src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Addon;
  15 use Friendica\Core\Cache;
  16 use Friendica\Core\Config;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Protocol;
  19 use Friendica\Core\System;
  20 use Friendica\Model\Contact;
  21 use Friendica\Model\Event;
  22 use Friendica\Network\Probe;
  23 use Friendica\Object\Image;
  24 use Friendica\Util\Map;
  25 use Friendica\Util\Network;
  26 use Friendica\Util\ParseUrl;
  27 use League\HTMLToMarkdown\HtmlConverter;
  28
  29 require_once "mod/proxy.php";
  30
  31 class BBCode extends BaseObject
  32 {
  33         /**
  34          * @brief Fetches attachment data that were generated the old way
  35          *
  36          * @param string $body Message body
  37          * @return array
  38          * 'type' -> Message type ("link", "video", "photo")
  39          * 'text' -> Text before the shared message
  40          * 'after' -> Text after the shared message
  41          * 'image' -> Preview image of the message
  42          * 'url' -> Url to the attached message
  43          * 'title' -> Title of the attachment
  44          * 'description' -> Description of the attachment
  45          */
  46         private static function getOldAttachmentData($body)
  47         {
  48                 $post = [];
  49
  50                 // Simplify image codes
  51                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  52
  53                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  54                         foreach ($attached as $data) {
  55                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  56                                         continue;
  57                                 }
  58
  59                                 $post["type"] = substr($data[1], 5);
  60
  61                                 $pos = strpos($body, $data[0]);
  62                                 if ($pos > 0) {
  63                                         $post["text"] = trim(substr($body, 0, $pos));
  64                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  65                                 } else {
  66                                         $post["text"] = trim(str_replace($data[0], "", $body));
  67                                 }
  68
  69                                 $attacheddata = $data[2];
  70
  71                                 $URLSearchString = "^\[\]";
  72
  73                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  74
  75                                         $picturedata = Image::getInfoFromURL($matches[1]);
  76
  77                                         if ($picturedata) {
  78                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  79                                                         $post["image"] = $matches[1];
  80                                                 } else {
  81                                                         $post["preview"] = $matches[1];
  82                                                 }
  83                                         }
  84                                 }
  85
  86                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  87                                         $post["url"] = $matches[1];
  88                                         $post["title"] = $matches[2];
  89                                 }
  90                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  91                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  92                                         $post["url"] = $matches[1];
  93                                 }
  94
  95                                 // Search for description
  96                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  97                                         $post["description"] = $matches[1];
  98                                 }
  99                         }
 100                 }
 101                 return $post;
 102         }
 103
 104         /**
 105          * @brief Fetches attachment data that were generated with the "attachment" element
 106          *
 107          * @param string $body Message body
 108          * @return array
 109          * 'type' -> Message type ("link", "video", "photo")
 110          * 'text' -> Text before the shared message
 111          * 'after' -> Text after the shared message
 112          * 'image' -> Preview image of the message
 113          * 'url' -> Url to the attached message
 114          * 'title' -> Title of the attachment
 115          * 'description' -> Description of the attachment
 116          */
 117         public static function getAttachmentData($body)
 118         {
 119                 $data = [];
 120
 121                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 122                         return self::getOldAttachmentData($body);
 123                 }
 124
 125                 $attributes = $match[2];
 126
 127                 $data["text"] = trim($match[1]);
 128
 129                 $type = "";
 130                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 131                 if (x($matches, 1)) {
 132                         $type = strtolower($matches[1]);
 133                 }
 134
 135                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 136                 if (x($matches, 1)) {
 137                         $type = strtolower($matches[1]);
 138                 }
 139
 140                 if ($type == "") {
 141                         return [];
 142                 }
 143
 144                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 145                         return [];
 146                 }
 147
 148                 if ($type != "") {
 149                         $data["type"] = $type;
 150                 }
 151
 152                 $url = "";
 153                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 154                 if (x($matches, 1)) {
 155                         $url = $matches[1];
 156                 }
 157
 158                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 159                 if (x($matches, 1)) {
 160                         $url = $matches[1];
 161                 }
 162
 163                 if ($url != "") {
 164                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 165                 }
 166
 167                 $title = "";
 168                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 169                 if (x($matches, 1)) {
 170                         $title = $matches[1];
 171                 }
 172
 173                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 174                 if (x($matches, 1)) {
 175                         $title = $matches[1];
 176                 }
 177
 178                 if ($title != "") {
 179                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 180                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 181                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 182                         $data["title"] = $title;
 183                 }
 184
 185                 $image = "";
 186                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 187                 if (x($matches, 1)) {
 188                         $image = $matches[1];
 189                 }
 190
 191                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 192                 if (x($matches, 1)) {
 193                         $image = $matches[1];
 194                 }
 195
 196                 if ($image != "") {
 197                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 198                 }
 199
 200                 $preview = "";
 201                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 202                 if (x($matches, 1)) {
 203                         $preview = $matches[1];
 204                 }
 205
 206                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 207                 if (x($matches, 1)) {
 208                         $preview = $matches[1];
 209                 }
 210
 211                 if ($preview != "") {
 212                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 213                 }
 214
 215                 $data["description"] = trim($match[3]);
 216
 217                 $data["after"] = trim($match[4]);
 218
 219                 return $data;
 220         }
 221
 222         public static function getAttachedData($body, $item = [])
 223         {
 224                 /*
 225                 - text:
 226                 - type: link, video, photo
 227                 - title:
 228                 - url:
 229                 - image:
 230                 - description:
 231                 - (thumbnail)
 232                 */
 233
 234                 $has_title = !empty($item['title']);
 235                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 236                 $post = self::getAttachmentData($body);
 237
 238                 // if nothing is found, it maybe having an image.
 239                 if (!isset($post["type"])) {
 240                         // Simplify image codes
 241                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 242
 243                         $URLSearchString = "^\[\]";
 244
 245                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 246
 247                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 248                                 if ((count($pictures) == 1) && !$has_title) {
 249                                         // Checking, if the link goes to a picture
 250                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 251
 252                                         // Workaround:
 253                                         // Sometimes photo posts to the own album are not detected at the start.
 254                                         // So we seem to cannot use the cache for these cases. That's strange.
