                         try {
 563                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 564                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 565                                 } else {
 566                                         throw new Exception('OEmbed is disabled for this attachment.');
 567                                 }
 568                         } catch (Exception $e) {
 569                                 $data["title"] = defaults($data, 'title', $data['url']);
 570
 571                                 if ($simplehtml != 4) {
 572                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 573                                 }
 574
 575                                 if (!empty($data["image"])) {
 576                                         $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), $data["title"]);
 577                                 } elseif (!empty($data["preview"])) {
 578                                         $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), $data["title"]);
 579                                 }
 580
 581                                 if (($data["type"] == "photo") && !empty($data["url"]) && !empty($data["image"])) {
 582                                         $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"]);
 583                                 } else {
 584                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 585                                 }
 586
 587                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 588                                         // Sanitize the HTML by converting it to BBCode
 589                                         $bbcode = HTML::toBBCode($data["description"]);
 590                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 591                                 }
 592                                 if ($data["type"] == "link") {
 593                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 594                                 }
 595
 596                                 if ($simplehtml != 4) {
 597                                         $return .= '</div>';
 598                                 }
 599                         }
 600                 }
 601
 602                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 603         }
 604
 605         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 606         {
 607                 $data = self::getAttachmentData($Text);
 608
 609                 if (!$data) {
 610                         return $Text;
 611                 } elseif ($nolink) {
 612                         return $data["text"] . defaults($data, 'after', '');
 613                 }
 614
 615                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 616                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 617                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 618                         $data["title"] = $data["url"];
 619                 } elseif (($text != "") && strstr($title, $text)) {
 620                         $data["text"] = $data["title"];
 621                         $data["title"] = $data["url"];
 622                 }
 623
 624                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 625                         return $data["title"] . $data["after"];
 626                 }
 627
 628                 // If the link already is included in the post, don't add it again
 629                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 630                         return $data["text"] . $data["after"];
 631                 }
 632
 633                 $text = $data["text"];
 634
 635                 if (!empty($data["url"]) && !empty($data["title"])) {
 636                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 637                 } elseif (!empty($data["url"])) {
 638                         $text .= "\n[url]" . $data["url"] . "[/url]";
 639                 }
 640
 641                 return $text . "\n" . $data["after"];
 642         }
 643
 644         /**
 645          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 646          *
 647          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 648          * @param array $match Array with the matching values
 649          * @return string reformatted link including HTML codes
 650          */
 651         private static function convertUrlForOStatusCallback($match)
 652         {
 653                 $url = $match[1];
 654
 655                 if (isset($match[2]) && ($match[1] != $match[2])) {
 656                         return $match[0];
 657                 }
 658
 659                 $parts = parse_url($url);
 660                 if (!isset($parts['scheme'])) {
 661                         return $match[0];
 662                 }
 663
 664                 return self::convertUrlForOStatus($url);
 665         }
 666
 667         /**
 668          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 669          * @param string $url URL that is about to be reformatted
 670          * @return string reformatted link including HTML codes
 671          */
 672         private static function convertUrlForOStatus($url)
 673         {
 674                 $parts = parse_url($url);
 675                 $scheme = $parts['scheme'] . '://';
 676                 $styled_url = str_replace($scheme, '', $url);
 677
 678                 if (strlen($styled_url) > 30) {
 679                         $styled_url = substr($styled_url, 0, 30) . "…";
 680                 }
 681
 682                 $html = '<a href="%s" target="_blank">%s</a>';
 683
 684                 return sprintf($html, $url, $styled_url);
 685         }
 686
 687         /*
 688          * [noparse][i]italic[/i][/noparse] turns into
 689          * [noparse][ i ]italic[ /i ][/noparse],
 690          * to hide them from parser.
