src/Content/Text/BBCode.php

   1 <?php
   2
   3 /**
   4  * @file src/Content/Text/BBCode.php
   5  */
   6
   7 namespace Friendica\Content\Text;
   8
   9 use DOMDocument;
  10 use DOMXPath;
  11 use Exception;
  12 use Friendica\BaseObject;
  13 use Friendica\Content\OEmbed;
  14 use Friendica\Content\Smilies;
  15 use Friendica\Core\Addon;
  16 use Friendica\Core\Cache;
  17 use Friendica\Core\Config;
  18 use Friendica\Core\L10n;
  19 use Friendica\Core\PConfig;
  20 use Friendica\Core\Protocol;
  21 use Friendica\Core\System;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Network\Probe;
  25 use Friendica\Object\Image;
  26 use Friendica\Util\Map;
  27 use Friendica\Util\Network;
  28 use Friendica\Util\ParseUrl;
  29 use League\HTMLToMarkdown\HtmlConverter;
  30
  31 require_once "mod/proxy.php";
  32
  33 class BBCode extends BaseObject
  34 {
  35         /**
  36          * @brief Fetches attachment data that were generated the old way
  37          *
  38          * @param string $body Message body
  39          * @return array
  40          * 'type' -> Message type ("link", "video", "photo")
  41          * 'text' -> Text before the shared message
  42          * 'after' -> Text after the shared message
  43          * 'image' -> Preview image of the message
  44          * 'url' -> Url to the attached message
  45          * 'title' -> Title of the attachment
  46          * 'description' -> Description of the attachment
  47          */
  48         private static function getOldAttachmentData($body)
  49         {
  50                 $post = [];
  51
  52                 // Simplify image codes
  53                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  54
  55                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  56                         foreach ($attached as $data) {
  57                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  58                                         continue;
  59                                 }
  60
  61                                 $post["type"] = substr($data[1], 5);
  62
  63                                 $pos = strpos($body, $data[0]);
  64                                 if ($pos > 0) {
  65                                         $post["text"] = trim(substr($body, 0, $pos));
  66                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  67                                 } else {
  68                                         $post["text"] = trim(str_replace($data[0], "", $body));
  69                                 }
  70
  71                                 $attacheddata = $data[2];
  72
  73                                 $URLSearchString = "^\[\]";
  74
  75                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  76
  77                                         $picturedata = Image::getInfoFromURL($matches[1]);
  78
  79                                         if ($picturedata) {
  80                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  81                                                         $post["image"] = $matches[1];
  82                                                 } else {
  83                                                         $post["preview"] = $matches[1];
  84                                                 }
  85                                         }
  86                                 }
  87
  88                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  89                                         $post["url"] = $matches[1];
  90                                         $post["title"] = $matches[2];
  91                                 }
  92                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  93                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  94                                         $post["url"] = $matches[1];
  95                                 }
  96
  97                                 // Search for description
  98                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  99                                         $post["description"] = $matches[1];
 100                                 }
 101                         }
 102                 }
 103                 return $post;
 104         }
 105
 106         /**
 107          * @brief Fetches attachment data that were generated with the "attachment" element
 108          *
 109          * @param string $body Message body
 110          * @return array
 111          * 'type' -> Message type ("link", "video", "photo")
 112          * 'text' -> Text before the shared message
 113          * 'after' -> Text after the shared message
 114          * 'image' -> Preview image of the message
 115          * 'url' -> Url to the attached message
 116          * 'title' -> Title of the attachment
 117          * 'description' -> Description of the attachment
 118          */
 119         public static function getAttachmentData($body)
 120         {
 121                 $data = [];
 122
 123                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 124                         return self::getOldAttachmentData($body);
 125                 }
 126
 127                 $attributes = $match[2];
 128
 129                 $data["text"] = trim($match[1]);
 130
 131                 $type = "";
 132                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 133                 if (x($matches, 1)) {
 134                         $type = strtolower($matches[1]);
 135                 }
 136
 137                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 138                 if (x($matches, 1)) {
 139                         $type = strtolower($matches[1]);
 140                 }
 141
 142                 if ($type == "") {
 143                         return [];
 144                 }
 145
 146                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 147                         return [];
 148                 }
 149
 150                 if ($type != "") {
 151                         $data["type"] = $type;
 152                 }
 153
 154                 $url = "";
 155                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 156                 if (x($matches, 1)) {
 157                         $url = $matches[1];
 158                 }
 159
 160                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 161                 if (x($matches, 1)) {
 162                         $url = $matches[1];
 163                 }
 164
 165                 if ($url != "") {
 166                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 167                 }
 168
 169                 $title = "";
 170                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 171                 if (x($matches, 1)) {
 172                         $title = $matches[1];
 173                 }
 174
 175                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 176                 if (x($matches, 1)) {
 177                         $title = $matches[1];
 178                 }
 179
 180                 if ($title != "") {
 181                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 182                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 183                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 184                         $data["title"] = $title;
 185                 }
 186
 187                 $image = "";
 188                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 189                 if (x($matches, 1)) {
 190                         $image = $matches[1];
 191                 }
 192
 193                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 194                 if (x($matches, 1)) {
 195                         $image = $matches[1];
 196                 }
 197
 198                 if ($image != "") {
 199                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 200                 }
 201
 202                 $preview = "";
 203                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 204                 if (x($matches, 1)) {
 205                         $preview = $matches[1];
 206                 }
 207
 208                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 209                 if (x($matches, 1)) {
 210                         $preview = $matches[1];
 211                 }
 212
 213                 if ($preview != "") {
 214                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 215                 }
 216
 217                 $data["description"] = trim($match[3]);
 218
 219                 $data["after"] = trim($match[4]);
 220
 221                 return $data;
 222         }
 223
 224         public static function getAttachedData($body, $item = [])
 225         {
 226                 /*
 227                 - text:
 228                 - type: link, video, photo
 229                 - title:
 230                 - url:
 231                 - image:
 232                 - description:
 233                 - (thumbnail)
 234                 */
 235
 236                 $has_title = !empty($item['title']);
 237                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 238                 $post = self::getAttachmentData($body);
 239
 240                 // if nothing is found, it maybe having an image.
 241                 if (!isset($post["type"])) {
 242                         // Simplify image codes
 243                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 244
 245                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 246
 247                         $URLSearchString = "^\[\]";
 248                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 249                                 if ((count($pictures) == 1) && !$has_title) {
 250                                         // Checking, if the link goes to a picture
 251                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 252
 253                                         // Workaround:
 254                                         // Sometimes photo posts to the own album are not detected at the start.
 255                                         // So we seem to cannot use the cache for these cases. That's strange.
