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\Network\Probe;
  24 use Friendica\Object\Image;
  25 use Friendica\Util\Map;
  26 use Friendica\Util\Network;
  27 use Friendica\Util\ParseUrl;
  28 use League\HTMLToMarkdown\HtmlConverter;
  29
  30 require_once "include/event.php";
  31 require_once "include/html2plain.php";
  32 require_once "mod/proxy.php";
  33
  34 class BBCode extends BaseObject
  35 {
  36         /**
  37          * @brief Fetches attachment data that were generated the old way
  38          *
  39          * @param string $body Message body
  40          * @return array
  41          * 'type' -> Message type ("link", "video", "photo")
  42          * 'text' -> Text before the shared message
  43          * 'after' -> Text after the shared message
  44          * 'image' -> Preview image of the message
  45          * 'url' -> Url to the attached message
  46          * 'title' -> Title of the attachment
  47          * 'description' -> Description of the attachment
  48          */
  49         private static function getOldAttachmentData($body)
  50         {
  51                 $post = [];
  52
  53                 // Simplify image codes
  54                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  55
  56                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  57                         foreach ($attached as $data) {
  58                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  59                                         continue;
  60                                 }
  61
  62                                 $post["type"] = substr($data[1], 5);
  63
  64                                 $pos = strpos($body, $data[0]);
  65                                 if ($pos > 0) {
  66                                         $post["text"] = trim(substr($body, 0, $pos));
  67                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  68                                 } else {
  69                                         $post["text"] = trim(str_replace($data[0], "", $body));
  70                                 }
  71
  72                                 $attacheddata = $data[2];
  73
  74                                 $URLSearchString = "^\[\]";
  75
  76                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  77
  78                                         $picturedata = Image::getInfoFromURL($matches[1]);
  79
  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                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  88                                         $post["url"] = $matches[1];
  89                                         $post["title"] = $matches[2];
  90                                 }
  91                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  92                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  93                                         $post["url"] = $matches[1];
  94                                 }
  95
  96                                 // Search for description
  97                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  98                                         $post["description"] = $matches[1];
  99                                 }
 100                         }
 101                 }
 102                 return $post;
 103         }
 104
 105         /**
 106          * @brief Fetches attachment data that were generated with the "attachment" element
 107          *
 108          * @param string $body Message body
 109          * @return array
 110          * 'type' -> Message type ("link", "video", "photo")
 111          * 'text' -> Text before the shared message
 112          * 'after' -> Text after the shared message
 113          * 'image' -> Preview image of the message
 114          * 'url' -> Url to the attached message
 115          * 'title' -> Title of the attachment
 116          * 'description' -> Description of the attachment
 117          */
 118         public static function getAttachmentData($body)
 119         {
 120                 $data = [];
 121
 122                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 123                         return self::getOldAttachmentData($body);
 124                 }
 125
 126                 $attributes = $match[2];
 127
 128                 $data["text"] = trim($match[1]);
 129
 130                 $type = "";
 131                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 132                 if (x($matches, 1)) {
 133                         $type = strtolower($matches[1]);
 134                 }
 135
 136                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 137                 if (x($matches, 1)) {
 138                         $type = strtolower($matches[1]);
 139                 }
 140
 141                 if ($type == "") {
 142                         return [];
 143                 }
 144
 145                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 146                         return [];
 147                 }
 148
 149                 if ($type != "") {
 150                         $data["type"] = $type;
 151                 }
 152
 153                 $url = "";
 154                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 155                 if (x($matches, 1)) {
 156                         $url = $matches[1];
 157                 }
 158
 159                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 160                 if (x($matches, 1)) {
 161                         $url = $matches[1];
 162                 }
 163
 164                 if ($url != "") {
 165                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 166                 }
 167
 168                 $title = "";
 169                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 170                 if (x($matches, 1)) {
 171                         $title = $matches[1];
 172                 }
 173
 174                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 175                 if (x($matches, 1)) {
 176                         $title = $matches[1];
 177                 }
 178
 179                 if ($title != "") {
 180                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 181                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 182                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 183                         $data["title"] = $title;
 184                 }
 185
 186                 $image = "";
 187                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 188                 if (x($matches, 1)) {
 189                         $image = $matches[1];
 190                 }
 191
 192                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 193                 if (x($matches, 1)) {
 194                         $image = $matches[1];
 195                 }
 196
 197                 if ($image != "") {
 198                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 199                 }
 200
 201                 $preview = "";
 202                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 203                 if (x($matches, 1)) {
 204                         $preview = $matches[1];
 205                 }
 206
 207                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 208                 if (x($matches, 1)) {
 209                         $preview = $matches[1];
 210                 }
 211
 212                 if ($preview != "") {
 213                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 214                 }
 215
 216                 $data["description"] = trim($match[3]);
 217
 218                 $data["after"] = trim($match[4]);
 219
 220                 return $data;
 221         }
 222
 223         public static function getAttachedData($body, $item = [])
 224         {
 225                 /*
 226                 - text:
 227                 - type: link, video, photo
 228                 - title:
 229                 - url:
 230                 - image:
 231                 - description:
 232                 - (thumbnail)
 233                 */
 234
 235                 $has_title = !empty($item['title']);
 236                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 237                 $post = self::getAttachmentData($body);
 238
 239                 // if nothing is found, it maybe having an image.
 240                 if (!isset($post["type"])) {
 241                         // Simplify image codes
 242                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 243
 244                         $URLSearchString = "^\[\]";
 245                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 246                                 if ((count($pictures) == 1) && !$has_title) {
 247                                         // Checking, if the link goes to a picture
 248                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 249
 250                                         // Workaround:
 251                                         // Sometimes photo posts to the own album are not detected at the start.
 252                                         // So we seem to cannot use the cache for these cases. That's strange.
 253                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 254                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 255                                         }
 256
 257                                         if ($data["type"] == "photo") {
 258                                                 $post["type"] = "photo";
 259                                                 if (isset($data["images"][0])) {
 260                                                         $post["image"] = $data["images"][0]["src"];
 261                                                         $post["url"] = $data["url"];
 262                                                 } else {
 263                                                         $post["image"] = $data["url"];
 264                                                 }
 265
 266                                                 $post["preview"] = $pictures[0][2];
 267                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 268                                         } else {
 269                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 270                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 271                                                         $post["type"] = "photo";
 272                                                         $post["image"] = $pictures[0][1];
 273                                                         $post["preview"] = $pictures[0][2];
 274                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 275                                                 }
 276                                         }
 277                                 } elseif (count($pictures) > 0) {
 278                                         $post["type"] = "link";
 279                                         $post["url"] = $plink;
 280                                         $post["image"] = $pictures[0][2];
 281                                         $post["text"] = $body;
 282                                 }
 283                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 284                                 if ((count($pictures) == 1) && !$has_title) {
 285                                         $post["type"] = "photo";
 286                                         $post["image"] = $pictures[0][1];
 287                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 288                                 } elseif (count($pictures) > 0) {
 289                                         $post["type"] = "link";
 290                                         $post["url"] = $plink;
 291                                         $post["image"] = $pictures[0][1];
 292                                         $post["text"] = $body;
 293                                 }
 294                         }
 295
 296                         // Test for the external links
 297                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 298                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 299
 300                         $links = array_merge($links1, $links2);
 301
 302                         // If there is only a single one, then use it.
