src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Addon;
  15 use Friendica\Core\Cache;
  16 use Friendica\Core\Config;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Logger;
  19 use Friendica\Core\Protocol;
  20 use Friendica\Core\Renderer;
  21 use Friendica\Core\System;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Network\Probe;
  25 use Friendica\Object\Image;
  26 use Friendica\Util\Map;
  27 use Friendica\Util\Network;
  28 use Friendica\Util\ParseUrl;
  29 use Friendica\Util\Proxy as ProxyUtils;
  30 use Friendica\Util\Strings;
  31
  32 class BBCode extends BaseObject
  33 {
  34         /**
  35          * @brief Fetches attachment data that were generated the old way
  36          *
  37          * @param string $body Message body
  38          * @return array
  39          * 'type' -> Message type ("link", "video", "photo")
  40          * 'text' -> Text before the shared message
  41          * 'after' -> Text after the shared message
  42          * 'image' -> Preview image of the message
  43          * 'url' -> Url to the attached message
  44          * 'title' -> Title of the attachment
  45          * 'description' -> Description of the attachment
  46          */
  47         private static function getOldAttachmentData($body)
  48         {
  49                 $post = [];
  50
  51                 // Simplify image codes
  52                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  53
  54                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  55                         foreach ($attached as $data) {
  56                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  57                                         continue;
  58                                 }
  59
  60                                 $post["type"] = substr($data[1], 5);
  61
  62                                 $pos = strpos($body, $data[0]);
  63                                 if ($pos > 0) {
  64                                         $post["text"] = trim(substr($body, 0, $pos));
  65                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  66                                 } else {
  67                                         $post["text"] = trim(str_replace($data[0], "", $body));
  68                                 }
  69
  70                                 $attacheddata = $data[2];
  71
  72                                 $URLSearchString = "^\[\]";
  73
  74                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  75
  76                                         $picturedata = Image::getInfoFromURL($matches[1]);
  77
  78                                         if ($picturedata) {
  79                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  80                                                         $post["image"] = $matches[1];
  81                                                 } else {
  82                                                         $post["preview"] = $matches[1];
  83                                                 }
  84                                         }
  85                                 }
  86
  87                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  88                                         $post["url"] = $matches[1];
  89                                         $post["title"] = $matches[2];
  90                                 }
  91                                 if (!empty($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
 246                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 247
 248                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 249                                 if ((count($pictures) == 1) && !$has_title) {
 250                                         // Checking, if the link goes to a picture
 251                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 252
 253                                         // Workaround:
 254                                         // Sometimes photo posts to the own album are not detected at the start.
 255                                         // So we seem to cannot use the cache for these cases. That's strange.
 256                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 257                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 258                                         }
 259
 260                                         if ($data["type"] == "photo") {
 261                                                 $post["type"] = "photo";
 262                                                 if (isset($data["images"][0])) {
 263                                                         $post["image"] = $data["images"][0]["src"];
 264                                                         $post["url"] = $data["url"];
 265                                                 } else {
 266                                                         $post["image"] = $data["url"];
 267                                                 }
 268
 269                                                 $post["preview"] = $pictures[0][2];
 270                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 271                                         } else {
 272                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 273                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 274                                                         $post["type"] = "photo";
 275                                                         $post["image"] = $pictures[0][1];
 276                                                         $post["preview"] = $pictures[0][2];
 277                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 278                                                 }
 279                                         }
 280                                 } elseif (count($pictures) > 0) {
 281                                         $post["type"] = "link";
 282                                         $post["url"] = $plink;
 283                                         $post["image"] = $pictures[0][2];
 284                                         $post["text"] = $body;
 285                                 }
 286                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 287                                 if ((count($pictures) == 1) && !$has_title) {
 288                                         $post["type"] = "photo";
 289                                         $post["image"] = $pictures[0][1];
 290                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 291                                 } elseif (count($pictures) > 0) {
 292                                         $post["type"] = "link";
 293                                         $post["url"] = $plink;
 294                                         $post["image"] = $pictures[0][1];
 295                                         $post["text"] = $body;
 296                                 }
 297                         }
 298
 299                         // Test for the external links
 300                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 301                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 302
 303                         $links = array_merge($links1, $links2);
 304
 305                         // If there is only a single one, then use it.
 306                         // This should cover link posts via API.
 307                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 308                                 $post["type"] = "link";
 309                                 $post["text"] = trim($body);
 310                                 $post["url"] = $links[0][1];
 311                         }
 312
 313                         // Now count the number of external media links
 314                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 315                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 316                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 317                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 318
 319                         // Add them to the other external links
 320                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 321
 322                         // Are there more than one?
 323                         if (count($links) > 1) {
 324                                 // The post will be the type "text", which means a blog post
 325                                 unset($post["type"]);
 326                                 $post["url"] = $plink;
 327                         }
 328
 329                         if (!isset($post["type"])) {
 330                                 $post["type"] = "text";
 331                                 $post["text"] = trim($body);
 332                         }
 333                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 334                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 335
 336                         if (isset($data["images"][0])) {
 337                                 $post["image"] = $data["images"][0]["src"];
 338                         }
 339                 }
 340
 341                 return $post;
 342         }
 343
 344         /**
 345          * @brief Converts a BBCode text into plaintext
 346          *
 347          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 348          *
 349          * @return string
 350          */
 351         public static function toPlaintext($text, $keep_urls = true)
 352         {
 353                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 354                 if (!$keep_urls) {
 355                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 356                 }
 357
 358                 return $naked_text;
 359         }
 360
 361         private static function proxyUrl($image, $simplehtml = false)
 362         {
 363                 // Only send proxied pictures to API and for internal display
 364                 if (in_array($simplehtml, [false, 2])) {
 365                         return ProxyUtils::proxifyUrl($image);
 366                 } else {
 367                         return $image;
 368                 }
 369         }
 370
 371         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 372         {
 373                 // Suppress "view full size"
 374                 if (intval(Config::get('system', 'no_view_full_size'))) {
 375                         $include_link = false;
 376                 }
 377
 378                 // Picture addresses can contain special characters
 379                 $s = htmlspecialchars_decode($srctext);
 380
 381                 $matches = null;
 382                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 383                 if ($c) {
 384                         foreach ($matches as $mtch) {
 385                                 Logger::log('scale_external_image: ' . $mtch[1]);
 386
 387                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 388                                 if (stristr($mtch[1], $hostname)) {
 389                                         continue;
 390                                 }
 391
 392                                 // $scale_replace, if passed, is an array of two elements. The
 393                                 // first is the name of the full-size image. The second is the
 394                                 // name of a remote, scaled-down version of the full size image.
