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\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\Hook;
  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 $body Message body
  39          * @return array
  40          *                     'type' -> Message type ("link", "video", "photo")
  41          *                     'text' -> Text before the shared message
  42          *                     'after' -> Text after the shared message
  43          *                     'image' -> Preview image of the message
  44          *                     'url' -> Url to the attached message
  45          *                     'title' -> Title of the attachment
  46          *                     'description' -> Description of the attachment
  47          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  48          */
  49         private static function getOldAttachmentData($body)
  50         {
  51                 $post = [];
  52
  53                 // Simplify image codes
  54                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  55
  56                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  57                         foreach ($attached as $data) {
  58                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  59                                         continue;
  60                                 }
  61
  62                                 $post["type"] = substr($data[1], 5);
  63
  64                                 $pos = strpos($body, $data[0]);
  65                                 if ($pos > 0) {
  66                                         $post["text"] = trim(substr($body, 0, $pos));
  67                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  68                                 } else {
  69                                         $post["text"] = trim(str_replace($data[0], "", $body));
  70                                 }
  71
  72                                 $attacheddata = $data[2];
  73
  74                                 $URLSearchString = "^\[\]";
  75
  76                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  77
  78                                         $picturedata = Image::getInfoFromURL($matches[1]);
  79
  80                                         if ($picturedata) {
  81                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  82                                                         $post["image"] = $matches[1];
  83                                                 } else {
  84                                                         $post["preview"] = $matches[1];
  85                                                 }
  86                                         }
  87                                 }
  88
  89                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  90                                         $post["url"] = $matches[1];
  91                                         $post["title"] = $matches[2];
  92                                 }
  93                                 if (!empty($post["url"]) && (in_array($post["type"], ["link", "video"]))
  94                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  95                                         $post["url"] = $matches[1];
  96                                 }
  97
  98                                 // Search for description
  99                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 100                                         $post["description"] = $matches[1];
 101                                 }
 102                         }
 103                 }
 104                 return $post;
 105         }
 106
 107         /**
 108          * @brief Fetches attachment data that were generated with the "attachment" element
 109          *
 110          * @param string $body Message body
 111          * @return array
 112          *                     'type' -> Message type ("link", "video", "photo")
 113          *                     'text' -> Text before the shared message
 114          *                     'after' -> Text after the shared message
 115          *                     'image' -> Preview image of the message
 116          *                     'url' -> Url to the attached message
 117          *                     'title' -> Title of the attachment
 118          *                     'description' -> Description of the attachment
 119          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 120          */
 121         public static function getAttachmentData($body)
 122         {
 123                 $data = [];
 124
 125                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 126                         return self::getOldAttachmentData($body);
 127                 }
 128
 129                 $attributes = $match[2];
 130
 131                 $data["text"] = trim($match[1]);
 132
 133                 $type = "";
 134                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 135                 if (!empty($matches[1])) {
 136                         $type = strtolower($matches[1]);
 137                 }
 138
 139                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 140                 if (!empty($matches[1])) {
 141                         $type = strtolower($matches[1]);
 142                 }
 143
 144                 if ($type == "") {
 145                         return [];
 146                 }
 147
 148                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 149                         return [];
 150                 }
 151
 152                 if ($type != "") {
 153                         $data["type"] = $type;
 154                 }
 155
 156                 $url = "";
 157                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 158                 if (!empty($matches[1])) {
 159                         $url = $matches[1];
 160                 }
 161
 162                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 163                 if (!empty($matches[1])) {
 164                         $url = $matches[1];
 165                 }
 166
 167                 if ($url != "") {
 168                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 169                 }
 170
 171                 $title = "";
 172                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 173                 if (!empty($matches[1])) {
 174                         $title = $matches[1];
 175                 }
 176
 177                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 178                 if (!empty($matches[1])) {
 179                         $title = $matches[1];
 180                 }
 181
 182                 if ($title != "") {
 183                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 184                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 185                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 186                         $data["title"] = $title;
 187                 }
 188
 189                 $image = "";
 190                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 191                 if (!empty($matches[1])) {
 192                         $image = $matches[1];
 193                 }
 194
 195                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 196                 if (!empty($matches[1])) {
 197                         $image = $matches[1];
 198                 }
 199
 200                 if ($image != "") {
 201                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 202                 }
 203
 204                 $preview = "";
 205                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 206                 if (!empty($matches[1])) {
 207                         $preview = $matches[1];
 208                 }
 209
 210                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 211                 if (!empty($matches[1])) {
 212                         $preview = $matches[1];
 213                 }
 214
 215                 if ($preview != "") {
 216                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 217                 }
 218
 219                 $data["description"] = trim($match[3]);
 220
 221                 $data["after"] = trim($match[4]);
 222
 223                 return $data;
 224         }
 225
 226         public static function getAttachedData($body, $item = [])
 227         {
 228                 /*
 229                 - text:
 230                 - type: link, video, photo
 231                 - title:
 232                 - url:
 233                 - image:
 234                 - description:
 235                 - (thumbnail)
 236                 */
 237
 238                 $has_title = !empty($item['title']);
 239                 $plink = defaults($item, 'plink', '');
 240                 $post = self::getAttachmentData($body);
 241
 242                 // if nothing is found, it maybe having an image.
 243                 if (!isset($post["type"])) {
 244                         // Simplify image codes
 245                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 246
 247                         $URLSearchString = "^\[\]";
 248
 249                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 250
 251                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 252                                 if ((count($pictures) == 1) && !$has_title) {
 253                                         // Checking, if the link goes to a picture
 254                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 255
 256                                         // Workaround:
 257                                         // Sometimes photo posts to the own album are not detected at the start.
 258                                         // So we seem to cannot use the cache for these cases. That's strange.
