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 use Friendica\Util\XML;
  32
  33 class BBCode extends BaseObject
  34 {
  35         /**
  36          * @brief Fetches attachment data that were generated the old way
  37          *
  38          * @param string $body Message body
  39          * @return array
  40          * 'type' -> Message type ("link", "video", "photo")
  41          * 'text' -> Text before the shared message
  42          * 'after' -> Text after the shared message
  43          * 'image' -> Preview image of the message
  44          * 'url' -> Url to the attached message
  45          * 'title' -> Title of the attachment
  46          * 'description' -> Description of the attachment
  47          */
  48         private static function getOldAttachmentData($body)
  49         {
  50                 $post = [];
  51
  52                 // Simplify image codes
  53                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  54
  55                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  56                         foreach ($attached as $data) {
  57                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  58                                         continue;
  59                                 }
  60
  61                                 $post["type"] = substr($data[1], 5);
  62
  63                                 $pos = strpos($body, $data[0]);
  64                                 if ($pos > 0) {
  65                                         $post["text"] = trim(substr($body, 0, $pos));
  66                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  67                                 } else {
  68                                         $post["text"] = trim(str_replace($data[0], "", $body));
  69                                 }
  70
  71                                 $attacheddata = $data[2];
  72
  73                                 $URLSearchString = "^\[\]";
  74
  75                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  76
  77                                         $picturedata = Image::getInfoFromURL($matches[1]);
  78
  79                                         if ($picturedata) {
  80                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  81                                                         $post["image"] = $matches[1];
  82                                                 } else {
  83                                                         $post["preview"] = $matches[1];
  84                                                 }
  85                                         }
  86                                 }
  87
  88                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  89                                         $post["url"] = $matches[1];
  90                                         $post["title"] = $matches[2];
  91                                 }
  92                                 if (!empty($post["url"]) && (in_array($post["type"], ["link", "video"]))
  93                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  94                                         $post["url"] = $matches[1];
  95                                 }
  96
  97                                 // Search for description
  98                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  99                                         $post["description"] = $matches[1];
 100                                 }
 101                         }
 102                 }
 103                 return $post;
 104         }
 105
 106         /**
 107          * @brief Fetches attachment data that were generated with the "attachment" element
 108          *
 109          * @param string $body Message body
 110          * @return array
 111          * 'type' -> Message type ("link", "video", "photo")
 112          * 'text' -> Text before the shared message
 113          * 'after' -> Text after the shared message
 114          * 'image' -> Preview image of the message
 115          * 'url' -> Url to the attached message
 116          * 'title' -> Title of the attachment
 117          * 'description' -> Description of the attachment
 118          */
 119         public static function getAttachmentData($body)
 120         {
 121                 $data = [];
 122
 123                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 124                         return self::getOldAttachmentData($body);
 125                 }
 126
 127                 $attributes = $match[2];
 128
 129                 $data["text"] = trim($match[1]);
 130
 131                 $type = "";
 132                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 133                 if (!empty($matches[1])) {
 134                         $type = strtolower($matches[1]);
 135                 }
 136
 137                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 138                 if (!empty($matches[1])) {
 139                         $type = strtolower($matches[1]);
 140                 }
 141
 142                 if ($type == "") {
 143                         return [];
 144                 }
 145
 146                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 147                         return [];
 148                 }
 149
 150                 if ($type != "") {
 151                         $data["type"] = $type;
 152                 }
 153
 154                 $url = "";
 155                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 156                 if (!empty($matches[1])) {
 157                         $url = $matches[1];
 158                 }
 159
 160                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 161                 if (!empty($matches[1])) {
 162                         $url = $matches[1];
 163                 }
 164
 165                 if ($url != "") {
 166                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 167                 }
 168
 169                 $title = "";
 170                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 171                 if (!empty($matches[1])) {
 172                         $title = $matches[1];
 173                 }
 174
 175                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 176                 if (!empty($matches[1])) {
 177                         $title = $matches[1];
 178                 }
 179
 180                 if ($title != "") {
 181                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 182                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 183                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 184                         $data["title"] = $title;
 185                 }
 186
 187                 $image = "";
 188                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 189                 if (!empty($matches[1])) {
 190                         $image = $matches[1];
 191                 }
 192
 193                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 194                 if (!empty($matches[1])) {
 195                         $image = $matches[1];
 196                 }
 197
 198                 if ($image != "") {
 199                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 200                 }
 201
 202                 $preview = "";
 203                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 204                 if (!empty($matches[1])) {
 205                         $preview = $matches[1];
 206                 }
 207
 208                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 209                 if (!empty($matches[1])) {
 210                         $preview = $matches[1];
 211                 }
 212
 213                 if ($preview != "") {
 214                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 215                 }
 216
 217                 $data["description"] = trim($match[3]);
 218
 219                 $data["after"] = trim($match[4]);
 220
 221                 return $data;
 222         }
 223
 224         public static function getAttachedData($body, $item = [])
 225         {
 226                 /*
 227                 - text:
 228                 - type: link, video, photo
 229                 - title:
 230                 - url:
 231                 - image:
 232                 - description:
 233                 - (thumbnail)
 234                 */
 235
 236                 $has_title = !empty($item['title']);
 237                 $plink = defaults($item, 'plink', '');
 238                 $post = self::getAttachmentData($body);
 239
 240                 // if nothing is found, it maybe having an image.
 241                 if (!isset($post["type"])) {
 242                         // Simplify image codes
 243                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 244
 245                         $URLSearchString = "^\[\]";
 246
 247                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 248
 249                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 250                                 if ((count($pictures) == 1) && !$has_title) {
 251                                         // Checking, if the link goes to a picture
 252                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 253
 254                                         // Workaround:
 255                                         // Sometimes photo posts to the own album are not detected at the start.
 256                                         // So we seem to cannot use the cache for these cases. That's strange.
