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                                         $post["after"] = '';
  71                                 }
  72
  73                                 $attacheddata = $data[2];
  74
  75                                 $URLSearchString = "^\[\]";
  76
  77                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  78
  79                                         $picturedata = Image::getInfoFromURL($matches[1]);
  80
  81                                         if ($picturedata) {
  82                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  83                                                         $post["image"] = $matches[1];
  84                                                 } else {
  85                                                         $post["preview"] = $matches[1];
  86                                                 }
  87                                         }
  88                                 }
  89
  90                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  91                                         $post["url"] = $matches[1];
  92                                         $post["title"] = $matches[2];
  93                                 }
  94                                 if (!empty($post["url"]) && (in_array($post["type"], ["link", "video"]))
  95                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  96                                         $post["url"] = $matches[1];
  97                                 }
  98
  99                                 // Search for description
 100                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 101                                         $post["description"] = $matches[1];
 102                                 }
 103                         }
 104                 }
 105                 return $post;
 106         }
 107
 108         /**
 109          * @brief Fetches attachment data that were generated with the "attachment" element
 110          *
 111          * @param string $body Message body
 112          * @return array
 113          *                     'type' -> Message type ("link", "video", "photo")
 114          *                     'text' -> Text before the shared message
 115          *                     'after' -> Text after the shared message
 116          *                     'image' -> Preview image of the message
 117          *                     'url' -> Url to the attached message
 118          *                     'title' -> Title of the attachment
 119          *                     'description' -> Description of the attachment
 120          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 121          */
 122         public static function getAttachmentData($body)
 123         {
 124                 $data = [];
 125
 126                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 127                         return self::getOldAttachmentData($body);
 128                 }
 129
 130                 $attributes = $match[2];
 131
 132                 $data["text"] = trim($match[1]);
 133
 134                 $type = "";
 135                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 136                 if (!empty($matches[1])) {
 137                         $type = strtolower($matches[1]);
 138                 }
 139
 140                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 141                 if (!empty($matches[1])) {
 142                         $type = strtolower($matches[1]);
 143                 }
 144
 145                 if ($type == "") {
 146                         return [];
 147                 }
 148
 149                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 150                         return [];
 151                 }
 152
 153                 if ($type != "") {
 154                         $data["type"] = $type;
 155                 }
 156
 157                 $url = "";
 158                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 159                 if (!empty($matches[1])) {
 160                         $url = $matches[1];
 161                 }
 162
 163                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 164                 if (!empty($matches[1])) {
 165                         $url = $matches[1];
 166                 }
 167
 168                 if ($url != "") {
 169                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 170                 }
 171
 172                 $title = "";
 173                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 174                 if (!empty($matches[1])) {
 175                         $title = $matches[1];
 176                 }
 177
 178                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 179                 if (!empty($matches[1])) {
 180                         $title = $matches[1];
 181                 }
 182
 183                 if ($title != "") {
 184                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 185                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 186                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 187                         $data["title"] = $title;
 188                 }
 189
 190                 $image = "";
 191                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 192                 if (!empty($matches[1])) {
 193                         $image = $matches[1];
 194                 }
 195
 196                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 197                 if (!empty($matches[1])) {
 198                         $image = $matches[1];
 199                 }
 200
 201                 if ($image != "") {
 202                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 203                 }
 204
 205                 $preview = "";
 206                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 207                 if (!empty($matches[1])) {
 208                         $preview = $matches[1];
 209                 }
 210
 211                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 212                 if (!empty($matches[1])) {
 213                         $preview = $matches[1];
 214                 }
 215
 216                 if ($preview != "") {
 217                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 218                 }
 219
 220                 $data["description"] = trim($match[3]);
 221
 222                 $data["after"] = trim($match[4]);
 223
 224                 return $data;
 225         }
 226
 227         public static function getAttachedData($body, $item = [])
 228         {
 229                 /*
 230                 - text:
 231                 - type: link, video, photo
 232                 - title:
 233                 - url:
 234                 - image:
 235                 - description:
 236                 - (thumbnail)
 237                 */
 238
 239                 $has_title = !empty($item['title']);
 240                 $plink = defaults($item, 'plink', '');
 241                 $post = self::getAttachmentData($body);
 242
 243                 // if nothing is found, it maybe having an image.
 244                 if (!isset($post["type"])) {
 245                         // Simplify image codes
 246                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 247
 248                         $URLSearchString = "^\[\]";
 249
 250                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 251
 252                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 253                                 if ((count($pictures) == 1) && !$has_title) {
 254                                         // Checking, if the link goes to a picture
 255                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 256
 257                                         // Workaround:
 258                                         // Sometimes photo posts to the own album are not detected at the start.
 259                                         // So we seem to cannot use the cache for these cases. That's strange.
 260                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 261                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 262                                         }
 263
 264                                         if ($data["type"] == "photo") {
 265                                                 $post["type"] = "photo";
 266                                                 if (isset($data["images"][0])) {
 267                                                         $post["image"] = $data["images"][0]["src"];
 268                                                         $post["url"] = $data["url"];
 269                                                 } else {
 270                                                         $post["image"] = $data["url"];
 271                                                 }
 272
 273                                                 $post["preview"] = $pictures[0][2];
 274                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 275                                         } else {
 276                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 277                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 278                                                         $post["type"] = "photo";
 279                                                         $post["image"] = $pictures[0][1];
 280                                                         $post["preview"] = $pictures[0][2];
 281                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 282                                                 }
 283                                         }
 284                                 } elseif (count($pictures) > 0) {
 285                                         $post["type"] = "link";
 286                                         $post["url"] = $plink;
 287                                         $post["image"] = $pictures[0][2];
 288                                         $post["text"] = $body;
 289                                 }
 290                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 291                                 if ((count($pictures) == 1) && !$has_title) {
 292                                         $post["type"] = "photo";
 293                                         $post["image"] = $pictures[0][1];
 294                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 295                                 } elseif (count($pictures) > 0) {
 296                                         $post["type"] = "link";
 297                                         $post["url"] = $plink;
 298                                         $post["image"] = $pictures[0][1];
 299                                         $post["text"] = $body;
 300                                 }
 301                         }
 302
 303                         // Test for the external links
 304                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 305                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 306
 307                         $links = array_merge($links1, $links2);
 308
 309                         // If there is only a single one, then use it.