 395                                 // This allows Friendica to display the smaller remote image if
 396                                 // one exists, while still linking to the full-size image
 397                                 if ($scale_replace) {
 398                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 399                                 } else {
 400                                         $scaled = $mtch[1];
 401                                 }
 402                                 $i = Network::fetchUrl($scaled);
 403                                 if (!$i) {
 404                                         return $srctext;
 405                                 }
 406
 407                                 // guess mimetype from headers or filename
 408                                 $type = Image::guessType($mtch[1], true);
 409
 410                                 if ($i) {
 411                                         $Image = new Image($i, $type);
 412                                         if ($Image->isValid()) {
 413                                                 $orig_width = $Image->getWidth();
 414                                                 $orig_height = $Image->getHeight();
 415
 416                                                 if ($orig_width > 640 || $orig_height > 640) {
 417                                                         $Image->scaleDown(640);
 418                                                         $new_width = $Image->getWidth();
 419                                                         $new_height = $Image->getHeight();
 420                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 421                                                         $s = str_replace(
 422                                                                 $mtch[0],
 423                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 424                                                                 . "\n" . (($include_link)
 425                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 426                                                                         : ''),
 427                                                                 $s
 428                                                         );
 429                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 430                                                 }
 431                                         }
 432                                 }
 433                         }
 434                 }
 435
 436                 // replace the special char encoding
 437                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 438                 return $s;
 439         }
 440
 441         /**
 442          * The purpose of this function is to apply system message length limits to
 443          * imported messages without including any embedded photos in the length
 444          *
 445          * @brief Truncates imported message body string length to max_import_size
 446          * @param string $body
 447          * @return string
 448          */
 449         public static function limitBodySize($body)
 450         {
 451                 $maxlen = Config::get('config', 'max_import_size', 0);
 452
 453                 // If the length of the body, including the embedded images, is smaller
 454                 // than the maximum, then don't waste time looking for the images
 455                 if ($maxlen && (strlen($body) > $maxlen)) {
 456
 457                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 458
 459                         $orig_body = $body;
 460                         $new_body = '';
 461                         $textlen = 0;
 462
 463                         $img_start = strpos($orig_body, '[img');
 464                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 465                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 466                         while (($img_st_close !== false) && ($img_end !== false)) {
 467
 468                                 $img_st_close++; // make it point to AFTER the closing bracket
 469                                 $img_end += $img_start;
 470                                 $img_end += strlen('[/img]');
 471
 472                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 473                                         // This is an embedded image
 474
 475                                         if (($textlen + $img_start) > $maxlen) {
 476                                                 if ($textlen < $maxlen) {
 477                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 478                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 479                                                         $textlen = $maxlen;
 480                                                 }
 481                                         } else {
 482                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 483                                                 $textlen += $img_start;
 484                                         }
 485
 486                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 487                                 } else {
 488
 489                                         if (($textlen + $img_end) > $maxlen) {
 490                                                 if ($textlen < $maxlen) {
 491                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 492                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 493                                                         $textlen = $maxlen;
 494                                                 }
 495                                         } else {
 496                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 497                                                 $textlen += $img_end;
 498                                         }
 499                                 }
 500                                 $orig_body = substr($orig_body, $img_end);
 501
 502                                 if ($orig_body === false) {
 503                                         // in case the body ends on a closing image tag
 504                                         $orig_body = '';
 505                                 }
 506
 507                                 $img_start = strpos($orig_body, '[img');
 508                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 509                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 510                         }
 511
 512                         if (($textlen + strlen($orig_body)) > $maxlen) {
 513                                 if ($textlen < $maxlen) {
 514                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 515                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 516                                 }
 517                         } else {
 518                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 519                                 $new_body = $new_body . $orig_body;
 520                         }
 521
 522                         return $new_body;
 523                 } else {
 524                         return $body;
 525                 }
 526         }
 527
 528         /**
 529          * Processes [attachment] tags
 530          *
 531          * Note: Can produce a [bookmark] tag in the returned string
 532          *
 533          * @brief Processes [attachment] tags
 534          * @param string $return
 535          * @param bool|int $simplehtml
 536          * @param bool $tryoembed
 537          * @return string
 538          */
 539         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 540         {
 541                 $data = self::getAttachmentData($return);
 542                 if (empty($data) || empty($data["url"])) {
 543                         return $return;
 544                 }
 545
 546                 if (isset($data["title"])) {
 547                         $data["title"] = strip_tags($data["title"]);
 548                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 549                 } else {
 550                         $data["title"] = null;
 551                 }
 552
 553                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 554                         $data["preview"] = $data["image"];
 555                         $data["image"] = "";
 556                 }
 557
 558                 $return = '';
 559                 if ($simplehtml == 7) {
 560                         $return = self::convertUrlForOStatus($data["url"]);
 561                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 562                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 563                 } else {
 564                         try {
 565                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 566                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 567                                 } else {
 568                                         throw new Exception('OEmbed is disabled for this attachment.');
 569                                 }
 570                         } catch (Exception $e) {
 571                                 $data["title"] = defaults($data, 'title', $data['url']);
 572
 573                                 if ($simplehtml != 4) {
 574                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 575                                 }
 576
 577                                 if (!empty($data['title']) && !empty($data['url'])) {
 578                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 579                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 580                                         } else {
 581                                                 if (!empty($data["image"])) {