 257                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 258                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 259                                         }
 260
 261                                         if ($data["type"] == "photo") {
 262                                                 $post["type"] = "photo";
 263                                                 if (isset($data["images"][0])) {
 264                                                         $post["image"] = $data["images"][0]["src"];
 265                                                         $post["url"] = $data["url"];
 266                                                 } else {
 267                                                         $post["image"] = $data["url"];
 268                                                 }
 269
 270                                                 $post["preview"] = $pictures[0][2];
 271                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 272                                         } else {
 273                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 274                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 275                                                         $post["type"] = "photo";
 276                                                         $post["image"] = $pictures[0][1];
 277                                                         $post["preview"] = $pictures[0][2];
 278                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 279                                                 }
 280                                         }
 281                                 } elseif (count($pictures) > 0) {
 282                                         $post["type"] = "link";
 283                                         $post["url"] = $plink;
 284                                         $post["image"] = $pictures[0][2];
 285                                         $post["text"] = $body;
 286                                 }
 287                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 288                                 if ((count($pictures) == 1) && !$has_title) {
 289                                         $post["type"] = "photo";
 290                                         $post["image"] = $pictures[0][1];
 291                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 292                                 } elseif (count($pictures) > 0) {
 293                                         $post["type"] = "link";
 294                                         $post["url"] = $plink;
 295                                         $post["image"] = $pictures[0][1];
 296                                         $post["text"] = $body;
 297                                 }
 298                         }
 299
 300                         // Test for the external links
 301                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 302                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 303
 304                         $links = array_merge($links1, $links2);
 305
 306                         // If there is only a single one, then use it.
 307                         // This should cover link posts via API.
 308                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 309                                 $post["type"] = "link";
 310                                 $post["text"] = trim($body);
 311                                 $post["url"] = $links[0][1];
 312                         }
 313
 314                         // Now count the number of external media links
 315                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 316                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 317                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 318                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 319
 320                         // Add them to the other external links
 321                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 322
 323                         // Are there more than one?
 324                         if (count($links) > 1) {
 325                                 // The post will be the type "text", which means a blog post
 326                                 unset($post["type"]);
 327                                 $post["url"] = $plink;
 328                         }
 329
 330                         if (!isset($post["type"])) {
 331                                 $post["type"] = "text";
 332                                 $post["text"] = trim($body);
 333                         }
 334                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 335                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 336
 337                         if (isset($data["images"][0])) {
 338                                 $post["image"] = $data["images"][0]["src"];
 339                         }
 340                 }
 341
 342                 return $post;
 343         }
 344
 345         /**
 346          * @brief Converts a BBCode text into plaintext
 347          *
 348          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 349          *
 350          * @return string
 351          */
 352         public static function toPlaintext($text, $keep_urls = true)
 353         {
 354                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 355                 if (!$keep_urls) {
 356                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 357                 }
 358
 359                 return $naked_text;
 360         }
 361
 362         private static function proxyUrl($image, $simplehtml = false)
 363         {
 364                 // Only send proxied pictures to API and for internal display
 365                 if (in_array($simplehtml, [false, 2])) {
 366                         return ProxyUtils::proxifyUrl($image);
 367                 } else {
 368                         return $image;
 369                 }
 370         }
 371
 372         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 373         {
 374                 // Suppress "view full size"
 375                 if (intval(Config::get('system', 'no_view_full_size'))) {
 376                         $include_link = false;
 377                 }
 378
 379                 // Picture addresses can contain special characters
 380                 $s = htmlspecialchars_decode($srctext);
 381
 382                 $matches = null;
 383                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 384                 if ($c) {
 385                         foreach ($matches as $mtch) {
 386                                 Logger::log('scale_external_image: ' . $mtch[1]);
 387
 388                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 389                                 if (stristr($mtch[1], $hostname)) {
 390                                         continue;
 391                                 }
 392
 393                                 // $scale_replace, if passed, is an array of two elements. The
 394                                 // first is the name of the full-size image. The second is the
 395                                 // name of a remote, scaled-down version of the full size image.
 396                                 // This allows Friendica to display the smaller remote image if
 397                                 // one exists, while still linking to the full-size image
 398                                 if ($scale_replace) {
 399                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 400                                 } else {
 401                                         $scaled = $mtch[1];
 402                                 }
 403                                 $i = Network::fetchUrl($scaled);
 404                                 if (!$i) {
 405                                         return $srctext;
 406                                 }
 407
 408                                 // guess mimetype from headers or filename
 409                                 $type = Image::guessType($mtch[1], true);
 410
 411                                 if ($i) {
 412                                         $Image = new Image($i, $type);
 413                                         if ($Image->isValid()) {
 414                                                 $orig_width = $Image->getWidth();
 415                                                 $orig_height = $Image->getHeight();
 416
 417                                                 if ($orig_width > 640 || $orig_height > 640) {
 418                                                         $Image->scaleDown(640);
 419                                                         $new_width = $Image->getWidth();
 420                                                         $new_height = $Image->getHeight();
 421                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 422                                                         $s = str_replace(
 423                                                                 $mtch[0],
 424                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 425                                                                 . "\n" . (($include_link)
 426                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 427                                                                         : ''),
 428                                                                 $s
 429                                                         );
 430                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 431                                                 }
 432                                         }
 433                                 }
 434                         }
 435                 }
 436
 437                 // replace the special char encoding
 438                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 439                 return $s;
 440         }
 441
 442         /**
 443          * The purpose of this function is to apply system message length limits to
 444          * imported messages without including any embedded photos in the length
 445          *
 446          * @brief Truncates imported message body string length to max_import_size
 447          * @param string $body
 448          * @return string
 449          */
 450         public static function limitBodySize($body)
 451         {
 452                 $maxlen = Config::get('config', 'max_import_size', 0);
 453
 454                 // If the length of the body, including the embedded images, is smaller
 455                 // than the maximum, then don't waste time looking for the images
 456                 if ($maxlen && (strlen($body) > $maxlen)) {
 457
 458                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 459
 460                         $orig_body = $body;
 461                         $new_body = '';
 462                         $textlen = 0;
 463
