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 "mod/proxy.php";
  29
  30 class BBCode
  31 {
  32         /**
  33          * @brief Fetches attachment data that were generated the old way
  34          *
  35          * @param string $body Message body
  36          * @return array
  37          * 'type' -> Message type ("link", "video", "photo")
  38          * 'text' -> Text before the shared message
  39          * 'after' -> Text after the shared message
  40          * 'image' -> Preview image of the message
  41          * 'url' -> Url to the attached message
  42          * 'title' -> Title of the attachment
  43          * 'description' -> Description of the attachment
  44          */
  45         private static function getOldAttachmentData($body)
  46         {
  47                 $post = [];
  48
  49                 // Simplify image codes
  50                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  51
  52                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  53                         foreach ($attached as $data) {
  54                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  55                                         continue;
  56                                 }
  57
  58                                 $post["type"] = substr($data[1], 5);
  59
  60                                 $pos = strpos($body, $data[0]);
  61                                 if ($pos > 0) {
  62                                         $post["text"] = trim(substr($body, 0, $pos));
  63                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  64                                 } else {
  65                                         $post["text"] = trim(str_replace($data[0], "", $body));
  66                                 }
  67
  68                                 $attacheddata = $data[2];
  69
  70                                 $URLSearchString = "^\[\]";
  71
  72                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  73
  74                                         $picturedata = Image::getInfoFromURL($matches[1]);
  75
  76                                         if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  77                                                 $post["image"] = $matches[1];
  78                                         } else {
  79                                                 $post["preview"] = $matches[1];
  80                                         }
  81                                 }
  82
  83                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  84                                         $post["url"] = $matches[1];
  85                                         $post["title"] = $matches[2];
  86                                 }
  87                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  88                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  89                                         $post["url"] = $matches[1];
  90                                 }
  91
  92                                 // Search for description
  93                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  94                                         $post["description"] = $matches[1];
  95                                 }
  96                         }
  97                 }
  98                 return $post;
  99         }
 100
 101         /**
 102          * @brief Fetches attachment data that were generated with the "attachment" element
 103          *
 104          * @param string $body Message body
 105          * @return array
 106          * 'type' -> Message type ("link", "video", "photo")
 107          * 'text' -> Text before the shared message
 108          * 'after' -> Text after the shared message
 109          * 'image' -> Preview image of the message
 110          * 'url' -> Url to the attached message
 111          * 'title' -> Title of the attachment
 112          * 'description' -> Description of the attachment
 113          */
 114         public static function getAttachmentData($body)
 115         {
 116                 $data = [];
 117
 118                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 119                         return self::getOldAttachmentData($body);
 120                 }
 121
 122                 $attributes = $match[2];
 123
 124                 $data["text"] = trim($match[1]);
 125
 126                 $type = "";
 127                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 128                 if (x($matches, 1)) {
 129                         $type = strtolower($matches[1]);
 130                 }
 131
 132                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 133                 if (x($matches, 1)) {
 134                         $type = strtolower($matches[1]);
 135                 }
 136
 137                 if ($type == "") {
 138                         return [];
 139                 }
 140
 141                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 142                         return [];
 143                 }
 144
 145                 if ($type != "") {
 146                         $data["type"] = $type;
 147                 }
 148
 149                 $url = "";
 150                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 151                 if (x($matches, 1)) {
 152                         $url = $matches[1];
 153                 }
 154
 155                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 156                 if (x($matches, 1)) {
 157                         $url = $matches[1];
 158                 }
 159
 160                 if ($url != "") {
 161                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 162                 }
 163
 164                 $title = "";
 165                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 166                 if (x($matches, 1)) {
 167                         $title = $matches[1];
 168                 }
 169
 170                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 171                 if (x($matches, 1)) {
 172                         $title = $matches[1];
 173                 }
 174
 175                 if ($title != "") {
 176                         $title = BBCode::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 177                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 178                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 179                         $data["title"] = $title;
 180                 }
 181
 182                 $image = "";
 183                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 184                 if (x($matches, 1)) {
 185                         $image = $matches[1];
 186                 }
 187
 188                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 189                 if (x($matches, 1)) {
 190                         $image = $matches[1];
 191                 }
 192
 193                 if ($image != "") {
 194                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 195                 }
 196
 197                 $preview = "";
 198                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 199                 if (x($matches, 1)) {
 200                         $preview = $matches[1];
 201                 }
 202
 203                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 204                 if (x($matches, 1)) {
 205                         $preview = $matches[1];
 206                 }
 207
 208                 if ($preview != "") {
 209                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 210                 }
 211
 212                 $data["description"] = trim($match[3]);
 213
 214                 $data["after"] = trim($match[4]);
 215
 216                 return $data;
 217         }
 218
 219         public static function getAttachedData($body, $item = [])
 220         {
 221                 /*
 222                 - text:
 223                 - type: link, video, photo
 224                 - title:
 225                 - url:
 226                 - image:
 227                 - description:
 228                 - (thumbnail)
 229                 */
 230
 231                 $has_title = !empty($item['title']);
 232                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 233                 $post = self::getAttachmentData($body);
 234
 235                 // if nothing is found, it maybe having an image.
 236                 if (!isset($post["type"])) {
 237                         // Simplify image codes
 238                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 239
 240                         $URLSearchString = "^\[\]";
 241                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 242                                 if ((count($pictures) == 1) && !$has_title) {
 243                                         // Checking, if the link goes to a picture
 244                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 245
 246                                         // Workaround:
 247                                         // Sometimes photo posts to the own album are not detected at the start.
 248                                         // So we seem to cannot use the cache for these cases. That's strange.
 249                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 250                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 251                                         }
 252
 253                                         if ($data["type"] == "photo") {
 254                                                 $post["type"] = "photo";
 255                                                 if (isset($data["images"][0])) {
 256                                                         $post["image"] = $data["images"][0]["src"];
 257                                                         $post["url"] = $data["url"];
 258                                                 } else {
 259                                                         $post["image"] = $data["url"];
 260                                                 }
 261
 262                                                 $post["preview"] = $pictures[0][2];
 263                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 264                                         } else {
 265                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 266                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 267                                                         $post["type"] = "photo";
 268                                                         $post["image"] = $pictures[0][1];
 269                                                         $post["preview"] = $pictures[0][2];
 270                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 271                                                 }
 272                                         }
 273                                 } elseif (count($pictures) > 0) {
 274                                         $post["type"] = "link";
 275                                         $post["url"] = $plink;
 276                                         $post["image"] = $pictures[0][2];
 277                                         $post["text"] = $body;
 278                                 }
 279                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 280                                 if ((count($pictures) == 1) && !$has_title) {
 281                                         $post["type"] = "photo";
 282                                         $post["image"] = $pictures[0][1];
 283                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 284                                 } elseif (count($pictures) > 0) {
 285                                         $post["type"] = "link";
 286                                         $post["url"] = $plink;
 287                                         $post["image"] = $pictures[0][1];
 288                                         $post["text"] = $body;
 289                                 }
 290                         }
 291
 292                         // Test for the external links
 293                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 294                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 295
 296                         $links = array_merge($links1, $links2);
 297
 298                         // If there is only a single one, then use it.