 259                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 260                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 261                                         }
 262
 263                                         if ($data["type"] == "photo") {
 264                                                 $post["type"] = "photo";
 265                                                 if (isset($data["images"][0])) {
 266                                                         $post["image"] = $data["images"][0]["src"];
 267                                                         $post["url"] = $data["url"];
 268                                                 } else {
 269                                                         $post["image"] = $data["url"];
 270                                                 }
 271
 272                                                 $post["preview"] = $pictures[0][2];
 273                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 274                                         } else {
 275                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 276                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 277                                                         $post["type"] = "photo";
 278                                                         $post["image"] = $pictures[0][1];
 279                                                         $post["preview"] = $pictures[0][2];
 280                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 281                                                 }
 282                                         }
 283                                 } elseif (count($pictures) > 0) {
 284                                         $post["type"] = "link";
 285                                         $post["url"] = $plink;
 286                                         $post["image"] = $pictures[0][2];
 287                                         $post["text"] = $body;
 288                                 }
 289                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 290                                 if ((count($pictures) == 1) && !$has_title) {
 291                                         $post["type"] = "photo";
 292                                         $post["image"] = $pictures[0][1];
 293                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 294                                 } elseif (count($pictures) > 0) {
 295                                         $post["type"] = "link";
 296                                         $post["url"] = $plink;
 297                                         $post["image"] = $pictures[0][1];
 298                                         $post["text"] = $body;
 299                                 }
 300                         }
 301
 302                         // Test for the external links
 303                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 304                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 305
 306                         $links = array_merge($links1, $links2);
 307
 308                         // If there is only a single one, then use it.
 309                         // This should cover link posts via API.
 310                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 311                                 $post["type"] = "link";
 312                                 $post["text"] = trim($body);
 313                                 $post["url"] = $links[0][1];
 314                         }
 315
 316                         // Now count the number of external media links
 317                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 318                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 319                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 320                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 321
 322                         // Add them to the other external links
 323                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 324
 325                         // Are there more than one?
 326                         if (count($links) > 1) {
 327                                 // The post will be the type "text", which means a blog post
 328                                 unset($post["type"]);
 329                                 $post["url"] = $plink;
 330                         }
 331
 332                         if (!isset($post["type"])) {
 333                                 $post["type"] = "text";
 334                                 $post["text"] = trim($body);
 335                         }
 336                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 337                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 338
 339                         if (isset($data["images"][0])) {
 340                                 $post["image"] = $data["images"][0]["src"];
 341                         }
 342                 }
 343
 344                 return $post;
 345         }
 346
 347         /**
 348          * @brief Converts a BBCode text into plaintext
 349          *
 350          * @param      $text
 351          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 352          *
 353          * @return string
 354          */
 355         public static function toPlaintext($text, $keep_urls = true)
 356         {
 357                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 358                 if (!$keep_urls) {
 359                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 360                 }
 361
 362                 return $naked_text;
 363         }
 364
 365         private static function proxyUrl($image, $simplehtml = false)
 366         {
 367                 // Only send proxied pictures to API and for internal display
 368                 if (in_array($simplehtml, [false, 2])) {
 369                         return ProxyUtils::proxifyUrl($image);
 370                 } else {
 371                         return $image;
 372                 }
 373         }
 374
 375         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 376         {
 377                 // Suppress "view full size"
 378                 if (intval(Config::get('system', 'no_view_full_size'))) {
 379                         $include_link = false;
 380                 }
 381
 382                 // Picture addresses can contain special characters
 383                 $s = htmlspecialchars_decode($srctext);
 384
 385                 $matches = null;
 386                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 387                 if ($c) {
 388                         foreach ($matches as $mtch) {
 389                                 Logger::log('scale_external_image: ' . $mtch[1]);
 390
 391                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 392                                 if (stristr($mtch[1], $hostname)) {
 393                                         continue;
 394                                 }
 395
 396                                 // $scale_replace, if passed, is an array of two elements. The
 397                                 // first is the name of the full-size image. The second is the
 398                                 // name of a remote, scaled-down version of the full size image.
 399                                 // This allows Friendica to display the smaller remote image if
 400                                 // one exists, while still linking to the full-size image
 401                                 if ($scale_replace) {
 402                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 403                                 } else {
 404                                         $scaled = $mtch[1];
 405                                 }
 406                                 $i = Network::fetchUrl($scaled);
 407                                 if (!$i) {
 408                                         return $srctext;
 409                                 }
 410
 411                                 // guess mimetype from headers or filename
 412                                 $type = Image::guessType($mtch[1], true);
 413
 414                                 if ($i) {
 415                                         $Image = new Image($i, $type);
 416                                         if ($Image->isValid()) {
 417                                                 $orig_width = $Image->getWidth();
 418                                                 $orig_height = $Image->getHeight();
 419
 420                                                 if ($orig_width > 640 || $orig_height > 640) {
 421                                                         $Image->scaleDown(640);
 422                                                         $new_width = $Image->getWidth();
 423                                                         $new_height = $Image->getHeight();
 424                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 425                                                         $s = str_replace(
 426                                                                 $mtch[0],
 427                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 428                                                                 . "\n" . (($include_link)
 429                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 430                                                                         : ''),
 431                                                                 $s
 432                                                         );
 433                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 434                                                 }
 435                                         }
 436                                 }
 437                         }
 438                 }
 439
 440                 // replace the special char encoding
 441                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 442                 return $s;
 443         }
 444
 445         /**
 446          * The purpose of this function is to apply system message length limits to
 447          * imported messages without including any embedded photos in the length
 448          *
 449          * @brief Truncates imported message body string length to max_import_size
 450          * @param string $body
 451          * @return string
 452          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 453          */
 454         public static function limitBodySize($body)
 455         {
 456                 $maxlen = Config::get('config', 'max_import_size', 0);
 457
 458                 // If the length of the body, including the embedded images, is smaller
 459                 // than the maximum, then don't waste time looking for the images
 460                 if ($maxlen && (strlen($body) > $maxlen)) {
 461
 462                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 463
 464                         $orig_body = $body;
 465                         $new_body = '';
 466                         $textlen = 0;
 467