 464                         $img_start = strpos($orig_body, '[img');
 465                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 466                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 467                         while (($img_st_close !== false) && ($img_end !== false)) {
 468
 469                                 $img_st_close++; // make it point to AFTER the closing bracket
 470                                 $img_end += $img_start;
 471                                 $img_end += strlen('[/img]');
 472
 473                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 474                                         // This is an embedded image
 475
 476                                         if (($textlen + $img_start) > $maxlen) {
 477                                                 if ($textlen < $maxlen) {
 478                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 479                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 480                                                         $textlen = $maxlen;
 481                                                 }
 482                                         } else {
 483                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 484                                                 $textlen += $img_start;
 485                                         }
 486
 487                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 488                                 } else {
 489
 490                                         if (($textlen + $img_end) > $maxlen) {
 491                                                 if ($textlen < $maxlen) {
 492                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 493                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 494                                                         $textlen = $maxlen;
 495                                                 }
 496                                         } else {
 497                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 498                                                 $textlen += $img_end;
 499                                         }
 500                                 }
 501                                 $orig_body = substr($orig_body, $img_end);
 502
 503                                 if ($orig_body === false) {
 504                                         // in case the body ends on a closing image tag
 505                                         $orig_body = '';
 506                                 }
 507
 508                                 $img_start = strpos($orig_body, '[img');
 509                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 510                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 511                         }
 512
 513                         if (($textlen + strlen($orig_body)) > $maxlen) {
 514                                 if ($textlen < $maxlen) {
 515                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 516                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 517                                 }
 518                         } else {
 519                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 520                                 $new_body = $new_body . $orig_body;
 521                         }
 522
 523                         return $new_body;
 524                 } else {
 525                         return $body;
 526                 }
 527         }
 528
 529         /**
 530          * Processes [attachment] tags
 531          *
 532          * Note: Can produce a [bookmark] tag in the returned string
 533          *
 534          * @brief Processes [attachment] tags
 535          * @param string $return
 536          * @param bool|int $simplehtml
 537          * @param bool $tryoembed
 538          * @return string
 539          */
 540         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 541         {
 542                 $data = self::getAttachmentData($return);
 543                 if (empty($data) || empty($data["url"])) {
 544                         return $return;
 545                 }
 546
 547                 if (isset($data["title"])) {
 548                         $data["title"] = strip_tags($data["title"]);
 549                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 550                 } else {
 551                         $data["title"] = null;
 552                 }
 553
 554                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 555                         $data["preview"] = $data["image"];
 556                         $data["image"] = "";
 557                 }
 558
 559                 $return = '';
 560                 if ($simplehtml == 7) {
 561                         $return = self::convertUrlForOStatus($data["url"]);
 562                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 563                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 564                 } else {
 565                         try {
 566                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 567                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 568                                 } else {
 569                                         throw new Exception('OEmbed is disabled for this attachment.');
 570                                 }
 571                         } catch (Exception $e) {
 572                                 $data["title"] = defaults($data, 'title', $data['url']);
 573
 574                                 if ($simplehtml != 4) {
 575                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 576                                 }
 577
 578                                 if (!empty($data['title']) && !empty($data['url'])) {
 579                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 580                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 581                                         } else {
 582                                                 if (!empty($data["image"])) {
 583                                                         $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), $data["title"]);
 584                                                 } elseif (!empty($data["preview"])) {
 585                                                         $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"]);
 586                                                 }
 587                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 588                                         }
 589                                 }
 590
 591                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 592                                         // Sanitize the HTML by converting it to BBCode
 593                                         $bbcode = HTML::toBBCode($data["description"]);
 594                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 595                                 }
 596
 597                                 if (!empty($data['url'])) {
 598                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 599                                 }
 600
 601                                 if ($simplehtml != 4) {
 602                                         $return .= '</div>';
 603                                 }
 604                         }
 605                 }
 606
 607                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 608         }
 609
 610         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 611         {
 612                 $data = self::getAttachmentData($Text);
 613
 614                 if (!$data) {
 615                         return $Text;
 616                 } elseif ($nolink) {
 617                         return $data["text"] . defaults($data, 'after', '');
 618                 }
 619
 620                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 621                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 622                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 623                         $data["title"] = $data["url"];
 624                 } elseif (($text != "") && strstr($title, $text)) {
 625                         $data["text"] = $data["title"];
 626                         $data["title"] = $data["url"];
 627                 }
 628
 629                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 630                         return $data["title"] . $data["after"];
 631                 }
 632
 633                 // If the link already is included in the post, don't add it again
 634                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 635                         return $data["text"] . $data["after"];
 636                 }
 637
 638                 $text = $data["text"];
 639
 640                 if (!empty($data["url"]) && !empty($data["title"])) {
 641                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 642                 } elseif (!empty($data["url"])) {
 643                         $text .= "\n[url]" . $data["url"] . "[/url]";
 644                 }
 645
 646                 return $text . "\n" . $data["after"];
 647         }
 648
 649         /**
 650          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 651          *
 652          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 653          * @param array $match Array with the matching values
 654          * @return string reformatted link including HTML codes
 655          */
 656         private static function convertUrlForOStatusCallback($match)
 657         {
 658                 $url = $match[1];
 659
 660                 if (isset($match[2]) && ($match[1] != $match[2])) {
 661                         return $match[0];
 662                 }
 663
 664                 $parts = parse_url($url);
 665                 if (!isset($parts['scheme'])) {
 666                         return $match[0];
 667                 }
 668
 669                 return self::convertUrlForOStatus($url);
 670         }
 671
 672         /**
 673          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 674          * @param string $url URL that is about to be reformatted
 675          * @return string reformatted link including HTML codes
 676          */
 677         private static function convertUrlForOStatus($url)
 678         {
 679                 $parts = parse_url($url);
 680                 $scheme = $parts['scheme'] . '://';
 681                 $styled_url = str_replace($scheme, '', $url);
 682
 683                 if (strlen($styled_url) > 30) {
 684                         $styled_url = substr($styled_url, 0, 30) . "…";
 685                 }
 686
 687                 $html = '<a href="%s" target="_blank">%s</a>';
 688
 689                 return sprintf($html, $url, $styled_url);
 690         }
 691
 692         /*
 693          * [noparse][i]italic[/i][/noparse] turns into
 694          * [noparse][ i ]italic[ /i ][/noparse],
 695          * to hide them from parser.