 299                         // This should cover link posts via API.
 300                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 301                                 $post["type"] = "link";
 302                                 $post["text"] = trim($body);
 303                                 $post["url"] = $links[0][1];
 304                         }
 305
 306                         // Now count the number of external media links
 307                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 308                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 309                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 310                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 311
 312                         // Add them to the other external links
 313                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 314
 315                         // Are there more than one?
 316                         if (count($links) > 1) {
 317                                 // The post will be the type "text", which means a blog post
 318                                 unset($post["type"]);
 319                                 $post["url"] = $plink;
 320                         }
 321
 322                         if (!isset($post["type"])) {
 323                                 $post["type"] = "text";
 324                                 $post["text"] = trim($body);
 325                         }
 326                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 327                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 328
 329                         if (isset($data["images"][0])) {
 330                                 $post["image"] = $data["images"][0]["src"];
 331                         }
 332                 }
 333
 334                 return $post;
 335         }
 336
 337         /**
 338          * @brief Convert a message into plaintext for connectors to other networks
 339          *
 340          * @param array $b The message array that is about to be posted
 341          * @param int $limit The maximum number of characters when posting to that network
 342          * @param bool $includedlinks Has an attached link to be included into the message?
 343          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 344          * @param string $target_network Name of the network where the post should go to.
 345          *
 346          * @return string The converted message
 347          */
 348         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 349         {
 350                 // Remove the hash tags
 351                 $URLSearchString = "^\[\]";
 352                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 353
 354                 // Add an URL element if the text contains a raw link
 355                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 356
 357                 // Remove the abstract
 358                 $body = self::stripAbstract($body);
 359
 360                 // At first look at data that is attached via "type-..." stuff
 361                 // This will hopefully replaced with a dedicated bbcode later
 362                 //$post = self::getAttachedData($b["body"]);
 363                 $post = self::getAttachedData($body, $b);
 364
 365                 if (($b["title"] != "") && ($post["text"] != "")) {
 366                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 367                 } elseif ($b["title"] != "") {
 368                         $post["text"] = trim($b["title"]);
 369                 }
 370
 371                 $abstract = "";
 372
 373                 // Fetch the abstract from the given target network
 374                 if ($target_network != "") {
 375                         $default_abstract = self::getAbstract($b["body"]);
 376                         $abstract = self::getAbstract($b["body"], $target_network);
 377
 378                         // If we post to a network with no limit we only fetch
 379                         // an abstract exactly for this network
 380                         if (($limit == 0) && ($abstract == $default_abstract)) {
 381                                 $abstract = "";
 382                         }
 383                 } else {// Try to guess the correct target network
 384                         switch ($htmlmode) {
 385                                 case 8:
 386                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
 387                                         break;
 388                                 case 7:
 389                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
 390                                         break;
 391                                 case 6:
 392                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
 393                                         break;
 394                                 default: // We don't know the exact target.
 395                                         // We fetch an abstract since there is a posting limit.
 396                                         if ($limit > 0) {
 397                                                 $abstract = self::getAbstract($b["body"]);
 398                                         }
 399                         }
 400                 }
 401
 402                 if ($abstract != "") {
 403                         $post["text"] = $abstract;
 404
 405                         if ($post["type"] == "text") {
 406                                 $post["type"] = "link";
 407                                 $post["url"] = $b["plink"];
 408                         }
 409                 }
 410
 411                 $html = BBCode::convert($post["text"].$post["after"], false, $htmlmode);
 412                 $msg = html2plain($html, 0, true);
 413                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 414
 415                 $link = "";
 416                 if ($includedlinks) {
 417                         if ($post["type"] == "link") {
 418                                 $link = $post["url"];
 419                         } elseif ($post["type"] == "text") {
 420                                 $link = $post["url"];
 421                         } elseif ($post["type"] == "video") {
 422                                 $link = $post["url"];
 423                         } elseif ($post["type"] == "photo") {
 424                                 $link = $post["image"];
 425                         }
 426
 427                         if (($msg == "") && isset($post["title"])) {
 428                                 $msg = trim($post["title"]);
 429                         }
 430
 431                         if (($msg == "") && isset($post["description"])) {
 432                                 $msg = trim($post["description"]);
 433                         }
 434
 435                         // If the link is already contained in the post, then it neeedn't to be added again
 436                         // But: if the link is beyond the limit, then it has to be added.
 437                         if (($link != "") && strstr($msg, $link)) {
 438                                 $pos = strpos($msg, $link);
 439
 440                                 // Will the text be shortened in the link?
 441                                 // Or is the link the last item in the post?
 442                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 443                                         $msg = trim(str_replace($link, "", $msg));
 444                                 } elseif (($limit == 0) || ($pos < $limit)) {
 445                                         // The limit has to be increased since it will be shortened - but not now
 446                                         // Only do it with Twitter (htmlmode = 8)
 447                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 448                                                 $limit = $limit - 23 + strlen($link);
 449                                         }
 450
 451                                         $link = "";
 452
 453                                         if ($post["type"] == "text") {
 454                                                 unset($post["url"]);
 455                                         }
 456                                 }
 457                         }
 458                 }
 459
 460                 if ($limit > 0) {
 461                         // Reduce multiple spaces
 462                         // When posted to a network with limited space, we try to gain space where possible
 463                         while (strpos($msg, "  ") !== false) {
 464                                 $msg = str_replace("  ", " ", $msg);
 465                         }
 466
 467                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 468                         if (iconv_strlen($link, "UTF-8") > 0) {
 469                                 $limit = $limit - 23;
 470                         }
 471
 472                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 473                                 if (($post["type"] == "text") && isset($post["url"])) {
 474                                         $post["url"] = $b["plink"];
 475                                 } elseif (!isset($post["url"])) {
 476                                         $limit = $limit - 23;
 477                                         $post["url"] = $b["plink"];
 478                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 479                                 //} elseif (strpos($b["body"], "[share") !== false) {
 480                                 //      $post["url"] = $b["plink"];
 481                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 482                                         $post["url"] = $b["plink"];
 483                                 }
 484                                 $msg = Plaintext::shorten($msg, $limit);
 485                         }
 486                 }
 487
 488                 $post["text"] = trim($msg);
 489
 490                 return($post);
 491         }
 492
 493         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 494         {
 495                 // Suppress "view full size"
 496                 if (intval(Config::get('system', 'no_view_full_size'))) {
 497                         $include_link = false;
 498                 }
 499
 500                 // Picture addresses can contain special characters
 501                 $s = htmlspecialchars_decode($srctext);
 502
 503                 $matches = null;
 504                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 505                 if ($c) {
 506                         foreach ($matches as $mtch) {
 507                                 logger('scale_external_image: ' . $mtch[1]);
 508
 509                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 510                                 if (stristr($mtch[1], $hostname)) {
 511                                         continue;
 512                                 }
 513
 514                                 // $scale_replace, if passed, is an array of two elements. The
 515                                 // first is the name of the full-size image. The second is the
 516                                 // name of a remote, scaled-down version of the full size image.