 256                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 257                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 258                                         }
 259
 260                                         if ($data["type"] == "photo") {
 261                                                 $post["type"] = "photo";
 262                                                 if (isset($data["images"][0])) {
 263                                                         $post["image"] = $data["images"][0]["src"];
 264                                                         $post["url"] = $data["url"];
 265                                                 } else {
 266                                                         $post["image"] = $data["url"];
 267                                                 }
 268
 269                                                 $post["preview"] = $pictures[0][2];
 270                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 271                                         } else {
 272                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 273                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 274                                                         $post["type"] = "photo";
 275                                                         $post["image"] = $pictures[0][1];
 276                                                         $post["preview"] = $pictures[0][2];
 277                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 278                                                 }
 279                                         }
 280                                 } elseif (count($pictures) > 0) {
 281                                         $post["type"] = "link";
 282                                         $post["url"] = $plink;
 283                                         $post["image"] = $pictures[0][2];
 284                                         $post["text"] = $body;
 285                                 }
 286                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 287                                 if ((count($pictures) == 1) && !$has_title) {
 288                                         $post["type"] = "photo";
 289                                         $post["image"] = $pictures[0][1];
 290                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 291                                 } elseif (count($pictures) > 0) {
 292                                         $post["type"] = "link";
 293                                         $post["url"] = $plink;
 294                                         $post["image"] = $pictures[0][1];
 295                                         $post["text"] = $body;
 296                                 }
 297                         }
 298
 299                         // Test for the external links
 300                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 301                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 302
 303                         $links = array_merge($links1, $links2);
 304
 305                         // If there is only a single one, then use it.
 306                         // This should cover link posts via API.
 307                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 308                                 $post["type"] = "link";
 309                                 $post["text"] = trim($body);
 310                                 $post["url"] = $links[0][1];
 311                         }
 312
 313                         // Now count the number of external media links
 314                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 315                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 316                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 317                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 318
 319                         // Add them to the other external links
 320                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 321
 322                         // Are there more than one?
 323                         if (count($links) > 1) {
 324                                 // The post will be the type "text", which means a blog post
 325                                 unset($post["type"]);
 326                                 $post["url"] = $plink;
 327                         }
 328
 329                         if (!isset($post["type"])) {
 330                                 $post["type"] = "text";
 331                                 $post["text"] = trim($body);
 332                         }
 333                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 334                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 335
 336                         if (isset($data["images"][0])) {
 337                                 $post["image"] = $data["images"][0]["src"];
 338                         }
 339                 }
 340
 341                 return $post;
 342         }
 343
 344         /**
 345          * @brief Convert a message into plaintext for connectors to other networks
 346          *
 347          * @param array $b The message array that is about to be posted
 348          * @param int $limit The maximum number of characters when posting to that network
 349          * @param bool $includedlinks Has an attached link to be included into the message?
 350          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 351          * @param string $target_network Name of the network where the post should go to.
 352          *
 353          * @return string The converted message
 354          */
 355         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 356         {
 357                 // Remove the hash tags
 358                 $URLSearchString = "^\[\]";
 359                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 360
 361                 // Add an URL element if the text contains a raw link
 362                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 363
 364                 // Remove the abstract
 365                 $body = self::stripAbstract($body);
 366
 367                 // At first look at data that is attached via "type-..." stuff
 368                 // This will hopefully replaced with a dedicated bbcode later
 369                 //$post = self::getAttachedData($b["body"]);
 370                 $post = self::getAttachedData($body, $b);
 371
 372                 if (($b["title"] != "") && ($post["text"] != "")) {
 373                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 374                 } elseif ($b["title"] != "") {
 375                         $post["text"] = trim($b["title"]);
 376                 }
 377
 378                 $abstract = "";
 379
 380                 // Fetch the abstract from the given target network
 381                 if ($target_network != "") {
 382                         $default_abstract = self::getAbstract($b["body"]);
 383                         $abstract = self::getAbstract($b["body"], $target_network);
 384
 385                         // If we post to a network with no limit we only fetch
 386                         // an abstract exactly for this network
 387                         if (($limit == 0) && ($abstract == $default_abstract)) {
 388                                 $abstract = "";
 389                         }
 390                 } else {// Try to guess the correct target network
 391                         switch ($htmlmode) {
 392                                 case 8:
 393                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
 394                                         break;
 395                                 case 7:
 396                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
 397                                         break;
 398                                 case 6:
 399                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
 400                                         break;
 401                                 default: // We don't know the exact target.
 402                                         // We fetch an abstract since there is a posting limit.
 403                                         if ($limit > 0) {
 404                                                 $abstract = self::getAbstract($b["body"]);
 405                                         }
 406                         }
 407                 }
 408
 409                 if ($abstract != "") {
 410                         $post["text"] = $abstract;
 411
 412                         if ($post["type"] == "text") {
 413                                 $post["type"] = "link";
 414                                 $post["url"] = $b["plink"];
 415                         }
 416                 }
 417
 418                 $html = self::convert($post["text"].$post["after"], false, $htmlmode);
 419                 $msg = HTML::toPlaintext($html, 0, true);
 420                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 421
 422                 $link = "";
 423                 if ($includedlinks) {
 424                         if ($post["type"] == "link") {
 425                                 $link = $post["url"];
 426                         } elseif ($post["type"] == "text") {
 427                                 $link = $post["url"];
 428                         } elseif ($post["type"] == "video") {
 429                                 $link = $post["url"];
 430                         } elseif ($post["type"] == "photo") {
 431                                 $link = $post["image"];
 432                         }
 433
 434                         if (($msg == "") && isset($post["title"])) {
 435                                 $msg = trim($post["title"]);
 436                         }
 437
 438                         if (($msg == "") && isset($post["description"])) {
 439                                 $msg = trim($post["description"]);
 440                         }
 441
 442                         // If the link is already contained in the post, then it neeedn't to be added again
 443                         // But: if the link is beyond the limit, then it has to be added.
 444                         if (($link != "") && strstr($msg, $link)) {
 445                                 $pos = strpos($msg, $link);
 446
 447                                 // Will the text be shortened in the link?
 448                                 // Or is the link the last item in the post?