 696          */
 697         private static function escapeNoparseCallback($match)
 698         {
 699                 $whole_match = $match[0];
 700                 $captured = $match[1];
 701                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 702                 $new_str = str_replace($captured, $spacefied, $whole_match);
 703                 return $new_str;
 704         }
 705
 706         /*
 707          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 708          * now turns back and the [noparse] tags are trimed
 709          * returning [i]italic[/i]
 710          */
 711         private static function unescapeNoparseCallback($match)
 712         {
 713                 $captured = $match[1];
 714                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 715                 return $unspacefied;
 716         }
 717
 718         /**
 719          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 720          * occurrences
 721          *
 722          * @param string $text        Text to search
 723          * @param string $name        Tag name
 724          * @param int    $occurrences Number of first occurrences to skip
 725          * @return boolean|array
 726          */
 727         public static function getTagPosition($text, $name, $occurrences = 0)
 728         {
 729                 if ($occurrences < 0) {
 730                         $occurrences = 0;
 731                 }
 732
 733                 $start_open = -1;
 734                 for ($i = 0; $i <= $occurrences; $i++) {
 735                         if ($start_open !== false) {
 736                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 737                         }
 738                 }
 739
 740                 if ($start_open === false) {
 741                         return false;
 742                 }
 743
 744                 $start_equal = strpos($text, '=', $start_open);
 745                 $start_close = strpos($text, ']', $start_open);
 746
 747                 if ($start_close === false) {
 748                         return false;
 749                 }
 750
 751                 $start_close++;
 752
 753                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 754
 755                 if ($end_open === false) {
 756                         return false;
 757                 }
 758
 759                 $res = [
 760                         'start' => [
 761                                 'open' => $start_open,
 762                                 'close' => $start_close
 763                         ],
 764                         'end' => [
 765                                 'open' => $end_open,
 766                                 'close' => $end_open + strlen('[/' . $name . ']')
 767                         ],
 768                 ];
 769
 770                 if ($start_equal !== false) {
 771                         $res['start']['equal'] = $start_equal + 1;
 772                 }
 773
 774                 return $res;
 775         }
 776
 777         /**
 778          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 779          *
 780          * @param type $pattern Preg pattern string
 781          * @param type $replace Preg replace string
 782          * @param type $name    BBCode tag name
 783          * @param type $text    Text to search
 784          * @return string
 785          */
 786         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 787         {
 788                 $occurrences = 0;
 789                 $pos = self::getTagPosition($text, $name, $occurrences);
 790                 while ($pos !== false && $occurrences++ < 1000) {
 791                         $start = substr($text, 0, $pos['start']['open']);
 792                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 793                         $end = substr($text, $pos['end']['close']);
 794                         if ($end === false) {
 795                                 $end = '';
 796                         }
 797
 798                         $subject = preg_replace($pattern, $replace, $subject);
 799                         $text = $start . $subject . $end;
 800
 801                         $pos = self::getTagPosition($text, $name, $occurrences);
 802                 }
 803
 804                 return $text;
 805         }
 806
 807         private static function extractImagesFromItemBody($body)
 808         {
 809                 $saved_image = [];
 810                 $orig_body = $body;
 811                 $new_body = '';
 812
 813                 $cnt = 0;
 814                 $img_start = strpos($orig_body, '[img');
 815                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 816                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 817                 while (($img_st_close !== false) && ($img_end !== false)) {
 818                         $img_st_close++; // make it point to AFTER the closing bracket
 819                         $img_end += $img_start;
 820
 821                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 822                                 // This is an embedded image
 823                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 824                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 825
 826                                 $cnt++;
 827                         } else {
 828                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 829                         }
 830
 831                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 832
 833                         if ($orig_body === false) {
 834                                 // in case the body ends on a closing image tag
 835                                 $orig_body = '';
 836                         }
 837
 838                         $img_start = strpos($orig_body, '[img');
 839                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 840                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 841                 }
 842
 843                 $new_body = $new_body . $orig_body;
 844
 845                 return ['body' => $new_body, 'images' => $saved_image];
 846         }
 847
 848         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 849         {
 850                 $newbody = $body;
 851
 852                 $cnt = 0;
 853                 foreach ($images as $image) {
 854                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 855                         // it loops over the array starting from the first element and going sequentially
 856                         // to the last element
 857                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 858                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 859                         $cnt++;
 860                 }
 861
 862                 return $newbody;
 863         }
 864
 865         /**
 866          * This function converts a [share] block to text according to a provided callback function whose signature is:
 867          *
 868          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 869          *
 870          * Where:
 871          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 872          * data lookup
 873          * - $author_contact is a contact record array
 874          * - $content is the inner content of the [share] block
 875          * - $is_quote_share indicates whether there's any content before the [share] block
 876          * - Return value is the string that should replace the [share] block in the provided text
 877          *
 878          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 879          *
 880          * @param  string   $text     A BBCode string
 881          * @param  callable $callback
 882          * @return string The BBCode string with all [share] blocks replaced
 883          */
 884         public static function convertShare($text, callable $callback)
 885         {
 886                 $return = preg_replace_callback(
 887                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 888                         function ($match) use ($callback) {
 889                                 $attribute_string = $match[2];
 890
 891                                 $attributes = [];
 892                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 893                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 894                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 895                                 }
 896
 897                                 // We only call this so that a previously unknown contact can be added.
 898                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 899                                 // This function then can fetch an entry from the contact table.
 900                                 Contact::getIdForURL($attributes['profile'], 0, true);
 901
 902                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 903                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 904
 905                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 906                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 907                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 908
 909                                 if ($attributes['avatar']) {
 910                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 911                                 }
 912
 913                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 914                         },
 915                         $text
 916                 );
 917
 918                 return $return;
 919         }
 920
 921         /**
 922          * Default [share] tag conversion callback
 923          *
 924          * Note: Can produce a [bookmark] tag in the output
 925          *
 926          * @see BBCode::convertShare()
 927          * @param array   $attributes     [share] block attribute values
 928          * @param array   $author_contact Contact row of the shared author
 929          * @param string  $content        Inner content of the [share] block
 930          * @param boolean $is_quote_share Whether there is content before the [share] block
 931          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 932          * @return string
 933          */
 934         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 935         {
 936                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 937
 938                 switch ($simplehtml) {
 939                         case 1:
 940                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 941                                 break;
 942                         case 2:
 943                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 944                                 break;
 945                         case 3: // Diaspora
 946                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . $mention . ':</b></p>' . "\n";
 947
 948                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 949                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 950                                 } else {
 951                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 952
 953                                         if ($attributes['link'] != '') {
 954                                                 $text .= '<p><a href="' . $attributes['link'] . '">[l]</a></p>' . "\n";