 517                                 // This allows Friendica to display the smaller remote image if
 518                                 // one exists, while still linking to the full-size image
 519                                 if ($scale_replace) {
 520                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 521                                 } else {
 522                                         $scaled = $mtch[1];
 523                                 }
 524                                 $i = Network::fetchUrl($scaled);
 525                                 if (!$i) {
 526                                         return $srctext;
 527                                 }
 528
 529                                 // guess mimetype from headers or filename
 530                                 $type = Image::guessType($mtch[1], true);
 531
 532                                 if ($i) {
 533                                         $Image = new Image($i, $type);
 534                                         if ($Image->isValid()) {
 535                                                 $orig_width = $Image->getWidth();
 536                                                 $orig_height = $Image->getHeight();
 537
 538                                                 if ($orig_width > 640 || $orig_height > 640) {
 539                                                         $Image->scaleDown(640);
 540                                                         $new_width = $Image->getWidth();
 541                                                         $new_height = $Image->getHeight();
 542                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 543                                                         $s = str_replace(
 544                                                                 $mtch[0],
 545                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 546                                                                 . "\n" . (($include_link)
 547                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 548                                                                         : ''),
 549                                                                 $s
 550                                                         );
 551                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 552                                                 }
 553                                         }
 554                                 }
 555                         }
 556                 }
 557
 558                 // replace the special char encoding
 559                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 560                 return $s;
 561         }
 562
 563         /**
 564          * The purpose of this function is to apply system message length limits to
 565          * imported messages without including any embedded photos in the length
 566          *
 567          * @brief Truncates imported message body string length to max_import_size
 568          * @param string $body
 569          * @return string
 570          */
 571         public static function limitBodySize($body)
 572         {
 573                 $maxlen = get_max_import_size();
 574
 575                 // If the length of the body, including the embedded images, is smaller
 576                 // than the maximum, then don't waste time looking for the images
 577                 if ($maxlen && (strlen($body) > $maxlen)) {
 578
 579                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 580
 581                         $orig_body = $body;
 582                         $new_body = '';
 583                         $textlen = 0;
 584
 585                         $img_start = strpos($orig_body, '[img');
 586                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 587                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 588                         while (($img_st_close !== false) && ($img_end !== false)) {
 589
 590                                 $img_st_close++; // make it point to AFTER the closing bracket
 591                                 $img_end += $img_start;
 592                                 $img_end += strlen('[/img]');
 593
 594                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 595                                         // This is an embedded image
 596
 597                                         if (($textlen + $img_start) > $maxlen) {
 598                                                 if ($textlen < $maxlen) {
 599                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 600                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 601                                                         $textlen = $maxlen;
 602                                                 }
 603                                         } else {
 604                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 605                                                 $textlen += $img_start;
 606                                         }
 607
 608                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 609                                 } else {
 610
 611                                         if (($textlen + $img_end) > $maxlen) {
 612                                                 if ($textlen < $maxlen) {
 613                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 614                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 615                                                         $textlen = $maxlen;
 616                                                 }
 617                                         } else {
 618                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 619                                                 $textlen += $img_end;
 620                                         }
 621                                 }
 622                                 $orig_body = substr($orig_body, $img_end);
 623
 624                                 if ($orig_body === false) {
 625                                         // in case the body ends on a closing image tag
 626                                         $orig_body = '';
 627                                 }
 628
 629                                 $img_start = strpos($orig_body, '[img');
 630                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 631                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 632                         }
 633
 634                         if (($textlen + strlen($orig_body)) > $maxlen) {
 635                                 if ($textlen < $maxlen) {
 636                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 637                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 638                                 }
 639                         } else {
 640                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 641                                 $new_body = $new_body . $orig_body;
 642                         }
 643
 644                         return $new_body;
 645                 } else {
 646                         return $body;
 647                 }
 648         }
 649
 650         /**
 651          * Processes [attachment] tags
 652          *
 653          * Note: Can produce a [bookmark] tag in the returned string
 654          *
 655          * @brief Processes [attachment] tags
 656          * @param string $return
 657          * @param bool|int $simplehtml
 658          * @param bool $tryoembed
 659          * @return string
 660          */
 661         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 662         {
 663                 $data = self::getAttachmentData($return);
 664                 if (!$data) {
 665                         return $return;
 666                 }
 667
 668                 if (isset($data["title"])) {
 669                         $data["title"] = strip_tags($data["title"]);
 670                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 671                 }
 672
 673                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
 674                         $data["preview"] = $data["image"];
 675                         $data["image"] = "";
 676                 }
 677
 678                 $return = '';
 679                 if ($simplehtml == 7) {
 680                         $return = self::convertUrlForMastodon($data["url"]);
 681                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 682                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 683                 } else {
 684                         try {
 685                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 686                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 687                                 } else {
 688                                         throw new Exception('OEmbed is disabled for this attachment.');
 689                                 }
 690                         } catch (Exception $e) {
 691                                 if ($simplehtml != 4) {
 692                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 693                                 }
 694
 695                                 if ($data["image"] != "") {
 696                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
 697                                 } elseif ($data["preview"] != "") {
 698                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
 699                                 }
 700
 701                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
 702                                         $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"]);
 703                                 } else {
 704                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 705                                 }
 706
 707                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 708                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(BBCode::convert($data["description"])));
 709                                 }
 710
 711                                 if ($data["type"] == "link") {
 712                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 713                                 }
 714
 715                                 if ($simplehtml != 4) {