 310                         // This should cover link posts via API.
 311                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 312                                 $post["type"] = "link";
 313                                 $post["text"] = trim($body);
 314                                 $post["url"] = $links[0][1];
 315                         }
 316
 317                         // Now count the number of external media links
 318                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 319                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 320                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 321                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 322
 323                         // Add them to the other external links
 324                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 325
 326                         // Are there more than one?
 327                         if (count($links) > 1) {
 328                                 // The post will be the type "text", which means a blog post
 329                                 unset($post["type"]);
 330                                 $post["url"] = $plink;
 331                         }
 332
 333                         if (!isset($post["type"])) {
 334                                 $post["type"] = "text";
 335                                 $post["text"] = trim($body);
 336                         }
 337                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 338                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 339
 340                         if (isset($data["images"][0])) {
 341                                 $post["image"] = $data["images"][0]["src"];
 342                         }
 343                 }
 344
 345                 return $post;
 346         }
 347
 348         /**
 349          * @brief Converts a BBCode text into plaintext
 350          *
 351          * @param      $text
 352          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 353          *
 354          * @return string
 355          */
 356         public static function toPlaintext($text, $keep_urls = true)
 357         {
 358                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 359                 if (!$keep_urls) {
 360                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 361                 }
 362
 363                 return $naked_text;
 364         }
 365
 366         private static function proxyUrl($image, $simplehtml = false)
 367         {
 368                 // Only send proxied pictures to API and for internal display
 369                 if (in_array($simplehtml, [false, 2])) {
 370                         return ProxyUtils::proxifyUrl($image);
 371                 } else {
 372                         return $image;
 373                 }
 374         }
 375
 376         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 377         {
 378                 // Suppress "view full size"
 379                 if (intval(Config::get('system', 'no_view_full_size'))) {
 380                         $include_link = false;
 381                 }
 382
 383                 // Picture addresses can contain special characters
 384                 $s = htmlspecialchars_decode($srctext);
 385
 386                 $matches = null;
 387                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 388                 if ($c) {
 389                         foreach ($matches as $mtch) {
 390                                 Logger::log('scale_external_image: ' . $mtch[1]);
 391
 392                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 393                                 if (stristr($mtch[1], $hostname)) {
 394                                         continue;
 395                                 }
 396
 397                                 // $scale_replace, if passed, is an array of two elements. The
 398                                 // first is the name of the full-size image. The second is the
 399                                 // name of a remote, scaled-down version of the full size image.
 400                                 // This allows Friendica to display the smaller remote image if
 401                                 // one exists, while still linking to the full-size image
 402                                 if ($scale_replace) {
 403                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 404                                 } else {
 405                                         $scaled = $mtch[1];
 406                                 }
 407                                 $i = Network::fetchUrl($scaled);
 408                                 if (!$i) {
 409                                         return $srctext;
 410                                 }
 411
 412                                 // guess mimetype from headers or filename
 413                                 $type = Image::guessType($mtch[1], true);
 414
 415                                 if ($i) {
 416                                         $Image = new Image($i, $type);
 417                                         if ($Image->isValid()) {
 418                                                 $orig_width = $Image->getWidth();
 419                                                 $orig_height = $Image->getHeight();
 420
 421                                                 if ($orig_width > 640 || $orig_height > 640) {
 422                                                         $Image->scaleDown(640);
 423                                                         $new_width = $Image->getWidth();
 424                                                         $new_height = $Image->getHeight();
 425                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 426                                                         $s = str_replace(
 427                                                                 $mtch[0],
 428                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 429                                                                 . "\n" . (($include_link)
 430                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 431                                                                         : ''),
 432                                                                 $s
 433                                                         );
 434                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 435                                                 }
 436                                         }
 437                                 }
 438                         }
 439                 }
 440
 441                 // replace the special char encoding
 442                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 443                 return $s;
 444         }
 445
 446         /**
 447          * The purpose of this function is to apply system message length limits to
 448          * imported messages without including any embedded photos in the length
 449          *
 450          * @brief Truncates imported message body string length to max_import_size
 451          * @param string $body
 452          * @return string
 453          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 454          */
 455         public static function limitBodySize($body)
 456         {
 457                 $maxlen = Config::get('config', 'max_import_size', 0);
 458
 459                 // If the length of the body, including the embedded images, is smaller
 460                 // than the maximum, then don't waste time looking for the images
 461                 if ($maxlen && (strlen($body) > $maxlen)) {
 462
 463                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 464
 465                         $orig_body = $body;
 466                         $new_body = '';
 467                         $textlen = 0;
 468
 469                         $img_start = strpos($orig_body, '[img');
 470                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 471                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 472                         while (($img_st_close !== false) && ($img_end !== false)) {
 473
 474                                 $img_st_close++; // make it point to AFTER the closing bracket
 475                                 $img_end += $img_start;
 476                                 $img_end += strlen('[/img]');
 477
 478                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 479                                         // This is an embedded image
 480
 481                                         if (($textlen + $img_start) > $maxlen) {
 482                                                 if ($textlen < $maxlen) {
 483                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 484                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 485                                                         $textlen = $maxlen;
 486                                                 }
 487                                         } else {
 488                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 489                                                 $textlen += $img_start;
 490                                         }
 491
 492                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 493                                 } else {
 494
 495                                         if (($textlen + $img_end) > $maxlen) {