 955                                         }
 956                                 }
 957
 958                                 break;
 959                         case 4:
 960                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 961                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 962                                 $headline .= ':</b></p>' . "\n";
 963
 964                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 965
 966                                 break;
 967                         case 5:
 968                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 969                                 break;
 970                         case 7: // statusnet/GNU Social
 971                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 972                                 break;
 973                         case 9: // Google+
 974                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 975                                 $text .= '<p>' . $content . '</p>' . "\n";
 976
 977                                 if ($attributes['link'] != '') {
 978                                         $text .= '<p>' . $attributes['link'] . '</p>';
 979                                 }
 980                                 break;
 981                         default:
 982                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
 983                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
 984                                         try {
 985                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
 986                                         } catch (Exception $e) {
 987                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
 988                                         }
 989                                 } else {
 990                                         $text = ($is_quote_share? "\n" : '');
 991
 992                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
 993                                         $text .= Renderer::replaceMacros($tpl, [
 994                                                 '$profile' => $attributes['profile'],
 995                                                 '$avatar'  => $attributes['avatar'],
 996                                                 '$author'  => $attributes['author'],
 997                                                 '$link'    => $attributes['link'],
 998                                                 '$posted'  => $attributes['posted'],
 999                                                 '$content' => trim($content)
1000                                         ]);
1001                                 }
1002                                 break;
1003                 }
1004
1005                 return $text;
1006         }
1007
1008         private static function removePictureLinksCallback($match)
1009         {
1010                 $text = Cache::get($match[1]);
1011
1012                 if (is_null($text)) {
1013                         $a = self::getApp();
1014
1015                         $stamp1 = microtime(true);
1016
1017                         $ch = @curl_init($match[1]);
1018                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1019                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1020                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1021                         @curl_exec($ch);
1022                         $curl_info = @curl_getinfo($ch);
1023
1024                         $a->saveTimestamp($stamp1, "network");
1025
1026                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1027                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1028                         } else {
1029                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1030
1031                                 // if its not a picture then look if its a page that contains a picture link
1032                                 $body = Network::fetchUrl($match[1]);
1033
1034                                 $doc = new DOMDocument();
1035                                 @$doc->loadHTML($body);
1036                                 $xpath = new DOMXPath($doc);
1037                                 $list = $xpath->query("//meta[@name]");
1038                                 foreach ($list as $node) {
1039                                         $attr = [];
1040
1041                                         if ($node->attributes->length) {
1042                                                 foreach ($node->attributes as $attribute) {
1043                                                         $attr[$attribute->name] = $attribute->value;
1044                                                 }
1045                                         }
1046
1047                                         if (strtolower($attr["name"]) == "twitter:image") {
1048                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1049                                         }
1050                                 }
1051                         }
1052                         Cache::set($match[1], $text);
1053                 }
1054
1055                 return $text;
1056         }
1057
1058         private static function expandLinksCallback($match)
1059         {
1060                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1061                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1062                 } else {
1063                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1064                 }
1065         }
1066
1067         private static function cleanPictureLinksCallback($match)
1068         {
1069                 $text = Cache::get($match[1]);
1070
1071                 if (is_null($text)) {
1072                         $a = self::getApp();
1073
1074                         $stamp1 = microtime(true);
1075
1076                         $ch = @curl_init($match[1]);
1077                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1078                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1079                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1080                         @curl_exec($ch);
1081                         $curl_info = @curl_getinfo($ch);
1082
1083                         $a->saveTimestamp($stamp1, "network");
1084
1085                         // if its a link to a picture then embed this picture
1086                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1087                                 $text = "[img]" . $match[1] . "[/img]";
1088                         } else {
1089                                 $text = "[img]" . $match[2] . "[/img]";
1090
1091                                 // if its not a picture then look if its a page that contains a picture link
1092                                 $body = Network::fetchUrl($match[1]);
1093
1094                                 $doc = new DOMDocument();
1095                                 @$doc->loadHTML($body);
1096                                 $xpath = new DOMXPath($doc);
1097                                 $list = $xpath->query("//meta[@name]");
1098                                 foreach ($list as $node) {
1099                                         $attr = [];
1100                                         if ($node->attributes->length) {
1101                                                 foreach ($node->attributes as $attribute) {
1102                                                         $attr[$attribute->name] = $attribute->value;
1103                                                 }
1104                                         }
1105
1106                                         if (strtolower($attr["name"]) == "twitter:image") {
1107                                                 $text = "[img]" . $attr["content"] . "[/img]";
1108                                         }
1109                                 }
1110                         }
1111                         Cache::set($match[1], $text);
1112                 }
1113
1114                 return $text;
1115         }
1116
1117         public static function cleanPictureLinks($text)
1118         {
1119                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1120                 return $return;
1121         }
1122
1123         /**
1124          * @brief Converts a BBCode message to HTML message
1125          *
1126          * BBcode 2 HTML was written by WAY2WEB.net
1127          * extended to work with Mistpark/Friendica - Mike Macgirvin
1128          *
1129          * Simple HTML values meaning:
1130          * - 0: Friendica display
1131          * - 1: Unused
1132          * - 2: Used for Google+, Windows Phone push, Friendica API
1133          * - 3: Used before converting to Markdown in bb2diaspora.php
1134          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1135          * - 5: Unused
1136          * - 6: Unused
1137          * - 7: Used for dfrn, OStatus
1138          * - 8: Used for WP backlink text setting
1139          *
1140          * @param string $text
1141          * @param bool   $try_oembed
1142          * @param int    $simple_html
1143          * @param bool   $for_plaintext
1144          * @return string
1145          */
1146         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1147         {
1148                 $a = self::getApp();
1149
1150                 /*
1151                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1152                  *
1153                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1154                  * $match[1] = $url
1155                  * $match[2] = $title or absent
1156                  */
1157                 $try_oembed_callback = function ($match)
1158                 {
1159                         $url = $match[1];
1160                         $title = defaults($match, 2, null);
1161
1162                         try {
1163                                 $return = OEmbed::getHTML($url, $title);
1164                         } catch (Exception $ex) {
1165                                 $return = $match[0];
1166                         }
1167
1168                         return $return;
1169                 };
1170
1171                 // Extracting multi-line code blocks before the whitespace processing
1172                 $codeblocks = [];
1173
1174                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1175                         function ($matches) use (&$codeblocks) {
1176                                 $return = $matches[0];
1177                                 if (strpos($matches[2], "\n") !== false) {
1178                                         $return = '#codeblock-' . count($codeblocks) . '#';
1179
1180                                         $codeblocks[] =  '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1181                                 }
1182                                 return $return;
1183                         },
1184                         $text
1185                 );
1186
1187                 // Hide all [noparse] contained bbtags by spacefying them
1188                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1189
1190                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1191                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1192                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1193
1194                 // Remove the abstract element. It is a non visible element.
