src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5 namespace Friendica\Content\Text;
   6
   7 use DOMDocument;
   8 use DomXPath;
   9 use Exception;
  10 use Friendica\Content\OEmbed;
  11 use Friendica\Content\Smilies;
  12 use Friendica\Content\Text\Plaintext;
  13 use Friendica\Core\Addon;
  14 use Friendica\Core\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\L10n;
  17 use Friendica\Core\Protocol;
  18 use Friendica\Core\PConfig;
  19 use Friendica\Core\System;
  20 use Friendica\Model\Contact;
  21 use Friendica\Object\Image;
  22 use Friendica\Util\Map;
  23 use Friendica\Util\Network;
  24 use Friendica\Util\ParseUrl;
  25
  26 require_once "include/event.php";
  27 require_once "include/html2plain.php";
  28 require_once "include/html2bbcode.php";
  29 require_once "mod/proxy.php";
  30
  31 class BBCode
  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 (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  78                                                 $post["image"] = $matches[1];
  79                                         } else {
  80                                                 $post["preview"] = $matches[1];
  81                                         }
  82                                 }
  83
  84                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  85                                         $post["url"] = $matches[1];
  86                                         $post["title"] = $matches[2];
  87                                 }
  88                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  89                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  90                                         $post["url"] = $matches[1];
  91                                 }
  92
  93                                 // Search for description
  94                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  95                                         $post["description"] = $matches[1];
  96                                 }
  97                         }
  98                 }
  99                 return $post;
 100         }
 101
 102         /**
 103          * @brief Fetches attachment data that were generated with the "attachment" element
 104          *
 105          * @param string $body Message body
 106          * @return array
 107          * 'type' -> Message type ("link", "video", "photo")
 108          * 'text' -> Text before the shared message
 109          * 'after' -> Text after the shared message
 110          * 'image' -> Preview image of the message
 111          * 'url' -> Url to the attached message
 112          * 'title' -> Title of the attachment
 113          * 'description' -> Description of the attachment
 114          */
 115         public static function getAttachmentData($body)
 116         {
 117                 $data = [];
 118
 119                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 120                         return self::getOldAttachmentData($body);
 121                 }
 122
 123                 $attributes = $match[2];
 124
 125                 $data["text"] = trim($match[1]);
 126
 127                 $type = "";
 128                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 129                 if (x($matches, 1)) {
 130                         $type = strtolower($matches[1]);
 131                 }
 132
 133                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 134                 if (x($matches, 1)) {
 135                         $type = strtolower($matches[1]);
 136                 }
 137
 138                 if ($type == "") {
 139                         return [];
 140                 }
 141
 142                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 143                         return [];
 144                 }
 145
 146                 if ($type != "") {
 147                         $data["type"] = $type;
 148                 }
 149
 150                 $url = "";
 151                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 152                 if (x($matches, 1)) {
 153                         $url = $matches[1];
 154                 }
 155
 156                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 157                 if (x($matches, 1)) {
 158                         $url = $matches[1];
 159                 }
 160
 161                 if ($url != "") {
 162                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 163                 }
 164
 165                 $title = "";
 166                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 167                 if (x($matches, 1)) {
 168                         $title = $matches[1];
 169                 }
 170
 171                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 172                 if (x($matches, 1)) {
 173                         $title = $matches[1];
 174                 }
 175
 176                 if ($title != "") {
 177                         $title = BBCode::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 178                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 179                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 180                         $data["title"] = $title;
 181                 }
 182
 183                 $image = "";
 184                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 185                 if (x($matches, 1)) {
 186                         $image = $matches[1];
 187                 }
 188
 189                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 190                 if (x($matches, 1)) {
 191                         $image = $matches[1];
 192                 }
 193
 194                 if ($image != "") {
 195                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 196                 }
 197
 198                 $preview = "";
 199                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 200                 if (x($matches, 1)) {
 201                         $preview = $matches[1];
 202                 }
 203
 204                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 205                 if (x($matches, 1)) {
 206                         $preview = $matches[1];
 207                 }
 208
 209                 if ($preview != "") {
 210                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 211                 }
 212
 213                 $data["description"] = trim($match[3]);
 214
 215                 $data["after"] = trim($match[4]);
 216
 217                 return $data;
 218         }
 219
 220         public static function getAttachedData($body, $item = [])
 221         {
 222                 /*
 223                 - text:
 224                 - type: link, video, photo
 225                 - title:
 226                 - url:
 227                 - image:
 228                 - description:
 229                 - (thumbnail)
 230                 */
 231
 232                 $has_title = !empty($item['title']);
 233                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 234                 $post = self::getAttachmentData($body);
 235
 236                 // if nothing is found, it maybe having an image.
 237                 if (!isset($post["type"])) {
 238                         // Simplify image codes
 239                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 240
 241                         $URLSearchString = "^\[\]";
 242                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 243                                 if ((count($pictures) == 1) && !$has_title) {
 244                                         // Checking, if the link goes to a picture
 245                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 246
 247                                         // Workaround:
 248                                         // Sometimes photo posts to the own album are not detected at the start.
 249                                         // So we seem to cannot use the cache for these cases. That's strange.
 250                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 251                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 252                                         }
 253
 254                                         if ($data["type"] == "photo") {
 255                                                 $post["type"] = "photo";
 256                                                 if (isset($data["images"][0])) {
 257                                                         $post["image"] = $data["images"][0]["src"];
 258                                                         $post["url"] = $data["url"];
 259                                                 } else {
 260                                                         $post["image"] = $data["url"];
 261                                                 }
 262
 263                                                 $post["preview"] = $pictures[0][2];
 264                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 265                                         } else {
 266                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 267                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 268                                                         $post["type"] = "photo";
 269                                                         $post["image"] = $pictures[0][1];
 270                                                         $post["preview"] = $pictures[0][2];
 271                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 272                                                 }
 273                                         }
 274                                 } elseif (count($pictures) > 0) {
 275                                         $post["type"] = "link";
 276                                         $post["url"] = $plink;
 277                                         $post["image"] = $pictures[0][2];
 278                                         $post["text"] = $body;
 279                                 }
 280                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 281                                 if ((count($pictures) == 1) && !$has_title) {
 282                                         $post["type"] = "photo";
 283                                         $post["image"] = $pictures[0][1];
 284                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 285                                 } elseif (count($pictures) > 0) {
 286                                         $post["type"] = "link";
 287                                         $post["url"] = $plink;
 288                                         $post["image"] = $pictures[0][1];
 289                                         $post["text"] = $body;
 290                                 }
 291                         }
 292
 293                         // Test for the external links
 294                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 295                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 296
 297                         $links = array_merge($links1, $links2);
 298
 299                         // If there is only a single one, then use it.