 691          */
 692         private static function escapeNoparseCallback($match)
 693         {
 694                 $whole_match = $match[0];
 695                 $captured = $match[1];
 696                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 697                 $new_str = str_replace($captured, $spacefied, $whole_match);
 698                 return $new_str;
 699         }
 700
 701         /*
 702          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 703          * now turns back and the [noparse] tags are trimed
 704          * returning [i]italic[/i]
 705          */
 706         private static function unescapeNoparseCallback($match)
 707         {
 708                 $captured = $match[1];
 709                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 710                 return $unspacefied;
 711         }
 712
 713         /**
 714          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 715          * occurrences
 716          *
 717          * @param string $text        Text to search
 718          * @param string $name        Tag name
 719          * @param int    $occurrences Number of first occurrences to skip
 720          * @return boolean|array
 721          */
 722         public static function getTagPosition($text, $name, $occurrences = 0)
 723         {
 724                 if ($occurrences < 0) {
 725                         $occurrences = 0;
 726                 }
 727
 728                 $start_open = -1;
 729                 for ($i = 0; $i <= $occurrences; $i++) {
 730                         if ($start_open !== false) {
 731                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 732                         }
 733                 }
 734
 735                 if ($start_open === false) {
 736                         return false;
 737                 }
 738
 739                 $start_equal = strpos($text, '=', $start_open);
 740                 $start_close = strpos($text, ']', $start_open);
 741
 742                 if ($start_close === false) {
 743                         return false;
 744                 }
 745
 746                 $start_close++;
 747
 748                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 749
 750                 if ($end_open === false) {
 751                         return false;
 752                 }
 753
 754                 $res = [
 755                         'start' => [
 756                                 'open' => $start_open,
 757                                 'close' => $start_close
 758                         ],
 759                         'end' => [
 760                                 'open' => $end_open,
 761                                 'close' => $end_open + strlen('[/' . $name . ']')
 762                         ],
 763                 ];
 764
 765                 if ($start_equal !== false) {
 766                         $res['start']['equal'] = $start_equal + 1;
 767                 }
 768
 769                 return $res;
 770         }
 771
 772         /**
 773          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 774          *
 775          * @param type $pattern Preg pattern string
 776          * @param type $replace Preg replace string
 777          * @param type $name    BBCode tag name
 778          * @param type $text    Text to search
 779          * @return string
 780          */
 781         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 782         {
 783                 $occurrences = 0;
 784                 $pos = self::getTagPosition($text, $name, $occurrences);
 785                 while ($pos !== false && $occurrences++ < 1000) {
 786                         $start = substr($text, 0, $pos['start']['open']);
 787                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 788                         $end = substr($text, $pos['end']['close']);
 789                         if ($end === false) {
 790                                 $end = '';
 791                         }
 792
 793                         $subject = preg_replace($pattern, $replace, $subject);
 794                         $text = $start . $subject . $end;
 795
 796                         $pos = self::getTagPosition($text, $name, $occurrences);
 797                 }
 798
 799                 return $text;
 800         }
 801
 802         private static function extractImagesFromItemBody($body)
 803         {
 804                 $saved_image = [];
 805                 $orig_body = $body;
 806                 $new_body = '';
 807
 808                 $cnt = 0;
 809                 $img_start = strpos($orig_body, '[img');
 810                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 811                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 812                 while (($img_st_close !== false) && ($img_end !== false)) {
 813                         $img_st_close++; // make it point to AFTER the closing bracket
 814                         $img_end += $img_start;
 815
 816                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 817                                 // This is an embedded image
 818                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 819                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 820
 821                                 $cnt++;
 822                         } else {
 823                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 824                         }
 825
 826                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 827
 828                         if ($orig_body === false) {
 829                                 // in case the body ends on a closing image tag
 830                                 $orig_body = '';
 831                         }
 832
 833                         $img_start = strpos($orig_body, '[img');
 834                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 835                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 836                 }
 837
 838                 $new_body = $new_body . $orig_body;
 839
 840                 return ['body' => $new_body, 'images' => $saved_image];
 841         }
 842
 843         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 844         {
 845                 $newbody = $body;
 846
 847                 $cnt = 0;
 848                 foreach ($images as $image) {
 849                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 850                         // it loops over the array starting from the first element and going sequentially
 851                         // to the last element
 852                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 853                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 854                         $cnt++;
 855                 }
 856
 857                 return $newbody;
 858         }
 859
 860         /**
 861          * Processes [share] tags
 862          *
 863          * Note: Can produce a [bookmark] tag in the output
 864          *
 865          * @brief Processes [share] tags
 866          * @param array    $share      preg_match_callback result array
 867          * @param bool|int $simplehtml
 868          * @return string
 869          */
 870         private static function convertShare($share, $simplehtml)
 871         {
 872                 $attributes = $share[2];
 873
 874                 $author = "";
 875                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
 876                 if (x($matches, 1)) {
 877                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
 878                 }
 879
 880                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
 881                 if (x($matches, 1)) {
 882                         $author = $matches[1];
 883                 }
 884
 885                 $profile = "";
 886                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
 887                 if (x($matches, 1)) {
 888                         $profile = $matches[1];
 889                 }
 890
 891                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
 892                 if (x($matches, 1)) {
 893                         $profile = $matches[1];
 894                 }
 895
 896                 $avatar = "";
 897                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
 898                 if (x($matches, 1)) {
 899                         $avatar = $matches[1];
 900                 }
 901
 902                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
 903                 if (x($matches, 1)) {
 904                         $avatar = $matches[1];
 905                 }
 906
 907                 $link = "";
 908                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
 909                 if (x($matches, 1)) {
 910                         $link = $matches[1];
 911                 }
 912
 913                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
 914                 if (x($matches, 1)) {
 915                         $link = $matches[1];
 916                 }
 917
 918                 $posted = "";
 919
 920                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
 921                 if (x($matches, 1)) {
 922                         $posted = $matches[1];
 923                 }
 924
 925                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
 926                 if (x($matches, 1)) {
 927                         $posted = $matches[1];
 928                 }
 929
 930                 // We only call this so that a previously unknown contact can be added.