 255                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 256                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 257                                         }
 258
 259                                         if ($data["type"] == "photo") {
 260                                                 $post["type"] = "photo";
 261                                                 if (isset($data["images"][0])) {
 262                                                         $post["image"] = $data["images"][0]["src"];
 263                                                         $post["url"] = $data["url"];
 264                                                 } else {
 265                                                         $post["image"] = $data["url"];
 266                                                 }
 267
 268                                                 $post["preview"] = $pictures[0][2];
 269                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 270                                         } else {
 271                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 272                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 273                                                         $post["type"] = "photo";
 274                                                         $post["image"] = $pictures[0][1];
 275                                                         $post["preview"] = $pictures[0][2];
 276                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 277                                                 }
 278                                         }
 279                                 } elseif (count($pictures) > 0) {
 280                                         $post["type"] = "link";
 281                                         $post["url"] = $plink;
 282                                         $post["image"] = $pictures[0][2];
 283                                         $post["text"] = $body;
 284                                 }
 285                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 286                                 if ((count($pictures) == 1) && !$has_title) {
 287                                         $post["type"] = "photo";
 288                                         $post["image"] = $pictures[0][1];
 289                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 290                                 } elseif (count($pictures) > 0) {
 291                                         $post["type"] = "link";
 292                                         $post["url"] = $plink;
 293                                         $post["image"] = $pictures[0][1];
 294                                         $post["text"] = $body;
 295                                 }
 296                         }
 297
 298                         // Test for the external links
 299                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 300                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 301
 302                         $links = array_merge($links1, $links2);
 303
 304                         // If there is only a single one, then use it.
 305                         // This should cover link posts via API.
 306                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 307                                 $post["type"] = "link";
 308                                 $post["text"] = trim($body);
 309                                 $post["url"] = $links[0][1];
 310                         }
 311
 312                         // Now count the number of external media links
 313                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 314                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 315                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 316                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 317
 318                         // Add them to the other external links
 319                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 320
 321                         // Are there more than one?
 322                         if (count($links) > 1) {
 323                                 // The post will be the type "text", which means a blog post
 324                                 unset($post["type"]);
 325                                 $post["url"] = $plink;
 326                         }
 327
 328                         if (!isset($post["type"])) {
 329                                 $post["type"] = "text";
 330                                 $post["text"] = trim($body);
 331                         }
 332                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 333                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 334
 335                         if (isset($data["images"][0])) {
 336                                 $post["image"] = $data["images"][0]["src"];
 337                         }
 338                 }
 339
 340                 return $post;
 341         }
 342
 343         /**
 344          * @brief Converts a BBCode text into plaintext
 345          *
 346          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 347          *
 348          * @return string
 349          */
 350         public static function toPlaintext($text, $keep_urls = true)
 351         {
 352                 $naked_text = preg_replace('/\[(.+?)\]/','', $text);
 353                 if (!$keep_urls) {
 354                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 355                 }
 356
 357                 return $naked_text;
 358         }
 359
 360         private static function proxyUrl($image, $simplehtml = false)
 361         {
 362                 // Only send proxied pictures to API and for internal display
 363                 if (in_array($simplehtml, [false, 2])) {
 364                         return proxy_url($image);
 365                 } else {
 366                         return $image;
 367                 }
 368         }
 369
 370         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 371         {
 372                 // Suppress "view full size"
 373                 if (intval(Config::get('system', 'no_view_full_size'))) {
 374                         $include_link = false;
 375                 }
 376
 377                 // Picture addresses can contain special characters
 378                 $s = htmlspecialchars_decode($srctext);
 379
 380                 $matches = null;
 381                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 382                 if ($c) {
 383                         foreach ($matches as $mtch) {
 384                                 logger('scale_external_image: ' . $mtch[1]);
 385
 386                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 387                                 if (stristr($mtch[1], $hostname)) {
 388                                         continue;
 389                                 }
 390
 391                                 // $scale_replace, if passed, is an array of two elements. The
 392                                 // first is the name of the full-size image. The second is the
 393                                 // name of a remote, scaled-down version of the full size image.
 394                                 // This allows Friendica to display the smaller remote image if
 395                                 // one exists, while still linking to the full-size image
 396                                 if ($scale_replace) {
 397                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 398                                 } else {
 399                                         $scaled = $mtch[1];
 400                                 }
 401                                 $i = Network::fetchUrl($scaled);
 402                                 if (!$i) {
 403                                         return $srctext;
 404                                 }
 405
 406                                 // guess mimetype from headers or filename
 407                                 $type = Image::guessType($mtch[1], true);
 408
 409                                 if ($i) {
 410                                         $Image = new Image($i, $type);
 411                                         if ($Image->isValid()) {
 412                                                 $orig_width = $Image->getWidth();
 413                                                 $orig_height = $Image->getHeight();
 414
 415                                                 if ($orig_width > 640 || $orig_height > 640) {
 416                                                         $Image->scaleDown(640);
 417                                                         $new_width = $Image->getWidth();
 418                                                         $new_height = $Image->getHeight();
 419                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 420                                                         $s = str_replace(
 421                                                                 $mtch[0],
 422                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 423                                                                 . "\n" . (($include_link)
 424                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 425                                                                         : ''),
 426                                                                 $s
 427                                                         );
 428                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 429                                                 }
 430                                         }
 431                                 }
 432                         }
 433                 }
 434
 435                 // replace the special char encoding
 436                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 437                 return $s;
 438         }
 439
 440         /**
 441          * The purpose of this function is to apply system message length limits to
 442          * imported messages without including any embedded photos in the length
 443          *
 444          * @brief Truncates imported message body string length to max_import_size
 445          * @param string $body
 446          * @return string
 447          */
 448         public static function limitBodySize($body)
 449         {
 450                 $maxlen = Config::get('config', 'max_import_size', 0);
 451
 452                 // If the length of the body, including the embedded images, is smaller
 453                 // than the maximum, then don't waste time looking for the images
 454                 if ($maxlen && (strlen($body) > $maxlen)) {
 455
 456                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 457
 458                         $orig_body = $body;
 459                         $new_body = '';