 582                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 583                                                 } elseif (!empty($data["preview"])) {
 584                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], self::proxyUrl($data["preview"], $simplehtml), $data["title"]);
 585                                                 }
 586                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 587                                         }
 588                                 }
 589
 590                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 591                                         // Sanitize the HTML by converting it to BBCode
 592                                         $bbcode = HTML::toBBCode($data["description"]);
 593                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 594                                 }
 595
 596                                 if (!empty($data['url'])) {
 597                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 598                                 }
 599
 600                                 if ($simplehtml != 4) {
 601                                         $return .= '</div>';
 602                                 }
 603                         }
 604                 }
 605
 606                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 607         }
 608
 609         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 610         {
 611                 $data = self::getAttachmentData($Text);
 612
 613                 if (!$data) {
 614                         return $Text;
 615                 } elseif ($nolink) {
 616                         return $data["text"] . defaults($data, 'after', '');
 617                 }
 618
 619                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 620                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 621                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 622                         $data["title"] = $data["url"];
 623                 } elseif (($text != "") && strstr($title, $text)) {
 624                         $data["text"] = $data["title"];
 625                         $data["title"] = $data["url"];
 626                 }
 627
 628                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 629                         return $data["title"] . $data["after"];
 630                 }
 631
 632                 // If the link already is included in the post, don't add it again
 633                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 634                         return $data["text"] . $data["after"];
 635                 }
 636
 637                 $text = $data["text"];
 638
 639                 if (!empty($data["url"]) && !empty($data["title"])) {
 640                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 641                 } elseif (!empty($data["url"])) {
 642                         $text .= "\n[url]" . $data["url"] . "[/url]";
 643                 }
 644
 645                 return $text . "\n" . $data["after"];
 646         }
 647
 648         /**
 649          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 650          *
 651          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 652          * @param array $match Array with the matching values
 653          * @return string reformatted link including HTML codes
 654          */
 655         private static function convertUrlForOStatusCallback($match)
 656         {
 657                 $url = $match[1];
 658
 659                 if (isset($match[2]) && ($match[1] != $match[2])) {
 660                         return $match[0];
 661                 }
 662
 663                 $parts = parse_url($url);
 664                 if (!isset($parts['scheme'])) {
 665                         return $match[0];
 666                 }
 667
 668                 return self::convertUrlForOStatus($url);
 669         }
 670
 671         /**
 672          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 673          * @param string $url URL that is about to be reformatted
 674          * @return string reformatted link including HTML codes
 675          */
 676         private static function convertUrlForOStatus($url)
 677         {
 678                 $parts = parse_url($url);
 679                 $scheme = $parts['scheme'] . '://';
 680                 $styled_url = str_replace($scheme, '', $url);
 681
 682                 if (strlen($styled_url) > 30) {
 683                         $styled_url = substr($styled_url, 0, 30) . "…";
 684                 }
 685
 686                 $html = '<a href="%s" target="_blank">%s</a>';
 687
 688                 return sprintf($html, $url, $styled_url);
 689         }
 690
 691         /*
 692          * [noparse][i]italic[/i][/noparse] turns into
 693          * [noparse][ i ]italic[ /i ][/noparse],
 694          * to hide them from parser.
 695          */
 696         private static function escapeNoparseCallback($match)
 697         {
 698                 $whole_match = $match[0];
 699                 $captured = $match[1];
 700                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 701                 $new_str = str_replace($captured, $spacefied, $whole_match);
 702                 return $new_str;
 703         }
 704
 705         /*
 706          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 707          * now turns back and the [noparse] tags are trimed
 708          * returning [i]italic[/i]
 709          */
 710         private static function unescapeNoparseCallback($match)
 711         {
 712                 $captured = $match[1];
 713                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 714                 return $unspacefied;
 715         }
 716
 717         /**
 718          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 719          * occurrences
 720          *
 721          * @param string $text        Text to search
 722          * @param string $name        Tag name
 723          * @param int    $occurrences Number of first occurrences to skip
 724          * @return boolean|array
 725          */
 726         public static function getTagPosition($text, $name, $occurrences = 0)
 727         {
 728                 if ($occurrences < 0) {
 729                         $occurrences = 0;
 730                 }
 731
 732                 $start_open = -1;
 733                 for ($i = 0; $i <= $occurrences; $i++) {
 734                         if ($start_open !== false) {
 735                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 736                         }
 737                 }
 738
 739                 if ($start_open === false) {
 740                         return false;
 741                 }
 742
 743                 $start_equal = strpos($text, '=', $start_open);
 744                 $start_close = strpos($text, ']', $start_open);
 745
 746                 if ($start_close === false) {
 747                         return false;
 748                 }
 749
 750                 $start_close++;
 751
 752                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 753
 754                 if ($end_open === false) {
 755                         return false;
 756                 }
 757
 758                 $res = [
 759                         'start' => [
 760                                 'open' => $start_open,
 761                                 'close' => $start_close
 762                         ],
 763                         'end' => [
 764                                 'open' => $end_open,
 765                                 'close' => $end_open + strlen('[/' . $name . ']')
 766                         ],
 767                 ];
 768
 769                 if ($start_equal !== false) {
 770                         $res['start']['equal'] = $start_equal + 1;
 771                 }
 772
 773                 return $res;
 774         }
 775
 776         /**
 777          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 778          *
 779          * @param type $pattern Preg pattern string
 780          * @param type $replace Preg replace string
 781          * @param type $name    BBCode tag name
 782          * @param type $text    Text to search
 783          * @return string
 784          */
 785         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 786         {
 787                 $occurrences = 0;
 788                 $pos = self::getTagPosition($text, $name, $occurrences);
 789                 while ($pos !== false && $occurrences++ < 1000) {
 790                         $start = substr($text, 0, $pos['start']['open']);
 791                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 792                         $end = substr($text, $pos['end']['close']);
 793                         if ($end === false) {
 794                                 $end = '';
 795                         }
 796
 797                         $subject = preg_replace($pattern, $replace, $subject);
 798                         $text = $start . $subject . $end;
 799
 800                         $pos = self::getTagPosition($text, $name, $occurrences);
 801                 }
 802
 803                 return $text;
 804         }
 805
 806         private static function extractImagesFromItemBody($body)
 807         {
 808                 $saved_image = [];
 809                 $orig_body = $body;
 810                 $new_body = '';
 811
 812                 $cnt = 0;
 813                 $img_start = strpos($orig_body, '[img');
 814                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 815                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 816                 while (($img_st_close !== false) && ($img_end !== false)) {
 817                         $img_st_close++; // make it point to AFTER the closing bracket
 818                         $img_end += $img_start;
 819