 303                         // This should cover link posts via API.
 304                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 305                                 $post["type"] = "link";
 306                                 $post["text"] = trim($body);
 307                                 $post["url"] = $links[0][1];
 308                         }
 309
 310                         // Now count the number of external media links
 311                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 312                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 313                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 314                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 315
 316                         // Add them to the other external links
 317                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 318
 319                         // Are there more than one?
 320                         if (count($links) > 1) {
 321                                 // The post will be the type "text", which means a blog post
 322                                 unset($post["type"]);
 323                                 $post["url"] = $plink;
 324                         }
 325
 326                         if (!isset($post["type"])) {
 327                                 $post["type"] = "text";
 328                                 $post["text"] = trim($body);
 329                         }
 330                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 331                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 332
 333                         if (isset($data["images"][0])) {
 334                                 $post["image"] = $data["images"][0]["src"];
 335                         }
 336                 }
 337
 338                 return $post;
 339         }
 340
 341         /**
 342          * @brief Convert a message into plaintext for connectors to other networks
 343          *
 344          * @param array $b The message array that is about to be posted
 345          * @param int $limit The maximum number of characters when posting to that network
 346          * @param bool $includedlinks Has an attached link to be included into the message?
 347          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 348          * @param string $target_network Name of the network where the post should go to.
 349          *
 350          * @return string The converted message
 351          */
 352         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 353         {
 354                 // Remove the hash tags
 355                 $URLSearchString = "^\[\]";
 356                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 357
 358                 // Add an URL element if the text contains a raw link
 359                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 360
 361                 // Remove the abstract
 362                 $body = self::stripAbstract($body);
 363
 364                 // At first look at data that is attached via "type-..." stuff
 365                 // This will hopefully replaced with a dedicated bbcode later
 366                 //$post = self::getAttachedData($b["body"]);
 367                 $post = self::getAttachedData($body, $b);
 368
 369                 if (($b["title"] != "") && ($post["text"] != "")) {
 370                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 371                 } elseif ($b["title"] != "") {
 372                         $post["text"] = trim($b["title"]);
 373                 }
 374
 375                 $abstract = "";
 376
 377                 // Fetch the abstract from the given target network
 378                 if ($target_network != "") {
 379                         $default_abstract = self::getAbstract($b["body"]);
 380                         $abstract = self::getAbstract($b["body"], $target_network);
 381
 382                         // If we post to a network with no limit we only fetch
 383                         // an abstract exactly for this network
 384                         if (($limit == 0) && ($abstract == $default_abstract)) {
 385                                 $abstract = "";
 386                         }
 387                 } else {// Try to guess the correct target network
 388                         switch ($htmlmode) {
 389                                 case 8:
 390                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
 391                                         break;
 392                                 case 7:
 393                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
 394                                         break;
 395                                 case 6:
 396                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
 397                                         break;
 398                                 default: // We don't know the exact target.
 399                                         // We fetch an abstract since there is a posting limit.
 400                                         if ($limit > 0) {
 401                                                 $abstract = self::getAbstract($b["body"]);
 402                                         }
 403                         }
 404                 }
 405
 406                 if ($abstract != "") {
 407                         $post["text"] = $abstract;
 408
 409                         if ($post["type"] == "text") {
 410                                 $post["type"] = "link";
 411                                 $post["url"] = $b["plink"];
 412                         }
 413                 }
 414
 415                 $html = self::convert($post["text"].$post["after"], false, $htmlmode);
 416                 $msg = html2plain($html, 0, true);
 417                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 418
 419                 $link = "";
 420                 if ($includedlinks) {
 421                         if ($post["type"] == "link") {
 422                                 $link = $post["url"];
 423                         } elseif ($post["type"] == "text") {
 424                                 $link = $post["url"];
 425                         } elseif ($post["type"] == "video") {
 426                                 $link = $post["url"];
 427                         } elseif ($post["type"] == "photo") {
 428                                 $link = $post["image"];
 429                         }
 430
 431                         if (($msg == "") && isset($post["title"])) {
 432                                 $msg = trim($post["title"]);
 433                         }
 434
 435                         if (($msg == "") && isset($post["description"])) {
 436                                 $msg = trim($post["description"]);
 437                         }
 438
 439                         // If the link is already contained in the post, then it neeedn't to be added again
 440                         // But: if the link is beyond the limit, then it has to be added.
 441                         if (($link != "") && strstr($msg, $link)) {
 442                                 $pos = strpos($msg, $link);
 443
 444                                 // Will the text be shortened in the link?
 445                                 // Or is the link the last item in the post?
 446                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 447                                         $msg = trim(str_replace($link, "", $msg));
 448                                 } elseif (($limit == 0) || ($pos < $limit)) {
 449                                         // The limit has to be increased since it will be shortened - but not now
 450                                         // Only do it with Twitter (htmlmode = 8)
 451                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 452                                                 $limit = $limit - 23 + strlen($link);
 453                                         }
 454
 455                                         $link = "";
 456
 457                                         if ($post["type"] == "text") {
 458                                                 unset($post["url"]);
 459                                         }
 460                                 }
 461                         }
 462                 }
 463
 464                 if ($limit > 0) {
 465                         // Reduce multiple spaces
 466                         // When posted to a network with limited space, we try to gain space where possible
 467                         while (strpos($msg, "  ") !== false) {
 468                                 $msg = str_replace("  ", " ", $msg);
 469                         }
 470
 471                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 472                         if (iconv_strlen($link, "UTF-8") > 0) {
 473                                 $limit = $limit - 23;
 474                         }
 475
 476                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 477                                 if (($post["type"] == "text") && isset($post["url"])) {
 478                                         $post["url"] = $b["plink"];
 479                                 } elseif (!isset($post["url"])) {
 480                                         $limit = $limit - 23;
 481                                         $post["url"] = $b["plink"];
 482                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 483                                 //} elseif (strpos($b["body"], "[share") !== false) {
 484                                 //      $post["url"] = $b["plink"];
 485                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 486                                         $post["url"] = $b["plink"];
 487                                 }
 488                                 $msg = Plaintext::shorten($msg, $limit);
 489                         }
 490                 }
 491
 492                 $post["text"] = trim($msg);
 493
 494                 return($post);
 495         }
 496
 497         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 498         {
 499                 // Suppress "view full size"
 500                 if (intval(Config::get('system', 'no_view_full_size'))) {
 501                         $include_link = false;
 502                 }
 503
 504                 // Picture addresses can contain special characters
 505                 $s = htmlspecialchars_decode($srctext);
 506
 507                 $matches = null;
 508                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 509                 if ($c) {
 510                         foreach ($matches as $mtch) {
 511                                 logger('scale_external_image: ' . $mtch[1]);
 512
 513                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 514                                 if (stristr($mtch[1], $hostname)) {
 515                                         continue;
 516                                 }
 517
 518                                 // $scale_replace, if passed, is an array of two elements. The
 519                                 // first is the name of the full-size image. The second is the
 520                                 // name of a remote, scaled-down version of the full size image.