 460                         $textlen = 0;
 461
 462                         $img_start = strpos($orig_body, '[img');
 463                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 464                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 465                         while (($img_st_close !== false) && ($img_end !== false)) {
 466
 467                                 $img_st_close++; // make it point to AFTER the closing bracket
 468                                 $img_end += $img_start;
 469                                 $img_end += strlen('[/img]');
 470
 471                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 472                                         // This is an embedded image
 473
 474                                         if (($textlen + $img_start) > $maxlen) {
 475                                                 if ($textlen < $maxlen) {
 476                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 477                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 478                                                         $textlen = $maxlen;
 479                                                 }
 480                                         } else {
 481                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 482                                                 $textlen += $img_start;
 483                                         }
 484
 485                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 486                                 } else {
 487
 488                                         if (($textlen + $img_end) > $maxlen) {
 489                                                 if ($textlen < $maxlen) {
 490                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 491                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 492                                                         $textlen = $maxlen;
 493                                                 }
 494                                         } else {
 495                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 496                                                 $textlen += $img_end;
 497                                         }
 498                                 }
 499                                 $orig_body = substr($orig_body, $img_end);
 500
 501                                 if ($orig_body === false) {
 502                                         // in case the body ends on a closing image tag
 503                                         $orig_body = '';
 504                                 }
 505
 506                                 $img_start = strpos($orig_body, '[img');
 507                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 508                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 509                         }
 510
 511                         if (($textlen + strlen($orig_body)) > $maxlen) {
 512                                 if ($textlen < $maxlen) {
 513                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 514                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 515                                 }
 516                         } else {
 517                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 518                                 $new_body = $new_body . $orig_body;
 519                         }
 520
 521                         return $new_body;
 522                 } else {
 523                         return $body;
 524                 }
 525         }
 526
 527         /**
 528          * Processes [attachment] tags
 529          *
 530          * Note: Can produce a [bookmark] tag in the returned string
 531          *
 532          * @brief Processes [attachment] tags
 533          * @param string $return
 534          * @param bool|int $simplehtml
 535          * @param bool $tryoembed
 536          * @return string
 537          */
 538         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 539         {
 540                 $data = self::getAttachmentData($return);
 541                 if (empty($data) || empty($data["url"])) {
 542                         return $return;
 543                 }
 544
 545                 if (isset($data["title"])) {
 546                         $data["title"] = strip_tags($data["title"]);
 547                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 548                 } else {
 549                         $data["title"] = null;
 550                 }
 551
 552                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 553                         $data["preview"] = $data["image"];
 554                         $data["image"] = "";
 555                 }
 556
 557                 $return = '';
 558                 if ($simplehtml == 7) {
 559                         $return = self::convertUrlForOStatus($data["url"]);
 560                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 561                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 562                 } else {
 563                         try {
 564                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 565                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 566                                 } else {
 567                                         throw new Exception('OEmbed is disabled for this attachment.');
 568                                 }
 569                         } catch (Exception $e) {
 570                                 $data["title"] = defaults($data, 'title', $data['url']);
 571
 572                                 if ($simplehtml != 4) {
 573                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 574                                 }
 575
 576                                 if (!empty($data["image"])) {
 577                                         $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), $data["title"]);
 578                                 } elseif (!empty($data["preview"])) {
 579                                         $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), $data["title"]);
 580                                 }
 581
 582                                 if (($data["type"] == "photo") && !empty($data["url"]) && !empty($data["image"])) {
 583                                         $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"]);
 584                                 } else {
 585                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 586                                 }
 587
 588                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 589                                         // Sanitize the HTML by converting it to BBCode
 590                                         $bbcode = HTML::toBBCode($data["description"]);
 591                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 592                                 }
 593                                 if ($data["type"] == "link") {
 594                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 595                                 }
 596
 597                                 if ($simplehtml != 4) {
 598                                         $return .= '</div>';
 599                                 }
 600                         }
 601                 }
 602
 603                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 604         }
 605
 606         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 607         {
 608                 $data = self::getAttachmentData($Text);
 609
 610                 if (!$data) {
 611                         return $Text;
 612                 } elseif ($nolink) {
 613                         return $data["text"] . $data["after"];
 614                 }
 615
 616                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 617                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 618                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 619                         $data["title"] = $data["url"];
 620                 } elseif (($text != "") && strstr($title, $text)) {
 621                         $data["text"] = $data["title"];
 622                         $data["title"] = $data["url"];
 623                 }
 624
 625                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 626                         return $data["title"] . $data["after"];
 627                 }
 628
 629                 // If the link already is included in the post, don't add it again
 630                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 631                         return $data["text"] . $data["after"];
 632                 }
 633
 634                 $text = $data["text"];
 635
 636                 if (($data["url"] != "") && ($data["title"] != "")) {
 637                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 638                 } elseif (($data["url"] != "")) {
 639                         $text .= "\n[url]" . $data["url"] . "[/url]";
 640                 }
 641
 642                 return $text . "\n" . $data["after"];
 643         }
 644
 645         /**
 646          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 647          *
 648          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 649          * @param array $match Array with the matching values
 650          * @return string reformatted link including HTML codes
 651          */
 652         private static function convertUrlForOStatusCallback($match)
 653         {
 654                 $url = $match[1];
 655
 656                 if (isset($match[2]) && ($match[1] != $match[2])) {
 657                         return $match[0];
 658                 }
 659
 660                 $parts = parse_url($url);
 661                 if (!isset($parts['scheme'])) {
 662                         return $match[0];
 663                 }
 664
 665                 return self::convertUrlForOStatus($url);
 666         }
 667
 668         /**
 669          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 670          * @param string $url URL that is about to be reformatted
 671          * @return string reformatted link including HTML codes
 672          */
 673         private static function convertUrlForOStatus($url)
 674         {
 675                 $parts = parse_url($url);
 676                 $scheme = $parts['scheme'] . '://';
 677                 $styled_url = str_replace($scheme, '', $url);
 678
 679                 if (strlen($styled_url) > 30) {
 680                         $styled_url = substr($styled_url, 0, 30) . "…";
 681                 }
 682
 683                 $html = '<a href="%s" target="_blank">%s</a>';
 684
 685                 return sprintf($html, $url, $styled_url);
 686         }
 687
 688         /*
 689          * [noparse][i]italic[/i][/noparse] turns into
 690          * [noparse][ i ]italic[ /i ][/noparse],
 691          * to hide them from parser.