 496                                                 if ($textlen < $maxlen) {
 497                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 498                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 499                                                         $textlen = $maxlen;
 500                                                 }
 501                                         } else {
 502                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 503                                                 $textlen += $img_end;
 504                                         }
 505                                 }
 506                                 $orig_body = substr($orig_body, $img_end);
 507
 508                                 if ($orig_body === false) {
 509                                         // in case the body ends on a closing image tag
 510                                         $orig_body = '';
 511                                 }
 512
 513                                 $img_start = strpos($orig_body, '[img');
 514                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 515                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 516                         }
 517
 518                         if (($textlen + strlen($orig_body)) > $maxlen) {
 519                                 if ($textlen < $maxlen) {
 520                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 521                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 522                                 }
 523                         } else {
 524                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 525                                 $new_body = $new_body . $orig_body;
 526                         }
 527
 528                         return $new_body;
 529                 } else {
 530                         return $body;
 531                 }
 532         }
 533
 534         /**
 535          * Processes [attachment] tags
 536          *
 537          * Note: Can produce a [bookmark] tag in the returned string
 538          *
 539          * @brief Processes [attachment] tags
 540          * @param string   $return
 541          * @param bool|int $simplehtml
 542          * @param bool     $tryoembed
 543          * @return string
 544          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 545          */
 546         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 547         {
 548                 $data = self::getAttachmentData($return);
 549                 if (empty($data) || empty($data["url"])) {
 550                         return $return;
 551                 }
 552
 553                 if (isset($data["title"])) {
 554                         $data["title"] = strip_tags($data["title"]);
 555                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 556                 } else {
 557                         $data["title"] = null;
 558                 }
 559
 560                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 561                         $data["preview"] = $data["image"];
 562                         $data["image"] = "";
 563                 }
 564
 565                 $return = '';
 566                 if ($simplehtml == 7) {
 567                         $return = self::convertUrlForOStatus($data["url"]);
 568                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 569                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 570                 } else {
 571                         try {
 572                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 573                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 574                                 } else {
 575                                         throw new Exception('OEmbed is disabled for this attachment.');
 576                                 }
 577                         } catch (Exception $e) {
 578                                 $data["title"] = defaults($data, 'title', $data['url']);
 579
 580                                 if ($simplehtml != 4) {
 581                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 582                                 }
 583
 584                                 if (!empty($data['title']) && !empty($data['url'])) {
 585                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 586                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 587                                         } else {
 588                                                 if (!empty($data["image"])) {
 589                                                         $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), $data["title"]);
 590                                                 } elseif (!empty($data["preview"])) {
 591                                                         $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"]);
 592                                                 }
 593                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 594                                         }
 595                                 }
 596
 597                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 598                                         // Sanitize the HTML by converting it to BBCode
 599                                         $bbcode = HTML::toBBCode($data["description"]);
 600                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 601                                 }
 602
 603                                 if (!empty($data['url'])) {
 604                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 605                                 }
 606
 607                                 if ($simplehtml != 4) {
 608                                         $return .= '</div>';
 609                                 }
 610                         }
 611                 }
 612
 613                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 614         }
 615
 616         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 617         {
 618                 $data = self::getAttachmentData($Text);
 619
 620                 if (!$data) {
 621                         return $Text;
 622                 } elseif ($nolink) {
 623                         return $data["text"] . defaults($data, 'after', '');
 624                 }
 625
 626                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 627                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 628                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 629                         $data["title"] = $data["url"];
 630                 } elseif (($text != "") && strstr($title, $text)) {
 631                         $data["text"] = $data["title"];
 632                         $data["title"] = $data["url"];
 633                 }
 634
 635                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 636                         return $data["title"] . $data["after"];
 637                 }
 638
 639                 // If the link already is included in the post, don't add it again
 640                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 641                         return $data["text"] . $data["after"];
 642                 }
 643
 644                 $text = $data["text"];
 645
 646                 if (!empty($data["url"]) && !empty($data["title"])) {
 647                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 648                 } elseif (!empty($data["url"])) {
 649                         $text .= "\n[url]" . $data["url"] . "[/url]";
 650                 }
 651
 652                 return $text . "\n" . $data["after"];
 653         }
 654
 655         /**
 656          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 657          *
 658          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 659          * @param array $match Array with the matching values
 660          * @return string reformatted link including HTML codes
 661          */
 662         private static function convertUrlForOStatusCallback($match)
 663         {
 664                 $url = $match[1];
 665
 666                 if (isset($match[2]) && ($match[1] != $match[2])) {
 667                         return $match[0];
 668                 }
 669
 670                 $parts = parse_url($url);
 671                 if (!isset($parts['scheme'])) {
 672                         return $match[0];
 673                 }
 674
 675                 return self::convertUrlForOStatus($url);
 676         }
 677
 678         /**
 679          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 680          * @param string $url URL that is about to be reformatted
 681          * @return string reformatted link including HTML codes
 682          */
 683         private static function convertUrlForOStatus($url)
 684         {
 685                 $parts = parse_url($url);
 686                 $scheme = $parts['scheme'] . '://';
 687                 $styled_url = str_replace($scheme, '', $url);
 688
 689                 if (strlen($styled_url) > 30) {
 690                         $styled_url = substr($styled_url, 0, 30) . "…";
 691                 }
 692
 693                 $html = '<a href="%s" target="_blank">%s</a>';
 694
 695                 return sprintf($html, $url, $styled_url);
 696         }
 697
 698         /*
 699          * [noparse][i]italic[/i][/noparse] turns into
 700          * [noparse][ i ]italic[ /i ][/noparse],
 701          * to hide them from parser.