 300                         // This should cover link posts via API.
 301                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 302                                 $post["type"] = "link";
 303                                 $post["text"] = trim($body);
 304                                 $post["url"] = $links[0][1];
 305                         }
 306
 307                         // Now count the number of external media links
 308                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 309                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 310                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 311                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 312
 313                         // Add them to the other external links
 314                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 315
 316                         // Are there more than one?
 317                         if (count($links) > 1) {
 318                                 // The post will be the type "text", which means a blog post
 319                                 unset($post["type"]);
 320                                 $post["url"] = $plink;
 321                         }
 322
 323                         if (!isset($post["type"])) {
 324                                 $post["type"] = "text";
 325                                 $post["text"] = trim($body);
 326                         }
 327                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 328                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 329
 330                         if (isset($data["images"][0])) {
 331                                 $post["image"] = $data["images"][0]["src"];
 332                         }
 333                 }
 334
 335                 return $post;
 336         }
 337
 338         /**
 339          * @brief Convert a message into plaintext for connectors to other networks
 340          *
 341          * @param array $b The message array that is about to be posted
 342          * @param int $limit The maximum number of characters when posting to that network
 343          * @param bool $includedlinks Has an attached link to be included into the message?
 344          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 345          * @param string $target_network Name of the network where the post should go to.
 346          *
 347          * @return string The converted message
 348          */
 349         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 350         {
 351                 // Remove the hash tags
 352                 $URLSearchString = "^\[\]";
 353                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 354
 355                 // Add an URL element if the text contains a raw link
 356                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 357
 358                 // Remove the abstract
 359                 $body = self::stripAbstract($body);
 360
 361                 // At first look at data that is attached via "type-..." stuff
 362                 // This will hopefully replaced with a dedicated bbcode later
 363                 //$post = self::getAttachedData($b["body"]);
 364                 $post = self::getAttachedData($body, $b);
 365
 366                 if (($b["title"] != "") && ($post["text"] != "")) {
 367                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 368                 } elseif ($b["title"] != "") {
 369                         $post["text"] = trim($b["title"]);
 370                 }
 371
 372                 $abstract = "";
 373
 374                 // Fetch the abstract from the given target network
 375                 if ($target_network != "") {
 376                         $default_abstract = self::getAbstract($b["body"]);
 377                         $abstract = self::getAbstract($b["body"], $target_network);
 378
 379                         // If we post to a network with no limit we only fetch
 380                         // an abstract exactly for this network
 381                         if (($limit == 0) && ($abstract == $default_abstract)) {
 382                                 $abstract = "";
 383                         }
 384                 } else {// Try to guess the correct target network
 385                         switch ($htmlmode) {
 386                                 case 8:
 387                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
 388                                         break;
 389                                 case 7:
 390                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
 391                                         break;
 392                                 case 6:
 393                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
 394                                         break;
 395                                 default: // We don't know the exact target.
 396                                         // We fetch an abstract since there is a posting limit.
 397                                         if ($limit > 0) {
 398                                                 $abstract = self::getAbstract($b["body"]);
 399                                         }
 400                         }
 401                 }
 402
 403                 if ($abstract != "") {
 404                         $post["text"] = $abstract;
 405
 406                         if ($post["type"] == "text") {
 407                                 $post["type"] = "link";
 408                                 $post["url"] = $b["plink"];
 409                         }
 410                 }
 411
 412                 $html = BBCode::convert($post["text"].$post["after"], false, $htmlmode);
 413                 $msg = html2plain($html, 0, true);
 414                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 415
 416                 $link = "";
 417                 if ($includedlinks) {
 418                         if ($post["type"] == "link") {
 419                                 $link = $post["url"];
 420                         } elseif ($post["type"] == "text") {
 421                                 $link = $post["url"];
 422                         } elseif ($post["type"] == "video") {
 423                                 $link = $post["url"];
 424                         } elseif ($post["type"] == "photo") {
 425                                 $link = $post["image"];
 426                         }
 427
 428                         if (($msg == "") && isset($post["title"])) {
 429                                 $msg = trim($post["title"]);
 430                         }
 431
 432                         if (($msg == "") && isset($post["description"])) {
 433                                 $msg = trim($post["description"]);
 434                         }
 435
 436                         // If the link is already contained in the post, then it neeedn't to be added again
 437                         // But: if the link is beyond the limit, then it has to be added.
 438                         if (($link != "") && strstr($msg, $link)) {
 439                                 $pos = strpos($msg, $link);
 440
 441                                 // Will the text be shortened in the link?
 442                                 // Or is the link the last item in the post?
 443                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 444                                         $msg = trim(str_replace($link, "", $msg));
 445                                 } elseif (($limit == 0) || ($pos < $limit)) {
 446                                         // The limit has to be increased since it will be shortened - but not now
 447                                         // Only do it with Twitter (htmlmode = 8)
 448                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 449                                                 $limit = $limit - 23 + strlen($link);
 450                                         }
 451
 452                                         $link = "";
 453
 454                                         if ($post["type"] == "text") {
 455                                                 unset($post["url"]);
 456                                         }
 457                                 }
 458                         }
 459                 }
 460
 461                 if ($limit > 0) {
 462                         // Reduce multiple spaces
 463                         // When posted to a network with limited space, we try to gain space where possible
 464                         while (strpos($msg, "  ") !== false) {
 465                                 $msg = str_replace("  ", " ", $msg);
 466                         }
 467
 468                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 469                         if (iconv_strlen($link, "UTF-8") > 0) {
 470                                 $limit = $limit - 23;
 471                         }
 472
 473                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 474                                 if (($post["type"] == "text") && isset($post["url"])) {
 475                                         $post["url"] = $b["plink"];
 476                                 } elseif (!isset($post["url"])) {
 477                                         $limit = $limit - 23;
 478                                         $post["url"] = $b["plink"];
 479                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 480                                 //} elseif (strpos($b["body"], "[share") !== false) {
 481                                 //      $post["url"] = $b["plink"];
 482                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 483                                         $post["url"] = $b["plink"];
 484                                 }
 485                                 $msg = Plaintext::shorten($msg, $limit);
 486                         }
 487                 }
 488
 489                 $post["text"] = trim($msg);
 490
 491                 return($post);
 492         }
 493
 494         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 495         {
 496                 // Suppress "view full size"
 497                 if (intval(Config::get('system', 'no_view_full_size'))) {
 498                         $include_link = false;
 499                 }
 500
 501                 // Picture addresses can contain special characters
 502                 $s = htmlspecialchars_decode($srctext);
 503
 504                 $matches = null;
 505                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 506                 if ($c) {
 507                         foreach ($matches as $mtch) {
 508                                 logger('scale_external_image: ' . $mtch[1]);
 509
 510                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 511                                 if (stristr($mtch[1], $hostname)) {
 512                                         continue;
 513                                 }
 514
 515                                 // $scale_replace, if passed, is an array of two elements. The
 516                                 // first is the name of the full-size image. The second is the
 517                                 // name of a remote, scaled-down version of the full size image.