 692          */
 693         private static function escapeNoparseCallback($match)
 694         {
 695                 $whole_match = $match[0];
 696                 $captured = $match[1];
 697                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 698                 $new_str = str_replace($captured, $spacefied, $whole_match);
 699                 return $new_str;
 700         }
 701
 702         /*
 703          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 704          * now turns back and the [noparse] tags are trimed
 705          * returning [i]italic[/i]
 706          */
 707         private static function unescapeNoparseCallback($match)
 708         {
 709                 $captured = $match[1];
 710                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 711                 return $unspacefied;
 712         }
 713
 714         /**
 715          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 716          * occurrences
 717          *
 718          * @param string $text        Text to search
 719          * @param string $name        Tag name
 720          * @param int    $occurrences Number of first occurrences to skip
 721          * @return boolean|array
 722          */
 723         public static function getTagPosition($text, $name, $occurrences = 0)
 724         {
 725                 if ($occurrences < 0) {
 726                         $occurrences = 0;
 727                 }
 728
 729                 $start_open = -1;
 730                 for ($i = 0; $i <= $occurrences; $i++) {
 731                         if ($start_open !== false) {
 732                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 733                         }
 734                 }
 735
 736                 if ($start_open === false) {
 737                         return false;
 738                 }
 739
 740                 $start_equal = strpos($text, '=', $start_open);
 741                 $start_close = strpos($text, ']', $start_open);
 742
 743                 if ($start_close === false) {
 744                         return false;
 745                 }
 746
 747                 $start_close++;
 748
 749                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 750
 751                 if ($end_open === false) {
 752                         return false;
 753                 }
 754
 755                 $res = [
 756                         'start' => [
 757                                 'open' => $start_open,
 758                                 'close' => $start_close
 759                         ],
 760                         'end' => [
 761                                 'open' => $end_open,
 762                                 'close' => $end_open + strlen('[/' . $name . ']')
 763                         ],
 764                 ];
 765
 766                 if ($start_equal !== false) {
 767                         $res['start']['equal'] = $start_equal + 1;
 768                 }
 769
 770                 return $res;
 771         }
 772
 773         /**
 774          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 775          *
 776          * @param type $pattern Preg pattern string
 777          * @param type $replace Preg replace string
 778          * @param type $name    BBCode tag name
 779          * @param type $text    Text to search
 780          * @return string
 781          */
 782         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 783         {
 784                 $occurrences = 0;
 785                 $pos = self::getTagPosition($text, $name, $occurrences);
 786                 while ($pos !== false && $occurrences++ < 1000) {
 787                         $start = substr($text, 0, $pos['start']['open']);
 788                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 789                         $end = substr($text, $pos['end']['close']);
 790                         if ($end === false) {
 791                                 $end = '';
 792                         }
 793
 794                         $subject = preg_replace($pattern, $replace, $subject);
 795                         $text = $start . $subject . $end;
 796
 797                         $pos = self::getTagPosition($text, $name, $occurrences);
 798                 }
 799
 800                 return $text;
 801         }
 802
 803         private static function extractImagesFromItemBody($body)
 804         {
 805                 $saved_image = [];
 806                 $orig_body = $body;
 807                 $new_body = '';
 808
 809                 $cnt = 0;
 810                 $img_start = strpos($orig_body, '[img');
 811                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 812                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 813                 while (($img_st_close !== false) && ($img_end !== false)) {
 814                         $img_st_close++; // make it point to AFTER the closing bracket
 815                         $img_end += $img_start;
 816
 817                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 818                                 // This is an embedded image
 819                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 820                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 821
 822                                 $cnt++;
 823                         } else {
 824                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 825                         }
 826
 827                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 828
 829                         if ($orig_body === false) {
 830                                 // in case the body ends on a closing image tag
 831                                 $orig_body = '';
 832                         }
 833
 834                         $img_start = strpos($orig_body, '[img');
 835                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 836                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 837                 }
 838
 839                 $new_body = $new_body . $orig_body;
 840
 841                 return ['body' => $new_body, 'images' => $saved_image];
 842         }
 843
 844         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 845         {
 846                 $newbody = $body;
 847
 848                 $cnt = 0;
 849                 foreach ($images as $image) {
 850                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 851                         // it loops over the array starting from the first element and going sequentially
 852                         // to the last element
 853                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 854                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 855                         $cnt++;
 856                 }
 857
 858                 return $newbody;
 859         }
 860
 861         /**
 862          * Processes [share] tags
 863          *
 864          * Note: Can produce a [bookmark] tag in the output
 865          *
 866          * @brief Processes [share] tags
 867          * @param array    $share      preg_match_callback result array
 868          * @param bool|int $simplehtml
 869          * @return string
 870          */
 871         private static function convertShare($share, $simplehtml)
 872         {
 873                 $attributes = $share[2];
 874
 875                 $author = "";
 876                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
 877                 if (x($matches, 1)) {
 878                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
 879                 }
 880
 881                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
 882                 if (x($matches, 1)) {
 883                         $author = $matches[1];
 884                 }
 885
 886                 $profile = "";
 887                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
 888                 if (x($matches, 1)) {
 889                         $profile = $matches[1];
 890                 }
 891
 892                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
 893                 if (x($matches, 1)) {
 894                         $profile = $matches[1];
 895                 }
 896
 897                 $avatar = "";
 898                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
 899                 if (x($matches, 1)) {
 900                         $avatar = $matches[1];
 901                 }
 902
 903                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
 904                 if (x($matches, 1)) {
 905                         $avatar = $matches[1];
 906                 }
 907
 908                 $link = "";
 909                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
 910                 if (x($matches, 1)) {
 911                         $link = $matches[1];
 912                 }
 913
 914                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
 915                 if (x($matches, 1)) {
 916                         $link = $matches[1];
 917                 }
 918
 919                 $posted = "";
 920
 921                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
 922                 if (x($matches, 1)) {
 923                         $posted = $matches[1];
 924                 }
 925
 926                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
 927                 if (x($matches, 1)) {
 928                         $posted = $matches[1];
 929                 }
 930
 931                 // We only call this so that a previously unknown contact can be added.
 932                 // This is important for the function "Model\Contact::getDetailsByURL()".
 933                 // This function then can fetch an entry from the contact table.