1195                 $text = self::stripAbstract($text);
1196
1197                 // Move all spaces out of the tags
1198                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1199                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1200
1201                 // Extract the private images which use data urls since preg has issues with
1202                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1203                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1204
1205                 $extracted = self::extractImagesFromItemBody($text);
1206                 $text = $extracted['body'];
1207                 $saved_image = $extracted['images'];
1208
1209                 // If we find any event code, turn it into an event.
1210                 // After we're finished processing the bbcode we'll
1211                 // replace all of the event code with a reformatted version.
1212
1213                 $ev = Event::fromBBCode($text);
1214
1215                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1216                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1217
1218                 $text = str_replace("<", "&lt;", $text);
1219                 $text = str_replace(">", "&gt;", $text);
1220
1221                 // remove some newlines before the general conversion
1222                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1223                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1224
1225                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1226                 if (!$try_oembed) {
1227                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1228                 }
1229
1230                 // Convert new line chars to html <br /> tags
1231
1232                 // nlbr seems to be hopelessly messed up
1233                 //      $Text = nl2br($Text);
1234
1235                 // We'll emulate it.
1236
1237                 $text = trim($text);
1238                 $text = str_replace("\r\n", "\n", $text);
1239
1240                 // removing multiplicated newlines
1241                 if (Config::get("system", "remove_multiplicated_lines")) {
1242                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1243                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1244                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1245                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1246                         do {
1247                                 $oldtext = $text;
1248                                 $text = str_replace($search, $replace, $text);
1249                         } while ($oldtext != $text);
1250                 }
1251
1252                 // Set up the parameters for a URL search string
1253                 $URLSearchString = "^\[\]";
1254                 // Set up the parameters for a MAIL search string
1255                 $MAILSearchString = $URLSearchString;
1256
1257                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1258                 if (!$for_plaintext) {
1259                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1260                         // Currently disabled, since the function is too greedy
1261                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1262                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1263                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1264                         if ($simple_html == 7) {
1265                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1266                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1267                         }
1268                 } else {
1269                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1270                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1271                 }
1272
1273
1274                 // Handle attached links or videos
1275                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1276
1277                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1278
1279                 // Remove all hashtag addresses
1280                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1281                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1282                 } elseif ($simple_html == 3) {
1283                         // The ! is converted to @ since Diaspora only understands the @
1284                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1285                                 '@<a href="$2">$3</a>',
1286                                 $text);
1287                 } elseif ($simple_html == 7) {
1288                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1289                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1290                                 $text);
1291                 } elseif (!$simple_html) {
1292                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1293                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1294                                 $text);
1295                 }
1296
1297                 // Bookmarks in red - will be converted to bookmarks in friendica
1298                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1299                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1300                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1301                                         "[bookmark=$1]$2[/bookmark]", $text);
1302
1303                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1304                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1305                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1306                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1307                 }
1308
1309                 if ($simple_html == 5) {
1310                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1311                 }
1312
1313                 // Perform URL Search
1314                 if ($try_oembed) {
1315                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1316                 }
1317
1318                 if ($simple_html == 5) {
1319                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1320                 } else {
1321                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1322                 }
1323
1324                 // Handle Diaspora posts
1325                 $text = preg_replace_callback(
1326                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1327                         function ($match) {
1328                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1329                         }, $text
1330                 );
1331
1332                 $text = preg_replace_callback(
1333                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1334                         function ($match) {
1335                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1336                         }, $text
1337                 );
1338
1339                 // Server independent link to posts and comments
1340                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1341                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1342                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1343
1344                 /* Tag conversion
1345                  * Supports:
1346                  * - #[url=<anything>]<term>[/url]
1347                  * - [url=<anything>]#<term>[/url]
1348                  */
1349                 $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1350                         return '#<a href="'
1351                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1352                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1353                                 . XML::escape($matches[1])
1354                                 . '</a>';
1355                 }, $text);
1356
1357                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1358                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1359
1360                 // Red compatibility, though the link can't be authenticated on Friendica
1361                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1362
1363
1364                 // we may need to restrict this further if it picks up too many strays
1365                 // link acct:user@host to a webfinger profile redirector
1366
1367                 $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);
1368
1369                 // Perform MAIL Search
1370                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1371                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1372
1373                 // leave open the posibility of [map=something]
1374                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1375
1376                 if (strpos($text, '[/map]') !== false) {
1377                         $text = preg_replace_callback(
1378                                 "/\[map\](.*?)\[\/map\]/ism",
1379                                 function ($match) use ($simple_html) {
1380                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1381                                 },
1382                                 $text
1383                         );
1384                 }
1385                 if (strpos($text, '[map=') !== false) {
1386                         $text = preg_replace_callback(
1387                                 "/\[map=(.*?)\]/ism",
1388                                 function ($match) use ($simple_html) {
1389                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1390                                 },
1391                                 $text
1392                         );
1393                 }
1394                 if (strpos($text, '[map]') !== false) {
1395                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1396                 }
1397
1398                 // Check for headers
1399                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1400                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1401                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1402                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1403                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1404                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1405
1406                 // Check for paragraph
1407                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1408
1409                 // Check for bold text
1410                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1411
1412                 // Check for Italics text
1413                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1414
1415                 // Check for Underline text
1416                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1417
1418                 // Check for strike-through text
1419                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1420
1421                 // Check for over-line text
1422                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1423
1424                 // Check for colored text
1425                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1426
1427                 // Check for sized text
1428                 // [size=50] --> font-size: 50px (with the unit).