 518                                 // This allows Friendica to display the smaller remote image if
 519                                 // one exists, while still linking to the full-size image
 520                                 if ($scale_replace) {
 521                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 522                                 } else {
 523                                         $scaled = $mtch[1];
 524                                 }
 525                                 $i = Network::fetchUrl($scaled);
 526                                 if (!$i) {
 527                                         return $srctext;
 528                                 }
 529
 530                                 // guess mimetype from headers or filename
 531                                 $type = Image::guessType($mtch[1], true);
 532
 533                                 if ($i) {
 534                                         $Image = new Image($i, $type);
 535                                         if ($Image->isValid()) {
 536                                                 $orig_width = $Image->getWidth();
 537                                                 $orig_height = $Image->getHeight();
 538
 539                                                 if ($orig_width > 640 || $orig_height > 640) {
 540                                                         $Image->scaleDown(640);
 541                                                         $new_width = $Image->getWidth();
 542                                                         $new_height = $Image->getHeight();
 543                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 544                                                         $s = str_replace(
 545                                                                 $mtch[0],
 546                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 547                                                                 . "\n" . (($include_link)
 548                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 549                                                                         : ''),
 550                                                                 $s
 551                                                         );
 552                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 553                                                 }
 554                                         }
 555                                 }
 556                         }
 557                 }
 558
 559                 // replace the special char encoding
 560                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 561                 return $s;
 562         }
 563
 564         /**
 565          * The purpose of this function is to apply system message length limits to
 566          * imported messages without including any embedded photos in the length
 567          *
 568          * @brief Truncates imported message body string length to max_import_size
 569          * @param string $body
 570          * @return string
 571          */
 572         public static function limitBodySize($body)
 573         {
 574                 $maxlen = get_max_import_size();
 575
 576                 // If the length of the body, including the embedded images, is smaller
 577                 // than the maximum, then don't waste time looking for the images
 578                 if ($maxlen && (strlen($body) > $maxlen)) {
 579
 580                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 581
 582                         $orig_body = $body;
 583                         $new_body = '';
 584                         $textlen = 0;
 585
 586                         $img_start = strpos($orig_body, '[img');
 587                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 588                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 589                         while (($img_st_close !== false) && ($img_end !== false)) {
 590
 591                                 $img_st_close++; // make it point to AFTER the closing bracket
 592                                 $img_end += $img_start;
 593                                 $img_end += strlen('[/img]');
 594
 595                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 596                                         // This is an embedded image
 597
 598                                         if (($textlen + $img_start) > $maxlen) {
 599                                                 if ($textlen < $maxlen) {
 600                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 601                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 602                                                         $textlen = $maxlen;
 603                                                 }
 604                                         } else {
 605                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 606                                                 $textlen += $img_start;
 607                                         }
 608
 609                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 610                                 } else {
 611
 612                                         if (($textlen + $img_end) > $maxlen) {
 613                                                 if ($textlen < $maxlen) {
 614                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 615                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 616                                                         $textlen = $maxlen;
 617                                                 }
 618                                         } else {
 619                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 620                                                 $textlen += $img_end;
 621                                         }
 622                                 }
 623                                 $orig_body = substr($orig_body, $img_end);
 624
 625                                 if ($orig_body === false) {
 626                                         // in case the body ends on a closing image tag
 627                                         $orig_body = '';
 628                                 }
 629
 630                                 $img_start = strpos($orig_body, '[img');
 631                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 632                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 633                         }
 634
 635                         if (($textlen + strlen($orig_body)) > $maxlen) {
 636                                 if ($textlen < $maxlen) {
 637                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 638                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 639                                 }
 640                         } else {
 641                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 642                                 $new_body = $new_body . $orig_body;
 643                         }
 644
 645                         return $new_body;
 646                 } else {
 647                         return $body;
 648                 }
 649         }
 650
 651         /**
 652          * Processes [attachment] tags
 653          *
 654          * Note: Can produce a [bookmark] tag in the returned string
 655          *
 656          * @brief Processes [attachment] tags
 657          * @param string $return
 658          * @param bool|int $simplehtml
 659          * @param bool $tryoembed
 660          * @return string
 661          */
 662         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 663         {
 664                 $data = self::getAttachmentData($return);
 665                 if (!$data) {
 666                         return $return;
 667                 }
 668
 669                 if (isset($data["title"])) {
 670                         $data["title"] = strip_tags($data["title"]);
 671                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 672                 }
 673
 674                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
 675                         $data["preview"] = $data["image"];
 676                         $data["image"] = "";
 677                 }
 678
 679                 $return = '';
 680                 if ($simplehtml == 7) {
 681                         $return = self::convertUrlForMastodon($data["url"]);
 682                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 683                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 684                 } else {
 685                         try {
 686                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 687                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 688                                 } else {
 689                                         throw new Exception('OEmbed is disabled for this attachment.');
 690                                 }
 691                         } catch (Exception $e) {
 692                                 if ($simplehtml != 4) {
 693                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 694                                 }
 695
 696                                 if ($data["image"] != "") {
 697                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
 698                                 } elseif ($data["preview"] != "") {
 699                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
 700                                 }
 701
 702                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
 703                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], proxy_url($data["image"]), $data["title"]);
 704                                 } else {
 705                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 706                                 }
 707
 708                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 709                                         // Sanitize the HTML by converting it to BBCode
 710                                         $bbcode = html2bbcode($data["description"]);
 711                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 712                                 }
 713                                 if ($data["type"] == "link") {
 714                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 715                                 }
 716
 717                                 if ($simplehtml != 4) {
 718                                         $return .= '</div>';
 719                                 }
 720                         }
 721                 }
 722
 723                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 724         }
 725
 726         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 727         {
 728                 $data = self::getAttachmentData($Text);
 729
 730                 if (!$data) {
 731                         return $Text;
 732                 } elseif ($nolink) {
 733                         return $data["text"] . $data["after"];
 734                 }
 735
 736                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
 737                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 738                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 739                         $data["title"] = $data["url"];
 740                 } elseif (($text != "") && strstr($title, $text)) {
 741                         $data["text"] = $data["title"];
 742                         $data["title"] = $data["url"];
 743                 }
 744
 745                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 746                         return $data["title"] . $data["after"];
 747                 }
 748
 749                 // If the link already is included in the post, don't add it again
 750                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 751                         return $data["text"] . $data["after"];
 752                 }
 753
 754                 $text = $data["text"];
 755
 756                 if (($data["url"] != "") && ($data["title"] != "")) {
 757                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 758                 } elseif (($data["url"] != "")) {
 759                         $text .= "\n[url]" . $data["url"] . "[/url]";
 760                 }
 761
 762                 return $text . "\n" . $data["after"];
 763         }
 764
 765         private static function cleanCss($input)
 766         {
 767                 $cleaned = "";
 768
 769                 $input = strtolower($input);
 770
 771                 for ($i = 0; $i < strlen($input); $i++) {
 772                         $char = substr($input, $i, 1);
 773
 774                         if (($char >= "a") && ($char <= "z")) {
 775                                 $cleaned .= $char;
 776                         }
 777
 778                         if (!(strpos(" #;:0123456789-_.%", $char) === false)) {
 779                                 $cleaned .= $char;
 780                         }
 781                 }
 782
 783                 return $cleaned;
 784         }
 785
 786         /**
 787          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 788          *
 789          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 790          * @param array $match Array with the matching values
 791          * @return string reformatted link including HTML codes
 792          */
 793         private static function convertUrlForMastodonCallback($match)
 794         {
 795                 $url = $match[1];
 796
 797                 if (isset($match[2]) && ($match[1] != $match[2])) {
 798                         return $match[0];
 799                 }
 800
 801                 $parts = parse_url($url);
 802                 if (!isset($parts['scheme'])) {
 803                         return $match[0];
 804                 }
 805
 806                 return self::convertUrlForMastodon($url);
 807         }
 808
 809         /**
 810          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon and GNU Social.