 931                 // This is important for the function "Model\Contact::getDetailsByURL()".
 932                 // This function then can fetch an entry from the contact table.
 933                 Contact::getIdForURL($profile, 0, true);
 934
 935                 $data = Contact::getDetailsByURL($profile);
 936
 937                 if (x($data, "name") && x($data, "addr")) {
 938                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
 939                 } else {
 940                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
 941                 }
 942
 943                 if (x($data, "addr")) {
 944                         $userid = $data["addr"];
 945                 } else {
 946                         $userid = Protocol::formatMention($profile, $author);
 947                 }
 948
 949                 if (x($data, "name")) {
 950                         $author = $data["name"];
 951                 }
 952
 953                 if (x($data, "micro")) {
 954                         $avatar = $data["micro"];
 955                 }
 956
 957                 $preshare = trim($share[1]);
 958                 if ($preshare != "") {
 959                         $preshare .= "<br />";
 960                 }
 961
 962                 switch ($simplehtml) {
 963                         case 1:
 964                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
 965                                 break;
 966                         case 2:
 967                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
 968                                 break;
 969                         case 3: // Diaspora
 970                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
 971
 972                                 $text = trim($share[1]);
 973
 974                                 if ($text != "") {
 975                                         $text .= "<hr />";
 976                                 }
 977
 978                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
 979                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
 980                                 } else {
 981                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
 982
 983                                         if ($link != "") {
 984                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
 985                                         }
 986                                 }
 987
 988                                 break;
 989                         case 4:
 990                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 991                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
 992                                 $headline .= ":</b><br />";
 993
 994                                 $text = trim($share[1]);
 995
 996                                 if ($text != "") {
 997                                         $text .= "<hr />";
 998                                 }
 999
1000                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1001
1002                                 break;
1003                         case 5:
1004                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1005                                 break;
1006                         case 6: // app.net
1007                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1008                                 break;
1009                         case 7: // statusnet/GNU Social
1010                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1011                                 break;
1012                         case 8: // twitter
1013                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1014                                 break;
1015                         case 9: // Google+/Facebook
1016                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1017
1018                                 if ($link != "") {
1019                                         $text .= "<br /><br />" . $link;
1020                                 }
1021                                 break;
1022                         default:
1023                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1024                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1025                                         try {
1026                                                 $oembed = OEmbed::getHTML($link, $preshare);
1027                                         } catch (Exception $e) {
1028                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1029                                         }
1030
1031                                         $text = $preshare . $oembed;
1032                                 } else {
1033                                         $text = trim($share[1]) . "\n";
1034
1035                                         $avatar = ProxyUtils::proxifyUrl($avatar, false, ProxyUtils::SIZE_THUMB);
1036
1037                                         $tpl = get_markup_template('shared_content.tpl');
1038                                         $text .= replace_macros($tpl, [
1039                                                 '$profile' => $profile,
1040                                                 '$avatar' => $avatar,
1041                                                 '$author' => $author,
1042                                                 '$link' => $link,
1043                                                 '$posted' => $posted,
1044                                                 '$content' => trim($share[3])
1045                                         ]);
1046                                 }
1047                                 break;
1048                 }
1049
1050                 return $text;
1051         }
1052
1053         private static function removePictureLinksCallback($match)
1054         {
1055                 $text = Cache::get($match[1]);
1056
1057                 if (is_null($text)) {
1058                         $a = self::getApp();
1059
1060                         $stamp1 = microtime(true);
1061
1062                         $ch = @curl_init($match[1]);
1063                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1064                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1065                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1066                         @curl_exec($ch);
1067                         $curl_info = @curl_getinfo($ch);
1068
1069                         $a->save_timestamp($stamp1, "network");
1070
1071                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1072                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1073                         } else {
1074                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1075
1076                                 // if its not a picture then look if its a page that contains a picture link
1077                                 $body = Network::fetchUrl($match[1]);
1078
1079                                 $doc = new DOMDocument();
1080                                 @$doc->loadHTML($body);