 820                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 821                                 // This is an embedded image
 822                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 823                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 824
 825                                 $cnt++;
 826                         } else {
 827                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 828                         }
 829
 830                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 831
 832                         if ($orig_body === false) {
 833                                 // in case the body ends on a closing image tag
 834                                 $orig_body = '';
 835                         }
 836
 837                         $img_start = strpos($orig_body, '[img');
 838                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 839                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 840                 }
 841
 842                 $new_body = $new_body . $orig_body;
 843
 844                 return ['body' => $new_body, 'images' => $saved_image];
 845         }
 846
 847         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 848         {
 849                 $newbody = $body;
 850
 851                 $cnt = 0;
 852                 foreach ($images as $image) {
 853                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 854                         // it loops over the array starting from the first element and going sequentially
 855                         // to the last element
 856                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 857                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 858                         $cnt++;
 859                 }
 860
 861                 return $newbody;
 862         }
 863
 864         /**
 865          * This function converts a [share] block to text according to a provided callback function whose signature is:
 866          *
 867          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 868          *
 869          * Where:
 870          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 871          * data lookup
 872          * - $author_contact is a contact record array
 873          * - $content is the inner content of the [share] block
 874          * - $is_quote_share indicates whether there's any content before the [share] block
 875          * - Return value is the string that should replace the [share] block in the provided text
 876          *
 877          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 878          *
 879          * @param  string   $text     A BBCode string
 880          * @param  callable $callback
 881          * @return string The BBCode string with all [share] blocks replaced
 882          */
 883         public static function convertShare($text, callable $callback)
 884         {
 885                 $return = preg_replace_callback(
 886                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 887                         function ($match) use ($callback) {
 888                                 $attribute_string = $match[2];
 889
 890                                 $attributes = [];
 891                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 892                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 893                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 894                                 }
 895
 896                                 // We only call this so that a previously unknown contact can be added.
 897                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 898                                 // This function then can fetch an entry from the contact table.
 899                                 Contact::getIdForURL($attributes['profile'], 0, true);
 900
 901                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 902                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 903
 904                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 905                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 906                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 907
 908                                 if ($attributes['avatar']) {
 909                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 910                                 }
 911
 912                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 913                         },
 914                         $text
 915                 );
 916
 917                 return $return;
 918         }
 919
 920         /**
 921          * Default [share] tag conversion callback
 922          *
 923          * Note: Can produce a [bookmark] tag in the output
 924          *
 925          * @see BBCode::convertShare()
 926          * @param array   $attributes     [share] block attribute values
 927          * @param array   $author_contact Contact row of the shared author
 928          * @param string  $content        Inner content of the [share] block
 929          * @param boolean $is_quote_share Whether there is content before the [share] block
 930          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 931          * @return string
 932          */
 933         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 934         {
 935                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 936
 937                 switch ($simplehtml) {
 938                         case 1:
 939                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 940                                 break;
 941                         case 2:
 942                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 943                                 break;
 944                         case 3: // Diaspora
 945                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . $mention . ':</b></p>' . "\n";
 946
 947                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 948                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 949                                 } else {
 950                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 951
 952                                         if ($attributes['link'] != '') {
 953                                                 $text .= '<p><a href="' . $attributes['link'] . '">[l]</a></p>' . "\n";
 954                                         }
 955                                 }
 956
 957                                 break;
 958                         case 4:
 959                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 960                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 961                                 $headline .= ':</b></p>' . "\n";
 962
 963                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 964
 965                                 break;
 966                         case 5:
 967                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 968                                 break;
 969                         case 7: // statusnet/GNU Social
 970                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 971                                 break;
 972                         case 9: // Google+
 973                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 974                                 $text .= '<p>' . $content . '</p>' . "\n";
 975
 976                                 if ($attributes['link'] != '') {
 977                                         $text .= '<p>' . $attributes['link'] . '</p>';
 978                                 }
 979                                 break;
 980                         default:
 981                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
 982                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
 983                                         try {
 984                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
 985                                         } catch (Exception $e) {
 986                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
 987                                         }
 988                                 } else {
 989                                         $text = ($is_quote_share? "\n" : '');
 990
 991                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
 992                                         $text .= Renderer::replaceMacros($tpl, [
 993                                                 '$profile' => $attributes['profile'],
 994                                                 '$avatar'  => $attributes['avatar'],
 995                                                 '$author'  => $attributes['author'],
 996                                                 '$link'    => $attributes['link'],
 997                                                 '$posted'  => $attributes['posted'],
 998                                                 '$content' => trim($content)
 999                                         ]);
1000                                 }
1001                                 break;
1002                 }
1003
1004                 return $text;
1005         }
1006
1007         private static function removePictureLinksCallback($match)
1008         {
1009                 $text = Cache::get($match[1]);
1010
1011                 if (is_null($text)) {
1012                         $a = self::getApp();
1013
1014                         $stamp1 = microtime(true);
1015
1016                         $ch = @curl_init($match[1]);
1017                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1018                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1019                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1020                         @curl_exec($ch);
1021                         $curl_info = @curl_getinfo($ch);
1022