 449                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 450                                         $msg = trim(str_replace($link, "", $msg));
 451                                 } elseif (($limit == 0) || ($pos < $limit)) {
 452                                         // The limit has to be increased since it will be shortened - but not now
 453                                         // Only do it with Twitter (htmlmode = 8)
 454                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 455                                                 $limit = $limit - 23 + strlen($link);
 456                                         }
 457
 458                                         $link = "";
 459
 460                                         if ($post["type"] == "text") {
 461                                                 unset($post["url"]);
 462                                         }
 463                                 }
 464                         }
 465                 }
 466
 467                 if ($limit > 0) {
 468                         // Reduce multiple spaces
 469                         // When posted to a network with limited space, we try to gain space where possible
 470                         while (strpos($msg, "  ") !== false) {
 471                                 $msg = str_replace("  ", " ", $msg);
 472                         }
 473
 474                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 475                         if (iconv_strlen($link, "UTF-8") > 0) {
 476                                 $limit = $limit - 23;
 477                         }
 478
 479                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 480                                 if (($post["type"] == "text") && isset($post["url"])) {
 481                                         $post["url"] = $b["plink"];
 482                                 } elseif (!isset($post["url"])) {
 483                                         $limit = $limit - 23;
 484                                         $post["url"] = $b["plink"];
 485                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 486                                 //} elseif (strpos($b["body"], "[share") !== false) {
 487                                 //      $post["url"] = $b["plink"];
 488                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 489                                         $post["url"] = $b["plink"];
 490                                 }
 491                                 $msg = Plaintext::shorten($msg, $limit);
 492                         }
 493                 }
 494
 495                 $post["text"] = trim($msg);
 496
 497                 return($post);
 498         }
 499
 500         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 501         {
 502                 // Suppress "view full size"
 503                 if (intval(Config::get('system', 'no_view_full_size'))) {
 504                         $include_link = false;
 505                 }
 506
 507                 // Picture addresses can contain special characters
 508                 $s = htmlspecialchars_decode($srctext);
 509
 510                 $matches = null;
 511                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 512                 if ($c) {
 513                         foreach ($matches as $mtch) {
 514                                 logger('scale_external_image: ' . $mtch[1]);
 515
 516                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 517                                 if (stristr($mtch[1], $hostname)) {
 518                                         continue;
 519                                 }
 520
 521                                 // $scale_replace, if passed, is an array of two elements. The
 522                                 // first is the name of the full-size image. The second is the
 523                                 // name of a remote, scaled-down version of the full size image.
 524                                 // This allows Friendica to display the smaller remote image if
 525                                 // one exists, while still linking to the full-size image
 526                                 if ($scale_replace) {
 527                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 528                                 } else {
 529                                         $scaled = $mtch[1];
 530                                 }
 531                                 $i = Network::fetchUrl($scaled);
 532                                 if (!$i) {
 533                                         return $srctext;
 534                                 }
 535
 536                                 // guess mimetype from headers or filename
 537                                 $type = Image::guessType($mtch[1], true);
 538
 539                                 if ($i) {
 540                                         $Image = new Image($i, $type);
 541                                         if ($Image->isValid()) {
 542                                                 $orig_width = $Image->getWidth();
 543                                                 $orig_height = $Image->getHeight();
 544
 545                                                 if ($orig_width > 640 || $orig_height > 640) {
 546                                                         $Image->scaleDown(640);
 547                                                         $new_width = $Image->getWidth();
 548                                                         $new_height = $Image->getHeight();
 549                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 550                                                         $s = str_replace(
 551                                                                 $mtch[0],
 552                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 553                                                                 . "\n" . (($include_link)
 554                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 555                                                                         : ''),
 556                                                                 $s
 557                                                         );
 558                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 559                                                 }
 560                                         }
 561                                 }
 562                         }
 563                 }
 564
 565                 // replace the special char encoding
 566                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 567                 return $s;
 568         }
 569
 570         /**
 571          * The purpose of this function is to apply system message length limits to
 572          * imported messages without including any embedded photos in the length
 573          *
 574          * @brief Truncates imported message body string length to max_import_size
 575          * @param string $body
 576          * @return string
 577          */
 578         public static function limitBodySize($body)
 579         {
 580                 $maxlen = get_max_import_size();
 581
 582                 // If the length of the body, including the embedded images, is smaller
 583                 // than the maximum, then don't waste time looking for the images
 584                 if ($maxlen && (strlen($body) > $maxlen)) {
 585
 586                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 587
 588                         $orig_body = $body;
 589                         $new_body = '';
 590                         $textlen = 0;
 591
 592                         $img_start = strpos($orig_body, '[img');
 593                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 594                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 595                         while (($img_st_close !== false) && ($img_end !== false)) {
 596
 597                                 $img_st_close++; // make it point to AFTER the closing bracket
 598                                 $img_end += $img_start;
 599                                 $img_end += strlen('[/img]');
 600
 601                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 602                                         // This is an embedded image
 603
 604                                         if (($textlen + $img_start) > $maxlen) {
 605                                                 if ($textlen < $maxlen) {
 606                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 607                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 608                                                         $textlen = $maxlen;
 609                                                 }
 610                                         } else {
 611                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 612                                                 $textlen += $img_start;
 613                                         }
 614
 615                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 616                                 } else {
 617
 618                                         if (($textlen + $img_end) > $maxlen) {
 619                                                 if ($textlen < $maxlen) {
 620                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 621                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 622                                                         $textlen = $maxlen;
 623                                                 }
 624                                         } else {
 625                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 626                                                 $textlen += $img_end;
 627                                         }
 628                                 }
 629                                 $orig_body = substr($orig_body, $img_end);
 630
 631                                 if ($orig_body === false) {
 632                                         // in case the body ends on a closing image tag
 633                                         $orig_body = '';
 634                                 }
 635
 636                                 $img_start = strpos($orig_body, '[img');
 637                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 638                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 639                         }
 640
 641                         if (($textlen + strlen($orig_body)) > $maxlen) {
 642                                 if ($textlen < $maxlen) {
 643                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 644                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 645                                 }
 646                         } else {
 647                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 648                                 $new_body = $new_body . $orig_body;
 649                         }
 650
 651                         return $new_body;
 652                 } else {
 653                         return $body;
 654                 }
 655         }
 656
 657         /**
 658          * Processes [attachment] tags
 659          *
 660          * Note: Can produce a [bookmark] tag in the returned string
 661          *
 662          * @brief Processes [attachment] tags
 663          * @param string $return
 664          * @param bool|int $simplehtml
 665          * @param bool $tryoembed
 666          * @return string
 667          */
 668         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 669         {
 670                 $data = self::getAttachmentData($return);
 671                 if (!$data) {
 672                         return $return;
 673                 }
 674
 675                 if (isset($data["title"])) {
 676                         $data["title"] = strip_tags($data["title"]);
 677                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 678                 }
 679
 680                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
 681                         $data["preview"] = $data["image"];
 682                         $data["image"] = "";
 683                 }
 684
 685                 $return = '';
 686                 if ($simplehtml == 7) {
 687                         $return = self::convertUrlForOStatus($data["url"]);
 688                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 689                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 690                 } else {
 691                         try {
 692                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 693                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 694                                 } else {
 695                                         throw new Exception('OEmbed is disabled for this attachment.');
 696                                 }
 697                         } catch (Exception $e) {
 698                                 if ($simplehtml != 4) {
 699                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 700                                 }
 701
 702                                 if ($data["image"] != "") {
 703                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
 704                                 } elseif ($data["preview"] != "") {
 705                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
 706                                 }
 707
 708                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
 709                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], proxy_url($data["image"]), $data["title"]);
 710                                 } else {
 711                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 712                                 }
 713