 716                                         $return .= '</div>';
 717                                 }
 718                         }
 719                 }
 720
 721                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 722         }
 723
 724         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 725         {
 726                 $data = self::getAttachmentData($Text);
 727
 728                 if (!$data) {
 729                         return $Text;
 730                 } elseif ($nolink) {
 731                         return $data["text"] . $data["after"];
 732                 }
 733
 734                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
 735                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 736                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 737                         $data["title"] = $data["url"];
 738                 } elseif (($text != "") && strstr($title, $text)) {
 739                         $data["text"] = $data["title"];
 740                         $data["title"] = $data["url"];
 741                 }
 742
 743                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 744                         return $data["title"] . $data["after"];
 745                 }
 746
 747                 // If the link already is included in the post, don't add it again
 748                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 749                         return $data["text"] . $data["after"];
 750                 }
 751
 752                 $text = $data["text"];
 753
 754                 if (($data["url"] != "") && ($data["title"] != "")) {
 755                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 756                 } elseif (($data["url"] != "")) {
 757                         $text .= "\n" . $data["url"];
 758                 }
 759
 760                 return $text . "\n" . $data["after"];
 761         }
 762
 763         private static function cleanCss($input)
 764         {
 765                 $cleaned = "";
 766
 767                 $input = strtolower($input);
 768
 769                 for ($i = 0; $i < strlen($input); $i++) {
 770                         $char = substr($input, $i, 1);
 771
 772                         if (($char >= "a") && ($char <= "z")) {
 773                                 $cleaned .= $char;
 774                         }
 775
 776                         if (!(strpos(" #;:0123456789-_.%", $char) === false)) {
 777                                 $cleaned .= $char;
 778                         }
 779                 }
 780
 781                 return $cleaned;
 782         }
 783
 784         /**
 785          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 786          *
 787          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 788          * @param array $match Array with the matching values
 789          * @return string reformatted link including HTML codes
 790          */
 791         private static function convertUrlForMastodonCallback($match)
 792         {
 793                 $url = $match[1];
 794
 795                 if (isset($match[2]) && ($match[1] != $match[2])) {
 796                         return $match[0];
 797                 }
 798
 799                 $parts = parse_url($url);
 800                 if (!isset($parts['scheme'])) {
 801                         return $match[0];
 802                 }
 803
 804                 return self::convertUrlForMastodon($url);
 805         }
 806
 807         /**
 808          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon and GNU Social.
 809          * @param string $url URL that is about to be reformatted
 810          * @return string reformatted link including HTML codes
 811          */
 812         private static function convertUrlForMastodon($url)
 813         {
 814                 $parts = parse_url($url);
 815                 $scheme = $parts['scheme'] . '://';
 816                 $styled_url = str_replace($scheme, '', $url);
 817
 818                 $html = '<a href="%s" class="attachment" rel="nofollow noopener" target="_blank">' .
 819                         '<span class="invisible">%s</span>';
 820
 821                 if (strlen($styled_url) > 30) {
 822                         $html .= '<span class="ellipsis">%s</span>' .
 823                                 '<span class="invisible">%s</span></a>';
 824
 825                         $ellipsis = substr($styled_url, 0, 30);
 826                         $rest = substr($styled_url, 30);
 827                         return sprintf($html, $url, $scheme, $ellipsis, $rest);
 828                 } else {
 829                         $html .= '%s</a>';
 830                         return sprintf($html, $url, $scheme, $styled_url);
 831                 }
 832         }
 833
 834         /*
 835          * [noparse][i]italic[/i][/noparse] turns into
 836          * [noparse][ i ]italic[ /i ][/noparse],
 837          * to hide them from parser.
 838          */
 839         private static function escapeNoparseCallback($match)
 840         {
 841                 $whole_match = $match[0];
 842                 $captured = $match[1];
 843                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 844                 $new_str = str_replace($captured, $spacefied, $whole_match);
 845                 return $new_str;
 846         }
 847
 848         /*
 849          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 850          * now turns back and the [noparse] tags are trimed
 851          * returning [i]italic[/i]
 852          */
 853         private static function unescapeNoparseCallback($match)
 854         {
 855                 $captured = $match[1];
 856                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 857                 return $unspacefied;
 858         }
 859
 860         /**
 861          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 862          * occurrences
 863          *
 864          * @param string $text        Text to search
 865          * @param string $name        Tag name
 866          * @param int    $occurrences Number of first occurrences to skip
 867          * @return boolean|array
 868          */
 869         public static function getTagPosition($text, $name, $occurrences = 0)
 870         {
 871                 if ($occurrences < 0) {
 872                         $occurrences = 0;
 873                 }
 874
 875                 $start_open = -1;
 876                 for ($i = 0; $i <= $occurrences; $i++) {
 877                         if ($start_open !== false) {
 878                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 879                         }
 880                 }
 881
 882                 if ($start_open === false) {
 883                         return false;
 884                 }
 885
 886                 $start_equal = strpos($text, '=', $start_open);
 887                 $start_close = strpos($text, ']', $start_open);
 888
 889                 if ($start_close === false) {
 890                         return false;
 891                 }
 892
 893                 $start_close++;
 894
 895                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 896
 897                 if ($end_open === false) {
 898                         return false;
 899                 }
 900
 901                 $res = [
 902                         'start' => [
 903                                 'open' => $start_open,
 904                                 'close' => $start_close
 905                         ],
 906                         'end' => [
 907                                 'open' => $end_open,
 908                                 'close' => $end_open + strlen('[/' . $name . ']')
 909                         ],
 910                 ];
 911
 912                 if ($start_equal !== false) {
 913                         $res['start']['equal'] = $start_equal + 1;
 914                 }
 915
 916                 return $res;
 917         }
 918
 919         /**
 920          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 921          *
 922          * @param type $pattern Preg pattern string
 923          * @param type $replace Preg replace string
 924          * @param type $name    BBCode tag name
 925          * @param type $text    Text to search
 926          * @return string
 927          */
 928         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 929         {
 930                 $occurrences = 0;
 931                 $pos = self::getTagPosition($text, $name, $occurrences);
 932                 while ($pos !== false && $occurrences++ < 1000) {
 933                         $start = substr($text, 0, $pos['start']['open']);
 934                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 935                         $end = substr($text, $pos['end']['close']);
 936                         if ($end === false) {
 937                                 $end = '';
 938                         }
 939
 940                         $subject = preg_replace($pattern, $replace, $subject);
 941                         $text = $start . $subject . $end;
 942
 943                         $pos = self::getTagPosition($text, $name, $occurrences);
 944                 }
 945
 946                 return $text;
 947         }
 948
 949         private static function extractImagesFromItemBody($body)
 950         {
 951                 $saved_image = [];
 952                 $orig_body = $body;
 953                 $new_body = '';
 954
 955                 $cnt = 0;
 956                 $img_start = strpos($orig_body, '[img');
 957                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 958                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 959                 while (($img_st_close !== false) && ($img_end !== false)) {
 960                         $img_st_close++; // make it point to AFTER the closing bracket
 961                         $img_end += $img_start;
 962
 963                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 964                                 // This is an embedded image
 965                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 966                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 967
 968                                 $cnt++;
 969                         } else {