 934                 Contact::getIdForURL($profile, 0, true);
 935
 936                 $data = Contact::getDetailsByURL($profile);
 937
 938                 if (x($data, "name") && x($data, "addr")) {
 939                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
 940                 } else {
 941                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
 942                 }
 943
 944                 if (x($data, "addr")) {
 945                         $userid = $data["addr"];
 946                 } else {
 947                         $userid = Protocol::formatMention($profile, $author);
 948                 }
 949
 950                 if (x($data, "name")) {
 951                         $author = $data["name"];
 952                 }
 953
 954                 if (x($data, "micro")) {
 955                         $avatar = $data["micro"];
 956                 }
 957
 958                 $preshare = trim($share[1]);
 959                 if ($preshare != "") {
 960                         $preshare .= "<br />";
 961                 }
 962
 963                 switch ($simplehtml) {
 964                         case 1:
 965                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
 966                                 break;
 967                         case 2:
 968                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
 969                                 break;
 970                         case 3: // Diaspora
 971                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
 972
 973                                 $text = trim($share[1]);
 974
 975                                 if ($text != "") {
 976                                         $text .= "<hr />";
 977                                 }
 978
 979                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
 980                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
 981                                 } else {
 982                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
 983
 984                                         if ($link != "") {
 985                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
 986                                         }
 987                                 }
 988
 989                                 break;
 990                         case 4:
 991                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 992                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
 993                                 $headline .= ":</b><br />";
 994
 995                                 $text = trim($share[1]);
 996
 997                                 if ($text != "") {
 998                                         $text .= "<hr />";
 999                                 }
1000
1001                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1002
1003                                 break;
1004                         case 5:
1005                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1006                                 break;
1007                         case 6: // app.net
1008                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1009                                 break;
1010                         case 7: // statusnet/GNU Social
1011                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1012                                 break;
1013                         case 8: // twitter
1014                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1015                                 break;
1016                         case 9: // Google+/Facebook
1017                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1018
1019                                 if ($link != "") {
1020                                         $text .= "<br /><br />" . $link;
1021                                 }
1022                                 break;
1023                         default:
1024                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1025                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1026                                         try {
1027                                                 $oembed = OEmbed::getHTML($link, $preshare);
1028                                         } catch (Exception $e) {
1029                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1030                                         }
1031
1032                                         $text = $preshare . $oembed;
1033                                 } else {
1034                                         $text = trim($share[1]) . "\n";
1035
1036                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1037
1038                                         $tpl = get_markup_template('shared_content.tpl');
1039                                         $text .= replace_macros($tpl, [
1040                                                 '$profile' => $profile,
1041                                                 '$avatar' => $avatar,
1042                                                 '$author' => $author,
1043                                                 '$link' => $link,
1044                                                 '$posted' => $posted,
1045                                                 '$content' => trim($share[3])
1046                                         ]);
1047                                 }
1048                                 break;
1049                 }
1050
1051                 return $text;
1052         }
1053
1054         private static function removePictureLinksCallback($match)
1055         {
1056                 $text = Cache::get($match[1]);
1057
1058                 if (is_null($text)) {
1059                         $a = self::getApp();
1060
1061                         $stamp1 = microtime(true);
1062
1063                         $ch = @curl_init($match[1]);
1064                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1065                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1066                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1067                         @curl_exec($ch);
1068                         $curl_info = @curl_getinfo($ch);
1069
1070                         $a->save_timestamp($stamp1, "network");
1071
1072                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1073                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1074                         } else {
1075                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1076
1077                                 // if its not a picture then look if its a page that contains a picture link
1078                                 $body = Network::fetchUrl($match[1]);
1079
1080                                 $doc = new DOMDocument();
1081                                 @$doc->loadHTML($body);
1082                                 $xpath = new DOMXPath($doc);
1083                                 $list = $xpath->query("//meta[@name]");
1084                                 foreach ($list as $node) {
1085                                         $attr = [];
1086
1087                                         if ($node->attributes->length) {
1088                                                 foreach ($node->attributes as $attribute) {
1089                                                         $attr[$attribute->name] = $attribute->value;
1090                                                 }
1091                                         }
1092
1093                                         if (strtolower($attr["name"]) == "twitter:image") {
1094                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1095                                         }
1096                                 }
1097                         }
1098                         Cache::set($match[1], $text);
1099                 }
1100
1101                 return $text;
1102         }
1103
1104         private static function expandLinksCallback($match)
1105         {
1106                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1107                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1108                 } else {
1109                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1110                 }
1111         }
1112
1113         private static function cleanPictureLinksCallback($match)
1114         {
1115                 $text = Cache::get($match[1]);
1116
1117                 if (is_null($text)) {
1118                         $a = self::getApp();
1119
1120                         $stamp1 = microtime(true);
1121
1122                         $ch = @curl_init($match[1]);
1123                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1124                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1125                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1126                         @curl_exec($ch);
1127                         $curl_info = @curl_getinfo($ch);
1128
1129                         $a->save_timestamp($stamp1, "network");
1130
1131                         // if its a link to a picture then embed this picture
1132                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1133                                 $text = "[img]" . $match[1] . "[/img]";
1134                         } else {
1135                                 $text = "[img]" . $match[2] . "[/img]";
1136
1137                                 // if its not a picture then look if its a page that contains a picture link
1138                                 $body = Network::fetchUrl($match[1]);
1139
1140                                 $doc = new DOMDocument();
1141                                 @$doc->loadHTML($body);
1142                                 $xpath = new DOMXPath($doc);
1143                                 $list = $xpath->query("//meta[@name]");
1144                                 foreach ($list as $node) {
1145                                         $attr = [];
1146                                         if ($node->attributes->length) {
1147                                                 foreach ($node->attributes as $attribute) {
1148                                                         $attr[$attribute->name] = $attribute->value;
1149                                                 }
1150                                         }
1151
1152                                         if (strtolower($attr["name"]) == "twitter:image") {
1153                                                 $text = "[img]" . $attr["content"] . "[/img]";
1154                                         }
1155                                 }
1156                         }
1157                         Cache::set($match[1], $text);
1158                 }
1159
1160                 return $text;
1161         }
1162
1163         public static function cleanPictureLinks($text)
1164         {
1165                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1166                 return $return;
1167         }
1168
1169         private static function textHighlightCallback($match)
1170         {
1171                 // Fallback in case the language doesn't exist
1172                 $return = '[code]' . $match[2] . '[/code]';
1173
1174                 if (in_array(strtolower($match[1]),
1175                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1176                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1177                 ) {
1178                         $return = text_highlight($match[2], strtolower($match[1]));
1179                 }
1180
1181                 return $return;
1182         }
1183
1184         /**
1185          * @brief Converts a BBCode message to HTML message
1186          *
1187          * BBcode 2 HTML was written by WAY2WEB.net
1188          * extended to work with Mistpark/Friendica - Mike Macgirvin
1189          *
1190          * Simple HTML values meaning:
1191          * - 0: Friendica display
1192          * - 1: Unused
1193          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1194          * - 3: Used before converting to Markdown in bb2diaspora.php
1195          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1196          * - 5: Unused
1197          * - 6: Used for Appnet
1198          * - 7: Used for dfrn, OStatus
1199          * - 8: Used for WP backlink text setting
1200          *
1201          * @param string $text
1202          * @param bool   $try_oembed
1203          * @param int    $simple_html
1204          * @param bool   $for_plaintext
1205          * @return string
1206          */
1207         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1208         {
1209                 $a = self::getApp();
1210
1211                 /*
1212                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1213                  *
1214                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1215                  * $match[1] = $url
1216                  * $match[2] = $title or absent
1217                  */
1218                 $try_oembed_callback = function ($match)
1219                 {
1220                         $url = $match[1];
1221                         $title = defaults($match, 2, null);
1222
1223                         try {
1224                                 $return = OEmbed::getHTML($url, $title);
1225                         } catch (Exception $ex) {
1226                                 $return = $match[0];
1227                         }
1228
1229                         return $return;
1230                 };
1231
1232                 // Hide all [noparse] contained bbtags by spacefying them
1233                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1234
1235                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1236                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1237                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1238
1239                 // Remove the abstract element. It is a non visible element.