 714                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 715                                         // Sanitize the HTML by converting it to BBCode
 716                                         $bbcode = HTML::toBBCode($data["description"]);
 717                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 718                                 }
 719                                 if ($data["type"] == "link") {
 720                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 721                                 }
 722
 723                                 if ($simplehtml != 4) {
 724                                         $return .= '</div>';
 725                                 }
 726                         }
 727                 }
 728
 729                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 730         }
 731
 732         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 733         {
 734                 $data = self::getAttachmentData($Text);
 735
 736                 if (!$data) {
 737                         return $Text;
 738                 } elseif ($nolink) {
 739                         return $data["text"] . $data["after"];
 740                 }
 741
 742                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
 743                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 744                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 745                         $data["title"] = $data["url"];
 746                 } elseif (($text != "") && strstr($title, $text)) {
 747                         $data["text"] = $data["title"];
 748                         $data["title"] = $data["url"];
 749                 }
 750
 751                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 752                         return $data["title"] . $data["after"];
 753                 }
 754
 755                 // If the link already is included in the post, don't add it again
 756                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 757                         return $data["text"] . $data["after"];
 758                 }
 759
 760                 $text = $data["text"];
 761
 762                 if (($data["url"] != "") && ($data["title"] != "")) {
 763                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 764                 } elseif (($data["url"] != "")) {
 765                         $text .= "\n[url]" . $data["url"] . "[/url]";
 766                 }
 767
 768                 return $text . "\n" . $data["after"];
 769         }
 770
 771         /**
 772          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 773          *
 774          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 775          * @param array $match Array with the matching values
 776          * @return string reformatted link including HTML codes
 777          */
 778         private static function convertUrlForOStatusCallback($match)
 779         {
 780                 $url = $match[1];
 781
 782                 if (isset($match[2]) && ($match[1] != $match[2])) {
 783                         return $match[0];
 784                 }
 785
 786                 $parts = parse_url($url);
 787                 if (!isset($parts['scheme'])) {
 788                         return $match[0];
 789                 }
 790
 791                 return self::convertUrlForOStatus($url);
 792         }
 793
 794         /**
 795          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 796          * @param string $url URL that is about to be reformatted
 797          * @return string reformatted link including HTML codes
 798          */
 799         private static function convertUrlForOStatus($url)
 800         {
 801                 $parts = parse_url($url);
 802                 $scheme = $parts['scheme'] . '://';
 803                 $styled_url = str_replace($scheme, '', $url);
 804
 805                 if (strlen($styled_url) > 30) {
 806                         $styled_url = substr($styled_url, 0, 30) . "…";
 807                 }
 808
 809                 $html = '<a href="%s" target="_blank">%s</a>';
 810
 811                 return sprintf($html, $url, $styled_url);
 812         }
 813
 814         /*
 815          * [noparse][i]italic[/i][/noparse] turns into
 816          * [noparse][ i ]italic[ /i ][/noparse],
 817          * to hide them from parser.
 818          */
 819         private static function escapeNoparseCallback($match)
 820         {
 821                 $whole_match = $match[0];
 822                 $captured = $match[1];
 823                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 824                 $new_str = str_replace($captured, $spacefied, $whole_match);
 825                 return $new_str;
 826         }
 827
 828         /*
 829          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 830          * now turns back and the [noparse] tags are trimed
 831          * returning [i]italic[/i]
 832          */
 833         private static function unescapeNoparseCallback($match)
 834         {
 835                 $captured = $match[1];
 836                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 837                 return $unspacefied;
 838         }
 839
 840         /**
 841          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 842          * occurrences
 843          *
 844          * @param string $text        Text to search
 845          * @param string $name        Tag name
 846          * @param int    $occurrences Number of first occurrences to skip
 847          * @return boolean|array
 848          */
 849         public static function getTagPosition($text, $name, $occurrences = 0)
 850         {
 851                 if ($occurrences < 0) {
 852                         $occurrences = 0;
 853                 }
 854
 855                 $start_open = -1;
 856                 for ($i = 0; $i <= $occurrences; $i++) {
 857                         if ($start_open !== false) {
 858                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 859                         }
 860                 }
 861
 862                 if ($start_open === false) {
 863                         return false;
 864                 }
 865
 866                 $start_equal = strpos($text, '=', $start_open);
 867                 $start_close = strpos($text, ']', $start_open);
 868
 869                 if ($start_close === false) {
 870                         return false;
 871                 }
 872
 873                 $start_close++;
 874
 875                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 876
 877                 if ($end_open === false) {
 878                         return false;
 879                 }
 880
 881                 $res = [
 882                         'start' => [
 883                                 'open' => $start_open,
 884                                 'close' => $start_close
 885                         ],
 886                         'end' => [
 887                                 'open' => $end_open,
 888                                 'close' => $end_open + strlen('[/' . $name . ']')
 889                         ],
 890                 ];
 891
 892                 if ($start_equal !== false) {
 893                         $res['start']['equal'] = $start_equal + 1;
 894                 }
 895
 896                 return $res;
 897         }
 898
 899         /**
 900          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 901          *
 902          * @param type $pattern Preg pattern string
 903          * @param type $replace Preg replace string
 904          * @param type $name    BBCode tag name
 905          * @param type $text    Text to search
 906          * @return string
 907          */
 908         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 909         {
 910                 $occurrences = 0;
 911                 $pos = self::getTagPosition($text, $name, $occurrences);
 912                 while ($pos !== false && $occurrences++ < 1000) {
 913                         $start = substr($text, 0, $pos['start']['open']);
 914                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 915                         $end = substr($text, $pos['end']['close']);
 916                         if ($end === false) {
 917                                 $end = '';
 918                         }
 919
 920                         $subject = preg_replace($pattern, $replace, $subject);
 921                         $text = $start . $subject . $end;
 922
 923                         $pos = self::getTagPosition($text, $name, $occurrences);
 924                 }
 925
 926                 return $text;
 927         }
 928
 929         private static function extractImagesFromItemBody($body)
 930         {
 931                 $saved_image = [];
 932                 $orig_body = $body;
 933                 $new_body = '';
 934
 935                 $cnt = 0;
 936                 $img_start = strpos($orig_body, '[img');
 937                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 938                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 939                 while (($img_st_close !== false) && ($img_end !== false)) {
 940                         $img_st_close++; // make it point to AFTER the closing bracket
 941                         $img_end += $img_start;
 942
 943                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 944                                 // This is an embedded image
 945                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 946                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 947
 948                                 $cnt++;
 949                         } else {
 950                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 951                         }
 952
 953                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 954
 955                         if ($orig_body === false) {
 956                                 // in case the body ends on a closing image tag
 957                                 $orig_body = '';
 958                         }
 959
 960                         $img_start = strpos($orig_body, '[img');
 961                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 962                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 963                 }
 964
 965                 $new_body = $new_body . $orig_body;
 966
 967                 return ['body' => $new_body, 'images' => $saved_image];
 968         }
 969
 970         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 971         {
 972                 $newbody = $body;
 973
 974                 $cnt = 0;
 975                 foreach ($images as $image) {
 976                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 977                         // it loops over the array starting from the first element and going sequentially
 978                         // to the last element
 979                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 980                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 981                         $cnt++;
 982                 }
 983
 984                 return $newbody;
 985         }
 986
 987         /**
 988          * Processes [share] tags
 989          *
 990          * Note: Can produce a [bookmark] tag in the output
 991          *
 992          * @brief Processes [share] tags
 993          * @param array    $share      preg_match_callback result array
 994          * @param bool|int $simplehtml
 995          * @return string
 996          */
 997         private static function convertShare($share, $simplehtml)
 998         {
 999                 $attributes = $share[2];
1000
1001                 $author = "";
1002                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1003                 if (x($matches, 1)) {
1004                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1005                 }
1006
1007                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1008                 if (x($matches, 1)) {
1009                         $author = $matches[1];
1010                 }
1011
1012                 $profile = "";
1013                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1014                 if (x($matches, 1)) {
1015                         $profile = $matches[1];
1016                 }
1017
1018                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1019                 if (x($matches, 1)) {
1020                         $profile = $matches[1];
1021                 }
1022
1023                 $avatar = "";
1024                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1025                 if (x($matches, 1)) {
1026                         $avatar = $matches[1];
1027                 }
1028
1029                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1030                 if (x($matches, 1)) {
1031                         $avatar = $matches[1];
1032                 }
1033
1034                 $link = "";
1035                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1036                 if (x($matches, 1)) {
1037                         $link = $matches[1];
1038                 }
1039
1040                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1041                 if (x($matches, 1)) {
1042                         $link = $matches[1];
1043                 }
1044
1045                 $posted = "";
1046
1047                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1048                 if (x($matches, 1)) {
1049                         $posted = $matches[1];
1050                 }
1051
1052                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1053                 if (x($matches, 1)) {
1054                         $posted = $matches[1];
1055                 }
1056
1057                 // We only call this so that a previously unknown contact can be added.