 521                                 // This allows Friendica to display the smaller remote image if
 522                                 // one exists, while still linking to the full-size image
 523                                 if ($scale_replace) {
 524                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 525                                 } else {
 526                                         $scaled = $mtch[1];
 527                                 }
 528                                 $i = Network::fetchUrl($scaled);
 529                                 if (!$i) {
 530                                         return $srctext;
 531                                 }
 532
 533                                 // guess mimetype from headers or filename
 534                                 $type = Image::guessType($mtch[1], true);
 535
 536                                 if ($i) {
 537                                         $Image = new Image($i, $type);
 538                                         if ($Image->isValid()) {
 539                                                 $orig_width = $Image->getWidth();
 540                                                 $orig_height = $Image->getHeight();
 541
 542                                                 if ($orig_width > 640 || $orig_height > 640) {
 543                                                         $Image->scaleDown(640);
 544                                                         $new_width = $Image->getWidth();
 545                                                         $new_height = $Image->getHeight();
 546                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 547                                                         $s = str_replace(
 548                                                                 $mtch[0],
 549                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 550                                                                 . "\n" . (($include_link)
 551                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 552                                                                         : ''),
 553                                                                 $s
 554                                                         );
 555                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 556                                                 }
 557                                         }
 558                                 }
 559                         }
 560                 }
 561
 562                 // replace the special char encoding
 563                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 564                 return $s;
 565         }
 566
 567         /**
 568          * The purpose of this function is to apply system message length limits to
 569          * imported messages without including any embedded photos in the length
 570          *
 571          * @brief Truncates imported message body string length to max_import_size
 572          * @param string $body
 573          * @return string
 574          */
 575         public static function limitBodySize($body)
 576         {
 577                 $maxlen = get_max_import_size();
 578
 579                 // If the length of the body, including the embedded images, is smaller
 580                 // than the maximum, then don't waste time looking for the images
 581                 if ($maxlen && (strlen($body) > $maxlen)) {
 582
 583                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 584
 585                         $orig_body = $body;
 586                         $new_body = '';
 587                         $textlen = 0;
 588
 589                         $img_start = strpos($orig_body, '[img');
 590                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 591                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 592                         while (($img_st_close !== false) && ($img_end !== false)) {
 593
 594                                 $img_st_close++; // make it point to AFTER the closing bracket
 595                                 $img_end += $img_start;
 596                                 $img_end += strlen('[/img]');
 597
 598                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 599                                         // This is an embedded image
 600
 601                                         if (($textlen + $img_start) > $maxlen) {
 602                                                 if ($textlen < $maxlen) {
 603                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 604                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 605                                                         $textlen = $maxlen;
 606                                                 }
 607                                         } else {
 608                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 609                                                 $textlen += $img_start;
 610                                         }
 611
 612                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 613                                 } else {
 614
 615                                         if (($textlen + $img_end) > $maxlen) {
 616                                                 if ($textlen < $maxlen) {
 617                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 618                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 619                                                         $textlen = $maxlen;
 620                                                 }
 621                                         } else {
 622                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 623                                                 $textlen += $img_end;
 624                                         }
 625                                 }
 626                                 $orig_body = substr($orig_body, $img_end);
 627
 628                                 if ($orig_body === false) {
 629                                         // in case the body ends on a closing image tag
 630                                         $orig_body = '';
 631                                 }
 632
 633                                 $img_start = strpos($orig_body, '[img');
 634                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 635                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 636                         }
 637
 638                         if (($textlen + strlen($orig_body)) > $maxlen) {
 639                                 if ($textlen < $maxlen) {
 640                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 641                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 642                                 }
 643                         } else {
 644                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 645                                 $new_body = $new_body . $orig_body;
 646                         }
 647
 648                         return $new_body;
 649                 } else {
 650                         return $body;
 651                 }
 652         }
 653
 654         /**
 655          * Processes [attachment] tags
 656          *
 657          * Note: Can produce a [bookmark] tag in the returned string
 658          *
 659          * @brief Processes [attachment] tags
 660          * @param string $return
 661          * @param bool|int $simplehtml
 662          * @param bool $tryoembed
 663          * @return string
 664          */
 665         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 666         {
 667                 $data = self::getAttachmentData($return);
 668                 if (!$data) {
 669                         return $return;
 670                 }
 671
 672                 if (isset($data["title"])) {
 673                         $data["title"] = strip_tags($data["title"]);
 674                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 675                 }
 676
 677                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
 678                         $data["preview"] = $data["image"];
 679                         $data["image"] = "";
 680                 }
 681
 682                 $return = '';
 683                 if ($simplehtml == 7) {
 684                         $return = self::convertUrlForMastodon($data["url"]);
 685                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 686                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 687                 } else {
 688                         try {
 689                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 690                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 691                                 } else {
 692                                         throw new Exception('OEmbed is disabled for this attachment.');
 693                                 }
 694                         } catch (Exception $e) {
 695                                 if ($simplehtml != 4) {
 696                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 697                                 }
 698
 699                                 if ($data["image"] != "") {
 700                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
 701                                 } elseif ($data["preview"] != "") {
 702                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
 703                                 }
 704
 705                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
 706                                         $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"]);
 707                                 } else {
 708                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 709                                 }
 710
 711                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 712                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($data["description"])));
 713                                 }
 714
 715                                 if ($data["type"] == "link") {
 716                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 717                                 }
 718
 719                                 if ($simplehtml != 4) {
 720                                         $return .= '</div>';
 721                                 }
 722                         }
 723                 }
 724
 725                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 726         }
 727
 728         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 729         {
 730                 $data = self::getAttachmentData($Text);
 731
 732                 if (!$data) {
 733                         return $Text;
 734                 } elseif ($nolink) {
 735                         return $data["text"] . $data["after"];
 736                 }
 737
 738                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
 739                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 740                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 741                         $data["title"] = $data["url"];
 742                 } elseif (($text != "") && strstr($title, $text)) {
 743                         $data["text"] = $data["title"];
 744                         $data["title"] = $data["url"];
 745                 }
 746
 747                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 748                         return $data["title"] . $data["after"];
 749                 }
 750
 751                 // If the link already is included in the post, don't add it again
 752                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 753                         return $data["text"] . $data["after"];
 754                 }
 755
 756                 $text = $data["text"];
 757
 758                 if (($data["url"] != "") && ($data["title"] != "")) {
 759                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 760                 } elseif (($data["url"] != "")) {
 761                         $text .= "\n" . $data["url"];
 762                 }
 763
 764                 return $text . "\n" . $data["after"];
 765         }
 766
 767         private static function cleanCss($input)
 768         {
 769                 $cleaned = "";
 770
 771                 $input = strtolower($input);
 772
 773                 for ($i = 0; $i < strlen($input); $i++) {
 774                         $char = substr($input, $i, 1);
 775
 776                         if (($char >= "a") && ($char <= "z")) {
 777                                 $cleaned .= $char;
 778                         }
 779
 780                         if (!(strpos(" #;:0123456789-_.%", $char) === false)) {
 781                                 $cleaned .= $char;
 782                         }
 783                 }
 784
 785                 return $cleaned;
 786         }
 787
 788         /**
 789          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 790          *
 791          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 792          * @param array $match Array with the matching values
 793          * @return string reformatted link including HTML codes
 794          */
 795         private static function convertUrlForMastodonCallback($match)
 796         {
 797                 $url = $match[1];
 798
 799                 if (isset($match[2]) && ($match[1] != $match[2])) {
 800                         return $match[0];
 801                 }
 802
 803                 $parts = parse_url($url);
 804                 if (!isset($parts['scheme'])) {
 805                         return $match[0];
 806                 }
 807
 808                 return self::convertUrlForMastodon($url);
 809         }
 810
 811         /**
 812          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon and GNU Social.