1023                         $a->saveTimestamp($stamp1, "network");
1024
1025                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1026                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1027                         } else {
1028                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1029
1030                                 // if its not a picture then look if its a page that contains a picture link
1031                                 $body = Network::fetchUrl($match[1]);
1032
1033                                 $doc = new DOMDocument();
1034                                 @$doc->loadHTML($body);
1035                                 $xpath = new DOMXPath($doc);
1036                                 $list = $xpath->query("//meta[@name]");
1037                                 foreach ($list as $node) {
1038                                         $attr = [];
1039
1040                                         if ($node->attributes->length) {
1041                                                 foreach ($node->attributes as $attribute) {
1042                                                         $attr[$attribute->name] = $attribute->value;
1043                                                 }
1044                                         }
1045
1046                                         if (strtolower($attr["name"]) == "twitter:image") {
1047                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1048                                         }
1049                                 }
1050                         }
1051                         Cache::set($match[1], $text);
1052                 }
1053
1054                 return $text;
1055         }
1056
1057         private static function expandLinksCallback($match)
1058         {
1059                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1060                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1061                 } else {
1062                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1063                 }
1064         }
1065
1066         private static function cleanPictureLinksCallback($match)
1067         {
1068                 $text = Cache::get($match[1]);
1069
1070                 if (is_null($text)) {
1071                         $a = self::getApp();
1072
1073                         $stamp1 = microtime(true);
1074
1075                         $ch = @curl_init($match[1]);
1076                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1077                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1078                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1079                         @curl_exec($ch);
1080                         $curl_info = @curl_getinfo($ch);
1081
1082                         $a->saveTimestamp($stamp1, "network");
1083
1084                         // if its a link to a picture then embed this picture
1085                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1086                                 $text = "[img]" . $match[1] . "[/img]";
1087                         } else {
1088                                 $text = "[img]" . $match[2] . "[/img]";
1089
1090                                 // if its not a picture then look if its a page that contains a picture link
1091                                 $body = Network::fetchUrl($match[1]);
1092
1093                                 $doc = new DOMDocument();
1094                                 @$doc->loadHTML($body);
1095                                 $xpath = new DOMXPath($doc);
1096                                 $list = $xpath->query("//meta[@name]");
1097                                 foreach ($list as $node) {
1098                                         $attr = [];
1099                                         if ($node->attributes->length) {
1100                                                 foreach ($node->attributes as $attribute) {
1101                                                         $attr[$attribute->name] = $attribute->value;
1102                                                 }
1103                                         }
1104
1105                                         if (strtolower($attr["name"]) == "twitter:image") {
1106                                                 $text = "[img]" . $attr["content"] . "[/img]";
1107                                         }
1108                                 }
1109                         }
1110                         Cache::set($match[1], $text);
1111                 }
1112
1113                 return $text;
1114         }
1115
1116         public static function cleanPictureLinks($text)
1117         {
1118                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1119                 return $return;
1120         }
1121
1122         /**
1123          * @brief Converts a BBCode message to HTML message
1124          *
1125          * BBcode 2 HTML was written by WAY2WEB.net
1126          * extended to work with Mistpark/Friendica - Mike Macgirvin
1127          *
1128          * Simple HTML values meaning:
1129          * - 0: Friendica display
1130          * - 1: Unused
1131          * - 2: Used for Google+, Windows Phone push, Friendica API
1132          * - 3: Used before converting to Markdown in bb2diaspora.php
1133          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1134          * - 5: Unused
1135          * - 6: Unused
1136          * - 7: Used for dfrn, OStatus
1137          * - 8: Used for WP backlink text setting
1138          *
1139          * @param string $text
1140          * @param bool   $try_oembed
1141          * @param int    $simple_html
1142          * @param bool   $for_plaintext
1143          * @return string
1144          */
1145         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1146         {
1147                 $a = self::getApp();
1148
1149                 /*
1150                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1151                  *
1152                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1153                  * $match[1] = $url
1154                  * $match[2] = $title or absent
1155                  */
1156                 $try_oembed_callback = function ($match)
1157                 {
1158                         $url = $match[1];
1159                         $title = defaults($match, 2, null);
1160
1161                         try {
1162                                 $return = OEmbed::getHTML($url, $title);
1163                         } catch (Exception $ex) {
1164                                 $return = $match[0];
1165                         }
1166
1167                         return $return;
1168                 };
1169
1170                 // Extracting multi-line code blocks before the whitespace processing
1171                 $codeblocks = [];
1172
1173                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1174                         function ($matches) use (&$codeblocks) {
1175                                 $return = $matches[0];
1176                                 if (strpos($matches[2], "\n") !== false) {
1177                                         $return = '#codeblock-' . count($codeblocks) . '#';
1178
1179                                         $codeblocks[] =  '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1180                                 }
1181                                 return $return;
1182                         },
1183                         $text
1184                 );
1185
1186                 // Hide all [noparse] contained bbtags by spacefying them
1187                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1188
1189                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1190                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1191                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1192
1193                 // Remove the abstract element. It is a non visible element.
1194                 $text = self::stripAbstract($text);
1195
1196                 // Move all spaces out of the tags
1197                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1198                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1199
1200                 // Extract the private images which use data urls since preg has issues with
1201                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1202                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1203
1204                 $extracted = self::extractImagesFromItemBody($text);
1205                 $text = $extracted['body'];
1206                 $saved_image = $extracted['images'];
1207
1208                 // If we find any event code, turn it into an event.
1209                 // After we're finished processing the bbcode we'll
1210                 // replace all of the event code with a reformatted version.
1211
1212                 $ev = Event::fromBBCode($text);
1213
1214                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1215                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1216
1217                 $text = str_replace("<", "&lt;", $text);
1218                 $text = str_replace(">", "&gt;", $text);
1219
1220                 // remove some newlines before the general conversion
1221                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1222                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1223
1224                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1225                 if (!$try_oembed) {
1226                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1227                 }
1228
1229                 // Convert new line chars to html <br /> tags
1230
1231                 // nlbr seems to be hopelessly messed up
1232                 //      $Text = nl2br($Text);
1233
1234                 // We'll emulate it.