1058                 // This is important for the function "Model\Contact::getDetailsByURL()".
1059                 // This function then can fetch an entry from the contact table.
1060                 Contact::getIdForURL($profile, 0, true);
1061
1062                 $data = Contact::getDetailsByURL($profile);
1063
1064                 if (x($data, "name") && x($data, "addr")) {
1065                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1066                 } else {
1067                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1068                 }
1069
1070                 if (x($data, "addr")) {
1071                         $userid = $data["addr"];
1072                 } else {
1073                         $userid = Protocol::formatMention($profile, $author);
1074                 }
1075
1076                 if (x($data, "name")) {
1077                         $author = $data["name"];
1078                 }
1079
1080                 if (x($data, "micro")) {
1081                         $avatar = $data["micro"];
1082                 }
1083
1084                 $preshare = trim($share[1]);
1085                 if ($preshare != "") {
1086                         $preshare .= "<br />";
1087                 }
1088
1089                 switch ($simplehtml) {
1090                         case 1:
1091                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1092                                 break;
1093                         case 2:
1094                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1095                                 break;
1096                         case 3: // Diaspora
1097                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1098
1099                                 $text = trim($share[1]);
1100
1101                                 if ($text != "") {
1102                                         $text .= "<hr />";
1103                                 }
1104
1105                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1106                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1107                                 } else {
1108                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1109
1110                                         if ($link != "") {
1111                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1112                                         }
1113                                 }
1114
1115                                 break;
1116                         case 4:
1117                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1118                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1119                                 $headline .= ":</b><br />";
1120
1121                                 $text = trim($share[1]);
1122
1123                                 if ($text != "") {
1124                                         $text .= "<hr />";
1125                                 }
1126
1127                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1128
1129                                 break;
1130                         case 5:
1131                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1132                                 break;
1133                         case 6: // app.net
1134                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1135                                 break;
1136                         case 7: // statusnet/GNU Social
1137                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1138                                 break;
1139                         case 8: // twitter
1140                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1141                                 break;
1142                         case 9: // Google+/Facebook
1143                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1144
1145                                 if ($link != "") {
1146                                         $text .= "<br /><br />" . $link;
1147                                 }
1148                                 break;
1149                         default:
1150                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1151                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1152                                         try {
1153                                                 $oembed = OEmbed::getHTML($link, $preshare);
1154                                         } catch (Exception $e) {
1155                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1156                                         }
1157
1158                                         $text = $preshare . $oembed;
1159                                 } else {
1160                                         $text = trim($share[1]) . "\n";
1161
1162                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1163
1164                                         $tpl = get_markup_template('shared_content.tpl');
1165                                         $text .= replace_macros($tpl, [
1166                                                 '$profile' => $profile,
1167                                                 '$avatar' => $avatar,
1168                                                 '$author' => $author,
1169                                                 '$link' => $link,
1170                                                 '$posted' => $posted,
1171                                                 '$content' => trim($share[3])
1172                                         ]);
1173                                 }
1174                                 break;
1175                 }
1176
1177                 return $text;
1178         }
1179
1180         private static function removePictureLinksCallback($match)
1181         {
1182                 $text = Cache::get($match[1]);
1183
1184                 if (is_null($text)) {
1185                         $a = self::getApp();
1186
1187                         $stamp1 = microtime(true);
1188
1189                         $ch = @curl_init($match[1]);
1190                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1191                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1192                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1193                         @curl_exec($ch);
1194                         $curl_info = @curl_getinfo($ch);
1195
1196                         $a->save_timestamp($stamp1, "network");
1197
1198                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1199                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1200                         } else {
1201                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1202
1203                                 // if its not a picture then look if its a page that contains a picture link
1204                                 $body = Network::fetchUrl($match[1]);
1205
1206                                 $doc = new DOMDocument();
1207                                 @$doc->loadHTML($body);
1208                                 $xpath = new DOMXPath($doc);
1209                                 $list = $xpath->query("//meta[@name]");
1210                                 foreach ($list as $node) {
1211                                         $attr = [];
1212
1213                                         if ($node->attributes->length) {
1214                                                 foreach ($node->attributes as $attribute) {
1215                                                         $attr[$attribute->name] = $attribute->value;
1216                                                 }
1217                                         }
1218
1219                                         if (strtolower($attr["name"]) == "twitter:image") {
1220                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1221                                         }
1222                                 }
1223                         }
1224                         Cache::set($match[1], $text);
1225                 }
1226
1227                 return $text;
1228         }
1229
1230         private static function expandLinksCallback($match)
1231         {
1232                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1233                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1234                 } else {
1235                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1236                 }
1237         }
1238
1239         private static function cleanPictureLinksCallback($match)
1240         {
1241                 $text = Cache::get($match[1]);
1242
1243                 if (is_null($text)) {
1244                         $a = self::getApp();
1245
1246                         $stamp1 = microtime(true);
1247
1248                         $ch = @curl_init($match[1]);
1249                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1250                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1251                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1252                         @curl_exec($ch);
1253                         $curl_info = @curl_getinfo($ch);
1254
1255                         $a->save_timestamp($stamp1, "network");
1256
1257                         // if its a link to a picture then embed this picture
1258                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1259                                 $text = "[img]" . $match[1] . "[/img]";
1260                         } else {
1261                                 $text = "[img]" . $match[2] . "[/img]";
1262
1263                                 // if its not a picture then look if its a page that contains a picture link
1264                                 $body = Network::fetchUrl($match[1]);
1265
1266                                 $doc = new DOMDocument();
1267                                 @$doc->loadHTML($body);
1268                                 $xpath = new DOMXPath($doc);
1269                                 $list = $xpath->query("//meta[@name]");
1270                                 foreach ($list as $node) {
1271                                         $attr = [];
1272                                         if ($node->attributes->length) {
1273                                                 foreach ($node->attributes as $attribute) {
1274                                                         $attr[$attribute->name] = $attribute->value;
1275                                                 }
1276                                         }
1277
1278                                         if (strtolower($attr["name"]) == "twitter:image") {
1279                                                 $text = "[img]" . $attr["content"] . "[/img]";
1280                                         }
1281                                 }
1282                         }
1283                         Cache::set($match[1], $text);
1284                 }
1285
1286                 return $text;
1287         }
1288
1289         public static function cleanPictureLinks($text)
1290         {
1291                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1292                 return $return;
1293         }
1294
1295         private static function textHighlightCallback($match)
1296         {
1297                 // Fallback in case the language doesn't exist
1298                 $return = '[code]' . $match[2] . '[/code]';
1299
1300                 if (in_array(strtolower($match[1]),
1301                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1302                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1303                 ) {
1304                         $return = text_highlight($match[2], strtolower($match[1]));
1305                 }
1306
1307                 return $return;
1308         }
1309
1310         /**
1311          * @brief Converts a BBCode message to HTML message
1312          *
1313          * BBcode 2 HTML was written by WAY2WEB.net
1314          * extended to work with Mistpark/Friendica - Mike Macgirvin
1315          *
1316          * Simple HTML values meaning:
1317          * - 0: Friendica display
1318          * - 1: Unused
1319          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1320          * - 3: Used before converting to Markdown in bb2diaspora.php
1321          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1322          * - 5: Unused
1323          * - 6: Used for Appnet
1324          * - 7: Used for dfrn, OStatus
1325          * - 8: Used for WP backlink text setting
1326          *
1327          * @param string $text
1328          * @param bool   $try_oembed
1329          * @param int    $simple_html
1330          * @param bool   $for_plaintext
1331          * @return string
1332          */