1240                 $text = self::stripAbstract($text);
1241
1242                 // Move all spaces out of the tags
1243                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1244                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1245
1246                 // Extract the private images which use data urls since preg has issues with
1247                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1248                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1249
1250                 $extracted = self::extractImagesFromItemBody($text);
1251                 $text = $extracted['body'];
1252                 $saved_image = $extracted['images'];
1253
1254                 // If we find any event code, turn it into an event.
1255                 // After we're finished processing the bbcode we'll
1256                 // replace all of the event code with a reformatted version.
1257
1258                 $ev = Event::fromBBCode($text);
1259
1260                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1261                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1262
1263                 $text = str_replace("<", "&lt;", $text);
1264                 $text = str_replace(">", "&gt;", $text);
1265
1266                 // remove some newlines before the general conversion
1267                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1268                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1269
1270                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1271                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1272
1273                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1274                 if (!$try_oembed) {
1275                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1276                 }
1277
1278                 // Check for [code] text here, before the linefeeds are messed with.
1279                 // The highlighter will unescape and re-escape the content.
1280                 if (strpos($text, '[code=') !== false) {
1281                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1282                 }
1283                 // Convert new line chars to html <br /> tags
1284
1285                 // nlbr seems to be hopelessly messed up
1286                 //      $Text = nl2br($Text);
1287
1288                 // We'll emulate it.
1289
1290                 $text = trim($text);
1291                 $text = str_replace("\r\n", "\n", $text);
1292
1293                 // removing multiplicated newlines
1294                 if (Config::get("system", "remove_multiplicated_lines")) {
1295                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1296                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1297                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1298                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1299                         do {
1300                                 $oldtext = $text;
1301                                 $text = str_replace($search, $replace, $text);
1302                         } while ($oldtext != $text);
1303                 }
1304
1305                 // Set up the parameters for a URL search string
1306                 $URLSearchString = "^\[\]";
1307                 // Set up the parameters for a MAIL search string
1308                 $MAILSearchString = $URLSearchString;
1309
1310                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1311                 if (!$for_plaintext) {
1312                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1313                         // Currently disabled, since the function is too greedy
1314                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1315                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1316                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1317                         if ($simple_html == 7) {
1318                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1319                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1320                         }
1321                 } else {
1322                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1323                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1324                 }
1325
1326
1327                 // Handle attached links or videos
1328                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1329
1330                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1331
1332                 // Remove all hashtag addresses
1333                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1334                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1335                 } elseif ($simple_html == 3) {
1336                         // The ! is converted to @ since Diaspora only understands the @
1337                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1338                                 '@<a href="$2">$3</a>',
1339                                 $text);
1340                 } elseif ($simple_html == 7) {
1341                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1342                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1343                                 $text);
1344                 } elseif (!$simple_html) {
1345                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1346                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1347                                 $text);
1348                 }
1349
1350                 // Bookmarks in red - will be converted to bookmarks in friendica
1351                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1352                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1353                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1354                                         "[bookmark=$1]$2[/bookmark]", $text);
1355
1356                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1357                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1358                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1359                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1360                 }
1361
1362                 if ($simple_html == 5) {
1363                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1364                 }
1365
1366                 // Perform URL Search
1367                 if ($try_oembed) {
1368                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1369                 }
1370
1371                 if ($simple_html == 5) {
1372                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1373                 } else {
1374                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1375                 }
1376
1377                 // Handle Diaspora posts
1378                 $text = preg_replace_callback(
1379                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1380                         function ($match) {
1381                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1382                         }, $text
1383                 );
1384
1385                 $text = preg_replace_callback(
1386                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1387                         function ($match) {
1388                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1389                         }, $text
1390                 );
1391
1392                 // Server independent link to posts and comments
1393                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1394                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1395                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1396
1397                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1398                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1399
1400                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1401                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1402
1403                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1404                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1405                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1406
1407                 // Red compatibility, though the link can't be authenticated on Friendica
1408                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1409
1410
1411                 // we may need to restrict this further if it picks up too many strays
1412                 // link acct:user@host to a webfinger profile redirector
1413
1414                 $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);
1415
1416                 // Perform MAIL Search
1417                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1418                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1419
1420                 // leave open the posibility of [map=something]
1421                 // this is replaced in prepare_body() which has knowledge of the item location
1422
1423                 if (strpos($text, '[/map]') !== false) {
1424                         $text = preg_replace_callback(
1425                                 "/\[map\](.*?)\[\/map\]/ism",
1426                                 function ($match) use ($simple_html) {
1427                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1428                                 },
1429                                 $text
1430                         );
1431                 }
1432                 if (strpos($text, '[map=') !== false) {
1433                         $text = preg_replace_callback(
1434                                 "/\[map=(.*?)\]/ism",
1435                                 function ($match) use ($simple_html) {
1436                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1437                                 },
1438                                 $text
1439                         );
1440                 }
1441                 if (strpos($text, '[map]') !== false) {
1442                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1443                 }
1444
1445                 // Check for headers
1446                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1447                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1448                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1449                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1450                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1451                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1452
1453                 // Check for paragraph
1454                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1455
1456                 // Check for bold text
1457                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1458
1459                 // Check for Italics text
1460                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1461
1462                 // Check for Underline text
1463                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1464
1465                 // Check for strike-through text
1466                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1467
1468                 // Check for over-line text
1469                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1470
1471                 // Check for colored text
1472                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1473
1474                 // Check for sized text
1475                 // [size=50] --> font-size: 50px (with the unit).