 970                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 971                         }
 972
 973                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 974
 975                         if ($orig_body === false) {
 976                                 // in case the body ends on a closing image tag
 977                                 $orig_body = '';
 978                         }
 979
 980                         $img_start = strpos($orig_body, '[img');
 981                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 982                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 983                 }
 984
 985                 $new_body = $new_body . $orig_body;
 986
 987                 return ['body' => $new_body, 'images' => $saved_image];
 988         }
 989
 990         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 991         {
 992                 $newbody = $body;
 993
 994                 $cnt = 0;
 995                 foreach ($images as $image) {
 996                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 997                         // it loops over the array starting from the first element and going sequentially
 998                         // to the last element
 999                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1000                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
1001                         $cnt++;
1002                 }
1003
1004                 return $newbody;
1005         }
1006
1007         /**
1008          * Processes [share] tags
1009          *
1010          * Note: Can produce a [bookmark] tag in the output
1011          *
1012          * @brief Processes [share] tags
1013          * @param array    $share      preg_match_callback result array
1014          * @param bool|int $simplehtml
1015          * @return string
1016          */
1017         private static function convertShare($share, $simplehtml)
1018         {
1019                 $attributes = $share[2];
1020
1021                 $author = "";
1022                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1023                 if (x($matches, 1)) {
1024                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1025                 }
1026
1027                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1028                 if (x($matches, 1)) {
1029                         $author = $matches[1];
1030                 }
1031
1032                 $profile = "";
1033                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1034                 if (x($matches, 1)) {
1035                         $profile = $matches[1];
1036                 }
1037
1038                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1039                 if (x($matches, 1)) {
1040                         $profile = $matches[1];
1041                 }
1042
1043                 $avatar = "";
1044                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1045                 if (x($matches, 1)) {
1046                         $avatar = $matches[1];
1047                 }
1048
1049                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1050                 if (x($matches, 1)) {
1051                         $avatar = $matches[1];
1052                 }
1053
1054                 $link = "";
1055                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1056                 if (x($matches, 1)) {
1057                         $link = $matches[1];
1058                 }
1059
1060                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1061                 if (x($matches, 1)) {
1062                         $link = $matches[1];
1063                 }
1064
1065                 $posted = "";
1066
1067                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1068                 if (x($matches, 1)) {
1069                         $posted = $matches[1];
1070                 }
1071
1072                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1073                 if (x($matches, 1)) {
1074                         $posted = $matches[1];
1075                 }
1076
1077                 // We only call this so that a previously unknown contact can be added.
1078                 // This is important for the function "Model\Contact::getDetailsByURL()".
1079                 // This function then can fetch an entry from the contact table.
1080                 Contact::getIdForURL($profile, 0);
1081
1082                 $data = Contact::getDetailsByURL($profile);
1083
1084                 if (x($data, "name") && x($data, "addr")) {
1085                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1086                 } else {
1087                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1088                 }
1089
1090                 if (x($data, "addr")) {
1091                         $userid = $data["addr"];
1092                 } else {
1093                         $userid = Protocol::formatMention($profile, $author);
1094                 }
1095
1096                 if (x($data, "name")) {
1097                         $author = $data["name"];
1098                 }
1099
1100                 if (x($data, "micro")) {
1101                         $avatar = $data["micro"];
1102                 }
1103
1104                 $preshare = trim($share[1]);
1105                 if ($preshare != "") {
1106                         $preshare .= "<br />";
1107                 }
1108
1109                 switch ($simplehtml) {
1110                         case 1:
1111                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1112                                 break;
1113                         case 2:
1114                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1115                                 break;
1116                         case 3: // Diaspora
1117                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1118
1119                                 $text = trim($share[1]);
1120
1121                                 if ($text != "") {
1122                                         $text .= "<hr />";
1123                                 }
1124
1125                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1126                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1127
1128                                         if ($link != "") {
1129                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1130                                         }
1131                                 } else {
1132                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1133                                 }
1134
1135                                 break;
1136                         case 4:
1137                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1138                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1139                                 $headline .= ":</b><br />";
1140
1141                                 $text = trim($share[1]);
1142
1143                                 if ($text != "") {
1144                                         $text .= "<hr />";
1145                                 }
1146
1147                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1148
1149                                 break;
1150                         case 5:
1151                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1152                                 break;
1153                         case 6: // app.net
1154                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1155                                 break;
1156                         case 7: // statusnet/GNU Social
1157                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1158                                 break;
1159                         case 8: // twitter
1160                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1161                                 break;
1162                         case 9: // Google+/Facebook
1163                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1164
1165                                 if ($link != "") {
1166                                         $text .= "<br /><br />" . $link;
1167                                 }
1168                                 break;
1169                         default:
1170                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1171                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1172                                         try {
1173                                                 $oembed = OEmbed::getHTML($link, $preshare);
1174                                         } catch (Exception $e) {
1175                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1176                                         }
1177
1178                                         $text = $preshare . $oembed;
1179                                 } else {
1180                                         $text = trim($share[1]) . "\n";
1181
1182                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1183
1184                                         $tpl = get_markup_template('shared_content.tpl');
1185                                         $text .= replace_macros($tpl, [
1186                                                 '$profile' => $profile,
1187                                                 '$avatar' => $avatar,
1188                                                 '$author' => $author,
1189                                                 '$link' => $link,
1190                                                 '$posted' => $posted,
1191                                                 '$content' => trim($share[3])
1192                                         ]);
1193                                 }
1194                                 break;
1195                 }
1196
1197                 return $text;
1198         }
1199
1200         private static function removePictureLinksCallback($match)
1201         {
1202                 $text = Cache::get($match[1]);
1203
1204                 if (is_null($text)) {
1205                         $a = get_app();
1206
1207                         $stamp1 = microtime(true);
1208
1209                         $ch = @curl_init($match[1]);