1333         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1334         {
1335                 $a = self::getApp();
1336
1337                 /*
1338                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1339                  *
1340                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1341                  * $match[1] = $url
1342                  * $match[2] = $title or absent
1343                  */
1344                 $try_oembed_callback = function ($match)
1345                 {
1346                         $url = $match[1];
1347                         $title = defaults($match, 2, null);
1348
1349                         try {
1350                                 $return = OEmbed::getHTML($url, $title);
1351                         } catch (Exception $ex) {
1352                                 $return = $match[0];
1353                         }
1354
1355                         return $return;
1356                 };
1357
1358                 // Hide all [noparse] contained bbtags by spacefying them
1359                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1360
1361                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1362                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1363                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1364
1365                 // Remove the abstract element. It is a non visible element.
1366                 $text = self::stripAbstract($text);
1367
1368                 // Move all spaces out of the tags
1369                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1370                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1371
1372                 // Extract the private images which use data urls since preg has issues with
1373                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1374                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1375
1376                 $extracted = self::extractImagesFromItemBody($text);
1377                 $text = $extracted['body'];
1378                 $saved_image = $extracted['images'];
1379
1380                 // If we find any event code, turn it into an event.
1381                 // After we're finished processing the bbcode we'll
1382                 // replace all of the event code with a reformatted version.
1383
1384                 $ev = Event::fromBBCode($text);
1385
1386                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1387                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1388
1389                 $text = str_replace("<", "&lt;", $text);
1390                 $text = str_replace(">", "&gt;", $text);
1391
1392                 // remove some newlines before the general conversion
1393                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1394                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1395
1396                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1397                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1398
1399                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1400                 if (!$try_oembed) {
1401                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1402                 }
1403
1404                 // Check for [code] text here, before the linefeeds are messed with.
1405                 // The highlighter will unescape and re-escape the content.
1406                 if (strpos($text, '[code=') !== false) {
1407                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1408                 }
1409                 // Convert new line chars to html <br /> tags
1410
1411                 // nlbr seems to be hopelessly messed up
1412                 //      $Text = nl2br($Text);
1413
1414                 // We'll emulate it.
1415
1416                 $text = trim($text);
1417                 $text = str_replace("\r\n", "\n", $text);
1418
1419                 // removing multiplicated newlines
1420                 if (Config::get("system", "remove_multiplicated_lines")) {
1421                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1422                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1423                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1424                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1425                         do {
1426                                 $oldtext = $text;
1427                                 $text = str_replace($search, $replace, $text);
1428                         } while ($oldtext != $text);
1429                 }
1430
1431                 // Set up the parameters for a URL search string
1432                 $URLSearchString = "^\[\]";
1433                 // Set up the parameters for a MAIL search string
1434                 $MAILSearchString = $URLSearchString;
1435
1436                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1437                 if (!$for_plaintext) {
1438                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1439                         // Currently disabled, since the function is too greedy
1440                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1441                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1442                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1443                         if ($simple_html == 7) {
1444                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1445                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1446                         }
1447                 } else {
1448                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1449                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1450                 }
1451
1452
1453                 // Handle attached links or videos
1454                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1455
1456                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1457
1458                 // Remove all hashtag addresses
1459                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1460                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1461                 } elseif ($simple_html == 3) {
1462                         // The ! is converted to @ since Diaspora only understands the @
1463                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1464                                 '@<a href="$2">$3</a>',
1465                                 $text);
1466                 } elseif ($simple_html == 7) {
1467                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1468                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1469                                 $text);
1470                 } elseif (!$simple_html) {
1471                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1472                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1473                                 $text);
1474                 }
1475
1476                 // Bookmarks in red - will be converted to bookmarks in friendica
1477                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1478                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1479                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1480                                         "[bookmark=$1]$2[/bookmark]", $text);
1481
1482                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1483                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1484                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1485                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1486                 }
1487
1488                 if ($simple_html == 5) {
1489                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1490                 }
1491
1492                 // Perform URL Search
1493                 if ($try_oembed) {
1494                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1495                 }
1496
1497                 if ($simple_html == 5) {
1498                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1499                 } else {
1500                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1501                 }
1502
1503                 // Handle Diaspora posts
1504                 $text = preg_replace_callback(
1505                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1506                         function ($match) {
1507                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1508                         }, $text
1509                 );
1510
1511                 // Server independent link to posts and comments
1512                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1513                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1514                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1515
1516                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1517                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1518
1519                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1520                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1521
1522                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1523                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1524                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1525
1526                 // Red compatibility, though the link can't be authenticated on Friendica
1527                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1528
1529
1530                 // we may need to restrict this further if it picks up too many strays
1531                 // link acct:user@host to a webfinger profile redirector
1532
1533                 $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);
1534
1535                 // Perform MAIL Search
1536                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1537                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1538
1539                 // leave open the posibility of [map=something]
1540                 // this is replaced in prepare_body() which has knowledge of the item location
1541
1542                 if (strpos($text, '[/map]') !== false) {
1543                         $text = preg_replace_callback(
1544                                 "/\[map\](.*?)\[\/map\]/ism",
1545                                 function ($match) use ($simple_html) {
1546                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1547                                 },
1548                                 $text
1549                         );
1550                 }
1551                 if (strpos($text, '[map=') !== false) {
1552                         $text = preg_replace_callback(
1553                                 "/\[map=(.*?)\]/ism",
1554                                 function ($match) use ($simple_html) {
1555                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1556                                 },
1557                                 $text
1558                         );
1559                 }
1560                 if (strpos($text, '[map]') !== false) {
1561                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1562                 }
1563
1564                 // Check for headers
1565                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1566                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1567                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1568                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1569                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1570                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1571
1572                 // Check for paragraph
1573                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1574
1575                 // Check for bold text
1576                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1577
1578                 // Check for Italics text
1579                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1580
1581                 // Check for Underline text
1582                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1583
1584                 // Check for strike-through text
1585                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1586
1587                 // Check for over-line text
1588                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1589
1590                 // Check for colored text
1591                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1592
1593                 // Check for sized text
1594                 // [size=50] --> font-size: 50px (with the unit).