 811          * @param string $url URL that is about to be reformatted
 812          * @return string reformatted link including HTML codes
 813          */
 814         private static function convertUrlForMastodon($url)
 815         {
 816                 $parts = parse_url($url);
 817                 $scheme = $parts['scheme'] . '://';
 818                 $styled_url = str_replace($scheme, '', $url);
 819
 820                 $html = '<a href="%s" class="attachment" rel="nofollow noopener" target="_blank">' .
 821                         '<span class="invisible">%s</span>';
 822
 823                 if (strlen($styled_url) > 30) {
 824                         $html .= '<span class="ellipsis">%s</span>' .
 825                                 '<span class="invisible">%s</span></a>';
 826
 827                         $ellipsis = substr($styled_url, 0, 30);
 828                         $rest = substr($styled_url, 30);
 829                         return sprintf($html, $url, $scheme, $ellipsis, $rest);
 830                 } else {
 831                         $html .= '%s</a>';
 832                         return sprintf($html, $url, $scheme, $styled_url);
 833                 }
 834         }
 835
 836         /**
 837          * @brief Shortens [url] BBCodes in a format that looks less ugly than the full address. (callback function)
 838          * @param array $match Array with the matching values
 839          * @return string reformatted link including HTML codes
 840          */
 841         private static function shortenVisibleUrlCallback($match)
 842         {
 843                 $url = $match[1];
 844
 845                 if (isset($match[2]) && ($match[1] != $match[2])) {
 846                         return $match[0];
 847                 }
 848
 849                 $parts = parse_url($url);
 850                 if (!isset($parts['scheme'])) {
 851                         return $match[0];
 852                 }
 853
 854                 return self::shortenVisibleUrl($url);
 855         }
 856
 857         /**
 858          * @brief Shortens [url] BBCodes in a format that looks less ugly than the full address.
 859          * @param string $url URL that is about to be reformatted
 860          * @return string reformatted link including HTML codes
 861          */
 862         private static function shortenVisibleUrl($url)
 863         {
 864                 $parts = parse_url($url);
 865                 $scheme = $parts['scheme'] . '://';
 866                 $styled_url = str_replace($scheme, '', $url);
 867
 868 //              Currently deactivated, due to preview problems inside of Diaspora
 869 //              if (strlen($styled_url) > 30) {
 870 //                      $styled_url = substr($styled_url, 0, 30) . "…";
 871 //              }
 872
 873                 $html = '<a href="%s" target="_blank">%s</a>';
 874
 875                 return sprintf($html, $url, $styled_url);
 876         }
 877
 878         /*
 879          * [noparse][i]italic[/i][/noparse] turns into
 880          * [noparse][ i ]italic[ /i ][/noparse],
 881          * to hide them from parser.
 882          */
 883         private static function escapeNoparseCallback($match)
 884         {
 885                 $whole_match = $match[0];
 886                 $captured = $match[1];
 887                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 888                 $new_str = str_replace($captured, $spacefied, $whole_match);
 889                 return $new_str;
 890         }
 891
 892         /*
 893          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 894          * now turns back and the [noparse] tags are trimed
 895          * returning [i]italic[/i]
 896          */
 897         private static function unescapeNoparseCallback($match)
 898         {
 899                 $captured = $match[1];
 900                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 901                 return $unspacefied;
 902         }
 903
 904         /**
 905          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 906          * occurrences
 907          *
 908          * @param string $text        Text to search
 909          * @param string $name        Tag name
 910          * @param int    $occurrences Number of first occurrences to skip
 911          * @return boolean|array
 912          */
 913         public static function getTagPosition($text, $name, $occurrences = 0)
 914         {
 915                 if ($occurrences < 0) {
 916                         $occurrences = 0;
 917                 }
 918
 919                 $start_open = -1;
 920                 for ($i = 0; $i <= $occurrences; $i++) {
 921                         if ($start_open !== false) {
 922                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 923                         }
 924                 }
 925
 926                 if ($start_open === false) {
 927                         return false;
 928                 }
 929
 930                 $start_equal = strpos($text, '=', $start_open);
 931                 $start_close = strpos($text, ']', $start_open);
 932
 933                 if ($start_close === false) {
 934                         return false;
 935                 }
 936
 937                 $start_close++;
 938
 939                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 940
 941                 if ($end_open === false) {
 942                         return false;
 943                 }
 944
 945                 $res = [
 946                         'start' => [
 947                                 'open' => $start_open,
 948                                 'close' => $start_close
 949                         ],
 950                         'end' => [
 951                                 'open' => $end_open,
 952                                 'close' => $end_open + strlen('[/' . $name . ']')
 953                         ],
 954                 ];
 955
 956                 if ($start_equal !== false) {
 957                         $res['start']['equal'] = $start_equal + 1;
 958                 }
 959
 960                 return $res;
 961         }
 962
 963         /**
 964          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 965          *
 966          * @param type $pattern Preg pattern string
 967          * @param type $replace Preg replace string
 968          * @param type $name    BBCode tag name
 969          * @param type $text    Text to search
 970          * @return string
 971          */
 972         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 973         {
 974                 $occurrences = 0;
 975                 $pos = self::getTagPosition($text, $name, $occurrences);
 976                 while ($pos !== false && $occurrences++ < 1000) {
 977                         $start = substr($text, 0, $pos['start']['open']);
 978                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 979                         $end = substr($text, $pos['end']['close']);
 980                         if ($end === false) {
 981                                 $end = '';
 982                         }
 983
 984                         $subject = preg_replace($pattern, $replace, $subject);
 985                         $text = $start . $subject . $end;
 986
 987                         $pos = self::getTagPosition($text, $name, $occurrences);
 988                 }
 989
 990                 return $text;
 991         }
 992
 993         private static function extractImagesFromItemBody($body)
 994         {
 995                 $saved_image = [];
 996                 $orig_body = $body;
 997                 $new_body = '';
 998
 999                 $cnt = 0;
1000                 $img_start = strpos($orig_body, '[img');
1001                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
1002                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
1003                 while (($img_st_close !== false) && ($img_end !== false)) {
1004                         $img_st_close++; // make it point to AFTER the closing bracket
1005                         $img_end += $img_start;
1006
1007                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
1008                                 // This is an embedded image
1009                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
1010                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
1011
1012                                 $cnt++;
1013                         } else {
1014                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
1015                         }
1016
1017                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
1018
1019                         if ($orig_body === false) {
1020                                 // in case the body ends on a closing image tag
1021                                 $orig_body = '';
1022                         }
1023
1024                         $img_start = strpos($orig_body, '[img');
1025                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
1026                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
1027                 }
1028
1029                 $new_body = $new_body . $orig_body;
1030
1031                 return ['body' => $new_body, 'images' => $saved_image];
1032         }
1033
1034         private static function interpolateSavedImagesIntoItemBody($body, array $images)
1035         {
1036                 $newbody = $body;
1037
1038                 $cnt = 0;
1039                 foreach ($images as $image) {
1040                         // We're depending on the property of 'foreach' (specified on the PHP website) that