1210                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1211                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1212                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1213                         @curl_exec($ch);
1214                         $curl_info = @curl_getinfo($ch);
1215
1216                         $a->save_timestamp($stamp1, "network");
1217
1218                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1219                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1220                         } else {
1221                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1222
1223                                 // if its not a picture then look if its a page that contains a picture link
1224                                 $body = Network::fetchUrl($match[1]);
1225
1226                                 $doc = new DOMDocument();
1227                                 @$doc->loadHTML($body);
1228                                 $xpath = new DomXPath($doc);
1229                                 $list = $xpath->query("//meta[@name]");
1230                                 foreach ($list as $node) {
1231                                         $attr = [];
1232
1233                                         if ($node->attributes->length) {
1234                                                 foreach ($node->attributes as $attribute) {
1235                                                         $attr[$attribute->name] = $attribute->value;
1236                                                 }
1237                                         }
1238
1239                                         if (strtolower($attr["name"]) == "twitter:image") {
1240                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1241                                         }
1242                                 }
1243                         }
1244                         Cache::set($match[1], $text);
1245                 }
1246
1247                 return $text;
1248         }
1249
1250         private static function expandLinksCallback($match)
1251         {
1252                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1253                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1254                 } else {
1255                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1256                 }
1257         }
1258
1259         private static function cleanPictureLinksCallback($match)
1260         {
1261                 $text = Cache::get($match[1]);
1262
1263                 if (is_null($text)) {
1264                         $a = get_app();
1265
1266                         $stamp1 = microtime(true);
1267
1268                         $ch = @curl_init($match[1]);
1269                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1270                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1271                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1272                         @curl_exec($ch);
1273                         $curl_info = @curl_getinfo($ch);
1274
1275                         $a->save_timestamp($stamp1, "network");
1276
1277                         // if its a link to a picture then embed this picture
1278                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1279                                 $text = "[img]" . $match[1] . "[/img]";
1280                         } else {
1281                                 $text = "[img]" . $match[2] . "[/img]";
1282
1283                                 // if its not a picture then look if its a page that contains a picture link
1284                                 $body = Network::fetchUrl($match[1]);
1285
1286                                 $doc = new DOMDocument();
1287                                 @$doc->loadHTML($body);
1288                                 $xpath = new DomXPath($doc);
1289                                 $list = $xpath->query("//meta[@name]");
1290                                 foreach ($list as $node) {
1291                                         $attr = [];
1292                                         if ($node->attributes->length) {
1293                                                 foreach ($node->attributes as $attribute) {
1294                                                         $attr[$attribute->name] = $attribute->value;
1295                                                 }
1296                                         }
1297
1298                                         if (strtolower($attr["name"]) == "twitter:image") {
1299                                                 $text = "[img]" . $attr["content"] . "[/img]";
1300                                         }
1301                                 }
1302                         }
1303                         Cache::set($match[1], $text);
1304                 }
1305
1306                 return $text;
1307         }
1308
1309         public static function cleanPictureLinks($text)
1310         {
1311                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1312                 return $return;
1313         }
1314
1315         private static function textHighlightCallback($match)
1316         {
1317                 if (in_array(strtolower($match[1]),
1318                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1319                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh'])
1320                 ) {
1321                         return text_highlight($match[2], strtolower($match[1]));
1322                 }
1323                 return $match[0];
1324         }
1325
1326         /**
1327          * @brief Converts a BBCode message to HTML message
1328          *
1329          * BBcode 2 HTML was written by WAY2WEB.net
1330          * extended to work with Mistpark/Friendica - Mike Macgirvin
1331          *
1332          * Simple HTML values meaning:
1333          * - 0: Friendica display
1334          * - 1: Unused
1335          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1336          * - 3: Used before converting to Markdown in bb2diaspora.php
1337          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1338          * - 5: Unused
1339          * - 6: Used for Appnet
1340          * - 7: Used for dfrn, OStatus
1341          * - 8: Used for WP backlink text setting
1342          *
1343          * @param string $text
1344          * @param bool   $try_oembed
1345          * @param int    $simple_html
1346          * @param bool   $for_plaintext
1347          * @return string
1348          */
1349         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1350         {
1351                 $a = get_app();
1352
1353                 /*
1354                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1355                  *
1356                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1357                  * $match[1] = $url
1358                  * $match[2] = $title or absent
1359                  */
1360                 $try_oembed_callback = function ($match)
1361                 {
1362                         $url = $match[1];
1363                         $title = defaults($match, 2, null);
1364
1365                         try {
1366                                 $return = OEmbed::getHTML($url, $title);
1367                         } catch (Exception $ex) {
1368                                 $return = $match[0];
1369                         }
1370
1371                         return $return;
1372                 };
1373
1374                 // Hide all [noparse] contained bbtags by spacefying them
1375                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1376
1377                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1378                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1379                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1380
1381                 // Remove the abstract element. It is a non visible element.
1382                 $text = self::stripAbstract($text);
1383
1384                 // Move all spaces out of the tags
1385                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1386                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1387
1388                 // Extract the private images which use data urls since preg has issues with
1389                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1390                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1391
1392                 $extracted = self::extractImagesFromItemBody($text);
1393                 $text = $extracted['body'];
1394                 $saved_image = $extracted['images'];
1395
1396                 // If we find any event code, turn it into an event.
1397                 // After we're finished processing the bbcode we'll
1398                 // replace all of the event code with a reformatted version.
1399
1400                 $ev = bbtoevent($text);
1401
1402                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1403                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1404
1405                 $text = str_replace("<", "&lt;", $text);
1406                 $text = str_replace(">", "&gt;", $text);
1407
1408                 // remove some newlines before the general conversion
1409                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1410                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1411
1412                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1413                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1414
1415                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1416                 if (!$try_oembed) {
1417                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1418                 }
1419
1420                 // Check for [code] text here, before the linefeeds are messed with.
1421                 // The highlighter will unescape and re-escape the content.
1422                 if (strpos($text, '[code=') !== false) {
1423                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1424                 }
1425                 // Convert new line chars to html <br /> tags
1426
1427                 // nlbr seems to be hopelessly messed up
1428                 //      $Text = nl2br($Text);
1429
1430                 // We'll emulate it.