 468                         $img_start = strpos($orig_body, '[img');
 469                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 470                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 471                         while (($img_st_close !== false) && ($img_end !== false)) {
 472
 473                                 $img_st_close++; // make it point to AFTER the closing bracket
 474                                 $img_end += $img_start;
 475                                 $img_end += strlen('[/img]');
 476
 477                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 478                                         // This is an embedded image
 479
 480                                         if (($textlen + $img_start) > $maxlen) {
 481                                                 if ($textlen < $maxlen) {
 482                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 483                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 484                                                         $textlen = $maxlen;
 485                                                 }
 486                                         } else {
 487                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 488                                                 $textlen += $img_start;
 489                                         }
 490
 491                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 492                                 } else {
 493
 494                                         if (($textlen + $img_end) > $maxlen) {
 495                                                 if ($textlen < $maxlen) {
 496                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 497                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 498                                                         $textlen = $maxlen;
 499                                                 }
 500                                         } else {
 501                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 502                                                 $textlen += $img_end;
 503                                         }
 504                                 }
 505                                 $orig_body = substr($orig_body, $img_end);
 506
 507                                 if ($orig_body === false) {
 508                                         // in case the body ends on a closing image tag
 509                                         $orig_body = '';
 510                                 }
 511
 512                                 $img_start = strpos($orig_body, '[img');
 513                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 514                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 515                         }
 516
 517                         if (($textlen + strlen($orig_body)) > $maxlen) {
 518                                 if ($textlen < $maxlen) {
 519                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 520                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 521                                 }
 522                         } else {
 523                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 524                                 $new_body = $new_body . $orig_body;
 525                         }
 526
 527                         return $new_body;
 528                 } else {
 529                         return $body;
 530                 }
 531         }
 532
 533         /**
 534          * Processes [attachment] tags
 535          *
 536          * Note: Can produce a [bookmark] tag in the returned string
 537          *
 538          * @brief Processes [attachment] tags
 539          * @param string   $return
 540          * @param bool|int $simplehtml
 541          * @param bool     $tryoembed
 542          * @return string
 543          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 544          */
 545         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 546         {
 547                 $data = self::getAttachmentData($return);
 548                 if (empty($data) || empty($data["url"])) {
 549                         return $return;
 550                 }
 551
 552                 if (isset($data["title"])) {
 553                         $data["title"] = strip_tags($data["title"]);
 554                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 555                 } else {
 556                         $data["title"] = null;
 557                 }
 558
 559                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 560                         $data["preview"] = $data["image"];
 561                         $data["image"] = "";
 562                 }
 563
 564                 $return = '';
 565                 if ($simplehtml == 7) {
 566                         $return = self::convertUrlForOStatus($data["url"]);
 567                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 568                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 569                 } else {
 570                         try {
 571                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 572                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 573                                 } else {
 574                                         throw new Exception('OEmbed is disabled for this attachment.');
 575                                 }
 576                         } catch (Exception $e) {
 577                                 $data["title"] = defaults($data, 'title', $data['url']);
 578
 579                                 if ($simplehtml != 4) {
 580                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 581                                 }
 582
 583                                 if (!empty($data['title']) && !empty($data['url'])) {
 584                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 585                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 586                                         } else {
 587                                                 if (!empty($data["image"])) {
 588                                                         $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), $data["title"]);
 589                                                 } elseif (!empty($data["preview"])) {
 590                                                         $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"]);
 591                                                 }
 592                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 593                                         }
 594                                 }
 595
 596                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 597                                         // Sanitize the HTML by converting it to BBCode
 598                                         $bbcode = HTML::toBBCode($data["description"]);
 599                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 600                                 }
 601
 602                                 if (!empty($data['url'])) {
 603                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 604                                 }
 605
 606                                 if ($simplehtml != 4) {
 607                                         $return .= '</div>';
 608                                 }
 609                         }
 610                 }
 611
 612                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 613         }
 614
 615         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 616         {
 617                 $data = self::getAttachmentData($Text);
 618
 619                 if (!$data) {
 620                         return $Text;
 621                 } elseif ($nolink) {
 622                         return $data["text"] . defaults($data, 'after', '');
 623                 }
 624
 625                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 626                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 627                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 628                         $data["title"] = $data["url"];
 629                 } elseif (($text != "") && strstr($title, $text)) {
 630                         $data["text"] = $data["title"];
 631                         $data["title"] = $data["url"];
 632                 }
 633
 634                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 635                         return $data["title"] . $data["after"];
 636                 }
 637
 638                 // If the link already is included in the post, don't add it again
 639                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 640                         return $data["text"] . $data["after"];
 641                 }
 642
 643                 $text = $data["text"];
 644
 645                 if (!empty($data["url"]) && !empty($data["title"])) {
 646                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 647                 } elseif (!empty($data["url"])) {
 648                         $text .= "\n[url]" . $data["url"] . "[/url]";
 649                 }
 650
 651                 return $text . "\n" . $data["after"];
 652         }
 653
 654         /**
 655          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 656          *
 657          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 658          * @param array $match Array with the matching values
 659          * @return string reformatted link including HTML codes
 660          */
 661         private static function convertUrlForOStatusCallback($match)
 662         {
 663                 $url = $match[1];
 664
 665                 if (isset($match[2]) && ($match[1] != $match[2])) {
 666                         return $match[0];
 667                 }
 668
 669                 $parts = parse_url($url);
 670                 if (!isset($parts['scheme'])) {
 671                         return $match[0];
 672                 }
 673
 674                 return self::convertUrlForOStatus($url);
 675         }
 676
 677         /**
 678          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 679          * @param string $url URL that is about to be reformatted
 680          * @return string reformatted link including HTML codes
 681          */
 682         private static function convertUrlForOStatus($url)
 683         {
 684                 $parts = parse_url($url);
 685                 $scheme = $parts['scheme'] . '://';
 686                 $styled_url = str_replace($scheme, '', $url);
 687
 688                 if (strlen($styled_url) > 30) {
 689                         $styled_url = substr($styled_url, 0, 30) . "…";
 690                 }
 691
 692                 $html = '<a href="%s" target="_blank">%s</a>';
 693
 694                 return sprintf($html, $url, $styled_url);
 695         }
 696
 697         /*
 698          * [noparse][i]italic[/i][/noparse] turns into
 699          * [noparse][ i ]italic[ /i ][/noparse],
 700          * to hide them from parser.