1041                         // it loops over the array starting from the first element and going sequentially
1042                         // to the last element
1043                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1044                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
1045                         $cnt++;
1046                 }
1047
1048                 return $newbody;
1049         }
1050
1051         /**
1052          * Processes [share] tags
1053          *
1054          * Note: Can produce a [bookmark] tag in the output
1055          *
1056          * @brief Processes [share] tags
1057          * @param array    $share      preg_match_callback result array
1058          * @param bool|int $simplehtml
1059          * @return string
1060          */
1061         private static function convertShare($share, $simplehtml)
1062         {
1063                 $attributes = $share[2];
1064
1065                 $author = "";
1066                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1067                 if (x($matches, 1)) {
1068                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1069                 }
1070
1071                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1072                 if (x($matches, 1)) {
1073                         $author = $matches[1];
1074                 }
1075
1076                 $profile = "";
1077                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1078                 if (x($matches, 1)) {
1079                         $profile = $matches[1];
1080                 }
1081
1082                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1083                 if (x($matches, 1)) {
1084                         $profile = $matches[1];
1085                 }
1086
1087                 $avatar = "";
1088                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1089                 if (x($matches, 1)) {
1090                         $avatar = $matches[1];
1091                 }
1092
1093                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1094                 if (x($matches, 1)) {
1095                         $avatar = $matches[1];
1096                 }
1097
1098                 $link = "";
1099                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1100                 if (x($matches, 1)) {
1101                         $link = $matches[1];
1102                 }
1103
1104                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1105                 if (x($matches, 1)) {
1106                         $link = $matches[1];
1107                 }
1108
1109                 $posted = "";
1110
1111                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1112                 if (x($matches, 1)) {
1113                         $posted = $matches[1];
1114                 }
1115
1116                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1117                 if (x($matches, 1)) {
1118                         $posted = $matches[1];
1119                 }
1120
1121                 // We only call this so that a previously unknown contact can be added.
1122                 // This is important for the function "Model\Contact::getDetailsByURL()".
1123                 // This function then can fetch an entry from the contact table.
1124                 Contact::getIdForURL($profile, 0, true);
1125
1126                 $data = Contact::getDetailsByURL($profile);
1127
1128                 if (x($data, "name") && x($data, "addr")) {
1129                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1130                 } else {
1131                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1132                 }
1133
1134                 if (x($data, "addr")) {
1135                         $userid = $data["addr"];
1136                 } else {
1137                         $userid = Protocol::formatMention($profile, $author);
1138                 }
1139
1140                 if (x($data, "name")) {
1141                         $author = $data["name"];
1142                 }
1143
1144                 if (x($data, "micro")) {
1145                         $avatar = $data["micro"];
1146                 }
1147
1148                 $preshare = trim($share[1]);
1149                 if ($preshare != "") {
1150                         $preshare .= "<br />";
1151                 }
1152
1153                 switch ($simplehtml) {
1154                         case 1:
1155                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1156                                 break;
1157                         case 2:
1158                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1159                                 break;
1160                         case 3: // Diaspora
1161                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1162
1163                                 $text = trim($share[1]);
1164
1165                                 if ($text != "") {
1166                                         $text .= "<hr />";
1167                                 }
1168
1169                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1170                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1171                                 } else {
1172                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1173
1174                                         if ($link != "") {
1175                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1176                                         }
1177                                 }
1178
1179                                 break;
1180                         case 4:
1181                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1182                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1183                                 $headline .= ":</b><br />";
1184
1185                                 $text = trim($share[1]);
1186
1187                                 if ($text != "") {
1188                                         $text .= "<hr />";
1189                                 }
1190
1191                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1192
1193                                 break;
1194                         case 5:
1195                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1196                                 break;
1197                         case 6: // app.net
1198                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1199                                 break;
1200                         case 7: // statusnet/GNU Social
1201                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1202                                 break;
1203                         case 8: // twitter
1204                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1205                                 break;
1206                         case 9: // Google+/Facebook
1207                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1208
1209                                 if ($link != "") {
1210                                         $text .= "<br /><br />" . $link;
1211                                 }
1212                                 break;
1213                         default:
1214                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1215                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1216                                         try {
1217                                                 $oembed = OEmbed::getHTML($link, $preshare);
1218                                         } catch (Exception $e) {
1219                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1220                                         }
1221
1222                                         $text = $preshare . $oembed;
1223                                 } else {
1224                                         $text = trim($share[1]) . "\n";
1225
1226                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1227
1228                                         $tpl = get_markup_template('shared_content.tpl');
1229                                         $text .= replace_macros($tpl, [
1230                                                 '$profile' => $profile,
1231                                                 '$avatar' => $avatar,
1232                                                 '$author' => $author,
1233                                                 '$link' => $link,
1234                                                 '$posted' => $posted,
1235                                                 '$content' => trim($share[3])
1236                                         ]);
1237                                 }
1238                                 break;
1239                 }
1240
1241                 return $text;
1242         }
1243
1244         private static function removePictureLinksCallback($match)
1245         {
1246                 $text = Cache::get($match[1]);
1247
1248                 if (is_null($text)) {
1249                         $a = get_app();
1250
1251                         $stamp1 = microtime(true);
1252
1253                         $ch = @curl_init($match[1]);
1254                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1255                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1256                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1257                         @curl_exec($ch);
1258                         $curl_info = @curl_getinfo($ch);