1476                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1477                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1478
1479                 // Check for centered text
1480                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1481
1482                 // Check for list text
1483                 $text = str_replace("[*]", "<li>", $text);
1484
1485                 // Check for style sheet commands
1486                 $text = preg_replace_callback(
1487                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1488                         function ($match) {
1489                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1490                         },
1491                         $text
1492                 );
1493
1494                 // Check for CSS classes
1495                 $text = preg_replace_callback(
1496                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1497                         function ($match) {
1498                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1499                         },
1500                         $text
1501                 );
1502
1503                 // handle nested lists
1504                 $endlessloop = 0;
1505
1506                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1507                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1508                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1509                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1510                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1511                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1512                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1513                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1514                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1515                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1516                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1517                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1518                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1519                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1520                 }
1521
1522                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1523                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1524                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1525                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1526
1527                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1528                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1529
1530                 $text = str_replace('[hr]', '<hr />', $text);
1531
1532                 // This is actually executed in prepare_body()
1533
1534                 $text = str_replace('[nosmile]', '', $text);
1535
1536                 // Check for font change text
1537                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1538
1539                 // Declare the format for [code] layout
1540
1541                 $CodeLayout = '<code>$1</code>';
1542                 // Check for [code] text
1543                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1544
1545                 // Declare the format for [spoiler] layout
1546                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1547
1548                 // Check for [spoiler] text
1549                 // handle nested quotes
1550                 $endlessloop = 0;
1551                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1552                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1553                 }
1554
1555                 // Check for [spoiler=Author] text
1556
1557                 $t_wrote = L10n::t('$1 wrote:');
1558
1559                 // handle nested quotes
1560                 $endlessloop = 0;
1561                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1562                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1563                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1564                                                  $text);
1565                 }
1566
1567                 // Declare the format for [quote] layout
1568                 $QuoteLayout = '<blockquote>$1</blockquote>';
1569
1570                 // Check for [quote] text
1571                 // handle nested quotes
1572                 $endlessloop = 0;
1573                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1574                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1575                 }
1576
1577                 // Check for [quote=Author] text
1578
1579                 $t_wrote = L10n::t('$1 wrote:');
1580
1581                 // handle nested quotes
1582                 $endlessloop = 0;
1583                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1584                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1585                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1586                                                  $text);
1587                 }
1588
1589
1590                 // [img=widthxheight]image source[/img]
1591                 $text = preg_replace_callback(
1592                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1593                         function ($matches) use ($simple_html) {
1594                                 if (strpos($matches[3], "data:image/") === 0) {
1595                                         return $matches[0];
1596                                 }
1597
1598                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1599                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1600                         },
1601                         $text
1602                 );
1603
1604                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1605                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1606
1607                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1608                         function ($matches) use ($simple_html) {
1609                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1610                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1611                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1612                         },
1613                         $text);
1614
1615                 // Images
1616                 // [img]pathtoimage[/img]
1617                 $text = preg_replace_callback(
1618                         "/\[img\](.*?)\[\/img\]/ism",
1619                         function ($matches) use ($simple_html) {
1620                                 if (strpos($matches[1], "data:image/") === 0) {
1621                                         return $matches[0];
1622                                 }
1623
1624                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1625                                 return "[img]" . $matches[1] . "[/img]";
1626                         },
1627                         $text
1628                 );
1629
1630                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1631                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1632
1633                 // Shared content
1634                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1635                         function ($match) use ($simple_html) {
1636                                 return self::convertShare($match, $simple_html);
1637                         }, $text);
1638
1639                 $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);
1640                 $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 />', $text);
1641                 //$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 />', $Text);
1642
1643                 // Try to Oembed
1644                 if ($try_oembed) {
1645                         $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);
1646                         $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);
1647
1648                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1649                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1650                 } else {
1651                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1652                                                 '<a href="$1" target="_blank">$1</a>', $text);
1653                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1654                                                 '<a href="$1" target="_blank">$1</a>', $text);
1655                 }
1656
1657                 // html5 video and audio
1658
1659
1660                 if ($try_oembed) {
1661                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1662                 } else {
1663                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1664                 }
1665
1666                 // Youtube extensions
1667                 if ($try_oembed) {
1668                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1669                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1670                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1671                 }
1672
1673                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1674                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1675                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1676
1677                 if ($try_oembed) {
1678                         $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);
1679                 } else {
1680                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1681                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1682                 }
1683
1684                 if ($try_oembed) {
1685                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1686                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1687                 }
1688
1689                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1690                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1691
1692                 if ($try_oembed) {
1693                         $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);
1694                 } else {
1695                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1696                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1697                 }
1698
1699                 // oembed tag
1700                 $text = OEmbed::BBCode2HTML($text);
1701
1702                 // Avoid triple linefeeds through oembed
1703                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1704
1705                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1706                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1707                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1708                 // start which is always required). Allow desc with a missing summary for compatibility.
1709
1710                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1711                         $sub = Event::getHTML($ev, $simple_html);
1712
1713                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1714                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1715                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1716                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1717                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1718                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1719                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1720                 }
1721
1722                 // Replace non graphical smilies for external posts
1723                 if ($simple_html) {
1724                         $text = Smilies::replace($text, false, true);
1725                 }
1726
1727                 // Replace inline code blocks
1728                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1729                         function ($match) use ($simple_html) {
1730                                 $return = '<key>' . $match[1] . '</key>';
1731                                 // Use <code> for Diaspora inline code blocks
1732                                 if ($simple_html === 3) {
1733                                         $return = '<code>' . $match[1] . '</code>';
1734                                 }
1735                                 return $return;
1736                         }
1737                 , $text);
1738
1739                 // Unhide all [noparse] contained bbtags unspacefying them
1740                 // and triming the [noparse] tag.
1741
1742                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1743                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1744                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1745
1746                 /// @todo What is the meaning of these lines?
1747                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1748                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1749
1750                 // Currently deactivated, it made problems with " inside of alt texts.
1751                 //$text = preg_replace('/\&quot\;/', '"', $text);
1752
1753                 // fix any escaped ampersands that may have been converted into links
1754                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1755
1756                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1757                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1758                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1759                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1760
1761                 // sanitize href attributes (only whitelisted protocols URLs)
1762                 // default value for backward compatibility
1763                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1764
1765                 // Always allowed protocol even if config isn't set or not including it
1766                 $allowed_link_protocols[] = 'http';
1767                 $allowed_link_protocols[] = 'redir/';
1768
1769                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1770                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1771
1772                 if ($saved_image) {
1773                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1774                 }
1775
1776                 // Clean up the HTML by loading and saving the HTML with the DOM.