1431
1432                 $text = trim($text);
1433                 $text = str_replace("\r\n", "\n", $text);
1434
1435                 // removing multiplicated newlines
1436                 if (Config::get("system", "remove_multiplicated_lines")) {
1437                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1438                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1439                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1440                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1441                         do {
1442                                 $oldtext = $text;
1443                                 $text = str_replace($search, $replace, $text);
1444                         } while ($oldtext != $text);
1445                 }
1446
1447                 // Set up the parameters for a URL search string
1448                 $URLSearchString = "^\[\]";
1449                 // Set up the parameters for a MAIL search string
1450                 $MAILSearchString = $URLSearchString;
1451
1452                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1453                 if (!$for_plaintext) {
1454                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1455                         // Currently disabled, since the function is too greedy
1456                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1457                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1458                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1459                         if ($simple_html == 7) {
1460                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1461                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1462                         }
1463                 } else {
1464                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1465                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1466                 }
1467
1468
1469                 // Handle attached links or videos
1470                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1471
1472                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1473
1474                 // Remove all hashtag addresses
1475                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1476                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1477                 } elseif ($simple_html == 3) {
1478                         // The ! is converted to @ since Diaspora only understands the @
1479                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1480                                 '@<a href="$2">$3</a>',
1481                                 $text);
1482                 } elseif ($simple_html == 7) {
1483                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1484                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1485                                 $text);
1486                 } elseif (!$simple_html) {
1487                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1488                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1489                                 $text);
1490                 }
1491
1492                 // Bookmarks in red - will be converted to bookmarks in friendica
1493                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1494                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1495                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1496                                         "[bookmark=$1]$2[/bookmark]", $text);
1497
1498                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1499                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1500                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1501                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1502                 }
1503
1504                 if ($simple_html == 5) {
1505                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1506                 }
1507
1508                 // Perform URL Search
1509                 if ($try_oembed) {
1510                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1511                 }
1512
1513                 if ($simple_html == 5) {
1514                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1515                 } else {
1516                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1517                 }
1518
1519                 // Handle Diaspora posts
1520                 $text = preg_replace_callback(
1521                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1522                         function ($match) {
1523                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1524                         }, $text
1525                 );
1526
1527                 // Server independent link to posts and comments
1528                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1529                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1530                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1531
1532                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1533                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1534
1535                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1536                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1537
1538                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1539                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1540                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1541
1542                 // Red compatibility, though the link can't be authenticated on Friendica
1543                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1544
1545
1546                 // we may need to restrict this further if it picks up too many strays
1547                 // link acct:user@host to a webfinger profile redirector
1548
1549                 $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);
1550
1551                 // Perform MAIL Search
1552                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1553                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1554
1555                 // leave open the posibility of [map=something]
1556                 // this is replaced in prepare_body() which has knowledge of the item location
1557
1558                 if (strpos($text, '[/map]') !== false) {
1559                         $text = preg_replace_callback(
1560                                 "/\[map\](.*?)\[\/map\]/ism",
1561                                 function ($match) {
1562                                         // the extra space in the following line is intentional
1563                                         // Whyyy? - @MrPetovan
1564                                         return str_replace($match[0], '<div class="map"  >' . Map::byLocation($match[1]) . '</div>', $match[0]);
1565                                 },
1566                                 $text
1567                         );
1568                 }
1569                 if (strpos($text, '[map=') !== false) {
1570                         $text = preg_replace_callback(
1571                                 "/\[map=(.*?)\]/ism",
1572                                 function ($match) {
1573                                         // the extra space in the following line is intentional
1574                                         // Whyyy? - @MrPetovan
1575                                         return str_replace($match[0], '<div class="map"  >' . Map::byCoordinates(str_replace('/', ' ', $match[1])) . '</div>', $match[0]);
1576                                 },
1577                                 $text
1578                         );
1579                 }
1580                 if (strpos($text, '[map]') !== false) {
1581                         $text = preg_replace("/\[map\]/", '<div class="map"></div>', $text);
1582                 }
1583
1584                 // Check for headers
1585                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1586                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1587                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1588                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1589                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1590                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1591
1592                 // Check for paragraph
1593                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1594
1595                 // Check for bold text
1596                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1597
1598                 // Check for Italics text
1599                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1600
1601                 // Check for Underline text
1602                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1603
1604                 // Check for strike-through text
1605                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1606
1607                 // Check for over-line text
1608                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1609
1610                 // Check for colored text
1611                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1612
1613                 // Check for sized text
1614                 // [size=50] --> font-size: 50px (with the unit).
1615                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1616                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1617
1618                 // Check for centered text
1619                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1620
1621                 // Check for list text
1622                 $text = str_replace("[*]", "<li>", $text);
1623
1624                 // Check for style sheet commands
1625                 $text = preg_replace_callback(
1626                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1627                         function ($match) {
1628                                 return "<span style=\"" . self::cleanCss($match[1]) . ";\">" . $match[2] . "</span>";
1629                         },
1630                         $text
1631                 );
1632
1633                 // Check for CSS classes
1634                 $text = preg_replace_callback(
1635                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1636                         function ($match) {
1637                                 return "<span class=\"" . self::cleanCss($match[1]) . "\">" . $match[2] . "</span>";
1638                         },
1639                         $text
1640                 );
1641
1642                 // handle nested lists
1643                 $endlessloop = 0;
1644
1645                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1646                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1647                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1648                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1649                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1650                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1651                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1652                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1653                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1654                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1655                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1656                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1657                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1658                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1659                 }
1660
1661                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1662                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1663                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1664                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1665
1666                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1667                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1668
1669                 $text = str_replace('[hr]', '<hr />', $text);
1670
1671                 // This is actually executed in prepare_body()
1672
1673                 $text = str_replace('[nosmile]', '', $text);
1674
1675                 // Check for font change text
1676                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1677
1678                 // Declare the format for [code] layout
1679
1680                 $CodeLayout = '<code>$1</code>';
1681                 // Check for [code] text
1682                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1683
1684                 // Declare the format for [spoiler] layout
1685                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1686
1687                 // Check for [spoiler] text
1688                 // handle nested quotes
1689                 $endlessloop = 0;
1690                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1691                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1692                 }
1693
1694                 // Check for [spoiler=Author] text
1695
1696                 $t_wrote = L10n::t('$1 wrote:');