1259
1260                         $a->save_timestamp($stamp1, "network");
1261
1262                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1263                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1264                         } else {
1265                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1266
1267                                 // if its not a picture then look if its a page that contains a picture link
1268                                 $body = Network::fetchUrl($match[1]);
1269
1270                                 $doc = new DOMDocument();
1271                                 @$doc->loadHTML($body);
1272                                 $xpath = new DomXPath($doc);
1273                                 $list = $xpath->query("//meta[@name]");
1274                                 foreach ($list as $node) {
1275                                         $attr = [];
1276
1277                                         if ($node->attributes->length) {
1278                                                 foreach ($node->attributes as $attribute) {
1279                                                         $attr[$attribute->name] = $attribute->value;
1280                                                 }
1281                                         }
1282
1283                                         if (strtolower($attr["name"]) == "twitter:image") {
1284                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1285                                         }
1286                                 }
1287                         }
1288                         Cache::set($match[1], $text);
1289                 }
1290
1291                 return $text;
1292         }
1293
1294         private static function expandLinksCallback($match)
1295         {
1296                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1297                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1298                 } else {
1299                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1300                 }
1301         }
1302
1303         private static function cleanPictureLinksCallback($match)
1304         {
1305                 $text = Cache::get($match[1]);
1306
1307                 if (is_null($text)) {
1308                         $a = get_app();
1309
1310                         $stamp1 = microtime(true);
1311
1312                         $ch = @curl_init($match[1]);
1313                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1314                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1315                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1316                         @curl_exec($ch);
1317                         $curl_info = @curl_getinfo($ch);
1318
1319                         $a->save_timestamp($stamp1, "network");
1320
1321                         // if its a link to a picture then embed this picture
1322                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1323                                 $text = "[img]" . $match[1] . "[/img]";
1324                         } else {
1325                                 $text = "[img]" . $match[2] . "[/img]";
1326
1327                                 // if its not a picture then look if its a page that contains a picture link
1328                                 $body = Network::fetchUrl($match[1]);
1329
1330                                 $doc = new DOMDocument();
1331                                 @$doc->loadHTML($body);
1332                                 $xpath = new DomXPath($doc);
1333                                 $list = $xpath->query("//meta[@name]");
1334                                 foreach ($list as $node) {
1335                                         $attr = [];
1336                                         if ($node->attributes->length) {
1337                                                 foreach ($node->attributes as $attribute) {
1338                                                         $attr[$attribute->name] = $attribute->value;
1339                                                 }
1340                                         }
1341
1342                                         if (strtolower($attr["name"]) == "twitter:image") {
1343                                                 $text = "[img]" . $attr["content"] . "[/img]";
1344                                         }
1345                                 }
1346                         }
1347                         Cache::set($match[1], $text);
1348                 }
1349
1350                 return $text;
1351         }
1352
1353         public static function cleanPictureLinks($text)
1354         {
1355                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1356                 return $return;
1357         }
1358
1359         private static function textHighlightCallback($match)
1360         {
1361                 if (in_array(strtolower($match[1]),
1362                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1363                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh'])
1364                 ) {
1365                         return text_highlight($match[2], strtolower($match[1]));
1366                 }
1367                 return $match[0];
1368         }
1369
1370         /**
1371          * @brief Converts a BBCode message to HTML message
1372          *
1373          * BBcode 2 HTML was written by WAY2WEB.net
1374          * extended to work with Mistpark/Friendica - Mike Macgirvin
1375          *
1376          * Simple HTML values meaning:
1377          * - 0: Friendica display
1378          * - 1: Unused
1379          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1380          * - 3: Used before converting to Markdown in bb2diaspora.php
1381          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1382          * - 5: Unused
1383          * - 6: Used for Appnet
1384          * - 7: Used for dfrn, OStatus
1385          * - 8: Used for WP backlink text setting
1386          *
1387          * @param string $text
1388          * @param bool   $try_oembed
1389          * @param int    $simple_html
1390          * @param bool   $for_plaintext
1391          * @return string
1392          */
1393         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1394         {
1395                 $a = get_app();
1396
1397                 /*
1398                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1399                  *
1400                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1401                  * $match[1] = $url
1402                  * $match[2] = $title or absent
1403                  */
1404                 $try_oembed_callback = function ($match)
1405                 {
1406                         $url = $match[1];
1407                         $title = defaults($match, 2, null);
1408
1409                         try {
1410                                 $return = OEmbed::getHTML($url, $title);
1411                         } catch (Exception $ex) {
1412                                 $return = $match[0];
1413                         }
1414
1415                         return $return;
1416                 };
1417
1418                 // Hide all [noparse] contained bbtags by spacefying them
1419                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1420
1421                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1422                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1423                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1424
1425                 // Remove the abstract element. It is a non visible element.
1426                 $text = self::stripAbstract($text);
1427
1428                 // Move all spaces out of the tags
1429                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1430                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1431
1432                 // Extract the private images which use data urls since preg has issues with
1433                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1434                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1435
1436                 $extracted = self::extractImagesFromItemBody($text);
1437                 $text = $extracted['body'];
1438                 $saved_image = $extracted['images'];
1439
1440                 // If we find any event code, turn it into an event.
1441                 // After we're finished processing the bbcode we'll
1442                 // replace all of the event code with a reformatted version.
1443
1444                 $ev = bbtoevent($text);
1445
1446                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1447                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1448
1449                 $text = str_replace("<", "&lt;", $text);
1450                 $text = str_replace(">", "&gt;", $text);
1451
1452                 // remove some newlines before the general conversion
1453                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1454                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1455
1456                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1457                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1458
1459                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1460                 if (!$try_oembed) {
1461                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1462                 }
1463
1464                 // Check for [code] text here, before the linefeeds are messed with.