 701          */
 702         private static function escapeNoparseCallback($match)
 703         {
 704                 $whole_match = $match[0];
 705                 $captured = $match[1];
 706                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 707                 $new_str = str_replace($captured, $spacefied, $whole_match);
 708                 return $new_str;
 709         }
 710
 711         /*
 712          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 713          * now turns back and the [noparse] tags are trimed
 714          * returning [i]italic[/i]
 715          */
 716         private static function unescapeNoparseCallback($match)
 717         {
 718                 $captured = $match[1];
 719                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 720                 return $unspacefied;
 721         }
 722
 723         /**
 724          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 725          * occurrences
 726          *
 727          * @param string $text        Text to search
 728          * @param string $name        Tag name
 729          * @param int    $occurrences Number of first occurrences to skip
 730          * @return boolean|array
 731          */
 732         public static function getTagPosition($text, $name, $occurrences = 0)
 733         {
 734                 if ($occurrences < 0) {
 735                         $occurrences = 0;
 736                 }
 737
 738                 $start_open = -1;
 739                 for ($i = 0; $i <= $occurrences; $i++) {
 740                         if ($start_open !== false) {
 741                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 742                         }
 743                 }
 744
 745                 if ($start_open === false) {
 746                         return false;
 747                 }
 748
 749                 $start_equal = strpos($text, '=', $start_open);
 750                 $start_close = strpos($text, ']', $start_open);
 751
 752                 if ($start_close === false) {
 753                         return false;
 754                 }
 755
 756                 $start_close++;
 757
 758                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 759
 760                 if ($end_open === false) {
 761                         return false;
 762                 }
 763
 764                 $res = [
 765                         'start' => [
 766                                 'open' => $start_open,
 767                                 'close' => $start_close
 768                         ],
 769                         'end' => [
 770                                 'open' => $end_open,
 771                                 'close' => $end_open + strlen('[/' . $name . ']')
 772                         ],
 773                 ];
 774
 775                 if ($start_equal !== false) {
 776                         $res['start']['equal'] = $start_equal + 1;
 777                 }
 778
 779                 return $res;
 780         }
 781
 782         /**
 783          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 784          *
 785          * @param string $pattern Preg pattern string
 786          * @param string $replace Preg replace string
 787          * @param string $name    BBCode tag name
 788          * @param string $text    Text to search
 789          * @return string
 790          */
 791         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 792         {
 793                 $occurrences = 0;
 794                 $pos = self::getTagPosition($text, $name, $occurrences);
 795                 while ($pos !== false && $occurrences++ < 1000) {
 796                         $start = substr($text, 0, $pos['start']['open']);
 797                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 798                         $end = substr($text, $pos['end']['close']);
 799                         if ($end === false) {
 800                                 $end = '';
 801                         }
 802
 803                         $subject = preg_replace($pattern, $replace, $subject);
 804                         $text = $start . $subject . $end;
 805
 806                         $pos = self::getTagPosition($text, $name, $occurrences);
 807                 }
 808
 809                 return $text;
 810         }
 811
 812         private static function extractImagesFromItemBody($body)
 813         {
 814                 $saved_image = [];
 815                 $orig_body = $body;
 816                 $new_body = '';
 817
 818                 $cnt = 0;
 819                 $img_start = strpos($orig_body, '[img');
 820                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 821                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 822                 while (($img_st_close !== false) && ($img_end !== false)) {
 823                         $img_st_close++; // make it point to AFTER the closing bracket
 824                         $img_end += $img_start;
 825
 826                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 827                                 // This is an embedded image
 828                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 829                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 830
 831                                 $cnt++;
 832                         } else {
 833                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 834                         }
 835
 836                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 837
 838                         if ($orig_body === false) {
 839                                 // in case the body ends on a closing image tag
 840                                 $orig_body = '';
 841                         }
 842
 843                         $img_start = strpos($orig_body, '[img');
 844                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 845                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 846                 }
 847
 848                 $new_body = $new_body . $orig_body;
 849
 850                 return ['body' => $new_body, 'images' => $saved_image];
 851         }
 852
 853         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 854         {
 855                 $newbody = $body;
 856
 857                 $cnt = 0;
 858                 foreach ($images as $image) {
 859                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 860                         // it loops over the array starting from the first element and going sequentially
 861                         // to the last element
 862                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 863                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 864                         $cnt++;
 865                 }
 866
 867                 return $newbody;
 868         }
 869
 870         /**
 871          * This function converts a [share] block to text according to a provided callback function whose signature is:
 872          *
 873          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 874          *
 875          * Where:
 876          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 877          * data lookup
 878          * - $author_contact is a contact record array
 879          * - $content is the inner content of the [share] block
 880          * - $is_quote_share indicates whether there's any content before the [share] block
 881          * - Return value is the string that should replace the [share] block in the provided text
 882          *
 883          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 884          *
 885          * @param  string   $text     A BBCode string
 886          * @param  callable $callback
 887          * @return string The BBCode string with all [share] blocks replaced
 888          */
 889         public static function convertShare($text, callable $callback)
 890         {
 891                 $return = preg_replace_callback(
 892                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 893                         function ($match) use ($callback) {
 894                                 $attribute_string = $match[2];
 895
 896                                 $attributes = [];
 897                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 898                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 899                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 900                                 }
 901
 902                                 // We only call this so that a previously unknown contact can be added.
 903                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 904                                 // This function then can fetch an entry from the contact table.
 905                                 Contact::getIdForURL($attributes['profile'], 0, true);
 906
 907                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 908                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 909
 910                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 911                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 912                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 913
 914                                 if ($attributes['avatar']) {
 915                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 916                                 }
 917
 918                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 919                         },
 920                         $text
 921                 );
 922
 923                 return $return;
 924         }
 925
 926         /**
 927          * Default [share] tag conversion callback
 928          *
 929          * Note: Can produce a [bookmark] tag in the output
 930          *
 931          * @see BBCode::convertShare()
 932          * @param array   $attributes     [share] block attribute values
 933          * @param array   $author_contact Contact row of the shared author
 934          * @param string  $content        Inner content of the [share] block
 935          * @param boolean $is_quote_share Whether there is content before the [share] block
 936          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 937          * @return string
 938          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 939          */
 940         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 941         {
 942                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 943
 944                 switch ($simplehtml) {
 945                         case 1:
 946                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 947                                 break;
 948                         case 2:
 949                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 950                                 break;
 951                         case 3: // Diaspora
 952                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . $mention . ':</b></p>' . "\n";
 953
 954                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 955                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 956                                 } else {
 957                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 958
 959                                         if ($attributes['link'] != '') {
 960                                                 $text .= '<p><a href="' . $attributes['link'] . '">[l]</a></p>' . "\n";
 961                                         }
 962                                 }
 963
 964                                 break;
 965                         case 4:
 966                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 967                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 968                                 $headline .= ':</b></p>' . "\n";
 969