1235
1236                 $text = trim($text);
1237                 $text = str_replace("\r\n", "\n", $text);
1238
1239                 // removing multiplicated newlines
1240                 if (Config::get("system", "remove_multiplicated_lines")) {
1241                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1242                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1243                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1244                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1245                         do {
1246                                 $oldtext = $text;
1247                                 $text = str_replace($search, $replace, $text);
1248                         } while ($oldtext != $text);
1249                 }
1250
1251                 // Set up the parameters for a URL search string
1252                 $URLSearchString = "^\[\]";
1253                 // Set up the parameters for a MAIL search string
1254                 $MAILSearchString = $URLSearchString;
1255
1256                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1257                 if (!$for_plaintext) {
1258                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1259                         // Currently disabled, since the function is too greedy
1260                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1261                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1262                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1263                         if ($simple_html == 7) {
1264                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1265                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1266                         }
1267                 } else {
1268                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1269                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1270                 }
1271
1272
1273                 // Handle attached links or videos
1274                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1275
1276                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1277
1278                 // Remove all hashtag addresses
1279                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1280                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1281                 } elseif ($simple_html == 3) {
1282                         // The ! is converted to @ since Diaspora only understands the @
1283                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1284                                 '@<a href="$2">$3</a>',
1285                                 $text);
1286                 } elseif ($simple_html == 7) {
1287                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1288                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1289                                 $text);
1290                 } elseif (!$simple_html) {
1291                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1292                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1293                                 $text);
1294                 }
1295
1296                 // Bookmarks in red - will be converted to bookmarks in friendica
1297                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1298                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1299                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1300                                         "[bookmark=$1]$2[/bookmark]", $text);
1301
1302                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1303                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1304                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1305                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1306                 }
1307
1308                 if ($simple_html == 5) {
1309                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1310                 }
1311
1312                 // Perform URL Search
1313                 if ($try_oembed) {
1314                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1315                 }
1316
1317                 if ($simple_html == 5) {
1318                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1319                 } else {
1320                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1321                 }
1322
1323                 // Handle Diaspora posts
1324                 $text = preg_replace_callback(
1325                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1326                         function ($match) {
1327                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1328                         }, $text
1329                 );
1330
1331                 $text = preg_replace_callback(
1332                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1333                         function ($match) {
1334                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1335                         }, $text
1336                 );
1337
1338                 // Server independent link to posts and comments
1339                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1340                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1341                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1342
1343                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1344                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1345
1346                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1347                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1348
1349                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1350                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1351                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1352
1353                 // Red compatibility, though the link can't be authenticated on Friendica
1354                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1355
1356
1357                 // we may need to restrict this further if it picks up too many strays
1358                 // link acct:user@host to a webfinger profile redirector
1359
1360                 $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);
1361
1362                 // Perform MAIL Search
1363                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1364                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1365
1366                 // leave open the posibility of [map=something]
1367                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1368
1369                 if (strpos($text, '[/map]') !== false) {
1370                         $text = preg_replace_callback(
1371                                 "/\[map\](.*?)\[\/map\]/ism",
1372                                 function ($match) use ($simple_html) {
1373                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1374                                 },
1375                                 $text
1376                         );
1377                 }
1378                 if (strpos($text, '[map=') !== false) {
1379                         $text = preg_replace_callback(
1380                                 "/\[map=(.*?)\]/ism",
1381                                 function ($match) use ($simple_html) {
1382                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1383                                 },
1384                                 $text
1385                         );
1386                 }
1387                 if (strpos($text, '[map]') !== false) {
1388                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1389                 }
1390
1391                 // Check for headers
1392                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1393                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1394                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1395                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1396                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1397                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1398
1399                 // Check for paragraph
1400                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1401
1402                 // Check for bold text
1403                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1404
1405                 // Check for Italics text
1406                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1407
1408                 // Check for Underline text
1409                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1410
1411                 // Check for strike-through text
1412                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1413
1414                 // Check for over-line text
1415                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1416
1417                 // Check for colored text
1418                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1419
1420                 // Check for sized text
1421                 // [size=50] --> font-size: 50px (with the unit).
1422                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1423                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1424
1425                 // Check for centered text
1426                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1427
1428                 // Check for list text
1429                 $text = str_replace("[*]", "<li>", $text);
1430
1431                 // Check for style sheet commands
1432                 $text = preg_replace_callback(
1433                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1434                         function ($match) {
1435                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1436                         },
1437                         $text
1438                 );
1439
1440                 // Check for CSS classes
1441                 $text = preg_replace_callback(
1442                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1443                         function ($match) {
1444                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1445                         },
1446                         $text
1447                 );
1448
1449                 // handle nested lists
1450                 $endlessloop = 0;
1451
1452                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1453                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1454                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1455                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1456                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1457                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1458                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1459                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1460                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1461                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1462                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1463                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1464                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1465                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1466                 }
1467
1468                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1469                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1470                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1471                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1472
1473                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1474                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1475