1465                 // The highlighter will unescape and re-escape the content.
1466                 if (strpos($text, '[code=') !== false) {
1467                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1468                 }
1469                 // Convert new line chars to html <br /> tags
1470
1471                 // nlbr seems to be hopelessly messed up
1472                 //      $Text = nl2br($Text);
1473
1474                 // We'll emulate it.
1475
1476                 $text = trim($text);
1477                 $text = str_replace("\r\n", "\n", $text);
1478
1479                 // removing multiplicated newlines
1480                 if (Config::get("system", "remove_multiplicated_lines")) {
1481                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1482                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1483                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1484                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1485                         do {
1486                                 $oldtext = $text;
1487                                 $text = str_replace($search, $replace, $text);
1488                         } while ($oldtext != $text);
1489                 }
1490
1491                 // Set up the parameters for a URL search string
1492                 $URLSearchString = "^\[\]";
1493                 // Set up the parameters for a MAIL search string
1494                 $MAILSearchString = $URLSearchString;
1495
1496                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1497                 if (!$for_plaintext) {
1498                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1499                         // Currently disabled, since the function is too greedy
1500                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1501                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1502                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1503                         if ($simple_html == 7) {
1504                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1505                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1506                         } else {
1507                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::shortenVisibleUrlCallback', $text);
1508                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::shortenVisibleUrlCallback', $text);
1509                         }
1510                 } else {
1511                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1512                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1513                 }
1514
1515
1516                 // Handle attached links or videos
1517                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1518
1519                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1520
1521                 // Remove all hashtag addresses
1522                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1523                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1524                 } elseif ($simple_html == 3) {
1525                         // The ! is converted to @ since Diaspora only understands the @
1526                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1527                                 '@<a href="$2">$3</a>',
1528                                 $text);
1529                 } elseif ($simple_html == 7) {
1530                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1531                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1532                                 $text);
1533                 } elseif (!$simple_html) {
1534                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1535                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1536                                 $text);
1537                 }
1538
1539                 // Bookmarks in red - will be converted to bookmarks in friendica
1540                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1541                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1542                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1543                                         "[bookmark=$1]$2[/bookmark]", $text);
1544
1545                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1546                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1547                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1548                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1549                 }
1550
1551                 if ($simple_html == 5) {
1552                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1553                 }
1554
1555                 // Perform URL Search
1556                 if ($try_oembed) {
1557                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1558                 }
1559
1560                 if ($simple_html == 5) {
1561                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1562                 } else {
1563                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1564                 }
1565
1566                 // Handle Diaspora posts
1567                 $text = preg_replace_callback(
1568                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1569                         function ($match) {
1570                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1571                         }, $text
1572                 );
1573
1574                 // Server independent link to posts and comments
1575                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1576                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1577                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1578
1579                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1580                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1581
1582                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1583                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1584
1585                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1586                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1587                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1588
1589                 // Red compatibility, though the link can't be authenticated on Friendica
1590                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1591
1592
1593                 // we may need to restrict this further if it picks up too many strays
1594                 // link acct:user@host to a webfinger profile redirector
1595
1596                 $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);
1597
1598                 // Perform MAIL Search
1599                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1600                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1601
1602                 // leave open the posibility of [map=something]
1603                 // this is replaced in prepare_body() which has knowledge of the item location
1604
1605                 if (strpos($text, '[/map]') !== false) {
1606                         $text = preg_replace_callback(
1607                                 "/\[map\](.*?)\[\/map\]/ism",
1608                                 function ($match) use ($simple_html) {
1609                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1610                                 },
1611                                 $text
1612                         );
1613                 }
1614                 if (strpos($text, '[map=') !== false) {
1615                         $text = preg_replace_callback(
1616                                 "/\[map=(.*?)\]/ism",
1617                                 function ($match) use ($simple_html) {
1618                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1619                                 },
1620                                 $text
1621                         );
1622                 }
1623                 if (strpos($text, '[map]') !== false) {
1624                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1625                 }
1626
1627                 // Check for headers
1628                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1629                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1630                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1631                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1632                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1633                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1634
1635                 // Check for paragraph
1636                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1637
1638                 // Check for bold text
1639                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1640
1641                 // Check for Italics text
1642                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1643
1644                 // Check for Underline text
1645                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1646
1647                 // Check for strike-through text
1648                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1649
1650                 // Check for over-line text
1651                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1652
1653                 // Check for colored text
1654                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1655
1656                 // Check for sized text
1657                 // [size=50] --> font-size: 50px (with the unit).
1658                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1659                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1660
1661                 // Check for centered text
1662                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1663
1664                 // Check for list text
1665                 $text = str_replace("[*]", "<li>", $text);
1666
1667                 // Check for style sheet commands
1668                 $text = preg_replace_callback(
1669                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1670                         function ($match) {
1671                                 return "<span style=\"" . self::cleanCss($match[1]) . ";\">" . $match[2] . "</span>";
1672                         },
1673                         $text
1674                 );
1675
1676                 // Check for CSS classes
1677                 $text = preg_replace_callback(
1678                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1679                         function ($match) {
1680                                 return "<span class=\"" . self::cleanCss($match[1]) . "\">" . $match[2] . "</span>";
1681                         },
1682                         $text
1683                 );
1684
1685                 // handle nested lists
1686                 $endlessloop = 0;
1687
1688                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1689                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1690                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1691                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1692                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1693                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1694                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1695                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1696                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1697                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1698                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1699                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1700                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1701                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1702                 }
1703
1704                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1705                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1706                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1707                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1708
1709                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1710                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1711
1712                 $text = str_replace('[hr]', '<hr />', $text);
1713
1714                 // This is actually executed in prepare_body()
1715
1716                 $text = str_replace('[nosmile]', '', $text);
1717
1718                 // Check for font change text
1719                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1720
1721                 // Declare the format for [code] layout
1722
1723                 $CodeLayout = '<code>$1</code>';
1724                 // Check for [code] text
1725                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1726
1727                 // Declare the format for [spoiler] layout
1728                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1729
1730                 // Check for [spoiler] text
1731                 // handle nested quotes
1732                 $endlessloop = 0;
1733                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1734                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1735                 }