 813          * @param string $url URL that is about to be reformatted
 814          * @return string reformatted link including HTML codes
 815          */
 816         private static function convertUrlForMastodon($url)
 817         {
 818                 $parts = parse_url($url);
 819                 $scheme = $parts['scheme'] . '://';
 820                 $styled_url = str_replace($scheme, '', $url);
 821
 822                 $html = '<a href="%s" class="attachment" rel="nofollow noopener" target="_blank">' .
 823                         '<span class="invisible">%s</span>';
 824
 825                 if (strlen($styled_url) > 30) {
 826                         $html .= '<span class="ellipsis">%s</span>' .
 827                                 '<span class="invisible">%s</span></a>';
 828
 829                         $ellipsis = substr($styled_url, 0, 30);
 830                         $rest = substr($styled_url, 30);
 831                         return sprintf($html, $url, $scheme, $ellipsis, $rest);
 832                 } else {
 833                         $html .= '%s</a>';
 834                         return sprintf($html, $url, $scheme, $styled_url);
 835                 }
 836         }
 837
 838         /*
 839          * [noparse][i]italic[/i][/noparse] turns into
 840          * [noparse][ i ]italic[ /i ][/noparse],
 841          * to hide them from parser.
 842          */
 843         private static function escapeNoparseCallback($match)
 844         {
 845                 $whole_match = $match[0];
 846                 $captured = $match[1];
 847                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 848                 $new_str = str_replace($captured, $spacefied, $whole_match);
 849                 return $new_str;
 850         }
 851
 852         /*
 853          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 854          * now turns back and the [noparse] tags are trimed
 855          * returning [i]italic[/i]
 856          */
 857         private static function unescapeNoparseCallback($match)
 858         {
 859                 $captured = $match[1];
 860                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 861                 return $unspacefied;
 862         }
 863
 864         /**
 865          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 866          * occurrences
 867          *
 868          * @param string $text        Text to search
 869          * @param string $name        Tag name
 870          * @param int    $occurrences Number of first occurrences to skip
 871          * @return boolean|array
 872          */
 873         public static function getTagPosition($text, $name, $occurrences = 0)
 874         {
 875                 if ($occurrences < 0) {
 876                         $occurrences = 0;
 877                 }
 878
 879                 $start_open = -1;
 880                 for ($i = 0; $i <= $occurrences; $i++) {
 881                         if ($start_open !== false) {
 882                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 883                         }
 884                 }
 885
 886                 if ($start_open === false) {
 887                         return false;
 888                 }
 889
 890                 $start_equal = strpos($text, '=', $start_open);
 891                 $start_close = strpos($text, ']', $start_open);
 892
 893                 if ($start_close === false) {
 894                         return false;
 895                 }
 896
 897                 $start_close++;
 898
 899                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 900
 901                 if ($end_open === false) {
 902                         return false;
 903                 }
 904
 905                 $res = [
 906                         'start' => [
 907                                 'open' => $start_open,
 908                                 'close' => $start_close
 909                         ],
 910                         'end' => [
 911                                 'open' => $end_open,
 912                                 'close' => $end_open + strlen('[/' . $name . ']')
 913                         ],
 914                 ];
 915
 916                 if ($start_equal !== false) {
 917                         $res['start']['equal'] = $start_equal + 1;
 918                 }
 919
 920                 return $res;
 921         }
 922
 923         /**
 924          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 925          *
 926          * @param type $pattern Preg pattern string
 927          * @param type $replace Preg replace string
 928          * @param type $name    BBCode tag name
 929          * @param type $text    Text to search
 930          * @return string
 931          */
 932         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 933         {
 934                 $occurrences = 0;
 935                 $pos = self::getTagPosition($text, $name, $occurrences);
 936                 while ($pos !== false && $occurrences++ < 1000) {
 937                         $start = substr($text, 0, $pos['start']['open']);
 938                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 939                         $end = substr($text, $pos['end']['close']);
 940                         if ($end === false) {
 941                                 $end = '';
 942                         }
 943
 944                         $subject = preg_replace($pattern, $replace, $subject);
 945                         $text = $start . $subject . $end;
 946
 947                         $pos = self::getTagPosition($text, $name, $occurrences);
 948                 }
 949
 950                 return $text;
 951         }
 952
 953         private static function extractImagesFromItemBody($body)
 954         {
 955                 $saved_image = [];
 956                 $orig_body = $body;
 957                 $new_body = '';
 958
 959                 $cnt = 0;
 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, $img_start), '[/img]') : false);
 963                 while (($img_st_close !== false) && ($img_end !== false)) {
 964                         $img_st_close++; // make it point to AFTER the closing bracket
 965                         $img_end += $img_start;
 966
 967                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 968                                 // This is an embedded image
 969                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 970                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 971
 972                                 $cnt++;
 973                         } else {
 974                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 975                         }
 976
 977                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 978
 979                         if ($orig_body === false) {
 980                                 // in case the body ends on a closing image tag
 981                                 $orig_body = '';
 982                         }
 983
 984                         $img_start = strpos($orig_body, '[img');
 985                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 986                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 987                 }
 988
 989                 $new_body = $new_body . $orig_body;
 990
 991                 return ['body' => $new_body, 'images' => $saved_image];
 992         }
 993
 994         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 995         {
 996                 $newbody = $body;
 997
 998                 $cnt = 0;
 999                 foreach ($images as $image) {
1000                         // We're depending on the property of 'foreach' (specified on the PHP website) that
1001                         // it loops over the array starting from the first element and going sequentially
1002                         // to the last element
1003                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1004                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
1005                         $cnt++;
1006                 }
1007
1008                 return $newbody;
1009         }
1010
1011         /**
1012          * Processes [share] tags
1013          *
1014          * Note: Can produce a [bookmark] tag in the output
1015          *
1016          * @brief Processes [share] tags
1017          * @param array    $share      preg_match_callback result array
1018          * @param bool|int $simplehtml
1019          * @return string
1020          */
1021         private static function convertShare($share, $simplehtml)
1022         {
1023                 $attributes = $share[2];
1024
1025                 $author = "";
1026                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1027                 if (x($matches, 1)) {
1028                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1029                 }
1030
1031                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1032                 if (x($matches, 1)) {
1033                         $author = $matches[1];
1034                 }
1035
1036                 $profile = "";
1037                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1038                 if (x($matches, 1)) {
1039                         $profile = $matches[1];
1040                 }
1041
1042                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1043                 if (x($matches, 1)) {
1044                         $profile = $matches[1];
1045                 }
1046
1047                 $avatar = "";
1048                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1049                 if (x($matches, 1)) {
1050                         $avatar = $matches[1];
1051                 }
1052
1053                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1054                 if (x($matches, 1)) {
1055                         $avatar = $matches[1];
1056                 }
1057
1058                 $link = "";
1059                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1060                 if (x($matches, 1)) {
1061                         $link = $matches[1];
1062                 }
1063
1064                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1065                 if (x($matches, 1)) {
1066                         $link = $matches[1];
1067                 }
1068
1069                 $posted = "";
1070
1071                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1072                 if (x($matches, 1)) {
1073                         $posted = $matches[1];
1074                 }
1075
1076                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1077                 if (x($matches, 1)) {
1078                         $posted = $matches[1];
1079                 }
1080
1081                 // We only call this so that a previously unknown contact can be added.