 970                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 971
 972                                 break;
 973                         case 5:
 974                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 975                                 break;
 976                         case 7: // statusnet/GNU Social
 977                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 978                                 break;
 979                         case 9: // Google+
 980                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 981                                 $text .= '<p>' . $content . '</p>' . "\n";
 982
 983                                 if ($attributes['link'] != '') {
 984                                         $text .= '<p>' . $attributes['link'] . '</p>';
 985                                 }
 986                                 break;
 987                         default:
 988                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
 989                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
 990                                         try {
 991                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
 992                                         } catch (Exception $e) {
 993                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
 994                                         }
 995                                 } else {
 996                                         $text = ($is_quote_share? "\n" : '');
 997
 998                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
 999                                         $text .= Renderer::replaceMacros($tpl, [
1000                                                 '$profile' => $attributes['profile'],
1001                                                 '$avatar'  => $attributes['avatar'],
1002                                                 '$author'  => $attributes['author'],
1003                                                 '$link'    => $attributes['link'],
1004                                                 '$posted'  => $attributes['posted'],
1005                                                 '$content' => trim($content)
1006                                         ]);
1007                                 }
1008                                 break;
1009                 }
1010
1011                 return $text;
1012         }
1013
1014         private static function removePictureLinksCallback($match)
1015         {
1016                 $text = Cache::get($match[1]);
1017
1018                 if (is_null($text)) {
1019                         $a = self::getApp();
1020
1021                         $stamp1 = microtime(true);
1022
1023                         $ch = @curl_init($match[1]);
1024                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1025                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1026                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1027                         @curl_exec($ch);
1028                         $curl_info = @curl_getinfo($ch);
1029
1030                         $a->saveTimestamp($stamp1, "network");
1031
1032                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1033                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1034                         } else {
1035                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1036
1037                                 // if its not a picture then look if its a page that contains a picture link
1038                                 $body = Network::fetchUrl($match[1]);
1039
1040                                 $doc = new DOMDocument();
1041                                 @$doc->loadHTML($body);
1042                                 $xpath = new DOMXPath($doc);
1043                                 $list = $xpath->query("//meta[@name]");
1044                                 foreach ($list as $node) {
1045                                         $attr = [];
1046
1047                                         if ($node->attributes->length) {
1048                                                 foreach ($node->attributes as $attribute) {
1049                                                         $attr[$attribute->name] = $attribute->value;
1050                                                 }
1051                                         }
1052
1053                                         if (strtolower($attr["name"]) == "twitter:image") {
1054                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1055                                         }
1056                                 }
1057                         }
1058                         Cache::set($match[1], $text);
1059                 }
1060
1061                 return $text;
1062         }
1063
1064         private static function expandLinksCallback($match)
1065         {
1066                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1067                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1068                 } else {
1069                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1070                 }
1071         }
1072
1073         private static function cleanPictureLinksCallback($match)
1074         {
1075                 $text = Cache::get($match[1]);
1076
1077                 if (is_null($text)) {
1078                         $a = self::getApp();
1079
1080                         $stamp1 = microtime(true);
1081
1082                         $ch = @curl_init($match[1]);
1083                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1084                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1085                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1086                         @curl_exec($ch);
1087                         $curl_info = @curl_getinfo($ch);
1088
1089                         $a->saveTimestamp($stamp1, "network");
1090
1091                         // if its a link to a picture then embed this picture
1092                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1093                                 $text = "[img]" . $match[1] . "[/img]";
1094                         } else {
1095                                 $text = "[img]" . $match[2] . "[/img]";
1096
1097                                 // if its not a picture then look if its a page that contains a picture link
1098                                 $body = Network::fetchUrl($match[1]);
1099
1100                                 $doc = new DOMDocument();
1101                                 @$doc->loadHTML($body);
1102                                 $xpath = new DOMXPath($doc);
1103                                 $list = $xpath->query("//meta[@name]");
1104                                 foreach ($list as $node) {
1105                                         $attr = [];
1106                                         if ($node->attributes->length) {
1107                                                 foreach ($node->attributes as $attribute) {
1108                                                         $attr[$attribute->name] = $attribute->value;
1109                                                 }
1110                                         }
1111
1112                                         if (strtolower($attr["name"]) == "twitter:image") {
1113                                                 $text = "[img]" . $attr["content"] . "[/img]";
1114                                         }
1115                                 }
1116                         }
1117                         Cache::set($match[1], $text);
1118                 }
1119
1120                 return $text;
1121         }
1122
1123         public static function cleanPictureLinks($text)
1124         {
1125                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1126                 return $return;
1127         }
1128
1129         /**
1130          * @brief Converts a BBCode message to HTML message
1131          *
1132          * BBcode 2 HTML was written by WAY2WEB.net
1133          * extended to work with Mistpark/Friendica - Mike Macgirvin
1134          *
1135          * Simple HTML values meaning:
1136          * - 0: Friendica display
1137          * - 1: Unused
1138          * - 2: Used for Google+, Windows Phone push, Friendica API
1139          * - 3: Used before converting to Markdown in bb2diaspora.php
1140          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1141          * - 5: Unused
1142          * - 6: Unused
1143          * - 7: Used for dfrn, OStatus
1144          * - 8: Used for WP backlink text setting
1145          *
1146          * @param string $text
1147          * @param bool   $try_oembed
1148          * @param int    $simple_html
1149          * @param bool   $for_plaintext
1150          * @return string
1151          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1152          */
1153         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1154         {
1155                 $a = self::getApp();
1156
1157                 /*
1158                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1159                  *
1160                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1161                  * $match[1] = $url
1162                  * $match[2] = $title or absent
1163                  */
1164                 $try_oembed_callback = function ($match)
1165                 {
1166                         $url = $match[1];
1167                         $title = defaults($match, 2, null);
1168
1169                         try {
1170                                 $return = OEmbed::getHTML($url, $title);
1171                         } catch (Exception $ex) {
1172                                 $return = $match[0];
1173                         }
1174
1175                         return $return;
1176                 };
1177
1178                 // Extracting multi-line code blocks before the whitespace processing
1179                 $codeblocks = [];
1180
1181                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1182                         function ($matches) use (&$codeblocks) {
1183                                 $return = $matches[0];
1184                                 if (strpos($matches[2], "\n") !== false) {
1185                                         $return = '#codeblock-' . count($codeblocks) . '#';
1186
1187                                         $codeblocks[] =  '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1188                                 }
1189                                 return $return;
1190                         },
1191                         $text
1192                 );
1193
1194                 // Hide all [noparse] contained bbtags by spacefying them
1195                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1196
1197                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1198                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1199                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1200
1201                 // Remove the abstract element. It is a non visible element.
1202                 $text = self::stripAbstract($text);
1203
1204                 // Move all spaces out of the tags
1205                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1206                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1207
1208                 // Extract the private images which use data urls since preg has issues with
1209                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1210                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1211
1212                 $extracted = self::extractImagesFromItemBody($text);
1213                 $text = $extracted['body'];
1214                 $saved_image = $extracted['images'];
1215
1216                 // If we find any event code, turn it into an event.