1429                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1430                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1431
1432                 // Check for centered text
1433                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1434
1435                 // Check for list text
1436                 $text = str_replace("[*]", "<li>", $text);
1437
1438                 // Check for style sheet commands
1439                 $text = preg_replace_callback(
1440                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1441                         function ($match) {
1442                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1443                         },
1444                         $text
1445                 );
1446
1447                 // Check for CSS classes
1448                 $text = preg_replace_callback(
1449                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1450                         function ($match) {
1451                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1452                         },
1453                         $text
1454                 );
1455
1456                 // handle nested lists
1457                 $endlessloop = 0;
1458
1459                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1460                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1461                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1462                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1463                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1464                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1465                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1466                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1467                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1468                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1469                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1470                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1471                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1472                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1473                 }
1474
1475                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1476                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1477                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1478                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1479
1480                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1481                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1482
1483                 $text = str_replace('[hr]', '<hr />', $text);
1484
1485                 // This is actually executed in Item::prepareBody()
1486
1487                 $text = str_replace('[nosmile]', '', $text);
1488
1489                 // Check for font change text
1490                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1491
1492                 // Declare the format for [code] layout
1493
1494                 $CodeLayout = '<code>$1</code>';
1495                 // Check for [code] text
1496                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1497
1498                 // Declare the format for [spoiler] layout
1499                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1500
1501                 // Check for [spoiler] text
1502                 // handle nested quotes
1503                 $endlessloop = 0;
1504                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1505                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1506                 }
1507
1508                 // Check for [spoiler=Author] text
1509
1510                 $t_wrote = L10n::t('$1 wrote:');
1511
1512                 // handle nested quotes
1513                 $endlessloop = 0;
1514                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1515                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1516                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1517                                                  $text);
1518                 }
1519
1520                 // Declare the format for [quote] layout
1521                 $QuoteLayout = '<blockquote>$1</blockquote>';
1522
1523                 // Check for [quote] text
1524                 // handle nested quotes
1525                 $endlessloop = 0;
1526                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1527                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1528                 }
1529
1530                 // Check for [quote=Author] text
1531
1532                 $t_wrote = L10n::t('$1 wrote:');
1533
1534                 // handle nested quotes
1535                 $endlessloop = 0;
1536                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1537                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1538                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1539                                                  $text);
1540                 }
1541
1542
1543                 // [img=widthxheight]image source[/img]
1544                 $text = preg_replace_callback(
1545                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1546                         function ($matches) use ($simple_html) {
1547                                 if (strpos($matches[3], "data:image/") === 0) {
1548                                         return $matches[0];
1549                                 }
1550
1551                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1552                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1553                         },
1554                         $text
1555                 );
1556
1557                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1558                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1559
1560                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1561                         function ($matches) use ($simple_html) {
1562                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1563                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1564                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1565                         },
1566                         $text);
1567
1568                 // Images
1569                 // [img]pathtoimage[/img]
1570                 $text = preg_replace_callback(
1571                         "/\[img\](.*?)\[\/img\]/ism",
1572                         function ($matches) use ($simple_html) {
1573                                 if (strpos($matches[1], "data:image/") === 0) {
1574                                         return $matches[0];
1575                                 }
1576
1577                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1578                                 return "[img]" . $matches[1] . "[/img]";
1579                         },
1580                         $text
1581                 );
1582
1583                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1584                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1585
1586                 // Shared content
1587                 $text = self::convertShare(
1588                         $text,
1589                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1590                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1591                         }
1592                 );
1593
1594                 $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);
1595                 $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 />', $text);
1596                 //$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 />', $Text);
1597
1598                 // Try to Oembed
1599                 if ($try_oembed) {
1600                         $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);
1601                         $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);
1602
1603                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1604                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1605                 } else {
1606                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1607                                                 '<a href="$1" target="_blank">$1</a>', $text);
1608                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1609                                                 '<a href="$1" target="_blank">$1</a>', $text);
1610                 }
1611
1612                 // html5 video and audio
1613
1614
1615                 if ($try_oembed) {
1616                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1617                 } else {
1618                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1619                 }
1620
1621                 // Youtube extensions
1622                 if ($try_oembed) {
1623                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1624                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1625                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1626                 }
1627
1628                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1629                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1630                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1631
1632                 if ($try_oembed) {
1633                         $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);
1634                 } else {
1635                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1636                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1637                 }
1638
1639                 if ($try_oembed) {
1640                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1641                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1642                 }
1643
1644                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1645                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1646
1647                 if ($try_oembed) {
1648                         $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);
1649                 } else {
1650                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1651                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1652                 }
1653
1654                 // oembed tag
1655                 $text = OEmbed::BBCode2HTML($text);
1656
1657                 // Avoid triple linefeeds through oembed
1658                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1659
1660                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1661                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1662                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1663                 // start which is always required). Allow desc with a missing summary for compatibility.
1664
1665                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1666                         $sub = Event::getHTML($ev, $simple_html);
1667
1668                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1669                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1670                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1671                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1672                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1673                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1674                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1675                 }
1676
1677                 // Replace non graphical smilies for external posts
1678                 if ($simple_html) {
1679                         $text = Smilies::replace($text, false, true);
1680                 }
1681
1682                 // Unhide all [noparse] contained bbtags unspacefying them
1683                 // and triming the [noparse] tag.
1684
1685                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1686                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1687                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1688
1689                 /// @todo What is the meaning of these lines?
1690                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1691                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1692
1693                 // Currently deactivated, it made problems with " inside of alt texts.
1694                 //$text = preg_replace('/\&quot\;/', '"', $text);
1695
1696                 // fix any escaped ampersands that may have been converted into links
1697                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1698
1699                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1700                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1701                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1702                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1703
1704                 // sanitize href attributes (only whitelisted protocols URLs)
1705                 // default value for backward compatibility
1706                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1707
1708                 // Always allowed protocol even if config isn't set or not including it
1709                 $allowed_link_protocols[] = 'http';
1710                 $allowed_link_protocols[] = 'redir/';
1711
1712                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1713                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1714
1715                 if ($saved_image) {
1716                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1717                 }
1718
1719                 // Restore code blocks
1720                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1721                         function ($matches) use ($codeblocks) {
1722                                 $return = $matches[0];
1723                                 if (isset($codeblocks[intval($matches[1])])) {
1724                                         $return = $codeblocks[$matches[1]];
1725                                 }
1726                                 return $return;
1727                         },
1728                         $text
1729                 );
1730
1731                 // Clean up the HTML by loading and saving the HTML with the DOM.
1732                 // Bad structured html can break a whole page.
1733                 // For performance reasons do it only with ativated item cache or at export.