1476                 $text = str_replace('[hr]', '<hr />', $text);
1477
1478                 // This is actually executed in Item::prepareBody()
1479
1480                 $text = str_replace('[nosmile]', '', $text);
1481
1482                 // Check for font change text
1483                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1484
1485                 // Declare the format for [code] layout
1486
1487                 $CodeLayout = '<code>$1</code>';
1488                 // Check for [code] text
1489                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1490
1491                 // Declare the format for [spoiler] layout
1492                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1493
1494                 // Check for [spoiler] text
1495                 // handle nested quotes
1496                 $endlessloop = 0;
1497                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1498                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1499                 }
1500
1501                 // Check for [spoiler=Author] text
1502
1503                 $t_wrote = L10n::t('$1 wrote:');
1504
1505                 // handle nested quotes
1506                 $endlessloop = 0;
1507                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1508                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1509                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1510                                                  $text);
1511                 }
1512
1513                 // Declare the format for [quote] layout
1514                 $QuoteLayout = '<blockquote>$1</blockquote>';
1515
1516                 // Check for [quote] text
1517                 // handle nested quotes
1518                 $endlessloop = 0;
1519                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1520                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1521                 }
1522
1523                 // Check for [quote=Author] text
1524
1525                 $t_wrote = L10n::t('$1 wrote:');
1526
1527                 // handle nested quotes
1528                 $endlessloop = 0;
1529                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1530                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1531                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1532                                                  $text);
1533                 }
1534
1535
1536                 // [img=widthxheight]image source[/img]
1537                 $text = preg_replace_callback(
1538                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1539                         function ($matches) use ($simple_html) {
1540                                 if (strpos($matches[3], "data:image/") === 0) {
1541                                         return $matches[0];
1542                                 }
1543
1544                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1545                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1546                         },
1547                         $text
1548                 );
1549
1550                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1551                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1552
1553                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1554                         function ($matches) use ($simple_html) {
1555                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1556                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1557                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1558                         },
1559                         $text);
1560
1561                 // Images
1562                 // [img]pathtoimage[/img]
1563                 $text = preg_replace_callback(
1564                         "/\[img\](.*?)\[\/img\]/ism",
1565                         function ($matches) use ($simple_html) {
1566                                 if (strpos($matches[1], "data:image/") === 0) {
1567                                         return $matches[0];
1568                                 }
1569
1570                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1571                                 return "[img]" . $matches[1] . "[/img]";
1572                         },
1573                         $text
1574                 );
1575
1576                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1577                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1578
1579                 // Shared content
1580                 $text = self::convertShare(
1581                         $text,
1582                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1583                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1584                         }
1585                 );
1586
1587                 $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);
1588                 $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 />', $text);
1589                 //$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 />', $Text);
1590
1591                 // Try to Oembed
1592                 if ($try_oembed) {
1593                         $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);
1594                         $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);
1595
1596                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1597                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1598                 } else {
1599                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1600                                                 '<a href="$1" target="_blank">$1</a>', $text);
1601                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1602                                                 '<a href="$1" target="_blank">$1</a>', $text);
1603                 }
1604
1605                 // html5 video and audio
1606
1607
1608                 if ($try_oembed) {
1609                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1610                 } else {
1611                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1612                 }
1613
1614                 // Youtube extensions
1615                 if ($try_oembed) {
1616                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1617                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1618                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1619                 }
1620
1621                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1622                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1623                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1624
1625                 if ($try_oembed) {
1626                         $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);
1627                 } else {
1628                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1629                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1630                 }
1631
1632                 if ($try_oembed) {
1633                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1634                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1635                 }
1636
1637                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1638                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1639
1640                 if ($try_oembed) {
1641                         $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);
1642                 } else {
1643                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1644                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1645                 }
1646
1647                 // oembed tag
1648                 $text = OEmbed::BBCode2HTML($text);
1649
1650                 // Avoid triple linefeeds through oembed
1651                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1652
1653                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1654                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1655                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1656                 // start which is always required). Allow desc with a missing summary for compatibility.
1657
1658                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1659                         $sub = Event::getHTML($ev, $simple_html);
1660
1661                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1662                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1663                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1664                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1665                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1666                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1667                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1668                 }
1669
1670                 // Replace non graphical smilies for external posts
1671                 if ($simple_html) {
1672                         $text = Smilies::replace($text, false, true);
1673                 }
1674
1675                 // Unhide all [noparse] contained bbtags unspacefying them
1676                 // and triming the [noparse] tag.
1677
1678                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1679                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1680                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1681
1682                 /// @todo What is the meaning of these lines?
1683                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1684                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1685
1686                 // Currently deactivated, it made problems with " inside of alt texts.
1687                 //$text = preg_replace('/\&quot\;/', '"', $text);
1688
1689                 // fix any escaped ampersands that may have been converted into links
1690                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1691
1692                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1693                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1694                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1695                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1696
1697                 // sanitize href attributes (only whitelisted protocols URLs)
1698                 // default value for backward compatibility
1699                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1700
1701                 // Always allowed protocol even if config isn't set or not including it
1702                 $allowed_link_protocols[] = 'http';
1703                 $allowed_link_protocols[] = 'redir/';
1704
1705                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1706                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1707
1708                 if ($saved_image) {
1709                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1710                 }
1711
1712                 // Restore code blocks
1713                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1714                         function ($matches) use ($codeblocks) {
1715                                 $return = $matches[0];
1716                                 if (isset($codeblocks[intval($matches[1])])) {
1717                                         $return = $codeblocks[$matches[1]];
1718                                 }
1719                                 return $return;
1720                         },
1721                         $text
1722                 );
1723
1724                 // Clean up the HTML by loading and saving the HTML with the DOM.
1725                 // Bad structured html can break a whole page.
1726                 // For performance reasons do it only with ativated item cache or at export.