 702          */
 703         private static function escapeNoparseCallback($match)
 704         {
 705                 $whole_match = $match[0];
 706                 $captured = $match[1];
 707                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 708                 $new_str = str_replace($captured, $spacefied, $whole_match);
 709                 return $new_str;
 710         }
 711
 712         /*
 713          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 714          * now turns back and the [noparse] tags are trimed
 715          * returning [i]italic[/i]
 716          */
 717         private static function unescapeNoparseCallback($match)
 718         {
 719                 $captured = $match[1];
 720                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 721                 return $unspacefied;
 722         }
 723
 724         /**
 725          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 726          * occurrences
 727          *
 728          * @param string $text        Text to search
 729          * @param string $name        Tag name
 730          * @param int    $occurrences Number of first occurrences to skip
 731          * @return boolean|array
 732          */
 733         public static function getTagPosition($text, $name, $occurrences = 0)
 734         {
 735                 if ($occurrences < 0) {
 736                         $occurrences = 0;
 737                 }
 738
 739                 $start_open = -1;
 740                 for ($i = 0; $i <= $occurrences; $i++) {
 741                         if ($start_open !== false) {
 742                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 743                         }
 744                 }
 745
 746                 if ($start_open === false) {
 747                         return false;
 748                 }
 749
 750                 $start_equal = strpos($text, '=', $start_open);
 751                 $start_close = strpos($text, ']', $start_open);
 752
 753                 if ($start_close === false) {
 754                         return false;
 755                 }
 756
 757                 $start_close++;
 758
 759                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 760
 761                 if ($end_open === false) {
 762                         return false;
 763                 }
 764
 765                 $res = [
 766                         'start' => [
 767                                 'open' => $start_open,
 768                                 'close' => $start_close
 769                         ],
 770                         'end' => [
 771                                 'open' => $end_open,
 772                                 'close' => $end_open + strlen('[/' . $name . ']')
 773                         ],
 774                 ];
 775
 776                 if ($start_equal !== false) {
 777                         $res['start']['equal'] = $start_equal + 1;
 778                 }
 779
 780                 return $res;
 781         }
 782
 783         /**
 784          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 785          *
 786          * @param string $pattern Preg pattern string
 787          * @param string $replace Preg replace string
 788          * @param string $name    BBCode tag name
 789          * @param string $text    Text to search
 790          * @return string
 791          */
 792         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 793         {
 794                 $occurrences = 0;
 795                 $pos = self::getTagPosition($text, $name, $occurrences);
 796                 while ($pos !== false && $occurrences++ < 1000) {
 797                         $start = substr($text, 0, $pos['start']['open']);
 798                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 799                         $end = substr($text, $pos['end']['close']);
 800                         if ($end === false) {
 801                                 $end = '';
 802                         }
 803
 804                         $subject = preg_replace($pattern, $replace, $subject);
 805                         $text = $start . $subject . $end;
 806
 807                         $pos = self::getTagPosition($text, $name, $occurrences);
 808                 }
 809
 810                 return $text;
 811         }
 812
 813         private static function extractImagesFromItemBody($body)
 814         {
 815                 $saved_image = [];
 816                 $orig_body = $body;
 817                 $new_body = '';
 818
 819                 $cnt = 0;
 820                 $img_start = strpos($orig_body, '[img');
 821                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 822                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 823                 while (($img_st_close !== false) && ($img_end !== false)) {
 824                         $img_st_close++; // make it point to AFTER the closing bracket
 825                         $img_end += $img_start;
 826
 827                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 828                                 // This is an embedded image
 829                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 830                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 831
 832                                 $cnt++;
 833                         } else {
 834                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 835                         }
 836
 837                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 838
 839                         if ($orig_body === false) {
 840                                 // in case the body ends on a closing image tag
 841                                 $orig_body = '';
 842                         }
 843
 844                         $img_start = strpos($orig_body, '[img');
 845                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 846                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 847                 }
 848
 849                 $new_body = $new_body . $orig_body;
 850
 851                 return ['body' => $new_body, 'images' => $saved_image];
 852         }
 853
 854         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 855         {
 856                 $newbody = $body;
 857
 858                 $cnt = 0;
 859                 foreach ($images as $image) {
 860                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 861                         // it loops over the array starting from the first element and going sequentially
 862                         // to the last element
 863                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 864                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 865                         $cnt++;
 866                 }
 867
 868                 return $newbody;
 869         }
 870
 871         /**
 872          * This function converts a [share] block to text according to a provided callback function whose signature is:
 873          *
 874          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 875          *
 876          * Where:
 877          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 878          * data lookup
 879          * - $author_contact is a contact record array
 880          * - $content is the inner content of the [share] block
 881          * - $is_quote_share indicates whether there's any content before the [share] block
 882          * - Return value is the string that should replace the [share] block in the provided text
 883          *
 884          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 885          *
 886          * @param  string   $text     A BBCode string
 887          * @param  callable $callback
 888          * @return string The BBCode string with all [share] blocks replaced
 889          */
 890         public static function convertShare($text, callable $callback)
 891         {
 892                 $return = preg_replace_callback(
 893                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 894                         function ($match) use ($callback) {
 895                                 $attribute_string = $match[2];
 896
 897                                 $attributes = [];
 898                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 899                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 900                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 901                                 }
 902
 903                                 // We only call this so that a previously unknown contact can be added.
 904                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 905                                 // This function then can fetch an entry from the contact table.
 906                                 Contact::getIdForURL($attributes['profile'], 0, true);
 907
 908                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 909                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 910
 911                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 912                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 913                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 914
 915                                 if ($attributes['avatar']) {
 916                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 917                                 }
 918
 919                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 920                         },
 921                         $text
 922                 );
 923
 924                 return $return;
 925         }
 926
 927         /**
 928          * Default [share] tag conversion callback
 929          *
 930          * Note: Can produce a [bookmark] tag in the output
 931          *
 932          * @see BBCode::convertShare()
 933          * @param array   $attributes     [share] block attribute values
 934          * @param array   $author_contact Contact row of the shared author
 935          * @param string  $content        Inner content of the [share] block
 936          * @param boolean $is_quote_share Whether there is content before the [share] block
 937          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 938          * @return string
 939          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 940          */
 941         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 942         {
 943                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 944
 945                 switch ($simplehtml) {
 946                         case 1:
 947                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 948                                 break;
 949                         case 2:
 950                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 951                                 break;
 952                         case 3: // Diaspora
 953                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . $mention . ':</b></p>' . "\n";
 954
 955                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 956                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 957                                 } else {
 958                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 959
 960                                         if ($attributes['link'] != '') {
 961                                                 $text .= '<p><a href="' . $attributes['link'] . '">[l]</a></p>' . "\n";
 962                                         }
 963                                 }
 964
 965                                 break;
 966                         case 4:
 967                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 968                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 969                                 $headline .= ':</b></p>' . "\n";
 970
 971                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 972