1082                 // This is important for the function "Model\Contact::getDetailsByURL()".
1083                 // This function then can fetch an entry from the contact table.
1084                 Contact::getIdForURL($profile, 0, true);
1085
1086                 $data = Contact::getDetailsByURL($profile);
1087
1088                 if (x($data, "name") && x($data, "addr")) {
1089                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1090                 } else {
1091                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1092                 }
1093
1094                 if (x($data, "addr")) {
1095                         $userid = $data["addr"];
1096                 } else {
1097                         $userid = Protocol::formatMention($profile, $author);
1098                 }
1099
1100                 if (x($data, "name")) {
1101                         $author = $data["name"];
1102                 }
1103
1104                 if (x($data, "micro")) {
1105                         $avatar = $data["micro"];
1106                 }
1107
1108                 $preshare = trim($share[1]);
1109                 if ($preshare != "") {
1110                         $preshare .= "<br />";
1111                 }
1112
1113                 switch ($simplehtml) {
1114                         case 1:
1115                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1116                                 break;
1117                         case 2:
1118                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1119                                 break;
1120                         case 3: // Diaspora
1121                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1122
1123                                 $text = trim($share[1]);
1124
1125                                 if ($text != "") {
1126                                         $text .= "<hr />";
1127                                 }
1128
1129                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1130                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1131
1132                                         if ($link != "") {
1133                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1134                                         }
1135                                 } else {
1136                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1137                                 }
1138
1139                                 break;
1140                         case 4:
1141                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1142                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1143                                 $headline .= ":</b><br />";
1144
1145                                 $text = trim($share[1]);
1146
1147                                 if ($text != "") {
1148                                         $text .= "<hr />";
1149                                 }
1150
1151                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1152
1153                                 break;
1154                         case 5:
1155                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1156                                 break;
1157                         case 6: // app.net
1158                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1159                                 break;
1160                         case 7: // statusnet/GNU Social
1161                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1162                                 break;
1163                         case 8: // twitter
1164                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1165                                 break;
1166                         case 9: // Google+/Facebook
1167                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1168
1169                                 if ($link != "") {
1170                                         $text .= "<br /><br />" . $link;
1171                                 }
1172                                 break;
1173                         default:
1174                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1175                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1176                                         try {
1177                                                 $oembed = OEmbed::getHTML($link, $preshare);
1178                                         } catch (Exception $e) {
1179                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1180                                         }
1181
1182                                         $text = $preshare . $oembed;
1183                                 } else {
1184                                         $text = trim($share[1]) . "\n";
1185
1186                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1187
1188                                         $tpl = get_markup_template('shared_content.tpl');
1189                                         $text .= replace_macros($tpl, [
1190                                                 '$profile' => $profile,
1191                                                 '$avatar' => $avatar,
1192                                                 '$author' => $author,
1193                                                 '$link' => $link,
1194                                                 '$posted' => $posted,
1195                                                 '$content' => trim($share[3])
1196                                         ]);
1197                                 }
1198                                 break;
1199                 }
1200
1201                 return $text;
1202         }
1203
1204         private static function removePictureLinksCallback($match)
1205         {
1206                 $text = Cache::get($match[1]);
1207
1208                 if (is_null($text)) {
1209                         $a = self::getApp();
1210
1211                         $stamp1 = microtime(true);
1212
1213                         $ch = @curl_init($match[1]);
1214                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1215                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1216                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1217                         @curl_exec($ch);
1218                         $curl_info = @curl_getinfo($ch);
1219
1220                         $a->save_timestamp($stamp1, "network");
1221
1222                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1223                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1224                         } else {
1225                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1226
1227                                 // if its not a picture then look if its a page that contains a picture link
1228                                 $body = Network::fetchUrl($match[1]);
1229
1230                                 $doc = new DOMDocument();
1231                                 @$doc->loadHTML($body);
1232                                 $xpath = new DomXPath($doc);
1233                                 $list = $xpath->query("//meta[@name]");
1234                                 foreach ($list as $node) {
1235                                         $attr = [];
1236
1237                                         if ($node->attributes->length) {
1238                                                 foreach ($node->attributes as $attribute) {
1239                                                         $attr[$attribute->name] = $attribute->value;
1240                                                 }
1241                                         }
1242
1243                                         if (strtolower($attr["name"]) == "twitter:image") {
1244                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1245                                         }
1246                                 }
1247                         }
1248                         Cache::set($match[1], $text);
1249                 }
1250
1251                 return $text;
1252         }
1253
1254         private static function expandLinksCallback($match)
1255         {
1256                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1257                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1258                 } else {
1259                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1260                 }
1261         }
1262
1263         private static function cleanPictureLinksCallback($match)
1264         {
1265                 $text = Cache::get($match[1]);
1266
1267                 if (is_null($text)) {
1268                         $a = self::getApp();
1269
1270                         $stamp1 = microtime(true);
1271
1272                         $ch = @curl_init($match[1]);
1273                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1274                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1275                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1276                         @curl_exec($ch);
1277                         $curl_info = @curl_getinfo($ch);
1278
1279                         $a->save_timestamp($stamp1, "network");
1280
1281                         // if its a link to a picture then embed this picture
1282                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1283                                 $text = "[img]" . $match[1] . "[/img]";
1284                         } else {
1285                                 $text = "[img]" . $match[2] . "[/img]";
1286
1287                                 // if its not a picture then look if its a page that contains a picture link
1288                                 $body = Network::fetchUrl($match[1]);
1289
1290                                 $doc = new DOMDocument();
1291                                 @$doc->loadHTML($body);
1292                                 $xpath = new DOMXPath($doc);
1293                                 $list = $xpath->query("//meta[@name]");
1294                                 foreach ($list as $node) {
1295                                         $attr = [];
1296                                         if ($node->attributes->length) {
1297                                                 foreach ($node->attributes as $attribute) {
1298                                                         $attr[$attribute->name] = $attribute->value;
1299                                                 }
1300                                         }
1301
1302                                         if (strtolower($attr["name"]) == "twitter:image") {
1303                                                 $text = "[img]" . $attr["content"] . "[/img]";
1304                                         }
1305                                 }
1306                         }
1307                         Cache::set($match[1], $text);
1308                 }
1309
1310                 return $text;
1311         }
1312
1313         public static function cleanPictureLinks($text)
1314         {
1315                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1316                 return $return;
1317         }
1318
1319         private static function textHighlightCallback($match)
1320         {
1321                 if (in_array(strtolower($match[1]),
1322                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1323                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh'])
1324                 ) {
1325                         return text_highlight($match[2], strtolower($match[1]));
1326                 }
1327                 return $match[0];
1328         }
1329
1330         /**
1331          * @brief Converts a BBCode message to HTML message
1332          *
1333          * BBcode 2 HTML was written by WAY2WEB.net
1334          * extended to work with Mistpark/Friendica - Mike Macgirvin
1335          *
1336          * Simple HTML values meaning:
1337          * - 0: Friendica display
1338          * - 1: Unused
1339          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1340          * - 3: Used before converting to Markdown in bb2diaspora.php
1341          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1342          * - 5: Unused
1343          * - 6: Used for Appnet
1344          * - 7: Used for dfrn, OStatus
1345          * - 8: Used for WP backlink text setting
1346          *
1347          * @param string $text
1348          * @param bool   $try_oembed
1349          * @param int    $simple_html
1350          * @param bool   $for_plaintext
1351          * @return string
1352          */
1353         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1354         {
1355                 $a = self::getApp();
1356
1357                 /*
1358                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1359                  *
1360                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1361                  * $match[1] = $url