1081                                 $xpath = new DOMXPath($doc);
1082                                 $list = $xpath->query("//meta[@name]");
1083                                 foreach ($list as $node) {
1084                                         $attr = [];
1085
1086                                         if ($node->attributes->length) {
1087                                                 foreach ($node->attributes as $attribute) {
1088                                                         $attr[$attribute->name] = $attribute->value;
1089                                                 }
1090                                         }
1091
1092                                         if (strtolower($attr["name"]) == "twitter:image") {
1093                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1094                                         }
1095                                 }
1096                         }
1097                         Cache::set($match[1], $text);
1098                 }
1099
1100                 return $text;
1101         }
1102
1103         private static function expandLinksCallback($match)
1104         {
1105                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1106                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1107                 } else {
1108                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1109                 }
1110         }
1111
1112         private static function cleanPictureLinksCallback($match)
1113         {
1114                 $text = Cache::get($match[1]);
1115
1116                 if (is_null($text)) {
1117                         $a = self::getApp();
1118
1119                         $stamp1 = microtime(true);
1120
1121                         $ch = @curl_init($match[1]);
1122                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1123                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1124                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1125                         @curl_exec($ch);
1126                         $curl_info = @curl_getinfo($ch);
1127
1128                         $a->save_timestamp($stamp1, "network");
1129
1130                         // if its a link to a picture then embed this picture
1131                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1132                                 $text = "[img]" . $match[1] . "[/img]";
1133                         } else {
1134                                 $text = "[img]" . $match[2] . "[/img]";
1135
1136                                 // if its not a picture then look if its a page that contains a picture link
1137                                 $body = Network::fetchUrl($match[1]);
1138
1139                                 $doc = new DOMDocument();
1140                                 @$doc->loadHTML($body);
1141                                 $xpath = new DOMXPath($doc);
1142                                 $list = $xpath->query("//meta[@name]");
1143                                 foreach ($list as $node) {
1144                                         $attr = [];
1145                                         if ($node->attributes->length) {
1146                                                 foreach ($node->attributes as $attribute) {
1147                                                         $attr[$attribute->name] = $attribute->value;
1148                                                 }
1149                                         }
1150
1151                                         if (strtolower($attr["name"]) == "twitter:image") {
1152                                                 $text = "[img]" . $attr["content"] . "[/img]";
1153                                         }
1154                                 }
1155                         }
1156                         Cache::set($match[1], $text);
1157                 }
1158
1159                 return $text;
1160         }
1161
1162         public static function cleanPictureLinks($text)
1163         {
1164                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1165                 return $return;
1166         }
1167
1168         private static function textHighlightCallback($match)
1169         {
1170                 // Fallback in case the language doesn't exist
1171                 $return = '[code]' . $match[2] . '[/code]';
1172
1173                 if (in_array(strtolower($match[1]),
1174                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1175                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1176                 ) {
1177                         $return = text_highlight($match[2], strtolower($match[1]));
1178                 }
1179
1180                 return $return;
1181         }
1182
1183         /**
1184          * @brief Converts a BBCode message to HTML message
1185          *
1186          * BBcode 2 HTML was written by WAY2WEB.net
1187          * extended to work with Mistpark/Friendica - Mike Macgirvin
1188          *
1189          * Simple HTML values meaning:
1190          * - 0: Friendica display
1191          * - 1: Unused
1192          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1193          * - 3: Used before converting to Markdown in bb2diaspora.php
1194          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1195          * - 5: Unused
1196          * - 6: Used for Appnet
1197          * - 7: Used for dfrn, OStatus
1198          * - 8: Used for WP backlink text setting
1199          *
1200          * @param string $text
1201          * @param bool   $try_oembed
1202          * @param int    $simple_html
1203          * @param bool   $for_plaintext
1204          * @return string
1205          */
1206         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1207         {
1208                 $a = self::getApp();
1209
1210                 /*
1211                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1212                  *
1213                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1214                  * $match[1] = $url
1215                  * $match[2] = $title or absent
1216                  */
1217                 $try_oembed_callback = function ($match)
1218                 {
1219                         $url = $match[1];
1220                         $title = defaults($match, 2, null);
1221
1222                         try {
1223                                 $return = OEmbed::getHTML($url, $title);
1224                         } catch (Exception $ex) {
1225                                 $return = $match[0];
1226                         }
1227
1228                         return $return;
1229                 };
1230
1231                 // Hide all [noparse] contained bbtags by spacefying them
1232                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1233
1234                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1235                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1236                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1237
1238                 // Remove the abstract element. It is a non visible element.