 973                                 break;
 974                         case 5:
 975                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 976                                 break;
 977                         case 7: // statusnet/GNU Social
 978                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 979                                 break;
 980                         case 9: // Google+
 981                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 982                                 $text .= '<p>' . $content . '</p>' . "\n";
 983
 984                                 if ($attributes['link'] != '') {
 985                                         $text .= '<p>' . $attributes['link'] . '</p>';
 986                                 }
 987                                 break;
 988                         default:
 989                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
 990                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
 991                                         try {
 992                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
 993                                         } catch (Exception $e) {
 994                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
 995                                         }
 996                                 } else {
 997                                         $text = ($is_quote_share? "\n" : '');
 998
 999                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1000                                         $text .= Renderer::replaceMacros($tpl, [
1001                                                 '$profile' => $attributes['profile'],
1002                                                 '$avatar'  => $attributes['avatar'],
1003                                                 '$author'  => $attributes['author'],
1004                                                 '$link'    => $attributes['link'],
1005                                                 '$posted'  => $attributes['posted'],
1006                                                 '$content' => trim($content)
1007                                         ]);
1008                                 }
1009                                 break;
1010                 }
1011
1012                 return $text;
1013         }
1014
1015         private static function removePictureLinksCallback($match)
1016         {
1017                 $text = Cache::get($match[1]);
1018
1019                 if (is_null($text)) {
1020                         $a = self::getApp();
1021
1022                         $stamp1 = microtime(true);
1023
1024                         $ch = @curl_init($match[1]);
1025                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1026                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1027                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1028                         @curl_exec($ch);
1029                         $curl_info = @curl_getinfo($ch);
1030
1031                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1032
1033                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1034                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1035                         } else {
1036                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1037
1038                                 // if its not a picture then look if its a page that contains a picture link
1039                                 $body = Network::fetchUrl($match[1]);
1040
1041                                 $doc = new DOMDocument();
1042                                 @$doc->loadHTML($body);
1043                                 $xpath = new DOMXPath($doc);
1044                                 $list = $xpath->query("//meta[@name]");
1045                                 foreach ($list as $node) {
1046                                         $attr = [];
1047
1048                                         if ($node->attributes->length) {
1049                                                 foreach ($node->attributes as $attribute) {
1050                                                         $attr[$attribute->name] = $attribute->value;
1051                                                 }
1052                                         }
1053
1054                                         if (strtolower($attr["name"]) == "twitter:image") {
1055                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1056                                         }
1057                                 }
1058                         }
1059                         Cache::set($match[1], $text);
1060                 }
1061
1062                 return $text;
1063         }
1064
1065         private static function expandLinksCallback($match)
1066         {
1067                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1068                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1069                 } else {
1070                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1071                 }
1072         }
1073
1074         private static function cleanPictureLinksCallback($match)
1075         {
1076                 $text = Cache::get($match[1]);
1077
1078                 if (is_null($text)) {
1079                         $a = self::getApp();
1080
1081                         $stamp1 = microtime(true);
1082
1083                         $ch = @curl_init($match[1]);
1084                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1085                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1086                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1087                         @curl_exec($ch);
1088                         $curl_info = @curl_getinfo($ch);
1089
1090                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1091
1092                         // if its a link to a picture then embed this picture
1093                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1094                                 $text = "[img]" . $match[1] . "[/img]";
1095                         } else {
1096                                 $text = "[img]" . $match[2] . "[/img]";
1097
1098                                 // if its not a picture then look if its a page that contains a picture link
1099                                 $body = Network::fetchUrl($match[1]);
1100
1101                                 $doc = new DOMDocument();
1102                                 @$doc->loadHTML($body);
1103                                 $xpath = new DOMXPath($doc);
1104                                 $list = $xpath->query("//meta[@name]");
1105                                 foreach ($list as $node) {
1106                                         $attr = [];
1107                                         if ($node->attributes->length) {
1108                                                 foreach ($node->attributes as $attribute) {
1109                                                         $attr[$attribute->name] = $attribute->value;
1110                                                 }
1111                                         }
1112
1113                                         if (strtolower($attr["name"]) == "twitter:image") {
1114                                                 $text = "[img]" . $attr["content"] . "[/img]";
1115                                         }
1116                                 }
1117                         }
1118                         Cache::set($match[1], $text);
1119                 }
1120
1121                 return $text;
1122         }
1123
1124         public static function cleanPictureLinks($text)
1125         {
1126                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1127                 return $return;
1128         }
1129
1130         /**
1131          * @brief Converts a BBCode message to HTML message
1132          *
1133          * BBcode 2 HTML was written by WAY2WEB.net
1134          * extended to work with Mistpark/Friendica - Mike Macgirvin
1135          *
1136          * Simple HTML values meaning:
1137          * - 0: Friendica display
1138          * - 1: Unused
1139          * - 2: Used for Google+, Windows Phone push, Friendica API
1140          * - 3: Used before converting to Markdown in bb2diaspora.php
1141          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1142          * - 5: Unused
1143          * - 6: Unused
1144          * - 7: Used for dfrn, OStatus
1145          * - 8: Used for WP backlink text setting
1146          *
1147          * @param string $text
1148          * @param bool   $try_oembed
1149          * @param int    $simple_html
1150          * @param bool   $for_plaintext
1151          * @return string
1152          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1153          */
1154         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1155         {
1156                 $a = self::getApp();
1157
1158                 /*
1159                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1160                  *
1161                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1162                  * $match[1] = $url
1163                  * $match[2] = $title or absent
1164                  */
1165                 $try_oembed_callback = function ($match)
1166                 {
1167                         $url = $match[1];
1168                         $title = defaults($match, 2, null);
1169
1170                         try {
1171                                 $return = OEmbed::getHTML($url, $title);
1172                         } catch (Exception $ex) {
1173                                 $return = $match[0];
1174                         }
1175
1176                         return $return;
1177                 };
1178
1179                 // Extracting multi-line code blocks before the whitespace processing
1180                 $codeblocks = [];
1181
1182                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1183                         function ($matches) use (&$codeblocks) {
1184                                 $return = $matches[0];
1185                                 if (strpos($matches[2], "\n") !== false) {
1186                                         $return = '#codeblock-' . count($codeblocks) . '#';
1187
1188                                         $codeblocks[] =  '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1189                                 }
1190                                 return $return;
1191                         },
1192                         $text
1193                 );
1194
1195                 // Hide all [noparse] contained bbtags by spacefying them
1196                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1197
1198                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1199                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1200                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1201
1202                 // Remove the abstract element. It is a non visible element.
1203                 $text = self::stripAbstract($text);
1204
1205                 // Move all spaces out of the tags
1206                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1207                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1208
1209                 // Extract the private images which use data urls since preg has issues with
1210                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1211                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1212
1213                 $extracted = self::extractImagesFromItemBody($text);
1214                 $text = $extracted['body'];
1215                 $saved_image = $extracted['images'];
1216
1217                 // If we find any event code, turn it into an event.
1218                 // After we're finished processing the bbcode we'll
1219                 // replace all of the event code with a reformatted version.