1736
1737                 // Check for [spoiler=Author] text
1738
1739                 $t_wrote = L10n::t('$1 wrote:');
1740
1741                 // handle nested quotes
1742                 $endlessloop = 0;
1743                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1744                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1745                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1746                                                  $text);
1747                 }
1748
1749                 // Declare the format for [quote] layout
1750                 $QuoteLayout = '<blockquote>$1</blockquote>';
1751
1752                 // Check for [quote] text
1753                 // handle nested quotes
1754                 $endlessloop = 0;
1755                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1756                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1757                 }
1758
1759                 // Check for [quote=Author] text
1760
1761                 $t_wrote = L10n::t('$1 wrote:');
1762
1763                 // handle nested quotes
1764                 $endlessloop = 0;
1765                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1766                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1767                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1768                                                  $text);
1769                 }
1770
1771
1772                 // [img=widthxheight]image source[/img]
1773                 $text = preg_replace_callback(
1774                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1775                         function ($matches) {
1776                                 if (strpos($matches[3], "data:image/") === 0) {
1777                                         return $matches[0];
1778                                 }
1779
1780                                 $matches[3] = proxy_url($matches[3]);
1781                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1782                         },
1783                         $text
1784                 );
1785
1786                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1787                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1788
1789                 // Images
1790                 // [img]pathtoimage[/img]
1791                 $text = preg_replace_callback(
1792                         "/\[img\](.*?)\[\/img\]/ism",
1793                         function ($matches) {
1794                                 if (strpos($matches[1], "data:image/") === 0) {
1795                                         return $matches[0];
1796                                 }
1797
1798                                 $matches[1] = proxy_url($matches[1]);
1799                                 return "[img]" . $matches[1] . "[/img]";
1800                         },
1801                         $text
1802                 );
1803
1804                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1805                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1806
1807                 // Shared content
1808                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1809                         function ($match) use ($simple_html) {
1810                                 return self::convertShare($match, $simple_html);
1811                         }, $text);
1812
1813                 $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);
1814                 $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 />', $text);
1815                 //$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 />', $Text);
1816
1817                 // Try to Oembed
1818                 if ($try_oembed) {
1819                         $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);
1820                         $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);
1821
1822                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1823                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1824                 } else {
1825                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1826                                                 '<a href="$1" target="_blank">$1</a>', $text);
1827                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1828                                                 '<a href="$1" target="_blank">$1</a>', $text);
1829                 }
1830
1831                 // html5 video and audio
1832
1833
1834                 if ($try_oembed) {
1835                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1836                 } else {
1837                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1838                 }
1839
1840                 // Youtube extensions
1841                 if ($try_oembed) {
1842                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1843                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1844                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1845                 }
1846
1847                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1848                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1849                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1850
1851                 if ($try_oembed) {
1852                         $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);
1853                 } else {
1854                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1855                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1856                 }
1857
1858                 if ($try_oembed) {
1859                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1860                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1861                 }
1862
1863                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1864                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1865
1866                 if ($try_oembed) {
1867                         $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);
1868                 } else {
1869                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1870                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1871                 }
1872
1873                 // oembed tag
1874                 $text = OEmbed::BBCode2HTML($text);
1875
1876                 // Avoid triple linefeeds through oembed
1877                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1878
1879                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1880                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1881                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1882                 // start which is always required). Allow desc with a missing summary for compatibility.
1883
1884                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1885                         $sub = format_event_html($ev, $simple_html);
1886
1887                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1888                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1889                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1890                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1891                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1892                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1893                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1894                 }
1895
1896                 // Replace non graphical smilies for external posts
1897                 if ($simple_html) {
1898                         $text = Smilies::replace($text, false, true);
1899                 }
1900
1901                 // Replace inline code blocks
1902                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1903                         function ($match) use ($simple_html) {
1904                                 $return = '<key>' . $match[1] . '</key>';
1905                                 // Use <code> for Diaspora inline code blocks
1906                                 if ($simple_html === 3) {
1907                                         $return = '<code>' . $match[1] . '</code>';
1908                                 }
1909                                 return $return;
1910                         }
1911                 , $text);
1912
1913                 // Unhide all [noparse] contained bbtags unspacefying them
1914                 // and triming the [noparse] tag.
1915
1916                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1917                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1918                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1919
1920
1921                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1922                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1923                 $text = preg_replace('/\&quot\;/', '"', $text);
1924
1925                 // fix any escaped ampersands that may have been converted into links
1926                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1927
1928                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1929                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1930                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1931                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1932
1933                 // sanitize href attributes (only whitelisted protocols URLs)
1934                 // default value for backward compatibility
1935                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1936
1937                 // Always allowed protocol even if config isn't set or not including it
1938                 $allowed_link_protocols[] = 'http';
1939                 $allowed_link_protocols[] = 'redir/';
1940
1941                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1942                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1943
1944                 if ($saved_image) {
1945                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1946                 }
1947
1948                 // Clean up the HTML by loading and saving the HTML with the DOM.
1949                 // Bad structured html can break a whole page.
1950                 // For performance reasons do it only with ativated item cache or at export.
1951                 if (!$try_oembed || (get_itemcachepath() != "")) {
1952                         $doc = new DOMDocument();
1953                         $doc->preserveWhiteSpace = false;
1954
1955                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1956
1957                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1958                         $encoding = '<?xml encoding="UTF-8">';
1959                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1960                         $doc->encoding = 'UTF-8';
1961                         $text = $doc->saveHTML();
1962                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1963
1964                         $text = str_replace('<br></li>', '</li>', $text);
1965
1966                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1967                 }
1968
1969                 // Clean up some useless linebreaks in lists
1970                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1971                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1972                 //$Text = str_replace('</li><br />', '</li>', $Text);
1973                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1974                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1975
1976                 Addon::callHooks('bbcode', $text);
1977
1978                 return trim($text);
1979         }
1980
1981         /**
1982          * @brief Strips the "abstract" tag from the provided text
1983          *
1984          * @param string $text The text with BBCode
1985          * @return string The same text - but without "abstract" element
1986          */
1987         public static function stripAbstract($text)
1988         {
1989                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1990                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1991
1992                 return $text;
1993         }
1994
1995         /**
1996          * @brief Returns the value of the "abstract" element
1997          *
1998          * @param string $text The text that maybe contains the element
1999          * @param string $addon The addon for which the abstract is meant for
2000          * @return string The abstract
2001          */
2002         private static function getAbstract($text, $addon = "")
2003         {
2004                 $abstract = "";
2005                 $abstracts = [];
2006                 $addon = strtolower($addon);
2007
2008                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
2009                         foreach ($results AS $result) {
2010                                 $abstracts[strtolower($result[1])] = $result[2];
2011                         }
2012                 }
2013
2014                 if (isset($abstracts[$addon])) {
2015                         $abstract = $abstracts[$addon];
2016                 }
2017
2018                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
2019                         $abstract = $result[1];
2020                 }
2021
2022                 return $abstract;
2023         }
2024 }