1697
1698                 // handle nested quotes
1699                 $endlessloop = 0;
1700                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1701                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1702                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1703                                                  $text);
1704                 }
1705
1706                 // Declare the format for [quote] layout
1707                 $QuoteLayout = '<blockquote>$1</blockquote>';
1708
1709                 // Check for [quote] text
1710                 // handle nested quotes
1711                 $endlessloop = 0;
1712                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1713                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1714                 }
1715
1716                 // Check for [quote=Author] text
1717
1718                 $t_wrote = L10n::t('$1 wrote:');
1719
1720                 // handle nested quotes
1721                 $endlessloop = 0;
1722                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1723                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1724                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1725                                                  $text);
1726                 }
1727
1728
1729                 // [img=widthxheight]image source[/img]
1730                 $text = preg_replace_callback(
1731                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1732                         function ($matches) {
1733                                 if (strpos($matches[3], "data:image/") === 0) {
1734                                         return $matches[0];
1735                                 }
1736
1737                                 $matches[3] = proxy_url($matches[3]);
1738                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1739                         },
1740                         $text
1741                 );
1742
1743                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1744                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1745
1746                 // Images
1747                 // [img]pathtoimage[/img]
1748                 $text = preg_replace_callback(
1749                         "/\[img\](.*?)\[\/img\]/ism",
1750                         function ($matches) {
1751                                 if (strpos($matches[1], "data:image/") === 0) {
1752                                         return $matches[0];
1753                                 }
1754
1755                                 $matches[1] = proxy_url($matches[1]);
1756                                 return "[img]" . $matches[1] . "[/img]";
1757                         },
1758                         $text
1759                 );
1760
1761                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1762                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1763
1764                 // Shared content
1765                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1766                         function ($match) use ($simple_html) {
1767                                 return self::convertShare($match, $simple_html);
1768                         }, $text);
1769
1770                 $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);
1771                 $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 />', $text);
1772                 //$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 />', $Text);
1773
1774                 // Try to Oembed
1775                 if ($try_oembed) {
1776                         $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);
1777                         $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);
1778
1779                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1780                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1781                 } else {
1782                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1783                                                 '<a href="$1" target="_blank">$1</a>', $text);
1784                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1785                                                 '<a href="$1" target="_blank">$1</a>', $text);
1786                 }
1787
1788                 // html5 video and audio
1789
1790
1791                 if ($try_oembed) {
1792                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1793                 } else {
1794                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1795                 }
1796
1797                 // Youtube extensions
1798                 if ($try_oembed) {
1799                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1800                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1801                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1802                 }
1803
1804                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1805                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1806                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1807
1808                 if ($try_oembed) {
1809                         $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);
1810                 } else {
1811                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1812                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1813                 }
1814
1815                 if ($try_oembed) {
1816                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1817                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1818                 }
1819
1820                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1821                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1822
1823                 if ($try_oembed) {
1824                         $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);
1825                 } else {
1826                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1827                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1828                 }
1829
1830                 // oembed tag
1831                 $text = OEmbed::BBCode2HTML($text);
1832
1833                 // Avoid triple linefeeds through oembed
1834                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1835
1836                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1837                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1838                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1839                 // start which is always required). Allow desc with a missing summary for compatibility.
1840
1841                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1842                         $sub = format_event_html($ev, $simple_html);
1843
1844                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1845                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1846                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1847                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1848                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1849                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1850                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1851                 }
1852
1853                 // Replace non graphical smilies for external posts
1854                 if ($simple_html) {
1855                         $text = Smilies::replace($text, false, true);
1856                 }
1857
1858                 // Replace inline code blocks
1859                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1860                         function ($match) use ($simple_html) {
1861                                 $return = '<key>' . $match[1] . '</key>';
1862                                 // Use <code> for Diaspora inline code blocks
1863                                 if ($simple_html === 3) {
1864                                         $return = '<code>' . $match[1] . '</code>';
1865                                 }
1866                                 return $return;
1867                         }
1868                 , $text);
1869
1870                 // Unhide all [noparse] contained bbtags unspacefying them
1871                 // and triming the [noparse] tag.
1872
1873                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1874                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1875                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1876
1877
1878                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1879                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1880                 $text = preg_replace('/\&quot\;/', '"', $text);
1881
1882                 // fix any escaped ampersands that may have been converted into links
1883                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1884
1885                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1886                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1887                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1888                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1889
1890                 // sanitize href attributes (only whitelisted protocols URLs)
1891                 // default value for backward compatibility
1892                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1893
1894                 // Always allowed protocol even if config isn't set or not including it
1895                 $allowed_link_protocols[] = 'http';
1896                 $allowed_link_protocols[] = 'redir/';
1897
1898                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1899                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1900
1901                 if ($saved_image) {
1902                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1903                 }
1904
1905                 // Clean up the HTML by loading and saving the HTML with the DOM.
1906                 // Bad structured html can break a whole page.
1907                 // For performance reasons do it only with ativated item cache or at export.
1908                 if (!$try_oembed || (get_itemcachepath() != "")) {
1909                         $doc = new DOMDocument();
1910                         $doc->preserveWhiteSpace = false;
1911
1912                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1913
1914                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1915                         $encoding = '<?xml encoding="UTF-8">';
1916                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1917                         $doc->encoding = 'UTF-8';
1918                         $text = $doc->saveHTML();
1919                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1920
1921                         $text = str_replace('<br></li>', '</li>', $text);
1922
1923                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1924                 }
1925
1926                 // Clean up some useless linebreaks in lists
1927                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1928                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1929                 //$Text = str_replace('</li><br />', '</li>', $Text);
1930                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1931                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1932
1933                 Addon::callHooks('bbcode', $text);
1934
1935                 return trim($text);
1936         }
1937
1938         /**
1939          * @brief Strips the "abstract" tag from the provided text
1940          *
1941          * @param string $text The text with BBCode
1942          * @return string The same text - but without "abstract" element
1943          */
1944         public static function stripAbstract($text)
1945         {
1946                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1947                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1948
1949                 return $text;
1950         }
1951
1952         /**
1953          * @brief Returns the value of the "abstract" element
1954          *
1955          * @param string $text The text that maybe contains the element
1956          * @param string $addon The addon for which the abstract is meant for
1957          * @return string The abstract
1958          */
1959         private static function getAbstract($text, $addon = "")
1960         {
1961                 $abstract = "";
1962                 $abstracts = [];
1963                 $addon = strtolower($addon);
1964
1965                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1966                         foreach ($results AS $result) {
1967                                 $abstracts[strtolower($result[1])] = $result[2];
1968                         }
1969                 }
1970
1971                 if (isset($abstracts[$addon])) {
1972                         $abstract = $abstracts[$addon];
1973                 }
1974
1975                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1976                         $abstract = $result[1];
1977                 }
1978
1979                 return $abstract;
1980         }
1981 }