1220
1221                 $ev = Event::fromBBCode($text);
1222
1223                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1224                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1225
1226                 $text = str_replace("<", "&lt;", $text);
1227                 $text = str_replace(">", "&gt;", $text);
1228
1229                 // remove some newlines before the general conversion
1230                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1231                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1232
1233                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1234                 if (!$try_oembed) {
1235                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1236                 }
1237
1238                 // Convert new line chars to html <br /> tags
1239
1240                 // nlbr seems to be hopelessly messed up
1241                 //      $Text = nl2br($Text);
1242
1243                 // We'll emulate it.
1244
1245                 $text = trim($text);
1246                 $text = str_replace("\r\n", "\n", $text);
1247
1248                 // removing multiplicated newlines
1249                 if (Config::get("system", "remove_multiplicated_lines")) {
1250                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1251                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1252                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1253                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1254                         do {
1255                                 $oldtext = $text;
1256                                 $text = str_replace($search, $replace, $text);
1257                         } while ($oldtext != $text);
1258                 }
1259
1260                 // Set up the parameters for a URL search string
1261                 $URLSearchString = "^\[\]";
1262                 // Set up the parameters for a MAIL search string
1263                 $MAILSearchString = $URLSearchString;
1264
1265                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1266                 if (!$for_plaintext) {
1267                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1268                         // Currently disabled, since the function is too greedy
1269                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1270                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1271                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1272                         if ($simple_html == 7) {
1273                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1274                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1275                         }
1276                 } else {
1277                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1278                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1279                 }
1280
1281
1282                 // Handle attached links or videos
1283                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1284
1285                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1286
1287                 // Remove all hashtag addresses
1288                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1289                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1290                 } elseif ($simple_html == 3) {
1291                         // The ! is converted to @ since Diaspora only understands the @
1292                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1293                                 '@<a href="$2">$3</a>',
1294                                 $text);
1295                 } elseif ($simple_html == 7) {
1296                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1297                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1298                                 $text);
1299                 } elseif (!$simple_html) {
1300                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1301                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1302                                 $text);
1303                 }
1304
1305                 // Bookmarks in red - will be converted to bookmarks in friendica
1306                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1307                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1308                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1309                                         "[bookmark=$1]$2[/bookmark]", $text);
1310
1311                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1312                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1313                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1314                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1315                 }
1316
1317                 if ($simple_html == 5) {
1318                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1319                 }
1320
1321                 // Perform URL Search
1322                 if ($try_oembed) {
1323                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1324                 }
1325
1326                 if ($simple_html == 5) {
1327                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1328                 } else {
1329                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1330                 }
1331
1332                 // Handle Diaspora posts
1333                 $text = preg_replace_callback(
1334                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1335                         function ($match) {
1336                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1337                         }, $text
1338                 );
1339
1340                 $text = preg_replace_callback(
1341                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1342                         function ($match) {
1343                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1344                         }, $text
1345                 );
1346
1347                 // Server independent link to posts and comments
1348                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1349                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1350                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1351
1352                 /* Tag conversion
1353                  * Supports:
1354                  * - #[url=<anything>]<term>[/url]
1355                  * - [url=<anything>]#<term>[/url]
1356                  */
1357                 $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1358                         return '#<a href="'
1359                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1360                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1361                                 . XML::escape($matches[1])
1362                                 . '</a>';
1363                 }, $text);
1364
1365                 // We need no target="_blank" for local links
1366                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1367                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1368                 $text = preg_replace("/\[url\](".$escapedBaseUrl."[$URLSearchString]*)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1369                 $text = preg_replace("/\[url\=(".$escapedBaseUrl."[$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1370
1371                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1372                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1373
1374                 // Red compatibility, though the link can't be authenticated on Friendica
1375                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1376
1377
1378                 // we may need to restrict this further if it picks up too many strays
1379                 // link acct:user@host to a webfinger profile redirector
1380
1381                 $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);
1382
1383                 // Perform MAIL Search
1384                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1385                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1386
1387                 // leave open the posibility of [map=something]
1388                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1389
1390                 if (strpos($text, '[/map]') !== false) {
1391                         $text = preg_replace_callback(
1392                                 "/\[map\](.*?)\[\/map\]/ism",
1393                                 function ($match) use ($simple_html) {
1394                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1395                                 },
1396                                 $text
1397                         );
1398                 }
1399                 if (strpos($text, '[map=') !== false) {
1400                         $text = preg_replace_callback(
1401                                 "/\[map=(.*?)\]/ism",
1402                                 function ($match) use ($simple_html) {
1403                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1404                                 },
1405                                 $text
1406                         );
1407                 }
1408                 if (strpos($text, '[map]') !== false) {
1409                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1410                 }
1411
1412                 // Check for headers
1413                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1414                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1415                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1416                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1417                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1418                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1419
1420                 // Check for paragraph
1421                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1422
1423                 // Check for bold text
1424                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1425
1426                 // Check for Italics text
1427                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1428
1429                 // Check for Underline text
1430                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1431
1432                 // Check for strike-through text
1433                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1434
1435                 // Check for over-line text
1436                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1437
1438                 // Check for colored text
1439                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1440
1441                 // Check for sized text
1442                 // [size=50] --> font-size: 50px (with the unit).