1727                 if (!$try_oembed || (get_itemcachepath() != "")) {
1728                         $doc = new DOMDocument();
1729                         $doc->preserveWhiteSpace = false;
1730
1731                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1732
1733                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1734                         $encoding = '<?xml encoding="UTF-8">';
1735                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1736                         $doc->encoding = 'UTF-8';
1737                         $text = $doc->saveHTML();
1738                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1739
1740                         $text = str_replace('<br></li>', '</li>', $text);
1741
1742                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1743                 }
1744
1745                 // Clean up some useless linebreaks in lists
1746                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1747                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1748                 //$Text = str_replace('</li><br />', '</li>', $Text);
1749                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1750                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1751
1752                 Addon::callHooks('bbcode', $text);
1753
1754                 return trim($text);
1755         }
1756
1757         /**
1758          * @brief Strips the "abstract" tag from the provided text
1759          *
1760          * @param string $text The text with BBCode
1761          * @return string The same text - but without "abstract" element
1762          */
1763         public static function stripAbstract($text)
1764         {
1765                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1766                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1767
1768                 return $text;
1769         }
1770
1771         /**
1772          * @brief Returns the value of the "abstract" element
1773          *
1774          * @param string $text The text that maybe contains the element
1775          * @param string $addon The addon for which the abstract is meant for
1776          * @return string The abstract
1777          */
1778         public static function getAbstract($text, $addon = "")
1779         {
1780                 $abstract = "";
1781                 $abstracts = [];
1782                 $addon = strtolower($addon);
1783
1784                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1785                         foreach ($results AS $result) {
1786                                 $abstracts[strtolower($result[1])] = $result[2];
1787                         }
1788                 }
1789
1790                 if (isset($abstracts[$addon])) {
1791                         $abstract = $abstracts[$addon];
1792                 }
1793
1794                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1795                         $abstract = $result[1];
1796                 }
1797
1798                 return $abstract;
1799         }
1800
1801         /**
1802          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1803          *
1804          * @param array $match Matching values for the callback
1805          * @return string Replaced mention
1806          */
1807         private static function bbCodeMention2DiasporaCallback($match)
1808         {
1809                 $contact = Contact::getDetailsByURL($match[3]);
1810
1811                 if (empty($contact['addr'])) {
1812                         $contact = Probe::uri($match[3]);
1813                 }
1814
1815                 if (empty($contact['addr'])) {
1816                         return $match[0];
1817                 }
1818
1819                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1820                 return $mention;
1821         }
1822
1823         /**
1824          * @brief Converts a BBCode text into Markdown
1825          *
1826          * This function converts a BBCode item body to be sent to Markdown-enabled
1827          * systems like Diaspora and Libertree
1828          *
1829          * @param string $text
1830          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1831          * @return string
1832          */
1833         public static function toMarkdown($text, $for_diaspora = true)
1834         {
1835                 $a = self::getApp();
1836
1837                 $original_text = $text;
1838
1839                 // Since Diaspora is creating a summary for links, this function removes them before posting
1840                 if ($for_diaspora) {
1841                         $text = self::removeShareInformation($text);
1842                 }
1843
1844                 /**
1845                  * Transform #tags, strip off the [url] and replace spaces with underscore
1846                  */
1847                 $url_search_string = "^\[\]";
1848                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1849                         function ($matches) {
1850                                 return '#' . str_replace(' ', '_', $matches[2]);
1851                         },
1852                         $text
1853                 );
1854
1855                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1856                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1857
1858                 // Convert it to HTML - don't try oembed
1859                 if ($for_diaspora) {
1860                         $text = self::convert($text, false, 3);
1861
1862                         // Add all tags that maybe were removed
1863                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1864                                 $tagline = "";
1865                                 foreach ($tags[2] as $tag) {
1866                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1867                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1868                                                 $tagline .= '#' . $tag . ' ';
1869                                         }
1870                                 }
1871                                 $text = $text . " " . $tagline;
1872                         }
1873                 } else {
1874                         $text = self::convert($text, false, 4);
1875                 }
1876
1877                 // mask some special HTML chars from conversation to markdown
1878                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1879
1880                 // If a link is followed by a quote then there should be a newline before it
1881                 // Maybe we should make this newline at every time before a quote.
1882                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1883
1884                 $stamp1 = microtime(true);
1885
1886                 // Now convert HTML to Markdown
1887                 $text = HTML::toMarkdown($text);
1888
1889                 // unmask the special chars back to HTML
1890                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1891
1892                 $a->saveTimestamp($stamp1, "parser");
1893
1894                 // Libertree has a problem with escaped hashtags.
1895                 $text = str_replace(['\#'], ['#'], $text);
1896
1897                 // Remove any leading or trailing whitespace, as this will mess up
1898                 // the Diaspora signature verification and cause the item to disappear
1899                 $text = trim($text);
1900
1901                 if ($for_diaspora) {
1902                         $url_search_string = "^\[\]";
1903                         $text = preg_replace_callback(
1904                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
1905                                 ['self', 'bbCodeMention2DiasporaCallback'],
1906                                 $text
1907                         );
1908                 }
1909
1910                 Addon::callHooks('bb2diaspora', $text);
1911
1912                 return $text;
1913         }
1914
1915         /**
1916      * @brief Pull out all #hashtags and @person tags from $string.
1917      *
1918      * We also get @person@domain.com - which would make
1919      * the regex quite complicated as tags can also
1920      * end a sentence. So we'll run through our results
1921      * and strip the period from any tags which end with one.
1922      * Returns array of tags found, or empty array.
1923      *
1924      * @param string $string Post content
1925      *
1926      * @return array List of tag and person names
1927      */
1928     public static function getTags($string)
1929     {
1930         $ret = [];
1931
1932         // Convert hashtag links to hashtags
1933         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1934
1935         // ignore anything in a code block
1936         $string = preg_replace('/\[code\](.*?)\[\/code\]/sm', '', $string);
1937
1938         // Force line feeds at bbtags
1939         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1940
1941         // ignore anything in a bbtag
1942         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1943
1944         // Match full names against @tags including the space between first and last
1945         // We will look these up afterward to see if they are full names or not recognisable.
1946
1947         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1948             foreach ($matches[1] as $match) {
1949                 if (strstr($match, ']')) {
1950                     // we might be inside a bbcode color tag - leave it alone
1951                     continue;
1952                 }
1953
1954                 if (substr($match, -1, 1) === '.') {
1955                     $ret[] = substr($match, 0, -1);
1956                 } else {
1957                     $ret[] = $match;
1958                 }
1959             }
1960         }
1961
1962         // Otherwise pull out single word tags. These can be @nickname, @first_last
1963         // and #hash tags.
1964
1965         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
1966             foreach ($matches[1] as $match) {
1967                 if (strstr($match, ']')) {
1968                     // we might be inside a bbcode color tag - leave it alone
1969                     continue;
1970                 }
1971                 if (substr($match, -1, 1) === '.') {
1972                     $match = substr($match,0,-1);
1973                 }
1974                 // ignore strictly numeric tags like #1
1975                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
1976                     continue;
1977                 }
1978                 // try not to catch url fragments
1979                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
1980                     continue;
1981                 }
1982                 $ret[] = $match;
1983             }
1984         }
1985
1986         return $ret;
1987     }
1988 }