1217                 // After we're finished processing the bbcode we'll
1218                 // replace all of the event code with a reformatted version.
1219
1220                 $ev = Event::fromBBCode($text);
1221
1222                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1223                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1224
1225                 $text = str_replace("<", "&lt;", $text);
1226                 $text = str_replace(">", "&gt;", $text);
1227
1228                 // remove some newlines before the general conversion
1229                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1230                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1231
1232                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1233                 if (!$try_oembed) {
1234                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1235                 }
1236
1237                 // Convert new line chars to html <br /> tags
1238
1239                 // nlbr seems to be hopelessly messed up
1240                 //      $Text = nl2br($Text);
1241
1242                 // We'll emulate it.
1243
1244                 $text = trim($text);
1245                 $text = str_replace("\r\n", "\n", $text);
1246
1247                 // removing multiplicated newlines
1248                 if (Config::get("system", "remove_multiplicated_lines")) {
1249                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1250                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1251                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1252                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1253                         do {
1254                                 $oldtext = $text;
1255                                 $text = str_replace($search, $replace, $text);
1256                         } while ($oldtext != $text);
1257                 }
1258
1259                 // Set up the parameters for a URL search string
1260                 $URLSearchString = "^\[\]";
1261                 // Set up the parameters for a MAIL search string
1262                 $MAILSearchString = $URLSearchString;
1263
1264                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1265                 if (!$for_plaintext) {
1266                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1267                         // Currently disabled, since the function is too greedy
1268                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1269                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1270                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1271                         if ($simple_html == 7) {
1272                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1273                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1274                         }
1275                 } else {
1276                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1277                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1278                 }
1279
1280
1281                 // Handle attached links or videos
1282                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1283
1284                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1285
1286                 // Remove all hashtag addresses
1287                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1288                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1289                 } elseif ($simple_html == 3) {
1290                         // The ! is converted to @ since Diaspora only understands the @
1291                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1292                                 '@<a href="$2">$3</a>',
1293                                 $text);
1294                 } elseif ($simple_html == 7) {
1295                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1296                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1297                                 $text);
1298                 } elseif (!$simple_html) {
1299                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1300                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1301                                 $text);
1302                 }
1303
1304                 // Bookmarks in red - will be converted to bookmarks in friendica
1305                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1306                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1307                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1308                                         "[bookmark=$1]$2[/bookmark]", $text);
1309
1310                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1311                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1312                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1313                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1314                 }
1315
1316                 if ($simple_html == 5) {
1317                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1318                 }
1319
1320                 // Perform URL Search
1321                 if ($try_oembed) {
1322                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1323                 }
1324
1325                 if ($simple_html == 5) {
1326                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1327                 } else {
1328                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1329                 }
1330
1331                 // Handle Diaspora posts
1332                 $text = preg_replace_callback(
1333                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1334                         function ($match) {
1335                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1336                         }, $text
1337                 );
1338
1339                 $text = preg_replace_callback(
1340                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1341                         function ($match) {
1342                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1343                         }, $text
1344                 );
1345
1346                 // Server independent link to posts and comments
1347                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1348                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1349                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1350
1351                 /* Tag conversion
1352                  * Supports:
1353                  * - #[url=<anything>]<term>[/url]
1354                  * - [url=<anything>]#<term>[/url]
1355                  */
1356                 $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1357                         return '#<a href="'
1358                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1359                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1360                                 . XML::escape($matches[1])
1361                                 . '</a>';
1362                 }, $text);
1363
1364                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1365                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1366
1367                 // Red compatibility, though the link can't be authenticated on Friendica
1368                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1369
1370
1371                 // we may need to restrict this further if it picks up too many strays
1372                 // link acct:user@host to a webfinger profile redirector
1373
1374                 $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);
1375
1376                 // Perform MAIL Search
1377                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1378                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1379
1380                 // leave open the posibility of [map=something]
1381                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1382
1383                 if (strpos($text, '[/map]') !== false) {
1384                         $text = preg_replace_callback(
1385                                 "/\[map\](.*?)\[\/map\]/ism",
1386                                 function ($match) use ($simple_html) {
1387                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1388                                 },
1389                                 $text
1390                         );
1391                 }
1392                 if (strpos($text, '[map=') !== false) {
1393                         $text = preg_replace_callback(
1394                                 "/\[map=(.*?)\]/ism",
1395                                 function ($match) use ($simple_html) {
1396                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1397                                 },
1398                                 $text
1399                         );
1400                 }
1401                 if (strpos($text, '[map]') !== false) {
1402                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1403                 }
1404
1405                 // Check for headers
1406                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1407                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1408                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1409                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1410                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1411                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1412
1413                 // Check for paragraph
1414                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1415
1416                 // Check for bold text
1417                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1418
1419                 // Check for Italics text
1420                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1421
1422                 // Check for Underline text
1423                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1424
1425                 // Check for strike-through text
1426                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1427
1428                 // Check for over-line text
1429                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1430
1431                 // Check for colored text
1432                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1433
1434                 // Check for sized text
1435                 // [size=50] --> font-size: 50px (with the unit).