1443                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1444                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1445
1446                 // Check for centered text
1447                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1448
1449                 // Check for list text
1450                 $text = str_replace("[*]", "<li>", $text);
1451
1452                 // Check for style sheet commands
1453                 $text = preg_replace_callback(
1454                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1455                         function ($match) {
1456                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1457                         },
1458                         $text
1459                 );
1460
1461                 // Check for CSS classes
1462                 $text = preg_replace_callback(
1463                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1464                         function ($match) {
1465                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1466                         },
1467                         $text
1468                 );
1469
1470                 // handle nested lists
1471                 $endlessloop = 0;
1472
1473                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1474                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1475                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1476                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1477                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1478                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1479                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1480                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1481                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1482                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1483                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1484                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1485                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1486                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1487                 }
1488
1489                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1490                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1491                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1492                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1493
1494                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1495                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1496
1497                 $text = str_replace('[hr]', '<hr />', $text);
1498
1499                 // This is actually executed in Item::prepareBody()
1500
1501                 $text = str_replace('[nosmile]', '', $text);
1502
1503                 // Check for font change text
1504                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1505
1506                 // Declare the format for [code] layout
1507
1508                 $CodeLayout = '<code>$1</code>';
1509                 // Check for [code] text
1510                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1511
1512                 // Declare the format for [spoiler] layout
1513                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1514
1515                 // Check for [spoiler] text
1516                 // handle nested quotes
1517                 $endlessloop = 0;
1518                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1519                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1520                 }
1521
1522                 // Check for [spoiler=Author] text
1523
1524                 $t_wrote = L10n::t('$1 wrote:');
1525
1526                 // handle nested quotes
1527                 $endlessloop = 0;
1528                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1529                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1530                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1531                                                  $text);
1532                 }
1533
1534                 // Declare the format for [quote] layout
1535                 $QuoteLayout = '<blockquote>$1</blockquote>';
1536
1537                 // Check for [quote] text
1538                 // handle nested quotes
1539                 $endlessloop = 0;
1540                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1541                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1542                 }
1543
1544                 // Check for [quote=Author] text
1545
1546                 $t_wrote = L10n::t('$1 wrote:');
1547
1548                 // handle nested quotes
1549                 $endlessloop = 0;
1550                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1551                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1552                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1553                                                  $text);
1554                 }
1555
1556
1557                 // [img=widthxheight]image source[/img]
1558                 $text = preg_replace_callback(
1559                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1560                         function ($matches) use ($simple_html) {
1561                                 if (strpos($matches[3], "data:image/") === 0) {
1562                                         return $matches[0];
1563                                 }
1564
1565                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1566                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1567                         },
1568                         $text
1569                 );
1570
1571                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1572                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1573
1574                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1575                         function ($matches) use ($simple_html) {
1576                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1577                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1578                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1579                         },
1580                         $text);
1581
1582                 // Images
1583                 // [img]pathtoimage[/img]
1584                 $text = preg_replace_callback(
1585                         "/\[img\](.*?)\[\/img\]/ism",
1586                         function ($matches) use ($simple_html) {
1587                                 if (strpos($matches[1], "data:image/") === 0) {
1588                                         return $matches[0];
1589                                 }
1590
1591                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1592                                 return "[img]" . $matches[1] . "[/img]";
1593                         },
1594                         $text
1595                 );
1596
1597                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1598                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1599
1600                 // Shared content
1601                 $text = self::convertShare(
1602                         $text,
1603                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1604                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1605                         }
1606                 );
1607
1608                 $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);
1609                 $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 />', $text);
1610                 //$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 />', $Text);
1611
1612                 // Try to Oembed
1613                 if ($try_oembed) {
1614                         $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);
1615                         $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);
1616
1617                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1618                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1619                 } else {
1620                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1621                                                 '<a href="$1" target="_blank">$1</a>', $text);
1622                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1623                                                 '<a href="$1" target="_blank">$1</a>', $text);
1624                 }
1625
1626                 // html5 video and audio
1627
1628
1629                 if ($try_oembed) {
1630                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1631                 } else {
1632                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1633                 }
1634
1635                 // Youtube extensions
1636                 if ($try_oembed) {
1637                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1638                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1639                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1640                 }
1641
1642                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1643                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1644                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1645
1646                 if ($try_oembed) {
1647                         $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);
1648                 } else {
1649                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1650                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1651                 }
1652
1653                 if ($try_oembed) {
1654                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1655                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1656                 }
1657
1658                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1659                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1660
1661                 if ($try_oembed) {
1662                         $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);
1663                 } else {
1664                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1665                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1666                 }
1667
1668                 // oembed tag
1669                 $text = OEmbed::BBCode2HTML($text);
1670
1671                 // Avoid triple linefeeds through oembed
1672                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1673
1674                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1675                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1676                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1677                 // start which is always required). Allow desc with a missing summary for compatibility.
1678
1679                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1680                         $sub = Event::getHTML($ev, $simple_html);
1681
1682                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1683                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1684                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1685                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1686                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1687                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1688                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1689                 }
1690
1691                 // Replace non graphical smilies for external posts
1692                 if ($simple_html) {
1693                         $text = Smilies::replace($text);
1694                 }
1695
1696                 // Unhide all [noparse] contained bbtags unspacefying them
1697                 // and triming the [noparse] tag.
1698
1699                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1700                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1701                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1702
1703                 /// @todo What is the meaning of these lines?
1704                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1705                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1706
1707                 // Currently deactivated, it made problems with " inside of alt texts.
1708                 //$text = preg_replace('/\&quot\;/', '"', $text);
1709
1710                 // fix any escaped ampersands that may have been converted into links
1711                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1712
1713                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1714                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1715                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1716                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1717
1718                 // sanitize href attributes (only whitelisted protocols URLs)
1719                 // default value for backward compatibility
1720                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1721
1722                 // Always allowed protocol even if config isn't set or not including it
1723                 $allowed_link_protocols[] = 'http';
1724                 $allowed_link_protocols[] = 'redir/';
1725
1726                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1727                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1728
1729                 if ($saved_image) {
1730                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1731                 }
1732
1733                 // Restore code blocks
1734                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1735                         function ($matches) use ($codeblocks) {
1736                                 $return = $matches[0];
1737                                 if (isset($codeblocks[intval($matches[1])])) {
1738                                         $return = $codeblocks[$matches[1]];
1739                                 }
1740                                 return $return;
1741                         },
1742                         $text
1743                 );
1744
1745                 // Clean up the HTML by loading and saving the HTML with the DOM.
1746                 // Bad structured html can break a whole page.
1747                 // For performance reasons do it only with activated item cache or at export.