1734                 if (!$try_oembed || (get_itemcachepath() != "")) {
1735                         $doc = new DOMDocument();
1736                         $doc->preserveWhiteSpace = false;
1737
1738                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1739
1740                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1741                         $encoding = '<?xml encoding="UTF-8">';
1742                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1743                         $doc->encoding = 'UTF-8';
1744                         $text = $doc->saveHTML();
1745                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1746
1747                         $text = str_replace('<br></li>', '</li>', $text);
1748
1749                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1750                 }
1751
1752                 // Clean up some useless linebreaks in lists
1753                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1754                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1755                 //$Text = str_replace('</li><br />', '</li>', $Text);
1756                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1757                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1758
1759                 Addon::callHooks('bbcode', $text);
1760
1761                 return trim($text);
1762         }
1763
1764         /**
1765          * @brief Strips the "abstract" tag from the provided text
1766          *
1767          * @param string $text The text with BBCode
1768          * @return string The same text - but without "abstract" element
1769          */
1770         public static function stripAbstract($text)
1771         {
1772                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1773                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1774
1775                 return $text;
1776         }
1777
1778         /**
1779          * @brief Returns the value of the "abstract" element
1780          *
1781          * @param string $text The text that maybe contains the element
1782          * @param string $addon The addon for which the abstract is meant for
1783          * @return string The abstract
1784          */
1785         public static function getAbstract($text, $addon = "")
1786         {
1787                 $abstract = "";
1788                 $abstracts = [];
1789                 $addon = strtolower($addon);
1790
1791                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1792                         foreach ($results AS $result) {
1793                                 $abstracts[strtolower($result[1])] = $result[2];
1794                         }
1795                 }
1796
1797                 if (isset($abstracts[$addon])) {
1798                         $abstract = $abstracts[$addon];
1799                 }
1800
1801                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1802                         $abstract = $result[1];
1803                 }
1804
1805                 return $abstract;
1806         }
1807
1808         /**
1809          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1810          *
1811          * @param array $match Matching values for the callback
1812          * @return string Replaced mention
1813          */
1814         private static function bbCodeMention2DiasporaCallback($match)
1815         {
1816                 $contact = Contact::getDetailsByURL($match[3]);
1817
1818                 if (empty($contact['addr'])) {
1819                         $contact = Probe::uri($match[3]);
1820                 }
1821
1822                 if (empty($contact['addr'])) {
1823                         return $match[0];
1824                 }
1825
1826                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1827                 return $mention;
1828         }
1829
1830         /**
1831          * @brief Converts a BBCode text into Markdown
1832          *
1833          * This function converts a BBCode item body to be sent to Markdown-enabled
1834          * systems like Diaspora and Libertree
1835          *
1836          * @param string $text
1837          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1838          * @return string
1839          */
1840         public static function toMarkdown($text, $for_diaspora = true)
1841         {
1842                 $a = self::getApp();
1843
1844                 $original_text = $text;
1845
1846                 // Since Diaspora is creating a summary for links, this function removes them before posting
1847                 if ($for_diaspora) {
1848                         $text = self::removeShareInformation($text);
1849                 }
1850
1851                 /**
1852                  * Transform #tags, strip off the [url] and replace spaces with underscore
1853                  */
1854                 $url_search_string = "^\[\]";
1855                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1856                         function ($matches) {
1857                                 return '#' . str_replace(' ', '_', $matches[2]);
1858                         },
1859                         $text
1860                 );
1861
1862                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1863                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1864
1865                 // Convert it to HTML - don't try oembed
1866                 if ($for_diaspora) {
1867                         $text = self::convert($text, false, 3);
1868
1869                         // Add all tags that maybe were removed
1870                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1871                                 $tagline = "";
1872                                 foreach ($tags[2] as $tag) {
1873                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1874                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1875                                                 $tagline .= '#' . $tag . ' ';
1876                                         }
1877                                 }
1878                                 $text = $text . " " . $tagline;
1879                         }
1880                 } else {
1881                         $text = self::convert($text, false, 4);
1882                 }
1883
1884                 // mask some special HTML chars from conversation to markdown
1885                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1886
1887                 // If a link is followed by a quote then there should be a newline before it
1888                 // Maybe we should make this newline at every time before a quote.
1889                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1890
1891                 $stamp1 = microtime(true);
1892
1893                 // Now convert HTML to Markdown
1894                 $text = HTML::toMarkdown($text);
1895
1896                 // unmask the special chars back to HTML
1897                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1898
1899                 $a->saveTimestamp($stamp1, "parser");
1900
1901                 // Libertree has a problem with escaped hashtags.
1902                 $text = str_replace(['\#'], ['#'], $text);
1903
1904                 // Remove any leading or trailing whitespace, as this will mess up
1905                 // the Diaspora signature verification and cause the item to disappear
1906                 $text = trim($text);
1907
1908                 if ($for_diaspora) {
1909                         $url_search_string = "^\[\]";
1910                         $text = preg_replace_callback(
1911                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
1912                                 ['self', 'bbCodeMention2DiasporaCallback'],
1913                                 $text
1914                         );
1915                 }
1916
1917                 Addon::callHooks('bb2diaspora', $text);
1918
1919                 return $text;
1920         }
1921
1922         /**
1923      * @brief Pull out all #hashtags and @person tags from $string.
1924      *
1925      * We also get @person@domain.com - which would make
1926      * the regex quite complicated as tags can also
1927      * end a sentence. So we'll run through our results
1928      * and strip the period from any tags which end with one.
1929      * Returns array of tags found, or empty array.
1930      *
1931      * @param string $string Post content
1932      *
1933      * @return array List of tag and person names
1934      */
1935     public static function getTags($string)
1936     {
1937         $ret = [];
1938
1939         // Convert hashtag links to hashtags
1940         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1941
1942         // ignore anything in a code block
1943         $string = preg_replace('/\[code\](.*?)\[\/code\]/sm', '', $string);
1944
1945         // Force line feeds at bbtags
1946         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1947
1948         // ignore anything in a bbtag
1949         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1950
1951         // Match full names against @tags including the space between first and last
1952         // We will look these up afterward to see if they are full names or not recognisable.
1953
1954         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1955             foreach ($matches[1] as $match) {
1956                 if (strstr($match, ']')) {
1957                     // we might be inside a bbcode color tag - leave it alone
1958                     continue;
1959                 }
1960
1961                 if (substr($match, -1, 1) === '.') {
1962                     $ret[] = substr($match, 0, -1);
1963                 } else {
1964                     $ret[] = $match;
1965                 }
1966             }
1967         }
1968
1969         // Otherwise pull out single word tags. These can be @nickname, @first_last
1970         // and #hash tags.
1971
1972         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
1973             foreach ($matches[1] as $match) {
1974                 if (strstr($match, ']')) {
1975                     // we might be inside a bbcode color tag - leave it alone
1976                     continue;
1977                 }
1978                 if (substr($match, -1, 1) === '.') {
1979                     $match = substr($match,0,-1);
1980                 }
1981                 // ignore strictly numeric tags like #1
1982                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
1983                     continue;
1984                 }
1985                 // try not to catch url fragments
1986                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
1987                     continue;
1988                 }
1989                 $ret[] = $match;
1990             }
1991         }
1992
1993         return $ret;
1994     }
1995 }