1436                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1437                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1438
1439                 // Check for centered text
1440                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1441
1442                 // Check for list text
1443                 $text = str_replace("[*]", "<li>", $text);
1444
1445                 // Check for style sheet commands
1446                 $text = preg_replace_callback(
1447                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1448                         function ($match) {
1449                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1450                         },
1451                         $text
1452                 );
1453
1454                 // Check for CSS classes
1455                 $text = preg_replace_callback(
1456                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1457                         function ($match) {
1458                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1459                         },
1460                         $text
1461                 );
1462
1463                 // handle nested lists
1464                 $endlessloop = 0;
1465
1466                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1467                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1468                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1469                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1470                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1471                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1472                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1473                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1474                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1475                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1476                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1477                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1478                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1479                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1480                 }
1481
1482                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1483                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1484                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1485                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1486
1487                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1488                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1489
1490                 $text = str_replace('[hr]', '<hr />', $text);
1491
1492                 // This is actually executed in Item::prepareBody()
1493
1494                 $text = str_replace('[nosmile]', '', $text);
1495
1496                 // Check for font change text
1497                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1498
1499                 // Declare the format for [code] layout
1500
1501                 $CodeLayout = '<code>$1</code>';
1502                 // Check for [code] text
1503                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1504
1505                 // Declare the format for [spoiler] layout
1506                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1507
1508                 // Check for [spoiler] text
1509                 // handle nested quotes
1510                 $endlessloop = 0;
1511                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1512                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1513                 }
1514
1515                 // Check for [spoiler=Author] text
1516
1517                 $t_wrote = L10n::t('$1 wrote:');
1518
1519                 // handle nested quotes
1520                 $endlessloop = 0;
1521                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1522                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1523                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1524                                                  $text);
1525                 }
1526
1527                 // Declare the format for [quote] layout
1528                 $QuoteLayout = '<blockquote>$1</blockquote>';
1529
1530                 // Check for [quote] text
1531                 // handle nested quotes
1532                 $endlessloop = 0;
1533                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1534                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1535                 }
1536
1537                 // Check for [quote=Author] text
1538
1539                 $t_wrote = L10n::t('$1 wrote:');
1540
1541                 // handle nested quotes
1542                 $endlessloop = 0;
1543                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1544                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1545                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1546                                                  $text);
1547                 }
1548
1549
1550                 // [img=widthxheight]image source[/img]
1551                 $text = preg_replace_callback(
1552                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1553                         function ($matches) use ($simple_html) {
1554                                 if (strpos($matches[3], "data:image/") === 0) {
1555                                         return $matches[0];
1556                                 }
1557
1558                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1559                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1560                         },
1561                         $text
1562                 );
1563
1564                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1565                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1566
1567                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1568                         function ($matches) use ($simple_html) {
1569                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1570                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1571                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1572                         },
1573                         $text);
1574
1575                 // Images
1576                 // [img]pathtoimage[/img]
1577                 $text = preg_replace_callback(
1578                         "/\[img\](.*?)\[\/img\]/ism",
1579                         function ($matches) use ($simple_html) {
1580                                 if (strpos($matches[1], "data:image/") === 0) {
1581                                         return $matches[0];
1582                                 }
1583
1584                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1585                                 return "[img]" . $matches[1] . "[/img]";
1586                         },
1587                         $text
1588                 );
1589
1590                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1591                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1592
1593                 // Shared content
1594                 $text = self::convertShare(
1595                         $text,
1596                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1597                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1598                         }
1599                 );
1600
1601                 $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);
1602                 $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 />', $text);
1603                 //$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 />', $Text);
1604
1605                 // Try to Oembed
1606                 if ($try_oembed) {
1607                         $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);
1608                         $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);
1609
1610                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1611                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1612                 } else {
1613                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1614                                                 '<a href="$1" target="_blank">$1</a>', $text);
1615                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1616                                                 '<a href="$1" target="_blank">$1</a>', $text);
1617                 }
1618
1619                 // html5 video and audio
1620
1621
1622                 if ($try_oembed) {
1623                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1624                 } else {
1625                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1626                 }
1627
1628                 // Youtube extensions
1629                 if ($try_oembed) {
1630                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1631                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1632                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1633                 }
1634
1635                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1636                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1637                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1638
1639                 if ($try_oembed) {
1640                         $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);
1641                 } else {
1642                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1643                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1644                 }
1645
1646                 if ($try_oembed) {
1647                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1648                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1649                 }
1650
1651                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1652                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1653
1654                 if ($try_oembed) {
1655                         $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);
1656                 } else {
1657                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1658                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1659                 }
1660
1661                 // oembed tag
1662                 $text = OEmbed::BBCode2HTML($text);
1663
1664                 // Avoid triple linefeeds through oembed
1665                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1666
1667                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1668                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1669                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1670                 // start which is always required). Allow desc with a missing summary for compatibility.
1671
1672                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1673                         $sub = Event::getHTML($ev, $simple_html);
1674
1675                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1676                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1677                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1678                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1679                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1680                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1681                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1682                 }
1683
1684                 // Replace non graphical smilies for external posts
1685                 if ($simple_html) {
1686                         $text = Smilies::replace($text, false, true);
1687                 }
1688
1689                 // Unhide all [noparse] contained bbtags unspacefying them
1690                 // and triming the [noparse] tag.
1691
1692                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1693                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1694                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1695
1696                 /// @todo What is the meaning of these lines?
1697                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1698                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1699
1700                 // Currently deactivated, it made problems with " inside of alt texts.
1701                 //$text = preg_replace('/\&quot\;/', '"', $text);
1702
1703                 // fix any escaped ampersands that may have been converted into links
1704                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1705
1706                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1707                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1708                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1709                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1710
1711                 // sanitize href attributes (only whitelisted protocols URLs)
1712                 // default value for backward compatibility
1713                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1714
1715                 // Always allowed protocol even if config isn't set or not including it
1716                 $allowed_link_protocols[] = 'http';
1717                 $allowed_link_protocols[] = 'redir/';
1718
1719                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1720                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1721
1722                 if ($saved_image) {
1723                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1724                 }
1725
1726                 // Restore code blocks
1727                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1728                         function ($matches) use ($codeblocks) {
1729                                 $return = $matches[0];
1730                                 if (isset($codeblocks[intval($matches[1])])) {
1731                                         $return = $codeblocks[$matches[1]];
1732                                 }
1733                                 return $return;
1734                         },
1735                         $text
1736                 );
1737
1738                 // Clean up the HTML by loading and saving the HTML with the DOM.
1739                 // Bad structured html can break a whole page.
1740                 // For performance reasons do it only with ativated item cache or at export.