1748                 if (!$try_oembed || (get_itemcachepath() != "")) {
1749                         $doc = new DOMDocument();
1750                         $doc->preserveWhiteSpace = false;
1751
1752                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1753
1754                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1755                         $encoding = '<?xml encoding="UTF-8">';
1756                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1757                         $doc->encoding = 'UTF-8';
1758                         $text = $doc->saveHTML();
1759                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1760
1761                         $text = str_replace('<br></li>', '</li>', $text);
1762
1763                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1764                 }
1765
1766                 // Clean up some useless linebreaks in lists
1767                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1768                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1769                 //$Text = str_replace('</li><br />', '</li>', $Text);
1770                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1771                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1772
1773                 Hook::callAll('bbcode', $text);
1774
1775                 return trim($text);
1776         }
1777
1778         /**
1779          * @brief Strips the "abstract" tag from the provided text
1780          *
1781          * @param string $text The text with BBCode
1782          * @return string The same text - but without "abstract" element
1783          */
1784         public static function stripAbstract($text)
1785         {
1786                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1787                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1788
1789                 return $text;
1790         }
1791
1792         /**
1793          * @brief Returns the value of the "abstract" element
1794          *
1795          * @param string $text The text that maybe contains the element
1796          * @param string $addon The addon for which the abstract is meant for
1797          * @return string The abstract
1798          */
1799         public static function getAbstract($text, $addon = "")
1800         {
1801                 $abstract = "";
1802                 $abstracts = [];
1803                 $addon = strtolower($addon);
1804
1805                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1806                         foreach ($results AS $result) {
1807                                 $abstracts[strtolower($result[1])] = $result[2];
1808                         }
1809                 }
1810
1811                 if (isset($abstracts[$addon])) {
1812                         $abstract = $abstracts[$addon];
1813                 }
1814
1815                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1816                         $abstract = $result[1];
1817                 }
1818
1819                 return $abstract;
1820         }
1821
1822         /**
1823          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1824          *
1825          * @param array $match Matching values for the callback
1826          *                     [1] = Mention type (! or @)
1827          *                     [2] = Name
1828          *                     [3] = Address
1829          * @return string Replaced mention
1830          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1831          * @throws \ImagickException
1832          */
1833         private static function bbCodeMention2DiasporaCallback($match)
1834         {
1835                 $contact = Contact::getDetailsByURL($match[3]);
1836
1837                 if (empty($contact['addr'])) {
1838                         $contact = Probe::uri($match[3]);
1839                 }
1840
1841                 if (empty($contact['addr'])) {
1842                         return $match[0];
1843                 }
1844
1845                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1846                 return $mention;
1847         }
1848
1849         /**
1850          * @brief Converts a BBCode text into Markdown
1851          *
1852          * This function converts a BBCode item body to be sent to Markdown-enabled
1853          * systems like Diaspora and Libertree
1854          *
1855          * @param string $text
1856          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1857          * @return string
1858          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1859          */
1860         public static function toMarkdown($text, $for_diaspora = true)
1861         {
1862                 $a = self::getApp();
1863
1864                 $original_text = $text;
1865
1866                 // Since Diaspora is creating a summary for links, this function removes them before posting
1867                 if ($for_diaspora) {
1868                         $text = self::removeShareInformation($text);
1869                 }
1870
1871                 /**
1872                  * Transform #tags, strip off the [url] and replace spaces with underscore
1873                  */
1874                 $url_search_string = "^\[\]";
1875                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1876                         function ($matches) {
1877                                 return '#' . str_replace(' ', '_', $matches[2]);
1878                         },
1879                         $text
1880                 );
1881
1882                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1883                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1884
1885                 // Convert it to HTML - don't try oembed
1886                 if ($for_diaspora) {
1887                         $text = self::convert($text, false, 3);
1888
1889                         // Add all tags that maybe were removed
1890                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1891                                 $tagline = "";
1892                                 foreach ($tags[2] as $tag) {
1893                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1894                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1895                                                 $tagline .= '#' . $tag . ' ';
1896                                         }
1897                                 }
1898                                 $text = $text . " " . $tagline;
1899                         }
1900                 } else {
1901                         $text = self::convert($text, false, 4);
1902                 }
1903
1904                 // mask some special HTML chars from conversation to markdown
1905                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1906
1907                 // If a link is followed by a quote then there should be a newline before it
1908                 // Maybe we should make this newline at every time before a quote.
1909                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1910
1911                 $stamp1 = microtime(true);
1912
1913                 // Now convert HTML to Markdown
1914                 $text = HTML::toMarkdown($text);
1915
1916                 // unmask the special chars back to HTML
1917                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1918
1919                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
1920
1921                 // Libertree has a problem with escaped hashtags.
1922                 $text = str_replace(['\#'], ['#'], $text);
1923
1924                 // Remove any leading or trailing whitespace, as this will mess up
1925                 // the Diaspora signature verification and cause the item to disappear
1926                 $text = trim($text);
1927
1928                 if ($for_diaspora) {
1929                         $url_search_string = "^\[\]";
1930                         $text = preg_replace_callback(
1931                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1932                                 ['self', 'bbCodeMention2DiasporaCallback'],
1933                                 $text
1934                         );
1935                 }
1936
1937                 Hook::callAll('bb2diaspora', $text);
1938
1939                 return $text;
1940         }
1941
1942         /**
1943      * @brief Pull out all #hashtags and @person tags from $string.
1944      *
1945      * We also get @person@domain.com - which would make
1946      * the regex quite complicated as tags can also
1947      * end a sentence. So we'll run through our results
1948      * and strip the period from any tags which end with one.
1949      * Returns array of tags found, or empty array.
1950      *
1951      * @param string $string Post content
1952      *
1953      * @return array List of tag and person names
1954      */
1955     public static function getTags($string)
1956     {
1957         $ret = [];
1958
1959         // Convert hashtag links to hashtags
1960         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1961
1962         // ignore anything in a code block
1963         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
1964
1965         // Force line feeds at bbtags
1966         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1967
1968         // ignore anything in a bbtag
1969         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1970
1971         // Match full names against @tags including the space between first and last
1972         // We will look these up afterward to see if they are full names or not recognisable.
1973
1974         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1975             foreach ($matches[1] as $match) {
1976                 if (strstr($match, ']')) {
1977                     // we might be inside a bbcode color tag - leave it alone
1978                     continue;
1979                 }
1980
1981                 if (substr($match, -1, 1) === '.') {
1982                     $ret[] = substr($match, 0, -1);
1983                 } else {
1984                     $ret[] = $match;
1985                 }
1986             }
1987         }
1988
1989         // Otherwise pull out single word tags. These can be @nickname, @first_last
1990         // and #hash tags.
1991
1992         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
1993             foreach ($matches[1] as $match) {
1994                 if (strstr($match, ']')) {
1995                     // we might be inside a bbcode color tag - leave it alone
1996                     continue;
1997                 }
1998                 if (substr($match, -1, 1) === '.') {
1999                     $match = substr($match,0,-1);
2000                 }
2001                 // ignore strictly numeric tags like #1
2002                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2003                     continue;
2004                 }
2005                 // try not to catch url fragments
2006                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2007                     continue;
2008                 }
2009                 $ret[] = $match;
2010             }
2011         }
2012
2013         return $ret;
2014     }
2015 }