1362                  * $match[2] = $title or absent
1363                  */
1364                 $try_oembed_callback = function ($match)
1365                 {
1366                         $url = $match[1];
1367                         $title = defaults($match, 2, null);
1368
1369                         try {
1370                                 $return = OEmbed::getHTML($url, $title);
1371                         } catch (Exception $ex) {
1372                                 $return = $match[0];
1373                         }
1374
1375                         return $return;
1376                 };
1377
1378                 // Hide all [noparse] contained bbtags by spacefying them
1379                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1380
1381                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1382                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1383                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1384
1385                 // Remove the abstract element. It is a non visible element.
1386                 $text = self::stripAbstract($text);
1387
1388                 // Move all spaces out of the tags
1389                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1390                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1391
1392                 // Extract the private images which use data urls since preg has issues with
1393                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1394                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1395
1396                 $extracted = self::extractImagesFromItemBody($text);
1397                 $text = $extracted['body'];
1398                 $saved_image = $extracted['images'];
1399
1400                 // If we find any event code, turn it into an event.
1401                 // After we're finished processing the bbcode we'll
1402                 // replace all of the event code with a reformatted version.
1403
1404                 $ev = bbtoevent($text);
1405
1406                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1407                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1408
1409                 $text = str_replace("<", "&lt;", $text);
1410                 $text = str_replace(">", "&gt;", $text);
1411
1412                 // remove some newlines before the general conversion
1413                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1414                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1415
1416                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1417                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1418
1419                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1420                 if (!$try_oembed) {
1421                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1422                 }
1423
1424                 // Check for [code] text here, before the linefeeds are messed with.
1425                 // The highlighter will unescape and re-escape the content.
1426                 if (strpos($text, '[code=') !== false) {
1427                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1428                 }
1429                 // Convert new line chars to html <br /> tags
1430
1431                 // nlbr seems to be hopelessly messed up
1432                 //      $Text = nl2br($Text);
1433
1434                 // We'll emulate it.
1435
1436                 $text = trim($text);
1437                 $text = str_replace("\r\n", "\n", $text);
1438
1439                 // removing multiplicated newlines
1440                 if (Config::get("system", "remove_multiplicated_lines")) {
1441                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1442                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1443                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1444                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1445                         do {
1446                                 $oldtext = $text;
1447                                 $text = str_replace($search, $replace, $text);
1448                         } while ($oldtext != $text);
1449                 }
1450
1451                 // Set up the parameters for a URL search string
1452                 $URLSearchString = "^\[\]";
1453                 // Set up the parameters for a MAIL search string
1454                 $MAILSearchString = $URLSearchString;
1455
1456                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1457                 if (!$for_plaintext) {
1458                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1459                         // Currently disabled, since the function is too greedy
1460                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1461                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1462                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1463                         if ($simple_html == 7) {
1464                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1465                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1466                         }
1467                 } else {
1468                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1469                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1470                 }
1471
1472
1473                 // Handle attached links or videos
1474                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1475
1476                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1477
1478                 // Remove all hashtag addresses
1479                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1480                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1481                 } elseif ($simple_html == 3) {
1482                         // The ! is converted to @ since Diaspora only understands the @
1483                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1484                                 '@<a href="$2">$3</a>',
1485                                 $text);
1486                 } elseif ($simple_html == 7) {
1487                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1488                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1489                                 $text);
1490                 } elseif (!$simple_html) {
1491                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1492                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1493                                 $text);
1494                 }
1495
1496                 // Bookmarks in red - will be converted to bookmarks in friendica
1497                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1498                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1499                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1500                                         "[bookmark=$1]$2[/bookmark]", $text);
1501
1502                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1503                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1504                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1505                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1506                 }
1507
1508                 if ($simple_html == 5) {
1509                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1510                 }
1511
1512                 // Perform URL Search
1513                 if ($try_oembed) {
1514                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1515                 }
1516
1517                 if ($simple_html == 5) {
1518                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1519                 } else {
1520                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1521                 }
1522
1523                 // Handle Diaspora posts
1524                 $text = preg_replace_callback(
1525                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1526                         function ($match) {
1527                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1528                         }, $text
1529                 );
1530
1531                 // Server independent link to posts and comments
1532                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1533                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1534                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1535
1536                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1537                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1538
1539                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1540                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1541
1542                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1543                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1544                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1545
1546                 // Red compatibility, though the link can't be authenticated on Friendica
1547                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1548
1549
1550                 // we may need to restrict this further if it picks up too many strays
1551                 // link acct:user@host to a webfinger profile redirector
1552
1553                 $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);
1554
1555                 // Perform MAIL Search
1556                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1557                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1558
1559                 // leave open the posibility of [map=something]
1560                 // this is replaced in prepare_body() which has knowledge of the item location
1561
1562                 if (strpos($text, '[/map]') !== false) {
1563                         $text = preg_replace_callback(
1564                                 "/\[map\](.*?)\[\/map\]/ism",
1565                                 function ($match) {
1566                                         // the extra space in the following line is intentional
1567                                         // Whyyy? - @MrPetovan
1568                                         return str_replace($match[0], '<div class="map"  >' . Map::byLocation($match[1]) . '</div>', $match[0]);
1569                                 },
1570                                 $text
1571                         );
1572                 }
1573                 if (strpos($text, '[map=') !== false) {
1574                         $text = preg_replace_callback(
1575                                 "/\[map=(.*?)\]/ism",
1576                                 function ($match) {
1577                                         // the extra space in the following line is intentional
1578                                         // Whyyy? - @MrPetovan
1579                                         return str_replace($match[0], '<div class="map"  >' . Map::byCoordinates(str_replace('/', ' ', $match[1])) . '</div>', $match[0]);
1580                                 },
1581                                 $text
1582                         );
1583                 }
1584                 if (strpos($text, '[map]') !== false) {
1585                         $text = preg_replace("/\[map\]/", '<div class="map"></div>', $text);
1586                 }
1587
1588                 // Check for headers
1589                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1590                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1591                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1592                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1593                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1594                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1595
1596                 // Check for paragraph
1597                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1598
1599                 // Check for bold text
1600                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1601
1602                 // Check for Italics text
1603                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1604
1605                 // Check for Underline text
1606                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1607
1608                 // Check for strike-through text
1609                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1610
1611                 // Check for over-line text
1612                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1613
1614                 // Check for colored text
1615                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1616
1617                 // Check for sized text
1618                 // [size=50] --> font-size: 50px (with the unit).