1741                 if (!$try_oembed || (get_itemcachepath() != "")) {
1742                         $doc = new DOMDocument();
1743                         $doc->preserveWhiteSpace = false;
1744
1745                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1746
1747                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1748                         $encoding = '<?xml encoding="UTF-8">';
1749                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1750                         $doc->encoding = 'UTF-8';
1751                         $text = $doc->saveHTML();
1752                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1753
1754                         $text = str_replace('<br></li>', '</li>', $text);
1755
1756                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1757                 }
1758
1759                 // Clean up some useless linebreaks in lists
1760                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1761                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1762                 //$Text = str_replace('</li><br />', '</li>', $Text);
1763                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1764                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1765
1766                 Hook::callAll('bbcode', $text);
1767
1768                 return trim($text);
1769         }
1770
1771         /**
1772          * @brief Strips the "abstract" tag from the provided text
1773          *
1774          * @param string $text The text with BBCode
1775          * @return string The same text - but without "abstract" element
1776          */
1777         public static function stripAbstract($text)
1778         {
1779                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1780                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1781
1782                 return $text;
1783         }
1784
1785         /**
1786          * @brief Returns the value of the "abstract" element
1787          *
1788          * @param string $text The text that maybe contains the element
1789          * @param string $addon The addon for which the abstract is meant for
1790          * @return string The abstract
1791          */
1792         public static function getAbstract($text, $addon = "")
1793         {
1794                 $abstract = "";
1795                 $abstracts = [];
1796                 $addon = strtolower($addon);
1797
1798                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1799                         foreach ($results AS $result) {
1800                                 $abstracts[strtolower($result[1])] = $result[2];
1801                         }
1802                 }
1803
1804                 if (isset($abstracts[$addon])) {
1805                         $abstract = $abstracts[$addon];
1806                 }
1807
1808                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1809                         $abstract = $result[1];
1810                 }
1811
1812                 return $abstract;
1813         }
1814
1815         /**
1816          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1817          *
1818          * @param array $match Matching values for the callback
1819          *                     [1] = Mention type (! or @)
1820          *                     [2] = Name
1821          *                     [3] = Address
1822          * @return string Replaced mention
1823          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1824          * @throws \ImagickException
1825          */
1826         private static function bbCodeMention2DiasporaCallback($match)
1827         {
1828                 $contact = Contact::getDetailsByURL($match[3]);
1829
1830                 if (empty($contact['addr'])) {
1831                         $contact = Probe::uri($match[3]);
1832                 }
1833
1834                 if (empty($contact['addr'])) {
1835                         return $match[0];
1836                 }
1837
1838                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1839                 return $mention;
1840         }
1841
1842         /**
1843          * @brief Converts a BBCode text into Markdown
1844          *
1845          * This function converts a BBCode item body to be sent to Markdown-enabled
1846          * systems like Diaspora and Libertree
1847          *
1848          * @param string $text
1849          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1850          * @return string
1851          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1852          */
1853         public static function toMarkdown($text, $for_diaspora = true)
1854         {
1855                 $a = self::getApp();
1856
1857                 $original_text = $text;
1858
1859                 // Since Diaspora is creating a summary for links, this function removes them before posting
1860                 if ($for_diaspora) {
1861                         $text = self::removeShareInformation($text);
1862                 }
1863
1864                 /**
1865                  * Transform #tags, strip off the [url] and replace spaces with underscore
1866                  */
1867                 $url_search_string = "^\[\]";
1868                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1869                         function ($matches) {
1870                                 return '#' . str_replace(' ', '_', $matches[2]);
1871                         },
1872                         $text
1873                 );
1874
1875                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1876                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1877
1878                 // Convert it to HTML - don't try oembed
1879                 if ($for_diaspora) {
1880                         $text = self::convert($text, false, 3);
1881
1882                         // Add all tags that maybe were removed
1883                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1884                                 $tagline = "";
1885                                 foreach ($tags[2] as $tag) {
1886                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1887                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1888                                                 $tagline .= '#' . $tag . ' ';
1889                                         }
1890                                 }
1891                                 $text = $text . " " . $tagline;
1892                         }
1893                 } else {
1894                         $text = self::convert($text, false, 4);
1895                 }
1896
1897                 // mask some special HTML chars from conversation to markdown
1898                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1899
1900                 // If a link is followed by a quote then there should be a newline before it
1901                 // Maybe we should make this newline at every time before a quote.
1902                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1903
1904                 $stamp1 = microtime(true);
1905
1906                 // Now convert HTML to Markdown
1907                 $text = HTML::toMarkdown($text);
1908
1909                 // unmask the special chars back to HTML
1910                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1911
1912                 $a->saveTimestamp($stamp1, "parser");
1913
1914                 // Libertree has a problem with escaped hashtags.
1915                 $text = str_replace(['\#'], ['#'], $text);
1916
1917                 // Remove any leading or trailing whitespace, as this will mess up
1918                 // the Diaspora signature verification and cause the item to disappear
1919                 $text = trim($text);
1920
1921                 if ($for_diaspora) {
1922                         $url_search_string = "^\[\]";
1923                         $text = preg_replace_callback(
1924                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1925                                 ['self', 'bbCodeMention2DiasporaCallback'],
1926                                 $text
1927                         );
1928                 }
1929
1930                 Hook::callAll('bb2diaspora', $text);
1931
1932                 return $text;
1933         }
1934
1935         /**
1936      * @brief Pull out all #hashtags and @person tags from $string.
1937      *
1938      * We also get @person@domain.com - which would make
1939      * the regex quite complicated as tags can also
1940      * end a sentence. So we'll run through our results
1941      * and strip the period from any tags which end with one.
1942      * Returns array of tags found, or empty array.
1943      *
1944      * @param string $string Post content
1945      *
1946      * @return array List of tag and person names
1947      */
1948     public static function getTags($string)
1949     {
1950         $ret = [];
1951
1952         // Convert hashtag links to hashtags
1953         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1954
1955         // ignore anything in a code block
1956         $string = preg_replace('/\[code\](.*?)\[\/code\]/sm', '', $string);
1957
1958         // Force line feeds at bbtags
1959         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1960
1961         // ignore anything in a bbtag
1962         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1963
1964         // Match full names against @tags including the space between first and last
1965         // We will look these up afterward to see if they are full names or not recognisable.
1966
1967         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1968             foreach ($matches[1] as $match) {
1969                 if (strstr($match, ']')) {
1970                     // we might be inside a bbcode color tag - leave it alone
1971                     continue;
1972                 }
1973
1974                 if (substr($match, -1, 1) === '.') {
1975                     $ret[] = substr($match, 0, -1);
1976                 } else {
1977                     $ret[] = $match;
1978                 }
1979             }
1980         }
1981
1982         // Otherwise pull out single word tags. These can be @nickname, @first_last
1983         // and #hash tags.
1984
1985         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
1986             foreach ($matches[1] as $match) {
1987                 if (strstr($match, ']')) {
1988                     // we might be inside a bbcode color tag - leave it alone
1989                     continue;
1990                 }
1991                 if (substr($match, -1, 1) === '.') {
1992                     $match = substr($match,0,-1);
1993                 }
1994                 // ignore strictly numeric tags like #1
1995                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
1996                     continue;
1997                 }
1998                 // try not to catch url fragments
1999                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2000                     continue;
2001                 }
2002                 $ret[] = $match;
2003             }
2004         }
2005
2006         return $ret;
2007     }
2008 }