1239                 $text = self::stripAbstract($text);
1240
1241                 // Move all spaces out of the tags
1242                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1243                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1244
1245                 // Extract the private images which use data urls since preg has issues with
1246                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1247                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1248
1249                 $extracted = self::extractImagesFromItemBody($text);
1250                 $text = $extracted['body'];
1251                 $saved_image = $extracted['images'];
1252
1253                 // If we find any event code, turn it into an event.
1254                 // After we're finished processing the bbcode we'll
1255                 // replace all of the event code with a reformatted version.
1256
1257                 $ev = Event::fromBBCode($text);
1258
1259                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1260                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1261
1262                 $text = str_replace("<", "&lt;", $text);
1263                 $text = str_replace(">", "&gt;", $text);
1264
1265                 // remove some newlines before the general conversion
1266                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1267                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1268
1269                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1270                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1271
1272                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1273                 if (!$try_oembed) {
1274                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1275                 }
1276
1277                 // Check for [code] text here, before the linefeeds are messed with.
1278                 // The highlighter will unescape and re-escape the content.
1279                 if (strpos($text, '[code=') !== false) {
1280                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1281                 }
1282                 // Convert new line chars to html <br /> tags
1283
1284                 // nlbr seems to be hopelessly messed up
1285                 //      $Text = nl2br($Text);
1286
1287                 // We'll emulate it.
1288
1289                 $text = trim($text);
1290                 $text = str_replace("\r\n", "\n", $text);
1291
1292                 // removing multiplicated newlines
1293                 if (Config::get("system", "remove_multiplicated_lines")) {
1294                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1295                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1296                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1297                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1298                         do {
1299                                 $oldtext = $text;
1300                                 $text = str_replace($search, $replace, $text);
1301                         } while ($oldtext != $text);
1302                 }
1303
1304                 // Set up the parameters for a URL search string
1305                 $URLSearchString = "^\[\]";
1306                 // Set up the parameters for a MAIL search string
1307                 $MAILSearchString = $URLSearchString;
1308
1309                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1310                 if (!$for_plaintext) {
1311                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1312                         // Currently disabled, since the function is too greedy
1313                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1314                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1315                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1316                         if ($simple_html == 7) {
1317                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1318                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1319                         }
1320                 } else {
1321                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1322                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1323                 }
1324
1325
1326                 // Handle attached links or videos
1327                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1328
1329                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1330
1331                 // Remove all hashtag addresses
1332                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1333                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1334                 } elseif ($simple_html == 3) {
1335                         // The ! is converted to @ since Diaspora only understands the @
1336                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1337                                 '@<a href="$2">$3</a>',
1338                                 $text);
1339                 } elseif ($simple_html == 7) {
1340                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1341                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1342                                 $text);
1343                 } elseif (!$simple_html) {
1344                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1345                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1346                                 $text);
1347                 }
1348
1349                 // Bookmarks in red - will be converted to bookmarks in friendica
1350                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1351                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1352                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1353                                         "[bookmark=$1]$2[/bookmark]", $text);
1354
1355                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1356                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1357                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1358                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1359                 }
1360
1361                 if ($simple_html == 5) {
1362                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1363                 }
1364
1365                 // Perform URL Search
1366                 if ($try_oembed) {
1367                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1368                 }
1369
1370                 if ($simple_html == 5) {
1371                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1372                 } else {
1373                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1374                 }
1375
1376                 // Handle Diaspora posts
1377                 $text = preg_replace_callback(
1378                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1379                         function ($match) {
1380                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1381                         }, $text
1382                 );
1383
1384                 $text = preg_replace_callback(
1385                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1386                         function ($match) {
1387                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1388                         }, $text
1389                 );
1390
1391                 // Server independent link to posts and comments
1392                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1393                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1394                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1395
1396                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1397                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1398
1399                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1400                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1401
1402                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1403                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1404                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1405
1406                 // Red compatibility, though the link can't be authenticated on Friendica
1407                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1408
1409
1410                 // we may need to restrict this further if it picks up too many strays
1411                 // link acct:user@host to a webfinger profile redirector
1412
1413                 $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);
1414
1415                 // Perform MAIL Search
1416                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1417                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1418
1419                 // leave open the posibility of [map=something]
1420                 // this is replaced in prepare_body() which has knowledge of the item location
1421
1422                 if (strpos($text, '[/map]') !== false) {
1423                         $text = preg_replace_callback(
1424                                 "/\[map\](.*?)\[\/map\]/ism",
1425                                 function ($match) use ($simple_html) {
1426                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1427                                 },
1428                                 $text
1429                         );
1430                 }
1431                 if (strpos($text, '[map=') !== false) {
1432                         $text = preg_replace_callback(
1433                                 "/\[map=(.*?)\]/ism",
1434                                 function ($match) use ($simple_html) {
1435                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1436                                 },
1437                                 $text
1438                         );
1439                 }
1440                 if (strpos($text, '[map]') !== false) {
1441                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1442                 }
1443
1444                 // Check for headers
1445                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1446                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1447                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1448                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1449                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1450                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1451
1452                 // Check for paragraph
1453                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1454
1455                 // Check for bold text
1456                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1457
1458                 // Check for Italics text
1459                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1460
1461                 // Check for Underline text
1462                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1463
1464                 // Check for strike-through text
1465                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1466
1467                 // Check for over-line text
1468                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1469
1470                 // Check for colored text
1471                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1472
1473                 // Check for sized text
1474                 // [size=50] --> font-size: 50px (with the unit).