1619                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1620                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1621
1622                 // Check for centered text
1623                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1624
1625                 // Check for list text
1626                 $text = str_replace("[*]", "<li>", $text);
1627
1628                 // Check for style sheet commands
1629                 $text = preg_replace_callback(
1630                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1631                         function ($match) {
1632                                 return "<span style=\"" . self::cleanCss($match[1]) . ";\">" . $match[2] . "</span>";
1633                         },
1634                         $text
1635                 );
1636
1637                 // Check for CSS classes
1638                 $text = preg_replace_callback(
1639                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1640                         function ($match) {
1641                                 return "<span class=\"" . self::cleanCss($match[1]) . "\">" . $match[2] . "</span>";
1642                         },
1643                         $text
1644                 );
1645
1646                 // handle nested lists
1647                 $endlessloop = 0;
1648
1649                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1650                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1651                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1652                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1653                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1654                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1655                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1656                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1657                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1658                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1659                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1660                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1661                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1662                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1663                 }
1664
1665                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1666                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1667                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1668                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1669
1670                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1671                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1672
1673                 $text = str_replace('[hr]', '<hr />', $text);
1674
1675                 // This is actually executed in prepare_body()
1676
1677                 $text = str_replace('[nosmile]', '', $text);
1678
1679                 // Check for font change text
1680                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1681
1682                 // Declare the format for [code] layout
1683
1684                 $CodeLayout = '<code>$1</code>';
1685                 // Check for [code] text
1686                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1687
1688                 // Declare the format for [spoiler] layout
1689                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1690
1691                 // Check for [spoiler] text
1692                 // handle nested quotes
1693                 $endlessloop = 0;
1694                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1695                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1696                 }
1697
1698                 // Check for [spoiler=Author] text
1699
1700                 $t_wrote = L10n::t('$1 wrote:');
1701
1702                 // handle nested quotes
1703                 $endlessloop = 0;
1704                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1705                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1706                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1707                                                  $text);
1708                 }
1709
1710                 // Declare the format for [quote] layout
1711                 $QuoteLayout = '<blockquote>$1</blockquote>';
1712
1713                 // Check for [quote] text
1714                 // handle nested quotes
1715                 $endlessloop = 0;
1716                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1717                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1718                 }
1719
1720                 // Check for [quote=Author] text
1721
1722                 $t_wrote = L10n::t('$1 wrote:');
1723
1724                 // handle nested quotes
1725                 $endlessloop = 0;
1726                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1727                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1728                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1729                                                  $text);
1730                 }
1731
1732
1733                 // [img=widthxheight]image source[/img]
1734                 $text = preg_replace_callback(
1735                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1736                         function ($matches) {
1737                                 if (strpos($matches[3], "data:image/") === 0) {
1738                                         return $matches[0];
1739                                 }
1740
1741                                 $matches[3] = proxy_url($matches[3]);
1742                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1743                         },
1744                         $text
1745                 );
1746
1747                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1748                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1749
1750                 // Images
1751                 // [img]pathtoimage[/img]
1752                 $text = preg_replace_callback(
1753                         "/\[img\](.*?)\[\/img\]/ism",
1754                         function ($matches) {
1755                                 if (strpos($matches[1], "data:image/") === 0) {
1756                                         return $matches[0];
1757                                 }
1758
1759                                 $matches[1] = proxy_url($matches[1]);
1760                                 return "[img]" . $matches[1] . "[/img]";
1761                         },
1762                         $text
1763                 );
1764
1765                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1766                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1767
1768                 // Shared content
1769                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1770                         function ($match) use ($simple_html) {
1771                                 return self::convertShare($match, $simple_html);
1772                         }, $text);
1773
1774                 $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);
1775                 $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 />', $text);
1776                 //$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 />', $Text);
1777
1778                 // Try to Oembed
1779                 if ($try_oembed) {
1780                         $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);
1781                         $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);
1782
1783                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1784                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1785                 } else {
1786                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1787                                                 '<a href="$1" target="_blank">$1</a>', $text);
1788                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1789                                                 '<a href="$1" target="_blank">$1</a>', $text);
1790                 }
1791
1792                 // html5 video and audio
1793
1794
1795                 if ($try_oembed) {
1796                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1797                 } else {
1798                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1799                 }
1800
1801                 // Youtube extensions
1802                 if ($try_oembed) {
1803                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1804                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1805                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1806                 }
1807
1808                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1809                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1810                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1811
1812                 if ($try_oembed) {
1813                         $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);
1814                 } else {
1815                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1816                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1817                 }
1818
1819                 if ($try_oembed) {
1820                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1821                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1822                 }
1823
1824                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1825                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1826
1827                 if ($try_oembed) {
1828                         $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);
1829                 } else {
1830                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1831                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1832                 }
1833
1834                 // oembed tag
1835                 $text = OEmbed::BBCode2HTML($text);
1836
1837                 // Avoid triple linefeeds through oembed
1838                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1839
1840                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1841                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1842                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1843                 // start which is always required). Allow desc with a missing summary for compatibility.
1844
1845                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1846                         $sub = format_event_html($ev, $simple_html);
1847
1848                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1849                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1850                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1851                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1852                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1853                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1854                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1855                 }
1856
1857                 // Replace non graphical smilies for external posts
1858                 if ($simple_html) {
1859                         $text = Smilies::replace($text, false, true);
1860                 }
1861
1862                 // Replace inline code blocks
1863                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1864                         function ($match) use ($simple_html) {
1865                                 $return = '<key>' . $match[1] . '</key>';
1866                                 // Use <code> for Diaspora inline code blocks
1867                                 if ($simple_html === 3) {
1868                                         $return = '<code>' . $match[1] . '</code>';
1869                                 }
1870                                 return $return;
1871                         }
1872                 , $text);
1873
1874                 // Unhide all [noparse] contained bbtags unspacefying them
1875                 // and triming the [noparse] tag.