1595                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1596                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1597
1598                 // Check for centered text
1599                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1600
1601                 // Check for list text
1602                 $text = str_replace("[*]", "<li>", $text);
1603
1604                 // Check for style sheet commands
1605                 $text = preg_replace_callback(
1606                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1607                         function ($match) {
1608                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1609                         },
1610                         $text
1611                 );
1612
1613                 // Check for CSS classes
1614                 $text = preg_replace_callback(
1615                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1616                         function ($match) {
1617                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1618                         },
1619                         $text
1620                 );
1621
1622                 // handle nested lists
1623                 $endlessloop = 0;
1624
1625                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1626                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1627                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1628                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1629                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1630                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1631                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1632                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1633                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1634                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1635                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1636                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1637                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1638                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1639                 }
1640
1641                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1642                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1643                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1644                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1645
1646                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1647                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1648
1649                 $text = str_replace('[hr]', '<hr />', $text);
1650
1651                 // This is actually executed in prepare_body()
1652
1653                 $text = str_replace('[nosmile]', '', $text);
1654
1655                 // Check for font change text
1656                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1657
1658                 // Declare the format for [code] layout
1659
1660                 $CodeLayout = '<code>$1</code>';
1661                 // Check for [code] text
1662                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1663
1664                 // Declare the format for [spoiler] layout
1665                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1666
1667                 // Check for [spoiler] text
1668                 // handle nested quotes
1669                 $endlessloop = 0;
1670                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1671                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1672                 }
1673
1674                 // Check for [spoiler=Author] text
1675
1676                 $t_wrote = L10n::t('$1 wrote:');
1677
1678                 // handle nested quotes
1679                 $endlessloop = 0;
1680                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1681                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1682                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1683                                                  $text);
1684                 }
1685
1686                 // Declare the format for [quote] layout
1687                 $QuoteLayout = '<blockquote>$1</blockquote>';
1688
1689                 // Check for [quote] text
1690                 // handle nested quotes
1691                 $endlessloop = 0;
1692                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1693                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1694                 }
1695
1696                 // Check for [quote=Author] text
1697
1698                 $t_wrote = L10n::t('$1 wrote:');
1699
1700                 // handle nested quotes
1701                 $endlessloop = 0;
1702                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1703                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1704                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1705                                                  $text);
1706                 }
1707
1708
1709                 // [img=widthxheight]image source[/img]
1710                 $text = preg_replace_callback(
1711                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1712                         function ($matches) {
1713                                 if (strpos($matches[3], "data:image/") === 0) {
1714                                         return $matches[0];
1715                                 }
1716
1717                                 $matches[3] = proxy_url($matches[3]);
1718                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1719                         },
1720                         $text
1721                 );
1722
1723                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1724                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1725
1726                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1727                         function ($matches) {
1728                                 $matches[1] = proxy_url($matches[1]);
1729                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1730                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1731                         },
1732                         $text);
1733
1734                 // Images
1735                 // [img]pathtoimage[/img]
1736                 $text = preg_replace_callback(
1737                         "/\[img\](.*?)\[\/img\]/ism",
1738                         function ($matches) {
1739                                 if (strpos($matches[1], "data:image/") === 0) {
1740                                         return $matches[0];
1741                                 }
1742
1743                                 $matches[1] = proxy_url($matches[1]);
1744                                 return "[img]" . $matches[1] . "[/img]";
1745                         },
1746                         $text
1747                 );
1748
1749                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1750                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1751
1752                 // Shared content
1753                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1754                         function ($match) use ($simple_html) {
1755                                 return self::convertShare($match, $simple_html);
1756                         }, $text);
1757
1758                 $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);
1759                 $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 />', $text);
1760                 //$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 />', $Text);
1761
1762                 // Try to Oembed
1763                 if ($try_oembed) {
1764                         $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);
1765                         $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);
1766
1767                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1768                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1769                 } else {
1770                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1771                                                 '<a href="$1" target="_blank">$1</a>', $text);
1772                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1773                                                 '<a href="$1" target="_blank">$1</a>', $text);
1774                 }
1775
1776                 // html5 video and audio
1777
1778
1779                 if ($try_oembed) {
1780                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1781                 } else {
1782                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1783                 }
1784
1785                 // Youtube extensions
1786                 if ($try_oembed) {
1787                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1788                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1789                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1790                 }
1791
1792                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1793                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1794                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1795
1796                 if ($try_oembed) {
1797                         $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);
1798                 } else {
1799                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1800                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1801                 }
1802
1803                 if ($try_oembed) {
1804                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1805                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1806                 }
1807
1808                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1809                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1810
1811                 if ($try_oembed) {
1812                         $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);
1813                 } else {
1814                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1815                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1816                 }
1817
1818                 // oembed tag
1819                 $text = OEmbed::BBCode2HTML($text);
1820
1821                 // Avoid triple linefeeds through oembed
1822                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1823
1824                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1825                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1826                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1827                 // start which is always required). Allow desc with a missing summary for compatibility.
1828
1829                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1830                         $sub = Event::getHTML($ev, $simple_html);
1831
1832                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1833                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1834                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1835                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1836                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1837                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1838                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1839                 }
1840
1841                 // Replace non graphical smilies for external posts
1842                 if ($simple_html) {
1843                         $text = Smilies::replace($text, false, true);
1844                 }
1845
1846                 // Replace inline code blocks
1847                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1848                         function ($match) use ($simple_html) {
1849                                 $return = '<key>' . $match[1] . '</key>';
1850                                 // Use <code> for Diaspora inline code blocks
1851                                 if ($simple_html === 3) {
1852                                         $return = '<code>' . $match[1] . '</code>';
1853                                 }
1854                                 return $return;
1855                         }
1856                 , $text);
1857
1858                 // Unhide all [noparse] contained bbtags unspacefying them
1859                 // and triming the [noparse] tag.