1475                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1476                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1477
1478                 // Check for centered text
1479                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1480
1481                 // Check for list text
1482                 $text = str_replace("[*]", "<li>", $text);
1483
1484                 // Check for style sheet commands
1485                 $text = preg_replace_callback(
1486                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1487                         function ($match) {
1488                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1489                         },
1490                         $text
1491                 );
1492
1493                 // Check for CSS classes
1494                 $text = preg_replace_callback(
1495                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1496                         function ($match) {
1497                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1498                         },
1499                         $text
1500                 );
1501
1502                 // handle nested lists
1503                 $endlessloop = 0;
1504
1505                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1506                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1507                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1508                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1509                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1510                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1511                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1512                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1513                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1514                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1515                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1516                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1517                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1518                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1519                 }
1520
1521                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1522                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1523                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1524                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1525
1526                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1527                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1528
1529                 $text = str_replace('[hr]', '<hr />', $text);
1530
1531                 // This is actually executed in prepare_body()
1532
1533                 $text = str_replace('[nosmile]', '', $text);
1534
1535                 // Check for font change text
1536                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1537
1538                 // Declare the format for [code] layout
1539
1540                 $CodeLayout = '<code>$1</code>';
1541                 // Check for [code] text
1542                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1543
1544                 // Declare the format for [spoiler] layout
1545                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1546
1547                 // Check for [spoiler] text
1548                 // handle nested quotes
1549                 $endlessloop = 0;
1550                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1551                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1552                 }
1553
1554                 // Check for [spoiler=Author] text
1555
1556                 $t_wrote = L10n::t('$1 wrote:');
1557
1558                 // handle nested quotes
1559                 $endlessloop = 0;
1560                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1561                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1562                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1563                                                  $text);
1564                 }
1565
1566                 // Declare the format for [quote] layout
1567                 $QuoteLayout = '<blockquote>$1</blockquote>';
1568
1569                 // Check for [quote] text
1570                 // handle nested quotes
1571                 $endlessloop = 0;
1572                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1573                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1574                 }
1575
1576                 // Check for [quote=Author] text
1577
1578                 $t_wrote = L10n::t('$1 wrote:');
1579
1580                 // handle nested quotes
1581                 $endlessloop = 0;
1582                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1583                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1584                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1585                                                  $text);
1586                 }
1587
1588
1589                 // [img=widthxheight]image source[/img]
1590                 $text = preg_replace_callback(
1591                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1592                         function ($matches) use ($simple_html) {
1593                                 if (strpos($matches[3], "data:image/") === 0) {
1594                                         return $matches[0];
1595                                 }
1596
1597                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1598                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1599                         },
1600                         $text
1601                 );
1602
1603                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1604                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1605
1606                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1607                         function ($matches) use ($simple_html) {
1608                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1609                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1610                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1611                         },
1612                         $text);
1613
1614                 // Images
1615                 // [img]pathtoimage[/img]
1616                 $text = preg_replace_callback(
1617                         "/\[img\](.*?)\[\/img\]/ism",
1618                         function ($matches) use ($simple_html) {
1619                                 if (strpos($matches[1], "data:image/") === 0) {
1620                                         return $matches[0];
1621                                 }
1622
1623                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1624                                 return "[img]" . $matches[1] . "[/img]";
1625                         },
1626                         $text
1627                 );
1628
1629                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1630                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1631
1632                 // Shared content
1633                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1634                         function ($match) use ($simple_html) {
1635                                 return self::convertShare($match, $simple_html);
1636                         }, $text);
1637
1638                 $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);
1639                 $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 />', $text);
1640                 //$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 />', $Text);
1641
1642                 // Try to Oembed
1643                 if ($try_oembed) {
1644                         $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);
1645                         $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);
1646
1647                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1648                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1649                 } else {
1650                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1651                                                 '<a href="$1" target="_blank">$1</a>', $text);
1652                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1653                                                 '<a href="$1" target="_blank">$1</a>', $text);
1654                 }
1655
1656                 // html5 video and audio
1657
1658
1659                 if ($try_oembed) {
1660                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1661                 } else {
1662                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1663                 }
1664
1665                 // Youtube extensions
1666                 if ($try_oembed) {
1667                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1668                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1669                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1670                 }
1671
1672                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1673                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1674                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1675
1676                 if ($try_oembed) {
1677                         $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);
1678                 } else {
1679                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1680                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1681                 }
1682
1683                 if ($try_oembed) {
1684                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1685                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1686                 }
1687
1688                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1689                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1690
1691                 if ($try_oembed) {
1692                         $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);
1693                 } else {
1694                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1695                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1696                 }
1697
1698                 // oembed tag
1699                 $text = OEmbed::BBCode2HTML($text);
1700
1701                 // Avoid triple linefeeds through oembed
1702                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1703
1704                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1705                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1706                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1707                 // start which is always required). Allow desc with a missing summary for compatibility.