1777                 // Bad structured html can break a whole page.
1778                 // For performance reasons do it only with ativated item cache or at export.
1779                 if (!$try_oembed || (get_itemcachepath() != "")) {
1780                         $doc = new DOMDocument();
1781                         $doc->preserveWhiteSpace = false;
1782
1783                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1784
1785                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1786                         $encoding = '<?xml encoding="UTF-8">';
1787                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1788                         $doc->encoding = 'UTF-8';
1789                         $text = $doc->saveHTML();
1790                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1791
1792                         $text = str_replace('<br></li>', '</li>', $text);
1793
1794                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1795                 }
1796
1797                 // Clean up some useless linebreaks in lists
1798                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1799                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1800                 //$Text = str_replace('</li><br />', '</li>', $Text);
1801                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1802                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1803
1804                 Addon::callHooks('bbcode', $text);
1805
1806                 return trim($text);
1807         }
1808
1809         /**
1810          * @brief Strips the "abstract" tag from the provided text
1811          *
1812          * @param string $text The text with BBCode
1813          * @return string The same text - but without "abstract" element
1814          */
1815         public static function stripAbstract($text)
1816         {
1817                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1818                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1819
1820                 return $text;
1821         }
1822
1823         /**
1824          * @brief Returns the value of the "abstract" element
1825          *
1826          * @param string $text The text that maybe contains the element
1827          * @param string $addon The addon for which the abstract is meant for
1828          * @return string The abstract
1829          */
1830         public static function getAbstract($text, $addon = "")
1831         {
1832                 $abstract = "";
1833                 $abstracts = [];
1834                 $addon = strtolower($addon);
1835
1836                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1837                         foreach ($results AS $result) {
1838                                 $abstracts[strtolower($result[1])] = $result[2];
1839                         }
1840                 }
1841
1842                 if (isset($abstracts[$addon])) {
1843                         $abstract = $abstracts[$addon];
1844                 }
1845
1846                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1847                         $abstract = $result[1];
1848                 }
1849
1850                 return $abstract;
1851         }
1852
1853         /**
1854          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1855          *
1856          * @param array $match Matching values for the callback
1857          * @return string Replaced mention
1858          */
1859         private static function bbCodeMention2DiasporaCallback($match)
1860         {
1861                 $contact = Contact::getDetailsByURL($match[3]);
1862
1863                 if (empty($contact['addr'])) {
1864                         $contact = Probe::uri($match[3]);
1865                 }
1866
1867                 if (empty($contact['addr'])) {
1868                         return $match[0];
1869                 }
1870
1871                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1872                 return $mention;
1873         }
1874
1875         /**
1876          * @brief Converts a BBCode text into Markdown
1877          *
1878          * This function converts a BBCode item body to be sent to Markdown-enabled
1879          * systems like Diaspora and Libertree
1880          *
1881          * @param string $text
1882          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1883          * @return string
1884          */
1885         public static function toMarkdown($text, $for_diaspora = true)
1886         {
1887                 $a = self::getApp();
1888
1889                 $original_text = $text;
1890
1891                 // Since Diaspora is creating a summary for links, this function removes them before posting
1892                 if ($for_diaspora) {
1893                         $text = self::removeShareInformation($text);
1894                 }
1895
1896                 /**
1897                  * Transform #tags, strip off the [url] and replace spaces with underscore
1898                  */
1899                 $url_search_string = "^\[\]";
1900                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1901                         function ($matches) {
1902                                 return '#' . str_replace(' ', '_', $matches[2]);
1903                         },
1904                         $text
1905                 );
1906
1907                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1908                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1909
1910                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
1911                 $codeblocks = [];
1912
1913                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1914                         function ($matches) use (&$codeblocks) {
1915                                 $return = $matches[0];
1916                                 if (strpos($matches[2], "\n") !== false) {
1917                                         $return = '#codeblock-' . count($codeblocks) . '#';
1918
1919                                         $prefix = '````' . $matches[1] . PHP_EOL;
1920                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
1921                                 }
1922                                 return $return;
1923                         },
1924                         $text
1925                 );
1926
1927                 // Convert it to HTML - don't try oembed
1928                 if ($for_diaspora) {
1929                         $text = self::convert($text, false, 3);
1930
1931                         // Add all tags that maybe were removed
1932                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1933                                 $tagline = "";
1934                                 foreach ($tags[2] as $tag) {
1935                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1936                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1937                                                 $tagline .= '#' . $tag . ' ';
1938                                         }
1939                                 }
1940                                 $text = $text . " " . $tagline;
1941                         }
1942                 } else {
1943                         $text = self::convert($text, false, 4);
1944                 }
1945
1946                 // mask some special HTML chars from conversation to markdown
1947                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1948
1949                 // If a link is followed by a quote then there should be a newline before it
1950                 // Maybe we should make this newline at every time before a quote.
1951                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1952
1953                 $stamp1 = microtime(true);
1954
1955                 // Now convert HTML to Markdown
1956                 $converter = new HtmlConverter();
1957                 $text = $converter->convert($text);
1958
1959                 // unmask the special chars back to HTML
1960                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1961
1962                 $a->save_timestamp($stamp1, "parser");
1963
1964                 // Libertree has a problem with escaped hashtags.
1965                 $text = str_replace(['\#'], ['#'], $text);
1966
1967                 // Remove any leading or trailing whitespace, as this will mess up
1968                 // the Diaspora signature verification and cause the item to disappear
1969                 $text = trim($text);
1970
1971                 if ($for_diaspora) {
1972                         $url_search_string = "^\[\]";
1973                         $text = preg_replace_callback(
1974                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
1975                                 ['self', 'bbCodeMention2DiasporaCallback'],
1976                                 $text
1977                         );
1978                 }
1979
1980                 // Restore code blocks
1981                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1982                         function ($matches) use ($codeblocks) {
1983                                 $return = '';
1984                                 if (isset($codeblocks[intval($matches[1])])) {
1985                                         $return = $codeblocks[$matches[1]];
1986                                 }
1987                                 return $return;
1988                         },
1989                         $text
1990                 );
1991
1992                 Addon::callHooks('bb2diaspora', $text);
1993
1994                 return $text;
1995         }
1996 }