1860
1861                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1862                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1863                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1864
1865                 /// @todo What is the meaning of these lines?
1866                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1867                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1868
1869                 // Currently deactivated, it made problems with " inside of alt texts.
1870                 //$text = preg_replace('/\&quot\;/', '"', $text);
1871
1872                 // fix any escaped ampersands that may have been converted into links
1873                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1874
1875                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1876                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1877                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1878                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1879
1880                 // sanitize href attributes (only whitelisted protocols URLs)
1881                 // default value for backward compatibility
1882                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1883
1884                 // Always allowed protocol even if config isn't set or not including it
1885                 $allowed_link_protocols[] = 'http';
1886                 $allowed_link_protocols[] = 'redir/';
1887
1888                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1889                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1890
1891                 if ($saved_image) {
1892                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1893                 }
1894
1895                 // Clean up the HTML by loading and saving the HTML with the DOM.
1896                 // Bad structured html can break a whole page.
1897                 // For performance reasons do it only with ativated item cache or at export.
1898                 if (!$try_oembed || (get_itemcachepath() != "")) {
1899                         $doc = new DOMDocument();
1900                         $doc->preserveWhiteSpace = false;
1901
1902                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1903
1904                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1905                         $encoding = '<?xml encoding="UTF-8">';
1906                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1907                         $doc->encoding = 'UTF-8';
1908                         $text = $doc->saveHTML();
1909                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1910
1911                         $text = str_replace('<br></li>', '</li>', $text);
1912
1913                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1914                 }
1915
1916                 // Clean up some useless linebreaks in lists
1917                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1918                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1919                 //$Text = str_replace('</li><br />', '</li>', $Text);
1920                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1921                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1922
1923                 Addon::callHooks('bbcode', $text);
1924
1925                 return trim($text);
1926         }
1927
1928         /**
1929          * @brief Strips the "abstract" tag from the provided text
1930          *
1931          * @param string $text The text with BBCode
1932          * @return string The same text - but without "abstract" element
1933          */
1934         public static function stripAbstract($text)
1935         {
1936                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1937                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1938
1939                 return $text;
1940         }
1941
1942         /**
1943          * @brief Returns the value of the "abstract" element
1944          *
1945          * @param string $text The text that maybe contains the element
1946          * @param string $addon The addon for which the abstract is meant for
1947          * @return string The abstract
1948          */
1949         private static function getAbstract($text, $addon = "")
1950         {
1951                 $abstract = "";
1952                 $abstracts = [];
1953                 $addon = strtolower($addon);
1954
1955                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1956                         foreach ($results AS $result) {
1957                                 $abstracts[strtolower($result[1])] = $result[2];
1958                         }
1959                 }
1960
1961                 if (isset($abstracts[$addon])) {
1962                         $abstract = $abstracts[$addon];
1963                 }
1964
1965                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1966                         $abstract = $result[1];
1967                 }
1968
1969                 return $abstract;
1970         }
1971
1972         /**
1973          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1974          *
1975          * @param array $match Matching values for the callback
1976          * @return string Replaced mention
1977          */
1978         private static function bbCodeMention2DiasporaCallback($match)
1979         {
1980                 $contact = Contact::getDetailsByURL($match[3]);
1981
1982                 if (empty($contact['addr'])) {
1983                         $contact = Probe::uri($match[3]);
1984                 }
1985
1986                 if (empty($contact['addr'])) {
1987                         return $match[0];
1988                 }
1989
1990                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1991                 return $mention;
1992         }
1993
1994         /**
1995          * @brief Converts a BBCode text into Markdown
1996          *
1997          * This function converts a BBCode item body to be sent to Markdown-enabled
1998          * systems like Diaspora and Libertree
1999          *
2000          * @param string $text
2001          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2002          * @return string
2003          */
2004         public static function toMarkdown($text, $for_diaspora = true)
2005         {
2006                 $a = self::getApp();
2007
2008                 $original_text = $text;
2009
2010                 // Since Diaspora is creating a summary for links, this function removes them before posting
2011                 if ($for_diaspora) {
2012                         $text = self::removeShareInformation($text);
2013                 }
2014
2015                 /**
2016                  * Transform #tags, strip off the [url] and replace spaces with underscore
2017                  */
2018                 $url_search_string = "^\[\]";
2019                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2020                         function ($matches) {
2021                                 return '#' . str_replace(' ', '_', $matches[2]);
2022                         },
2023                         $text
2024                 );
2025
2026                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2027                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2028
2029                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2030                 $codeblocks = [];
2031
2032                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2033                         function ($matches) use (&$codeblocks) {
2034                                 $return = $matches[0];
2035                                 if (strpos($matches[2], "\n") !== false) {
2036                                         $return = '#codeblock-' . count($codeblocks) . '#';
2037
2038                                         $prefix = '````' . $matches[1] . PHP_EOL;
2039                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2040                                 }
2041                                 return $return;
2042                         },
2043                         $text
2044                 );
2045
2046                 // Convert it to HTML - don't try oembed
2047                 if ($for_diaspora) {
2048                         $text = self::convert($text, false, 3);
2049
2050                         // Add all tags that maybe were removed
2051                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2052                                 $tagline = "";
2053                                 foreach ($tags[2] as $tag) {
2054                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2055                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2056                                                 $tagline .= '#' . $tag . ' ';
2057                                         }
2058                                 }
2059                                 $text = $text . " " . $tagline;
2060                         }
2061                 } else {
2062                         $text = self::convert($text, false, 4);
2063                 }
2064
2065                 // mask some special HTML chars from conversation to markdown
2066                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2067
2068                 // If a link is followed by a quote then there should be a newline before it
2069                 // Maybe we should make this newline at every time before a quote.
2070                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2071
2072                 $stamp1 = microtime(true);
2073
2074                 // Now convert HTML to Markdown
2075                 $converter = new HtmlConverter();
2076                 $text = $converter->convert($text);
2077
2078                 // unmask the special chars back to HTML
2079                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2080
2081                 $a->save_timestamp($stamp1, "parser");
2082
2083                 // Libertree has a problem with escaped hashtags.
2084                 $text = str_replace(['\#'], ['#'], $text);
2085
2086                 // Remove any leading or trailing whitespace, as this will mess up
2087                 // the Diaspora signature verification and cause the item to disappear
2088                 $text = trim($text);
2089
2090                 if ($for_diaspora) {
2091                         $url_search_string = "^\[\]";
2092                         $text = preg_replace_callback(
2093                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2094                                 ['self', 'bbCodeMention2DiasporaCallback'],
2095                                 $text
2096                         );
2097                 }
2098
2099                 // Restore code blocks
2100                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2101                         function ($matches) use ($codeblocks) {
2102                                 $return = '';
2103                                 if (isset($codeblocks[intval($matches[1])])) {
2104                                         $return = $codeblocks[$matches[1]];
2105                                 }
2106                                 return $return;
2107                         },
2108                         $text
2109                 );
2110
2111                 Addon::callHooks('bb2diaspora', $text);
2112
2113                 return $text;
2114         }
2115 }