1708
1709                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1710                         $sub = Event::getHTML($ev, $simple_html);
1711
1712                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1713                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1714                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1715                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1716                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1717                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1718                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1719                 }
1720
1721                 // Replace non graphical smilies for external posts
1722                 if ($simple_html) {
1723                         $text = Smilies::replace($text, false, true);
1724                 }
1725
1726                 // Replace inline code blocks
1727                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1728                         function ($match) use ($simple_html) {
1729                                 $return = '<key>' . $match[1] . '</key>';
1730                                 // Use <code> for Diaspora inline code blocks
1731                                 if ($simple_html === 3) {
1732                                         $return = '<code>' . $match[1] . '</code>';
1733                                 }
1734                                 return $return;
1735                         }
1736                 , $text);
1737
1738                 // Unhide all [noparse] contained bbtags unspacefying them
1739                 // and triming the [noparse] tag.
1740
1741                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1742                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1743                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1744
1745                 /// @todo What is the meaning of these lines?
1746                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1747                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1748
1749                 // Currently deactivated, it made problems with " inside of alt texts.
1750                 //$text = preg_replace('/\&quot\;/', '"', $text);
1751
1752                 // fix any escaped ampersands that may have been converted into links
1753                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1754
1755                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1756                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1757                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1758                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1759
1760                 // sanitize href attributes (only whitelisted protocols URLs)
1761                 // default value for backward compatibility
1762                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1763
1764                 // Always allowed protocol even if config isn't set or not including it
1765                 $allowed_link_protocols[] = 'http';
1766                 $allowed_link_protocols[] = 'redir/';
1767
1768                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1769                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1770
1771                 if ($saved_image) {
1772                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1773                 }
1774
1775                 // Clean up the HTML by loading and saving the HTML with the DOM.
1776                 // Bad structured html can break a whole page.
1777                 // For performance reasons do it only with ativated item cache or at export.
1778                 if (!$try_oembed || (get_itemcachepath() != "")) {
1779                         $doc = new DOMDocument();
1780                         $doc->preserveWhiteSpace = false;
1781
1782                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1783
1784                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1785                         $encoding = '<?xml encoding="UTF-8">';
1786                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1787                         $doc->encoding = 'UTF-8';
1788                         $text = $doc->saveHTML();
1789                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1790
1791                         $text = str_replace('<br></li>', '</li>', $text);
1792
1793                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1794                 }
1795
1796                 // Clean up some useless linebreaks in lists
1797                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1798                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1799                 //$Text = str_replace('</li><br />', '</li>', $Text);
1800                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1801                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1802
1803                 Addon::callHooks('bbcode', $text);
1804
1805                 return trim($text);
1806         }
1807
1808         /**
1809          * @brief Strips the "abstract" tag from the provided text
1810          *
1811          * @param string $text The text with BBCode
1812          * @return string The same text - but without "abstract" element
1813          */
1814         public static function stripAbstract($text)
1815         {
1816                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1817                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1818
1819                 return $text;
1820         }
1821
1822         /**
1823          * @brief Returns the value of the "abstract" element
1824          *
1825          * @param string $text The text that maybe contains the element
1826          * @param string $addon The addon for which the abstract is meant for
1827          * @return string The abstract
1828          */
1829         public static function getAbstract($text, $addon = "")
1830         {
1831                 $abstract = "";
1832                 $abstracts = [];
1833                 $addon = strtolower($addon);
1834
1835                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1836                         foreach ($results AS $result) {
1837                                 $abstracts[strtolower($result[1])] = $result[2];
1838                         }
1839                 }
1840
1841                 if (isset($abstracts[$addon])) {
1842                         $abstract = $abstracts[$addon];
1843                 }
1844