1876
1877                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1878                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1879                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1880
1881
1882                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1883                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1884                 $text = preg_replace('/\&quot\;/', '"', $text);
1885
1886                 // fix any escaped ampersands that may have been converted into links
1887                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1888
1889                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1890                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1891                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1892                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1893
1894                 // sanitize href attributes (only whitelisted protocols URLs)
1895                 // default value for backward compatibility
1896                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1897
1898                 // Always allowed protocol even if config isn't set or not including it
1899                 $allowed_link_protocols[] = 'http';
1900                 $allowed_link_protocols[] = 'redir/';
1901
1902                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1903                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1904
1905                 if ($saved_image) {
1906                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1907                 }
1908
1909                 // Clean up the HTML by loading and saving the HTML with the DOM.
1910                 // Bad structured html can break a whole page.
1911                 // For performance reasons do it only with ativated item cache or at export.
1912                 if (!$try_oembed || (get_itemcachepath() != "")) {
1913                         $doc = new DOMDocument();
1914                         $doc->preserveWhiteSpace = false;
1915
1916                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1917
1918                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1919                         $encoding = '<?xml encoding="UTF-8">';
1920                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1921                         $doc->encoding = 'UTF-8';
1922                         $text = $doc->saveHTML();
1923                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1924
1925                         $text = str_replace('<br></li>', '</li>', $text);
1926
1927                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1928                 }
1929
1930                 // Clean up some useless linebreaks in lists
1931                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1932                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1933                 //$Text = str_replace('</li><br />', '</li>', $Text);
1934                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1935                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1936
1937                 Addon::callHooks('bbcode', $text);
1938
1939                 return trim($text);
1940         }
1941
1942         /**
1943          * @brief Strips the "abstract" tag from the provided text
1944          *
1945          * @param string $text The text with BBCode
1946          * @return string The same text - but without "abstract" element
1947          */
1948         public static function stripAbstract($text)
1949         {
1950                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1951                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1952
1953                 return $text;
1954         }
1955
1956         /**
1957          * @brief Returns the value of the "abstract" element
1958          *
1959          * @param string $text The text that maybe contains the element
1960          * @param string $addon The addon for which the abstract is meant for
1961          * @return string The abstract
1962          */
1963         private static function getAbstract($text, $addon = "")
1964         {
1965                 $abstract = "";
1966                 $abstracts = [];
1967                 $addon = strtolower($addon);
1968
1969                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1970                         foreach ($results AS $result) {
1971                                 $abstracts[strtolower($result[1])] = $result[2];
1972                         }
1973                 }
1974
1975                 if (isset($abstracts[$addon])) {
1976                         $abstract = $abstracts[$addon];
1977                 }
1978
1979                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1980                         $abstract = $result[1];
1981                 }
1982
1983                 return $abstract;
1984         }
1985
1986         /**
1987          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1988          *
1989          * @param array $match Matching values for the callback
1990          * @return string Replaced mention
1991          */
1992         private static function bbCodeMention2DiasporaCallback($match)
1993         {
1994                 $contact = Contact::getDetailsByURL($match[3]);
1995
1996                 if (empty($contact['addr'])) {
1997                         $contact = Probe::uri($match[3]);
1998                 }
1999
2000                 if (empty($contact['addr'])) {
2001                         return $match[0];
2002                 }
2003
2004                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
2005                 return $mention;
2006         }
2007
2008         /**
2009          * @brief Converts a BBCode text into Markdown
2010          *
2011          * This function converts a BBCode item body to be sent to Markdown-enabled
2012          * systems like Diaspora and Libertree
2013          *
2014          * @param string $text
2015          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2016          * @return string
2017          */
2018         public static function toMarkdown($text, $for_diaspora = true)
2019         {
2020                 $a = self::getApp();
2021
2022                 $original_text = $text;
2023
2024                 // Since Diaspora is creating a summary for links, this function removes them before posting
2025                 if ($for_diaspora) {
2026                         $text = self::removeShareInformation($text);
2027                 }
2028
2029                 /**
2030                  * Transform #tags, strip off the [url] and replace spaces with underscore
2031                  */
2032                 $url_search_string = "^\[\]";
2033                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2034                         function ($matches) {
2035                                 return '#' . str_replace(' ', '_', $matches[2]);
2036                         },
2037                         $text
2038                 );
2039
2040                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2041                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2042
2043                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2044                 $codeblocks = [];
2045
2046                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2047                         function ($matches) use (&$codeblocks) {
2048                                 $return = $matches[0];
2049                                 if (strpos($matches[2], "\n") !== false) {
2050                                         $return = '#codeblock-' . count($codeblocks) . '#';
2051
2052                                         $prefix = '````' . $matches[1] . PHP_EOL;
2053                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2054                                 }
2055                                 return $return;
2056                         },
2057                         $text
2058                 );
2059
2060                 // Convert it to HTML - don't try oembed
2061                 if ($for_diaspora) {
2062                         $text = self::convert($text, false, 3);
2063
2064                         // Add all tags that maybe were removed
2065                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2066                                 $tagline = "";
2067                                 foreach ($tags[2] as $tag) {
2068                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2069                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2070                                                 $tagline .= '#' . $tag . ' ';
2071                                         }
2072                                 }
2073                                 $text = $text . " " . $tagline;
2074                         }
2075                 } else {
2076                         $text = self::convert($text, false, 4);
2077                 }
2078
2079                 // mask some special HTML chars from conversation to markdown
2080                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2081
2082                 // If a link is followed by a quote then there should be a newline before it
2083                 // Maybe we should make this newline at every time before a quote.
2084                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2085
2086                 $stamp1 = microtime(true);
2087
2088                 // Now convert HTML to Markdown
2089                 $converter = new HtmlConverter();
2090                 $text = $converter->convert($text);
2091
2092                 // unmask the special chars back to HTML
2093                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2094
2095                 $a->save_timestamp($stamp1, "parser");
2096
2097                 // Libertree has a problem with escaped hashtags.
2098                 $text = str_replace(['\#'], ['#'], $text);
2099
2100                 // Remove any leading or trailing whitespace, as this will mess up
2101                 // the Diaspora signature verification and cause the item to disappear
2102                 $text = trim($text);
2103
2104                 if ($for_diaspora) {
2105                         $url_search_string = "^\[\]";
2106                         $text = preg_replace_callback(
2107                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2108                                 ['self', 'bbCodeMention2DiasporaCallback'],
2109                                 $text
2110                         );
2111                 }
2112
2113                 // Restore code blocks
2114                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2115                         function ($matches) use ($codeblocks) {
2116                                 $return = '';
2117                                 if (isset($codeblocks[intval($matches[1])])) {
2118                                         $return = $codeblocks[$matches[1]];
2119                                 }
2120                                 return $return;
2121                         },
2122                         $text
2123                 );
2124
2125                 Addon::callHooks('bb2diaspora', $text);
2126
2127                 return $text;
2128         }
2129 }