1845                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1846                         $abstract = $result[1];
1847                 }
1848
1849                 return $abstract;
1850         }
1851
1852         /**
1853          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1854          *
1855          * @param array $match Matching values for the callback
1856          * @return string Replaced mention
1857          */
1858         private static function bbCodeMention2DiasporaCallback($match)
1859         {
1860                 $contact = Contact::getDetailsByURL($match[3]);
1861
1862                 if (empty($contact['addr'])) {
1863                         $contact = Probe::uri($match[3]);
1864                 }
1865
1866                 if (empty($contact['addr'])) {
1867                         return $match[0];
1868                 }
1869
1870                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1871                 return $mention;
1872         }
1873
1874         /**
1875          * @brief Converts a BBCode text into Markdown
1876          *
1877          * This function converts a BBCode item body to be sent to Markdown-enabled
1878          * systems like Diaspora and Libertree
1879          *
1880          * @param string $text
1881          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1882          * @return string
1883          */
1884         public static function toMarkdown($text, $for_diaspora = true)
1885         {
1886                 $a = self::getApp();
1887
1888                 $original_text = $text;
1889
1890                 // Since Diaspora is creating a summary for links, this function removes them before posting
1891                 if ($for_diaspora) {
1892                         $text = self::removeShareInformation($text);
1893                 }
1894
1895                 /**
1896                  * Transform #tags, strip off the [url] and replace spaces with underscore
1897                  */
1898                 $url_search_string = "^\[\]";
1899                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1900                         function ($matches) {
1901                                 return '#' . str_replace(' ', '_', $matches[2]);
1902                         },
1903                         $text
1904                 );
1905
1906                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1907                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1908
1909                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
1910                 $codeblocks = [];
1911
1912                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1913                         function ($matches) use (&$codeblocks) {
1914                                 $return = $matches[0];
1915                                 if (strpos($matches[2], "\n") !== false) {
1916                                         $return = '#codeblock-' . count($codeblocks) . '#';
1917
1918                                         $prefix = '````' . $matches[1] . PHP_EOL;
1919                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
1920                                 }
1921                                 return $return;
1922                         },
1923                         $text
1924                 );
1925
1926                 // Convert it to HTML - don't try oembed
1927                 if ($for_diaspora) {
1928                         $text = self::convert($text, false, 3);
1929
1930                         // Add all tags that maybe were removed
1931                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1932                                 $tagline = "";
1933                                 foreach ($tags[2] as $tag) {
1934                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1935                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1936                                                 $tagline .= '#' . $tag . ' ';
1937                                         }
1938                                 }
1939                                 $text = $text . " " . $tagline;
1940                         }
1941                 } else {
1942                         $text = self::convert($text, false, 4);
1943                 }
1944
1945                 // mask some special HTML chars from conversation to markdown
1946                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1947
1948                 // If a link is followed by a quote then there should be a newline before it
1949                 // Maybe we should make this newline at every time before a quote.
1950                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1951
1952                 $stamp1 = microtime(true);
1953
1954                 // Now convert HTML to Markdown
1955                 $converter = new HtmlConverter();
1956                 $text = $converter->convert($text);
1957
1958                 // unmask the special chars back to HTML
1959                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1960
1961                 $a->save_timestamp($stamp1, "parser");
1962
1963                 // Libertree has a problem with escaped hashtags.
1964                 $text = str_replace(['\#'], ['#'], $text);
1965
1966                 // Remove any leading or trailing whitespace, as this will mess up
1967                 // the Diaspora signature verification and cause the item to disappear
1968                 $text = trim($text);
1969
1970                 if ($for_diaspora) {
1971                         $url_search_string = "^\[\]";
1972                         $text = preg_replace_callback(
1973                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
1974                                 ['self', 'bbCodeMention2DiasporaCallback'],
1975                                 $text
1976                         );
1977                 }
1978
1979                 // Restore code blocks
1980                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1981                         function ($matches) use ($codeblocks) {
1982                                 $return = '';
1983                                 if (isset($codeblocks[intval($matches[1])])) {
1984                                         $return = $codeblocks[$matches[1]];
1985                                 }
1986                                 return $return;
1987                         },
1988                         $text
1989                 );
1990
1991                 Addon::callHooks('bb2diaspora', $text);
1992
1993                 return $text;
1994         }
1995 }